close

Вход

Забыли?

вход по аккаунту

?

Microwave-assisted Extraction of Tomato Peels and Physicochemical Stability, In Vitro Bioaccessibility, and Cellular Uptake of Lycopene-Loaded Emulsions

код для вставкиСкачать
MICROWAVE-ASSISTED EXTRACTION OF TOMATO PEELS
AND PHYSICOCHEMICAL STABILITY, IN VITRO
BIOACCESSIBILITY, AND CELLULAR UPTAKE OF LYCOPENELOADED EMULSIONS
by
Kacie K.H.Y. Ho
A Dissertation
Submitted to the Faculty of Purdue University
In Partial Fulfillment of the Requirements for the degree of
Doctor of Philosophy
Department of Food Science
West Lafayette, Indiana
August 2017
ProQuest Number: 10615866
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
ProQuest 10615866
Published by ProQuest LLC (2017 ). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
ii
THE PURDUE UNIVERSITY GRADUATE SCHOOL
STATEMENT OF COMMITTEE APPROVAL
Dr. M. Fernanda San Martin-González, Chair
Department of Food Science
Dr. Andrea M. Liceaga
Department of Food Science
Dr. Ganesan Narsimhan
Department of Agricultural and Biological Engineering
Dr. Mario G. Ferruzzi
Department of Food, Bioprocessing and Nutrition Science
North Carolina State University
Dr. Bernhard van Lengerich
Former Chief Science Officer, VP Technology Strategy
General Mills
Approved by:
Dr. Carlos M. Corvalan
Head of the Graduate Program
iii
I dedicate this to:
my parents who have sacrificed, in more ways than I can imagine, so I could grow into
the person I am today.
I love you, Mom, and I 錫爸爸.
and Jordan Oshiro, who literally fed me during grad school. Thanks for cooking, packing
me food whenever I worked late, and making sure I never went hungry.
おいしかった.
iv
ACKNOWLEDGEMENTS
Fellowship support was provided by the National Science Foundation Graduate
Research Fellowship (Grant No. DGE-1333468), the Purdue Department of Food
Science Industry Fellows Program, and the Purdue Doctoral Fellowship. Project funds
were supported by the National Science Foundation Graduate Research Opportunities
World Wide (GROW) program in support with the Netherlands Organisation for
Scientific Research (NWO), and the Wageningen University Graduate School (VLAG) as
well as the Purdue Department of Food Science Industry Fellows Program
I would like to express my deepest gratitude to my major professor, Dr. M.
Fernanda San Martín-González, who has been incredibly patient and supportive of all my
academic, research, and career-related pursuits. Thank you for being my trusted mentor
and advisor during my career as a Ph.D. student. I’d also like to thank my committee
member, Dr. Mario G. Ferruzzi, for his guidance with research, professional
development, and also for welcoming me into his cell lab and allowing me to work with
his group at North Carolina State University. I am also very grateful for my committee
members, Dr. Andrea M. Liceaga, Dr. Bernhard van Lengerich, and Dr. Ganesan
Narsimhan, who have guided me throughout the Ph.D. process and with my dissertation
research. My research supervisors from Wageningen University, Dr. Claire BertonCarabin and Dr. Karin Schroën, have been instrumental in my research pursuits and my
scholarly development. Thank you for taking a chance on the random American who
emailed you to initiate a collaboration, for welcoming me into your research group, and
for all your time and sharing your expertise.
I probably never would have attended graduate school if it were not for Dr.
Wayne Iwaoka, my former academic advisor at the University of Hawaiʻi. Thank you for
investing so much care and time into mentoring me, for providing so many amazing
learning opportunities, and for pushing me out of my comfort zone. Thanks to you, that
quiet girl from Waikoloa Village is not afraid to venture to all corners of the world.
Mahalo nui loa to my University of Hawaiʻi CTAHR ʻohana, in particular Dr.
Soojin Jun. Thank you for giving me my first lab research experience and for your
v
continual guidance and support (I might not have had graduate funding if you had not
encouraged me to apply for the NSF Fellowship). Working with you and your graduate
students not only opened my eyes to the world of Food Engineering and Processing, but
also taught me perseverance and dedication.
I have a long list of people at Purdue University who have helped support me
throughout my graduate school journey. I would like to thank my lab mates, in
particular, Veronica Rodriguez-Martinez and Simran Kaur who were, and to this day still
are, my lab support system. Thanks for making life in a yellow-lighted lab pleasant and
for helping West Lafayette, Indiana feel a little bit more like a home. Outside of our lab,
I also received valuable technical assistance from the Ferruzzi and Hamaker lab members
(past and present). Thanks for always being kind and helpful carotenoid gurus
(especially Hawi Debelo, Ingrid Aragon, Darwin Ortiz, Tristan Lipkie, and Milena Leon
Garcia) and for sharing cell culture wisdom (Mohammad Chegeni, Sydney Moser, and
Marwa Mohamed). Thank you to my officemates, in particular Ximena Yepez, Kaitlin
Kaczay, and Shanleigh Thompson, for moral support and for keeping me company during
those late working and studying nights. Thank you to all my Purdue friends, especially
Beth Pletsch, Ben Redan, Chrisitian Wright, Ethan Braun, Pablo Torres, Celina To, and
Tomasz Wilmanski for all the support over the years.
Dank u wel to the Food Process Engineering Group at Wageningen University
and Research Center (WUR) for a gezellig experience. Although my stay was less than a
year, the experience was life-changing. I was blessed to be welcomed into a group of
brilliant researchers who shared their resources and time with me. Anja Schröder, thank
you for all your guidance with Lissajous-plots, especially while I was just starting my
research with interfacial rheology. Kelly Muijlwijk and Meinou Corstens, thank you for
your friendship and patience, especially with the ADT. I’d also like to acknowledge my
student, Jettie Faber, for all her hard work on lycopene-loaded solid lipid nanoparticles,
which parallels this dissertation research. Bedankt to all my WUR friends, especially, the
former occupants of the K1.24 office and the FPE coffee break attendees. Thank you for
your friendship and for helping me grow as a scientist and as an individual.
I’d like to thank my parents, Patrick and Edna Ho, and my brother, Andrew Ho,
for being supportive through the years and putting up with the awkward 5 or 6-hour time
vi
difference. Thanks for always being there for me. Last, but certainly not least, I’d like to
thank Jordan Oshiro. Thank you for putting up with my late night laboratory sessions,
taking care of me whenever I got sick, listening to all my practice presentations, and for
being so understanding even though I always seem to find some other country to run off
to.
vii
TABLE OF CONTENTS
LIST OF TABLES ............................................................................................................ xii
LIST OF FIGURES ......................................................................................................... xiii
LIST OF ABBREVIATIONS .......................................................................................... xix
ABSTRACT.......................................................................................................................xx
CHAPTER 1 REVIEW OF THE LITERATURE ...............................................................1
1.1 Introduction ............................................................................................................... 1
1.2 Phytochemicals and their functionality in food ........................................................ 2
1.2.1 Natural colors ..................................................................................................... 2
1.2.2 Antioxidants ....................................................................................................... 6
1.2.3 Bioactive ingredients and bioavailibility ........................................................... 6
1.4 Extraction techniques for bioactive phytochemical extraction ................................. 8
1.4.1 Conventional extraction techniques ................................................................... 8
1.4.2 Ultrasound-assisted extraction ........................................................................... 9
1.4.3 Supercritical fluid extraction............................................................................ 10
1.4.4 Microwave-assisted extraction......................................................................... 11
1.5 Delivery systems for lipophillic bioactive compounds........................................... 13
1.5.1 Emulsion-based delivery systems .................................................................... 14
1.5.2 Emulsifiers and droplet stabilization ............................................................... 17
1.5.3 Factors affecting stability ................................................................................. 18
1.5.4 Fabrication and droplet size reduction ............................................................. 20
1.5.5 Stabilization of phytochemicals and potential enhanced bioavailibility.......... 24
1.6 Digestion, bioaccessibility, and bioavailability of carotenoids and emulsion
systems .......................................................................................................................... 25
1.5.4 Techniques for assessing in vitro bioaccessibility and bioavailability ............ 26
1.5.5 Carotenoid digestion and absorption ............................................................... 28
1.6.3 Potential fate of emulsion delivery systems in the gastrointestinal tract ......... 29
1.7 Aims of the research ............................................................................................... 30
1.8 References ............................................................................................................... 32
viii
CHAPTER 2 MICROWAVE-ASSISTED EXTRACTION OF LYCOPENE IN
TOMATO PEELS: EFFECT OF EXTRACTION CONDITIONS ON ALL-TRANS AND
CIS- ISOMER YIELDS .....................................................................................................51
2.1 Introduction ............................................................................................................. 51
2.2 Materials and methods ............................................................................................ 53
2.2.1 Reagents and standards .................................................................................... 53
2.2.3 Raw materials and sample preparation ............................................................ 53
2.2.4 Experimental design......................................................................................... 54
2.2.5 Microwave-assisted extraction of lycopene ..................................................... 55
2.2.6 Conventional extraction of lycopene ............................................................... 56
2.2.7 Quantification with high performance liquid chromatography (HPLC-DAD) 56
2.2.8 Electron microscopy imaging .......................................................................... 57
2.3 Results and discussion ............................................................................................ 58
2.3.1 Lycopene recovery of low EA extractions ....................................................... 58
2.3.2 Lycopene recovery of high EA extractions...................................................... 64
2.3.3 Transmission electron microscopy of tomato peels ......................................... 70
2.4 Conclusions ............................................................................................................. 71
2.5 Acknowledgements ................................................................................................. 72
2.6 References ............................................................................................................... 73
CHAPTER 3 PHYSICOCHEMICAL STABILITY OF LYCOPENE-LOADED
EMULSIONS STABILIZED BY PLANT OR DAIRY PROTEINS ................................76
3.1 Introduction ............................................................................................................. 76
3.2 Materials and methods ............................................................................................ 79
3.2.1 Materials .......................................................................................................... 79
3.2.2 Preparation of lycopene oil stock ..................................................................... 79
3.2.3 Preparation of the aqueous phase ..................................................................... 80
3.2.4 Preparation of lycopene-loaded emulsions ...................................................... 80
3.2.5 Particle size measurment .................................................................................. 81
3.2.6 Zeta potential measurement ............................................................................. 82
3.2.8 Adsorption kinetics of interfacial films ........................................................... 83
3.2.9 Interfacial rheology of interfacial films ........................................................... 83
ix
2.2.5. Statistical analysis ........................................................................................... 84
3.3 Results and discussion ............................................................................................ 84
3.3.1 Physical stability of lycopene-loaded emulsions ............................................. 84
3.3.2 Encapsulation stability of lycopene-loaded emulsions .................................... 87
3.3.3 Adsorption kinetics .......................................................................................... 88
3.3.4 Interfacial rheology .......................................................................................... 89
3.3.5 Comparison and design considerations for protein-stabilized emulsions ........ 90
3.4 Conclusions ............................................................................................................. 92
3.5 Acknowledgements ................................................................................................. 93
3.6 References ............................................................................................................... 95
CHAPTER 4 SYNERGISTIC AND ANTAGONISTIC EFFECTS OF PLANT AND
DAIRY PROTEIN BLENDS ON THE PHYSICOCHEMICAL STABILITY OF
LYCOPENE-LOADED EMULSIONS ...........................................................................102
4.1 Introduction ........................................................................................................... 102
4.2 Materials and methods .......................................................................................... 104
4.2.1 Materials ........................................................................................................ 104
4.2.2 Lycopene extraction and oil phase preparation.............................................. 104
4.2.3 Aqueous phase preparation ............................................................................ 105
4.2.4 Emulsion preparation ..................................................................................... 105
4.2.6 Zeta potential measurement ........................................................................... 106
4.2.7 Determination of lycopene retention in emulsions ........................................ 106
4.2.8 Adsorption kinetics of interfacial films ......................................................... 107
4.2.9 Interfacial rheology of interfacial films ......................................................... 108
4.2.10 Statistical analysis ........................................................................................ 109
4.3 Results and discussion .......................................................................................... 109
4.3.1 Physical stability of lycopene-loaded emulsions ........................................... 109
4.3.2 Chemical stability of lycopene in emulsions ................................................. 113
4.3.3 Adsorption kinetics ........................................................................................ 115
4.3.4 Interfacial rheology ........................................................................................ 117
4.3.5 Comparison of protein-protein blends and potential droplet stabilization
mechanisms ............................................................................................................. 120
x
4.4 Conclusions ........................................................................................................... 122
4.5 Acknowledgements ............................................................................................... 123
4.6 References ............................................................................................................. 124
CHAPTER 5 BIOACCESSIBILITY AND CELLULAR UPTAKE OF WHEY AND
PEA PROTEIN LYCOPENE-LOADED EMULSIONS.................................................130
5.1 Introduction ........................................................................................................... 130
5.2 Materials and Methods.......................................................................................... 130
5.2.1 Materials ........................................................................................................ 130
5.2.2 Preparation of emulsion aqueous phase ......................................................... 131
5.2.3 Preparation of emulsion oil phase .................................................................. 131
5.2.4 Emulsion fabrication ...................................................................................... 132
5.2.5 Droplet size and surface charge measurement ............................................... 132
5.2.6 In vitro digestion of emulsions ...................................................................... 132
5.2.7 Cell culture and treatments ............................................................................ 133
5.2.8 Lycopene extraction and quantification via HPLC ........................................ 134
5.2.9 Statistical analysis .......................................................................................... 135
5.3 Results and discussion .......................................................................................... 135
5.3.1 Emulsion characterization .............................................................................. 135
5.3.2 Lycopene content and digestive stability ....................................................... 136
5.3.3 In vitro bioaccessibility of lycopene .............................................................. 138
5.3.4 Cellular uptake of lycopene ........................................................................... 141
5.4 Conclusions ........................................................................................................... 143
5.5 Acknowledgements ............................................................................................... 144
5.6 References ............................................................................................................. 145
CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS ....................................150
6.1 Summary and overall conclusions ........................................................................ 150
6.2 Future directions ................................................................................................... 151
6.3 References ............................................................................................................. 153
APPENDIX A SUPPLEMENTARY TABLES ......................................................... 154
APPENDIX B SUPPLEMENTARY FIGURES ........................................................ 159
APPENDIX C PUBLISHED ABSTRACTS .............................................................. 162
xi
VITA ................................................................................................................................165
xii
LIST OF TABLES
Table 1-1 Common color pigments from plant and animal sources ....................................3
Table 2-1 Response surface methodology parameters.....................................................55
Table 2-2 Cis-, trans-, and total lycopene yields from low EA MAE ...............................61
Table 2-3 Cis-, trans-, and total lycopene yields from high EA MAE .............................66
Table 3-1 Summary comparison of physical and chemical properties lycopene-loaded
emulsions stabilized with WPI, SC, SPI, or PPI. Proteins that strongly demonstrated
relatively high (++++) values for a given characteristic are compared against those
with intermediate (+++ or ++) and lower (+) values. ................................................93
Table A- 1 Treatment times (seconds) for MAE of tomato peels based off of power and
energy equivalents ...................................................................................................154
Table A- 2 Particle size, d3,2 of lycopene-loaded emulsions over time. ..........................155
Table A- 3 Span of lycopene-loaded emulsions overtime ...............................................156
Table A- 4 Encapsulation efficiency (%) of lycopene-loaded emulsions fabricated with
protein and protein blends. ......................................................................................157
Table A- 5 Lycopene retention in emulsions stabilized with SC-blends .........................158
xiii
LIST OF FIGURES
Figure 1-1 Chemical structure of all-trans-lycopene and its isomers. .................................8
Figure 1-2 Schematic of droplet behavior that leads to destabilization. Small, stable
droplets can undergo reversible creaming or flocculation, both of which can lead to
coalesence if the interfacial film is ruptured. Small droplets are especially
susceptible to Ostwald ripening and will diffuse through the continuous phase to
become incorporated into larger droplets. Both coalesced droplets and large droplets
resulting from Ostwald ripening can lead to phase separation. .................................19
Figure 1-3 Single- and two-compartment model for in vitro cellular uptake and
transport/chylomicron secretion, respectively. In the two-compartment model, the
cells are grown on an insert within the apical compartment. Carotenoids that are
collected from the basolateral compartment are assumed to be transported out of the
enterocyte. ..................................................................................................................28
Figure 1-4 Schematic of lycopene absorption. Food and lipids are digested allowing for
release of mono- and diglycerols, free fatty acids, and lycopene. Bile salts, monoand diglycerides, and cholesterol stabilize mixed micelles, which can then transport
lipophilic compounds across the unstirred water layer to cross into the enterocyte
passively (diffusion) or actively through a membrane protein receptor (SR-BI). .....29
Figure 1-5 Schematic of potential delivery system behavior and bioactive compound
uptake. A formulated delivery system may be resistant or partially resistant to
digestion. Complexes formed with biopolymers or stable delivery systems may be
capable of transporting bioactive compounds across the unstirred water layer
without needing to be reorganized into a mixed micelle. Uptake into the enterocyte
is expected to follow similar pathways as that of conventional carotenoid uptake
(passive diffusion or facilitiated difussion via a membrane protein). If particles,
complexes, or delivery systems are small enough, they may be capable of passing
between the enterocytes (paracelluar). .......................................................................30
xiv
Figure 2-1 Representative chromatogram of carotenoid extract from MAE of tomato
peels at 470 nm. Suspected peak identies are as follow: (a) β-carotene, (b) cislycopene isomer, (c) cis-lycopene isomer, (d) all-trans-lycopene, (e) 5-cis-lycopene.59
Figure 2-2 Response surface plots for all-trans-lycopene yield from low EA MAE with
solvent ratio vs. power (top row) and energy vs. power (bottom row) plotted. Power
levels are fixed at (a) 24kJ, (b) 36 kJ, and (c) 48 kJ and solvent ratios are fixed at (d)
1:0 hexane:EA, (e) 1.5:0.5 hexane:EA, and (f) 1:1 mL hexane : mL EA solvent ratio.
The maximum predicted extraction yield was (g) 10.362 mg/100g with a treatment
comprising of: 1:1 mL hexane : mL EA solvent ratio, 1600 W, 24 kJ. Plotted
response values represent predicted values from the model. .....................................60
Figure 2-3 Response surface plot for cis-lycopene yield from low EA MAE. The
maximum cis-isomer extraction yield (a) was predicted to be 4.450 mg/100g with a
solvent ratio of 1:1 mL hexane: mL EA and a 1:20 solid-liquid ratio. .....................63
Figure 2-4 Comparison of control (conventional) methods vs. optimized low EA MAE.
The MAE conditions used (1:1 solvent ratio, 1:20 solid-liquid ratio, 1600 W, 24 kJ
equivalents for 15 seconds) were determined as optimal by RSM. Extraction yields
of cis, trans, and total lycopene are shown where same letters denote values that are
not significantly different at the α=0.05 level based on the Tukey Kramer method
for pairwise comparisons. Response values shown represent the mean + SD (n=3).64
Figure 2-5 Response surface plot for all-trans-lycopene yield from high EA MAE. The
maximum all-trans-extraction yield (a) was predicted to be 13.872 mg/100g with a
full EA solvent and when treated at 400 W. Solvent ratio significantly affected the
extraction yield (P=0.004) while power did not (P=0.210). Plotted response values
indicate mean + SD (n=3). .........................................................................................67
Figure 2-6 Comparison of control (conventional) methods vs. the high EA MAE
treatment with the highest all-trans-lycopene yield. The MAE conditions used (0:1
solvent ratio, 1:20 solid-liquid ratio, 400 W, 24 kJ equivalents for 60 seconds) were
determined as optimal by RSM. Extraction yields of cis, trans, and total lycopene
are shown where same letters denote values that are not significantly different at the
α=0.05 level based on the Tukey Kramer method for pairwise comparisons.
Response values shown represent the mean + SD (n=3). ..........................................68
xv
Figure 2-7 TEM images of a fresh tomato peel with no extraction (a), byproduct tomato
peel with no extraction (b), byproduct tomato peel subjected to control extraction for
30 minutes (c), and byproduct tomato peel subjected to MAE (1:1 solvent ratio, 1:20
solid-liquid ratio, 1600 W, 24 kJ, for 15 seconds) (d). Visibly more holes and
fissures are present in extracted samples, thus suggesting that MAE, and to some
extent conventional extraction, cause structural disruption. Scale bars indicate 1 µm.71
Figure 3-1 Particle size (d4,3; left y-axis) of lycopene-loaded emulsions over time.
Response values shown represent the mean + SD (n=3), with letters denoting
samples that are significantly different at a given storage time (α=0.05) ..................85
Figure 3-2 Comparison of particle size distribution of lycopene-loaded emulsions
stabilized with WPI (A), SC (B), SPI (C), and PPI (D) at day 0 and day 14, with and
without 1% SDS. Identical distributions with and without SDS dilution suggest that
flocculation did not occur in such samples. When Day 0 and Day 14 distributions
are identical the emulsions are stable. ........................................................................86
Figure 3-3 Relative retention of lycopene, as a function of time for lycopene-loaded
emulsions. Response values shown represent the mean + SD (n=3), with same letters
denoting values that are not significantly different (α=0.05). ....................................87
Figure 3-4 Adsorption kinetics of WPI (A), SC (B), SPI (C), and PPI (D) at the O/W
interface as a function of time (log scale). The slope of the line correlates with the
rate of adsorption to the interface. The dashed line represents the interfacial tension
of the stripped O/W interface in the absence of protein at ~36 mN/m. .....................88
Figure 3-5 Elastic (filled shapes) and loss (open shapes) moduli (A) and loss tangent (B)
of proteins at deformations between 0.03-0.35. Higher loss tangent values indicate a
more viscous response, while lower values indicate a more elastic behavior.
Response values shown represent the mean + SD (n=3). Statistical differences
amongst protein films are shown (B) with same letters denoting values that are not
significantly different (α=0.05). .................................................................................90
Figure 4-1 Examples of Lissajous-Bowditch curves depicting viscous (a), viscoelastic (b),
elastic (c), and non-linear viscoelastic (d) interfaces. Figure adapted from
Deshpande (2010) and Sagis & Scholten (2014). ....................................................109
xvi
Figure 4-2 Visual appearance of lycopene-loaded emulsions stabilized by WPI (a), SC
(b), SPI (c), PPI (d) or protein blends, 1:1 SPI-WPI (e), 3:1 SPI-WPI (f), 1:1 PPIWPI (g), 3:1 PPI-WPI (h), 1:1 SPI-SC (i), 1:1 PPI-SC (j), 1:1 SPI-PPI (k) on day 14
of storage. SC-blend samples (i, j) exhibit a lighter color compared to the other
emulsions and an orange creamed layer (highlighted in the dashed line box). .......110
Figure 4-3 Comparison of droplet size distributions of lycopene-loaded emulsions
stabilized with 1:1 SPI-WPI (A), 1:1 PPI-WPI (B), 3:1 SPI-WPI (C), 3:1 PPI-WPI
(D), 1:1 SPI-SC (E), 1:1 PPI-SC (F), and 1:1 SPI-PPI (G) at day 0 and 14 with and
without 1% SDS. Identical distributions with and without SDS dilution suggest that
flocculation did not occur. When Day 0 and Day 14 distributions are identical the
emulsions are physically stable. ...............................................................................110
Figure 4-4 Droplet size, d3,2 of lycopene-loaded emulsions over time. Response values
shown represent the mean + SD (n=3). At day 14, an asterisk (*) denotes a value that
is significantly (p > 0.05) different. Statistical differences for values at days all
other time points (days 0, 3, and 7) are listed in Appendix A,Table A-2 ................111
Figure 4-5 Initial zeta potential of lycopene-loaded emulsions fabricated with proteins
and protein blends. The results previously obtained for emulsions stabilized with the
individual proteins are displayed as reference (within the gray dashed-line box),
while the results obtained for emulsions stabilized with protein blends are shown to
the right. Data shown represent the mean + SD (n=3), with same superscript letters
denoting values that are not significantly different (p > 0.05). ................................113
Figure 4-6 Percent lycopene retention after 14 days of storage. The results previously
obtained for emulsions stabilized with the individual proteins are displayed as
reference (within the gray dashed-line box), while the results obtained for emulsions
stabilized with protein blends are shown to the right. Values shown represent the
mean + SD (n=3), with same superscript letters denoting values that are not
significantly different (p > 0.05). .............................................................................114
Figure 4-7 Adsorption kinetics of the three stages of adsorption, typically expected for
the dynamic interfacial tension response of proteins (A) is shown as an example
adapted from Beverung, Radke, & Blanch (1999) and the observed adsorption
kinetics for SPI and WPI (B), PPI and WPI (C), and SC and plant protein blends (D)
xvii
at the O-W interface. The interfacial tension as function of time for 3:1 SPI-WPI (a),
1:1 SPI-WPI (b), 3:1 PPI-WPI (c), 1:1 PPI-WPI (d), 1:1 SPI-PPI (e), 1:1 SPI-SC (f),
and 1:1 PPI-SC (g) are plotted as data points connected by solid lines. Dashed lines
correspond to individual proteins (WPI, SPI, PPI, and SC) and the interfacial tension
of oil-water interface is represented by a blue dashed horizontal line for reference.116
Figure 4-8 Comparison of (A) WPI blends and (B) SC blends + SPI-PPI blend elastic
moduli, E’d (a) and loss moduli, E’’d (b) for protein films at the oil–water interface;
and (C) loss tangent (E’’d/E’d) at 0.15 deformation. Results for individual proteins
are shown as: (A, B) solid lines (E’d), dashed lines (E’’d) and (C) black bars for
reference. Data shown represent mean + SD (n=3). Statistical differences amongst
protein films are shown (C) with same letters denoting values that are not
significantly different (α = 0.05). .............................................................................118
Figure 4-9 Lissajous-Bowditch plots of interfacial films prepared with proteins and
protein blends at the oil-water interface (amplitude = 0.1). Surface pressure (π) is
plotted against the applied deformation ...................................................................119
Figure 4-10 Lissajous-Bowditch plots of interfacial films prepared with proteins and
protein blends at the oil-water interface, at deformations (∆A/A0 (-)) 0.03 (green),
0.05 (blue), 0.1 (yellow), 0.15 (orange), 0.25 (gray), 0.3 (light blue). Surface
pressure (π) is plotted against the applied deformation. ..........................................120
Figure 4-11 Schematic of possible scenarios for protein blend behavior at the oil-water
interface....................................................................................................................121
Figure 5-1 Droplet size distributions of lycopene-loaded whey and pea protein emulsions
(solid lines) and diluted with 10% SDS (dashed lines) to estimate the size of nonflocculated droplets. .................................................................................................136
Figure 5-2 Representative chromatograms of lycopene extracted from protein-stabilized
emulsions (whey or pea), lycopene oil, and tomato paste. Peaks for cis-isomers (a)
and all-trans-lycopene (b) are shown for all samples. The 5-cis-isomer can be
observed as a right shoulder off of the all-trans-lycopene peak. .............................137
Figure 5-3 Lycopene content of emulsions (whey or pea protein-stabilized), lycopene oil
(LO), and tomato paste (TP) before (A) and after (B) digestion. Response values
xviii
shown represent the mean + SD (n=3), with same letters denoting values that are not
significantly different (α=0.05). ...............................................................................138
Figure 5-4 Micellarization efficiency of lycopene from emulsions (whey or pea proteinstabilized), lycopene oil (LO), or tomato paste (TP). Response values shown
represent the mean + SD (n=3), with same letters denoting values that are not
significantly different (α=0.05). ...............................................................................140
Figure 5-5 Cellular uptake efficiency (A) and absolute accumulation (B) of cis- and
trans-lycopene after 3 hours of treatment with emulsions (whey or pea proteinstabilized), lycopene oil (LO), or tomato paste (TP). Response values shown
represent the mean + standard deviation (n=3), with same letters denoting values
that are not significantly different (α=0.05). ............................................................141
Figure B- 1 Determination of optimal protein concentration. Particle size (left y-axis) and
correlating percent of excess protein (right y-axis) versus protein concentration
added to the emulsion for WPI (A), SC (B), SPI (C), and PPI (D). Dashed line
denotes the selected protein concentration. ........................................................... 159
Figure B- 2 Span of lycopene-loaded emulsions over time. Response values shown
represent the mean + SD (n=3), with same letters denoting values that are not
significantly different (α=0.05). ..............................................................................160
Figure B- 3 Initial zeta potential of lycopene-loaded emulsions fabricated with proteins
and protein blends. Response values shown represent the mean + SD (n=3), with
same letters denoting values that are not significantly different (α=0.05)..............160
Figure B- 4 Encapsulation efficiency of lycopene in protein stabilized emulsions at t=14
days. Response values shown represent the mean + SD (n=3), with same letters
denoting values that are not significantly different (α=0.05)..................................161
Figure B- 5 Cis- (A) and all-trans-lycopene (B) uptake from emulsions and tomato paste
overtime. Response values shown represent the mean + SD (n=4), with same letters
denoting values that are not significantly different (α=0.05)..................................161
xix
LIST OF ABBREVIATIONS
H2 O
CO2
UAE
SFE
MAE
EA
RSM
BHT
SD
HPLC-DAD
LOD
LOQ
TEM
ANOVA
O/W
WPI
SC
SPI
PPI
CLA
BCA
SDS
LO
TP
Water
Carbon dioxide
Ultrasound-assisted extraction
Supercritical fluid extraction
Microwave-assisted extraction
Ethyl acetate
Response surface methodology
Butylated hydroxytoluene
Standard deviation
High performance liquid chromatography with a diode array
detector
Limit of detection
Limit of quantiation
Transmission electron microscopy
Analysis of variance
Oil-in-water
Whey protein isolate
Sodium caseinate
Soy protein isolate
Pea protein isolate
Conjugated linoleic acid
Bicinchoninic acid
Sodium dodecyl sulfate
Lycopene oil
Tomato paste
xx
ABSTRACT
Author: Ho, Kacie, K.H.Y. PhD
Institution: Purdue University
Degree Received: August 2017
Title: Microwave-assisted Extraction of Tomato Peels and Physicochemical Stability, In
Vitro Bioaccessibility, and Cellular Uptake of Lycopene-Loaded Emulsions.
Major Professor: M. Fernanda San Martin-González,
Chronic disease is responsible for roughly two-thirds of deaths globally. Although a
variety of factors contribute, phytochemical consumption may lower the risk of several
diseases. Many phytochemicals act as functional pigments or antioxidants, can be found
in manufacturing waste streams, and are often underutilized or discarded. Unfortunately,
many phytochemicals are poorly extractable, unstable, and have limited bioavailability
and efficient strategies must be developed in order to use these compounds. Lycopene, a
hydrophobic and labile compound, was selected as a model carotenoid to investigate
microwave-assisted extraction (MAE) and for developing a protein-stabilized emulsion
delivery system. Dairy proteins are commonly used as emulsifiers due to their favorable
functional properties; however, plant proteins offer a relatively sustainable alternative to
reduce the dependence on animal-derived products. Thus, the objectives of this research
were to: 1) determine the effect of MAE conditions on cis- and trans-lycopene recovery
from tomato peels, a processing byproduct, 2) determine the effect of dairy or plant
protein on lycopene-loaded emulsion physicochemical stability, 3) investigate the effect
of mixed protein interfaces on interfacial rheology, and 4) assess in vitro bioaccessibility
and cellular uptake of lycopene from emulsions. Findings indicated that MAE saved time
by extracting significantly more all-trans-lycopene from tomato peels compared to a
slower, conventional extraction. MAE-processed tomato peels exhibited more structural
disruption compared to others, which may have allowed for better lycopene recovery.
Pea protein stabilized emulsions by forming viscoelastic, interfacial protein films and
retained lycopene similarly to casein, an industry standard. Interestingly, binary blends
containing whey with soy or pea protein improved emulsion stability compared to those
stabilized with only one protein. Despite having the best individual stability, casein
exhibited antagonistic effects when blended with plant proteins. In vitro studies indicated
xxi
that pea protein emulsions had a significantly lower micellarization efficiency, but
significantly higher uptake efficiencies compared to whey emulsions, non-emulsified
lycopene oil, and tomato paste. Mechanisms for interfacial film behavior and
bioaccessibility were proposed as these findings open opportunities for future research
endeavors. Overall, these results indicate that MAE reduces lycopene recovery time and
that plant proteins have favorable functionality for the delivery of lipophilic bioactives.
1
CHAPTER 1 REVIEW OF THE LITERATURE
1.1 Introduction
Phytochemicals are a diverse class of compounds that are secondary plant
metabolites, meaning that they are not directly needed for basic growth or reproduction of
plants, but are involved with other critical processes such as photosynthesis, pollination,
defense, or survival (Howitt & Pogson, 2006; Springob & Kutchan, 2009). Unlike
essential vitamins and minerals, which are needed in the human diet for basic nutrition,
phytochemicals do not have recommended dietary levels as the absence of them does not
directly linked with specific deficiencies (Alminger et al., 2014). However, the
consumption of phytochemicals, such as carotenoids, has been associated with a reduced
risk of various diseases (Gann et al., 1999; Jha, Flather, Lonn, Farkouh, & Yusuf, 1995;
Sesso, Buring, Norkus, & Gaziano, 2004; Sesso, Liu, Gaziano, & Buring, 2003; Vainioa
& Rautalahti, 1998). Additionally, many phytochemicals are pigmented and/or have
antioxidant capabilities, making them attractive candidates as naturally-sourced, edible
ingredients for the food industry. Although there is value to consuming phytochemicals
within the original food matrix, in some applications, specialized processing or addition of
the target phytochemical may allow for enhanced functionality or improved product
quality (Calvo, García, & Selgas, 2008; Reboul et al., 2005). Additionally, many
agricultural byproducts, such as tomato peels, apple pomace, and okara (a tofu byproduct)
contain substantial amounts of functional phytochemicals that are typically underutilized
or disposed of after processing (Jankowiak, Trifunovic, Boom, & Van Der Goot, 2014;
Knoblich, Anderson, & Latshaw, 2005; Lu & Foo, 2000). In order to better utilize these
resources, efficient extraction strategies should be employed to effectively separate
valuable phytochemicals from waste streams. In order to stabilize labile phytochemical
compounds for potential use as food ingredients, encapsulation or delivery systems can be
designed to enhance physicochemical stability and to possibly enhance the oral
bioavailability and nutritional properties.
2
1.2 Phytochemicals and their functionality in food
Phytochemicals have numerous functional roles in plants, but also serve useful
purposes for people. Throughout history, humans have used phytochemicals for food,
pigment, ceremonial and medicinal purposes (Okigbo, Anuagasi, & Amadi, 2009; Sopher
& David Sopher, 1964; Springob & Kutchan, 2009). For food applications, relevant
functionalities for phytochemicals include their abilities as colorants, antioxidants, or as
health-promoting, bioactive ingredients.
1.2.1 Natural colors
Color is a primary characteristic that can determine acceptability of a food product.
While synthetic colorants have provided a wide array of hue and saturation for edible
applications, naturally sourced pigments may be more appealing to the modern consumer.
The food industry as a whole has evolved to incorporate more ‘wholesome and healthy’
ingredients for the current health-concerned consumer. Paralleling this theme, many
naturally-sourced pigments are label-friendly and may provide an added health benefit
when consumed. Color is often the first thing consumers notice when they see a food
product. In some cases the perceived color has an obvious association with the quality of
the product (e.g., color as an indication of fruit ripeness or freshness). Unfortunately,
physicochemical stability can be a challenge for natural colorant utilization as many of
these pigments degrade when exposed to heat, light, oxygen, or undergo changes in pH
(Table 1.1). Since natural pigments vary in source, structure, and have different
properties, it is important to consider the optimum conditions for extraction and how the
particular color compound will interact with the food matrix and external environment.
3
Table 1-1 Common color pigments from plant and animal sources
Pigment
Class
Water
Color
Natural Sources
Susceptibility
Reddish-
Salmon, shellfish,
Light, heat, oxygen1
orange
algae
Solubility
Astaxanthin
Carotenoid
Insoluble
(Haematococcus
pluvialis)
β-Cryptoxanthin
Carotenoid
Insoluble
Yellow
Avocadoes,
Light, heat, oxygen2
persimmons,
peppers, oranges,
papaya
Bixin
Carotenoid
Insoluble
Red, orange
Annatto
Oxygen, exposure to sulfur
dioxide, less susceptible to
light and heat <100°C2
Carotenes (α, β)
Carotenoid
Insoluble
Yellow, orange Carrots, peppers,
Light, heat, oxygen2,3
and other yelloworange vegetables
Crocin
Carotenoid
Soluble
Yellow
Saffron, gardenia,
Light, oxygen3
Cape jasmine plant
3
4
Table 1-1 continued
Lutein and
Carotenoid
Insoluble
zeaxanthin
Yellow,
Spinach, lettuce,
yellow-orange
corn, egg yolk,
Light, heat, oxygen2
broccoli, marigold
Lycopene
Carotenoid
Insoluble
Red
Tomatoes, papaya,
Light, heat, oxygen2,4
grapefruit,
watermelon, guava
Norbixin
Carotenoid
Soluble
Yellow, orange Annatto
Oxygen, pH, exposure to
sulfur dioxide5, less
susceptible to light and
heat <100°C10
Anthocyanin
Betalain
Phenolic
Phenolic
Soluble
Soluble
Reds, purples,
Cranberries, grapes,
pH, temperature, light,
blues
apples
enzymes, sulfites, sugars6
Yellow, red
Red beets, cactus
Light, heat, oxygen7
pear
Carminic acid
(carmine)
Phenolic
Insoluble
Yellow,
Female cochineal
pH, susceptible to
orange, red
insects (cochineal
microbiological attack,
extract)
stable against oxidation
and light4,8
4
5
Table 1-1 continued
Curcumin
Chlorophyll
Phenolic
Tetrapyrrol
Insoluble
Greenish-
Turmeric
(polar solvent-
yellow,
soluble)
yellow-orange
Insoluble
Greens, blues,
Plants, algae,
yellows, greys,
bacteria
Light, pH, chemical
oxidants, metal ions9
Light, temperature1,10
browns
1
Niamnuy, C., Devahastin, S., Soponronnarit, S., & Vijaya Raghavan, G. (2008). Kinetics of astaxanthin degradation and color changes of dried shrimp during
storage. Journal of Food Engineering, 87-691-600.
2
Rodriquez-Amaya. (2001). A guide to carotenoid analysis in foods. Washington, DC, OMNI Research. pp. 1-10.
3
Henry, B.S. (1996). Annatto. In: Natural Food Colorants. London, UK, Chapman & Hall pp. 47-52.
4
Fuller, G.W. (1994). Colorants. In: New Food Product Development: From Concept to Marketplace. Boca Raton, FL, CRC Press, pp. 123-124.
5
Henry, B.S. (1996). Annatto. In: Natural Food Colorants. London, UK, Chapman & Hall pp. 47-52.
6
Frank, Kerstin, et al. (2012). Stability of anthocyanin-rich W/O/W- emulsions designed for intestinal release in gastrointestinal environment. Journal of Food
Science, 77(12): N50-N57.
7
Herbach, K.M., Stinzing, F.C., Carle, R. (2004). Impact of thermal treatment on color and pigment pattern of red beet (Beta vulgaris L.) preparations. Journal
of Food Science, 69: C491.
8
Marmion, D.M. (1991). Colorants exempt from certification. In: Handbook of U.S. Colorants: Foods, Drugs, Cosmetics, and Medical Devices. United States,
John Wiley & Sons, Inc. pp. 119-148.
9
Buescher, R., Yang, L. (2000). Turmeric. In: Natural Food Colorants: Science and Technology. New York, NY, Marcel Dekker, Inc. pp. 205-226.
10
Delgado-Vargas, Francisco, Paredes-Lopez, Octavio. (2003). Natural Colorants for Food and Nutraceutical Use. Boca Raton, FL, CRC Press. pp. 221-251.
5
6
1.2.2 Antioxidants
Lipid oxidation is a major issue that affects the quality and nutrition of food
products. Fat depots exist in a variety of food products and can lead to oxidative rancidity
over time. Polyunsaturated fatty acids are especially susceptible to this as the rate of
oxidation drastically increases as a function of the number of unsaturated bonds (Labuza
& Dugan, 1971). The food industry typically combats lipid oxidation via packaging
solutions (i.e., protecting the product against oxygen and/or light exposure), controlling
temperature/storage conditions, or by incorporating an antioxidant in the formulation.
Traditionally, synthetic antioxidants (e.g., butylated hydroxyanisole) have been used to
combat oxidation in susceptible foods. Yet, potential, but not necessarily well supported,
health risks associated with synthetic antioxidants, in combination with market trends
towards “natural” products or “clean” labels have led to an increased interest in using
naturally-sourced antioxidants (Frankel, 1993). Many plant pigments and phytochemicals
are capable of scavenging free-radicals and could potentially be used as antioxidant
ingredients.
1.2.3 Bioactive ingredients and bioavailibility
Aside from preventing lipid oxidation in food products, phytochemicals can
potentially limit deleterious oxidative reactions in the human body. Phenolics and
carotenoids can scavenge free-radicals to mitigate oxidative stress (Jha et al., 1995),
reduce the risk of cardiovascular disease. Other potential benefits independent from
antioxidant activity have also been observed, such as tumor suppression (Hagiwara et al.,
2002; Livny et al., 2002; Y.J. Surh, 2003).
The human gastrointestinal system treats phytochemicals as xenobiotics (Lampe &
Chang, 2007) in that consumed compounds are absorbed, metabolized, distributed to
select tissues, and then excreted to prevent accumulation (Undevia, Gomez-Abuin, &
Ratain, 2005). From a toxicity standpoint, this process helps to prevent overdose of
phytochemicals, vitamins, and minerals from our diet. However, depending on the types
of food one consumes, it may be difficult for adequate phytochemical amounts to reach
7
the target tissues. In many cases, phytochemicals have shown to have health benefits
when consumed in whole foods but this will vary depending on the compound and the
matrix. For example, β-carotene is less bioavailable from leafy vegetables compared to
carrots and other food sources (Castenmiller & West, 1998; Chung, Rasmussen, &
Johnson, 2004). Fruits such as pumpkin and mango have carotenoids that are dissolved
within dispersed oil droplets, which allows for better digestive release and subsequent
absorption compared to crystalline carotenoids (Williams, Boileau, & Erdman, 1998).
Low bioavailibility may be due to the entrapment of carotenoids, complexation with
proteins, or competition with other carotenoids for inclusion in the mixed micelle.
Processing can potentially enhance bioavailability of carotenoids and other
phytochemicals by disrupting surrounding plant cell walls and/or by breaking down
protein-carotenoid complexes. A well known example of this is the increase in lycopene
bioavailability in processed tomato paste compared unprocessed juice (Stahl & Sies,
1992) and raw tomatoes (Gärtner, Stahl, & Sies, 1997).
1.2.4. Lycopene
Carotenoids represent a specific class of phytochemicals that are commonly found
in fruits and vegetables such as watermelon and tomatoes (Rodriguez-Amaya, 2001). Of
the carotenoids, lycopene has been reported to have the highest singlet oxygen quenching
ability (DiMascio, Devasagayam, Kaiser, & Sies, 1990). Lycopene is the dominant
carotenoid pigment present in tomatoes. Due to its highly unsaturated structure—C40
acyclic isoprenoid with 13 double bonds where all but two are conjugated (Figure 1-1)—
this nonpolar, non-provitamin A compound is biologically active as an antioxidant, singlet
oxygen quencher, gap-junction communication inducer, and immune response modulator
(Roldán-Gutiérrez & Dolores Luque de Castro, 2007; J Shi & Xue, 2010).
8
Figure 1-1 Chemical structure of all-trans-lycopene and its isomers.
Due to its poor solubility, antioxidant, and health-associated properties, lycopene
was selected as a model carotenoid to investigate for investigating extraction and
encapsulation technologies. Although the major focus of this dissertation will be on
lycopene, the discussed technologies could be applied to other phytochemicals and
hydrophobic compounds. Additionally, since the literature involving lycopene extraction
and delivery is a narrow and specific subset of research, this chapter will also review some
relevant knowledge related to non-lycopene phytochemicals.
1.4 Extraction techniques for bioactive phytochemical extraction
1.4.1 Conventional extraction techniques
Conventional extraction methods (i.e. solvent extraction, soxhlet, water
percolation, soaking) may be employed to extract phytochemicals. However, these
traditional methods have drawbacks, including low efficiency, long extraction times, and
large amounts of solvent waste (Chan, Yusoff, Ngoh, & Kung, 2011). A major limitation
is that several phytochemicals are thermally labile or susceptible to oxidation. This poses
a problem as some conditions in traditional extraction may be too harsh and cause the
target compound to degrade. Emerging technologies, such as microwave-assisted
9
extraction and supercritical fluid extraction, may be more efficient and provide better
quality phytochemical ingredients. Thus, several efforts have been made to apply these
techniques to biomaterials to improve extraction strategies (Lijun Wang & Weller, 2006).
1.4.2 Ultrasound-assisted extraction
Ultrasound-assisted extraction (UAE) has been readily used in the pharmaceutical
and supplement industry to produce plant-based bioactive products. Ultrasound waves
assist the extraction process by producing a compression wave perpendicular to the
surface (Fellows, 2000). As this wave attenuates into the food it produces localized
changes in pressure and temperature that induce cavitation and compromise cell
membranes to allow for enhanced extraction. This non-thermal effect may be
advantageous when thermal labile compounds are not readily released or solubilized.
Ultrasound technology has been used as a pre-treatment to better facilitate mass transfer
during the extraction process. El Darra, Grimi, Maroun, Louka, & Vorobiev (2013)El
Darra et al., investigated the effect of pretreatments on red grapes for red wine
fermentation. Wine extracts treated with ultrasound for 15 minutes exhibited higher
anthocyanins content compared to the control and those treated with moderate heat.
However, samples pretreated with pulsed electric field (PEF) were approximately 50-100
mg/L greater than the ultrasound treated samples. Similarly, ultrasound treatment
demonstrated an improvement in color intensity of red wine samples, although PEF
treatments were superior in this experiment—with a 16% improvement over the control
for ultrasound vs. a 23% improvement for PEF. As expected, a favorable solvent for UAE
is one that can readily solubilize the target compound. However, the chosen solvent
should also have a low vapor pressure to better induce cavitation (Pingret, Fabiano-Tixier,
& Chemat, 2013) . Although UAE has delivered promising results in the laboratory,
scaling up may be challenging as extraction efficiencies may be low without pretreatments (Vilkhu, Mawson, Simons, & Bates, 2008). However, if this technology is
adapted for the particular application, it may provide benefits when used at the industrial
scale. Patist & Bates (2008) report that improved design—resulting in increased
efficiency, easy installation, reduced energy costs, and less maintenance—make
ultrasound technology an attractive candidate for the food industry.
10
1.4.3 Supercritical fluid extraction
Another emerging technology that has been of interest recently is supercritical
fluid extraction (SFE). Unlike MAE, SFE does not rely on traditional organic solvents.
Rather, supercritical fluids (typically supercritical water and carbon dioxide), which are of
less concern to the environment, are employed. SFE is of interest to the food industry as
samples are not heated, byproducts are not produced, and harmful solvents are not
required (da Silva, Rocha-Santos, & Duarte, 2016). In this process, extraction vessels are
pressurized so that the liquid solvent (i.e. CO2 or H2O) is above the critical point. Above
a given critical temperature and pressure, fluids exhibit a combination of properties
characteristic of liquid and gas that allow for improved diffusion into food samples for
extraction (Pan, Wang, Chen, & Chang, 2012). These supercritical fluids exhibit low
surface tension allowing for improved matrix penetration that is complimented by the
superior mass transfer characteristics due to their liquid-like densities (J Shi & Xue,
2010). The process conditions and combination of supercritical fluid solvents can be
adjusted so that compounds can be selectively isolated. When selecting co-solvents, one
should consider the interaction between the selected solvents as well as the effect the
mixture will have on the sample material (Joslin, Gray, Goldman, Tomberli, & Li, 1996).
Richins et al. (2010) found that SFE was able to selectively extract and separate
Capsicum carotenoids from present chlorophylls. Although they achieved ~85%
extraction rates of carotenoids from Capsicum, the authors note that decreased extraction
was observed for carrots and corn containing non-acylated carotenoids. Similarly,
Macías-Sánchez, Fernandez-Sevilla, Fernández, García, & Grima (2010) found that
chlorophylls were not extracted during carotenoid SFE from the algal species
Scenedesmus almeriensis. However, their findings also indicated that β-carotene and
lutein yields were only 50% and <10% of the reference method. Studies have been done
to assess the impact of using “modifiers” or additional solvents to improve the solubility
and yield of SFE. (Pan et al., 2012) indicated that a greater yield of astaxanthin from
Haematococcus pluvialis was obtained when greater concentrations of ethanol were used
as a modifier compared to water. The authors speculated that this improvement may have
been due to the ethanol inducing cell swelling in Haematococcus pluvialis and an overall
11
increase in mass transfer compared to water, which would cause resistance with ice
formation.
Within the past few decades, SFE has gained popularity for carotenoid extraction
(Barth, Zhou, Kute, & Rosenthal, 1995; Careri et al., 2001; de la Fuente, Oyarzún,
Quezada, & del Valle, 2006; Nobre et al., 2006; Rozzi, Singh, Vierling, & Watkins,
2002). In theory, SFE may reduce the risk of carotenoid degradation during extraction
compared to conventional solvent extraction, which may require more thermal input,
however, Baysal, Ersus, & Starmans (2000) reported low lycopene and β-carotene yields
(20 and 40%, respectively) due to thermal degradation at 65°C. Improved recoveries up
to ~50% were observed when 5% ethanol was incorporated as a co-solvent. Similarly,
Vaughn Katherine, Clausen Edgar, King Jerry, Howard Luke, & Julie (2008) reported
lycopene recoveries of ~40% for SFE-treated watermelon with the addition of an ethanol
co-solvent (15%) at 70°C. The authors speculate that this optimal extraction struck a
balance between higher temperatures, which resulted in better extractions but more
degradation and lower temperatures, which were less efficient extractions but better
compound retention. In a study to explore SFE of carotenoids from Haematococcus
pluvialis algae, Nobre et al. (2006) were able to extract astaxanthin, β-carotene, and lutein
at recoveries >90% of conventional (acetone at room temperature) extractions at 300 bar
and 60°C with an ethanol co-solvent. Higher carotenoid recoveries may thus be achieved
with SFE if the conditions are optimized.
1.4.4 Microwave-assisted extraction
Microwave-assisted extraction (MAE) has demonstrated improved efficiency over
other extraction techniques, including conventional (maceration and heat reflux
extraction) and ultrasonic extraction (Chan et al., 2011; Hao, Han, Huang, Xue, & Deng,
2002). This method couples solvents with microwave heating to increase the extraction
kinetics, thus reducing the amount of organic chemicals required and extraction time
compared to traditional procedures (Delazar et al., 2012). MAE may provide a rapid
alternative to tedious solvent-liquid extraction methods for bioactive compounds.
During microwave heating, there are two mechanisms by which heat is generated.
The first is related to the oscillating change in charge or dipole rotation. During
12
microwave heating, an alternating electric field is applied, causing food molecules to align
with the electric field and rotate/realign when the field alternates. This rotation causes
friction in the system, resulting in a generation of heat (Neas & Collins, 1988). The
second mechanism is ionic drifting or polarization, in which the applied electric field
causes ions present in the food to move rapidly and collide. This increase in movement
and collisions causes an increase in kinetic energy, thus resulting in heat generation
(Singh & Heldman, 2009). Higher temperatures can enhance extraction in plant materials
by allowing the solvent to penetrate through cell walls faster and by improving solubility
of the target compound in the heated liquid (Dandekar & Gaikar, 2002). Also, due to the
rapid heating effect, shorter MAE treatments allow for higher temperatures while limiting
the exposure of thermolabile samples to high temperatures.
In addition to the rapid heating effect, MAE can also improve the extraction via
physical alteration of the sample material. Polar components in plant materials experience
localized superheating, which can cause cells to swell and rupture, allowing for the release
of pigments (Dandekar & Gaikar, 2002; Jassie, Revesz, Kierstead, Hasty, & Matz, 1997;
Kaufmann & Christen, 2002; Paré, 1994). Therefore, MAE may also improve the
extraction yield of pigment compounds from natural sources by allowing them to escape
otherwise intact cell barriers. As discussed previously, naturally sourced pigments are
often found in enclosed biological structures (i.e. photosyntehtic organelles). Thus, efforts
have been made to study the effect of physically disruptive pretreatments on MAE.
Hiranvarachat, Devahastin, Chiewchan, & Vijaya Raghavan (2013) found that carrots that
were pre-blanched or pre-acidified exhibited higher carotenoid content and antioxidant
activity after MAE compared to untreated carrots. Similarly, soaking pre-treatments have
also demonstrated the potentional to improve extraction. In particular water, being a polar
compound, can greatly affect heating kinetics and extraction in a microwave system.
Dandekar & Gaikar (2002) found that MAE efficiency of curcumoid from hydrated plant
samples improved with higher moisture contents. In this case, the increased amount of
water enabled increased localized heating, allowing for better release of curcumoid from
the sample.
When extracting compounds, solvent selection is critical. Ideally, the solvent used
should be able to selectively solubilize the compound of interest. Overall, the choice of
13
solvent depends on the desired compound and the form the raw sample material. Since
the solvent combinations can be easily manipulated, MAE has potential for a variety of
different color pigments with varying polarity (Jain, Jain, Pandey, Vyas, & Shukla, 2009).
Aside from solvent type, the solvent-to-solid ratio is a factor that should be
considered for MAE. In a comparison of extractions methods for mulberry fruit extract,
Zou et al. (2012) demonstrated that a MAE treatment of 132 seconds yielded ~1.5 times
the amount of anthocyanins compared to a conventional solvent extraction done for 15
minutes with an irradiation power of 366 W. For all experiments, the solvent-to-solid
ratio was kept constant at 25:1 to ensure consistent microwave penetration depth into the
sample. Sun, Liao, Wang, Hu, & Chen (2007) also found MAE to be more efficient for
extracting anthocyanins from their raspberry samples, however, the optimal conditions
involved a longer extraction time of 12.1 minutes with a solvent-to-solid ratio of 4:1 and
an irradiation power of 425 W. HPLC-MS results indicated that the resulting extracts
from conventional extraction and MAE were similar, implying that the used MAE
protocol did not induce additional degradation to the anthocyanin pigments. Differences
in optimized conditions between the two previously mentioned studies may be due to the
matrix, microwave unit, or irradiation power, the difference in solvent-to-solid ratio is a
likely factor to explain the difference in optimal extraction times. Generally, higher
solvent-to-solid ratios exhibit improved extraction efficiencies, however extremely high
ratios may result in lower extractions if the large amount of solvent does not achieve the
target temperature. Considering this, when varying the solvent-to-solid ratio with an
experiment, it may be beneficial to adjust the amount of solid material in relation to a
constant solvent volume to maintain a consistent temperature (Spigno & De Faveri, 2009).
Aside from finding the optimal ratio for functionality reasons, using conservative amounts
of solvent should be considered as the rationale for using alternative extraction methods is
often to reduce solvent waste.
1.5 Delivery systems for lipophillic bioactive compounds
Following extraction, phytochemicals should be handled in a manner that does not
induce degradation. Aside from considering the susceptibility to environmental factors,
one must also consider the solubility of the phytochemical, especially when considering
14
incorporation into the food matrix. In particular, carotenoids and other lipid-soluble
pigments are not readily soluble in aqueous solutions. A strategy to overcome this is to
use encapsulation or delivery systems.
Encapsulation can improve the shelf-life, dispersability, stability, and delivery of
compounds and may also prevent unwanted ingredient reactions such as off-flavor
development (Shahidi & Han, 1993). Originally introduced as a method to enhance
metabolite separation in the biotechnology industry, encapsulation is the process of
entrapping a target compound within a layer of protective coating (Nedovic, Kalusevic,
Manojlovic, Levic, & Bugarski, 2011). This technology has also been used extensively in
the pharmaceutical industry for targeted delivery and in the food industry for protecting
labile compounds and enhancing solubility in the food matrix (Wandrey, Bartkowiak, &
Harding, 2010). Zuidam & Shimoni (2010) divide encapsulation into two main
categories: 1) matrix and 2) reservoir type. Emulsification can be considered an example
of matrix type encapsulation since the target compound (phytochemical) is distributed
within a carrier material (oil droplet), whereas coating-based methods, which develop a
shell or outer layer, are examples of reservoir encapsulation techniques.
1.5.1 Emulsion-based delivery systems
In many cases, the compound of interest is in liquid form and can be dispersed in
another liquid that is compatible with the food matrix. For example, dispersing a lipidsoluble liquid throughout an aqueous solution can enable an otherwise insoluble
phytochemical to be incorporated into a water-based product. The process of dispersing
an insoluble liquid (e.g., oil) throughout a continuous bulk liquid (e.g., water) is called
emulsification. Under normal circumstances, oil and water are not readily mixed due to
the high Gibbs free energy or surface energy at the oil-water interface. The two phases
favor a state with less tension (lower free energy) by minimizing the interfacial area
between oil and water. This is done via coalescence of oil droplets, or with an added
emulsifier, which can lower the interfacial tension at the oil-water interface to facilitate
formation of the dispersed phase (Dalgleish, 2006; Walstra, 2003). While this state is not
thermodynamically favored, it can be kinetically or meta-stable in that there is a sufficient
15
energy barrier to overcome (e.g., energy input during droplet breakup) and that the
emulsifier provides electrostatics or steric forces to prevent droplet collision.
Emulsions can be classified by the nature of their dispersed and continuous phases,
where oil-in-water (O/W) emulsions have oil droplets dispersed throughout a water phase.
In general, O/W emulsions can be manipulated by varying the emulsifier
type/concentration and by altering the components of the aqueous phase (Kalnin, Schafer,
Amenitsch, & Ollivon, 2004). Since O/W emulsions are more suitable for carotenoid
delivery as compared to water-in-oil (W/O) emulsions, the remainder of this chapter
focuses on discussion of the former.
Droplet size plays an instrumental role in the encapsulation and delivery of
compounds. Ultra fine emulsions—nano and microemulsions—have been used to
disperse otherwise insoluble compounds in food matrices (Anarjan, Tan, Nehdi, & Ling,
2012). Although nanoemulsions and microemulsions are technically both emulsions in the
sense that they are dispersed liquid-liquid systems, there is a distinct difference that
separates nanoemulsions and emulsions from microemulsions. Unfortunately, differences
between nanoemulsions and microemulsions are often unclear as these two systems are
similar in several ways. Similar to microemulsions, the reduced sized nanoemulsion
droplets provide a variety of benefits compared to conventional emulsions. Both
nanoemulsions and microemulsions share the advantage of improved stability against
gravitational separation due to small droplet size, which would otherwise lead to creaming
and are can be optically transparent, since the small droplets are smaller than the
wavelength of light and thus scatter light weakly (Julian McClements, 2012; Sagalowicz
& Leser, 2010). This allows for a broad range of food and beverage applications
compared to conventional emulsions, as the smaller droplets will not alter the color and
opacity of the finished food product. Also, attractive forces between droplets decrease as
size decreases, while steric hindrances continue to provide a repulsive force. Therefore,
better stability against flocculation and coalescence is observed with finer emulsions
(Qian & McClements, 2011). However, there are distinct differences between micro and
nanoemulsions that should be understood. Specifically, microemulsions are
thermodynamically stable whereas nanoemulsions and emulsions are kinetically stable.
Thus, although the terms “micro” and “nano” may insinuate that the magnitude of droplet
16
diameter differentiates the two, thermodynamic instability rather than size is the main
characteristic that sets distinguishes these emulsion systems (Anton & Vandamme, 2011;
Rao & Mcclements, 2011) as microemulsions can indeed be within the nanoscale range.
Compared to microemulsions, stabilized nanoemulsions have a higher free energy.
Strictly speaking, for an emulsion to be considered a “nanoemulsion”, some
authors have proposed diameter size restrictions such as <200 nm (Ostertag, Weiss, &
McClements, 2012) or <500 nm (Mahdi Jafari, He, & Bhandari, 2006). There is no
singular consensus on a size limit that could distinguish nanoemulsions and emulsions
based off of their properties although most investigators describe nanoemulsions as having
droplet diameters between 20-200 nm, which is reasonable since bulk properties of liquid
droplets start to change below 200 nm due to an dramatic increase in surface area-tovolume ratio (Fathi, Mozafari, & Mohebbi, 2012; Lijuan Wang et al., 2009)(Fathi et al.,
2012; Lijuan Wang et al., 2009).
The advantage of nanoemulsions over microemulsions for food applications relates
to the thermodynamics of each system. Specifically, since microemulsions are
thermodynamically stable, their inherent stabilities are dependent on the composition and
storage conditions. Theoretically speaking, a microemulsion could remain stable as long
as the environmental conditions remain constant (Rao & McClements, 2012). However,
in the food industry, products/ingredients are subjected to fluctuations in temperature,
could be diluted or combined with other ingredients, or may experience other changes as
the item is transported, handled, and stored. Thus, microemulsions are likely to become
unstable as extrinsic thermodynamic factors change (Ostertag et al., 2012) and must be
formulated carefully in order to maximize shelf-life. Conversely, nanoemulsions are
thermodynamically unstable, so the system will break down over time due to exceeded
energy barriers rather than dilution or conservative changes in temperature.
Physicochemical interactions, such as steric and electrostatic forces, will repel droplets
and consequentially increase the energy barrier. In order to optmize the energy barrier
and the related stabilizing interactions, care should be taken during formulation to select
appropriate ingredients, oil-surfactant ratio, and processing methods to ensure that particle
size and distribution is optimal. Since food manufacturers have more control over the
fabrication and incorporation of emulsions than they do over storage and handling
17
conditions of finished products, nanoemulsions may be more appropriate for food
applications.
1.5.2 Emulsifiers and droplet stabilization
Emulsifiers are surface-active compounds that are used to stabilize the dispersed
phase of emulsion systems. At the interface, they act to reduce the interfacial tension and
can alter the viscoelastic properties of the interface to inhibit coalescence. There are
various types of emulsifiers, including small-molecule emulsifiers, polysaccharides, and
proteins, which may be charged or nonionic (Acosta, 2009; Narang, Delmarre, & Gao,
2007). Nonionic emulsifiers are stabilized by hydrogen bonding and dipole interactions,
while charged ones are further stabilized by the ionic interactions (Narang et al., 2007).
Also, surface charges from the adsorbed material can repel, preventing the droplets from
approaching others (Chang, Tu, Ghosh, & Nickerson, 2015).
The term surfactant refers to a small, amphiphatic, surface-active emulsifer (J.
Surh, Decker, & McClements, 2006; Walstra, 2008). Surfactants encompass both polar
and nonpolar properties, and can be characterized by the degree of hyrdophile-to-lipophile
balance (HLB) values (Narang et al., 2007). A greater HLB value indicates it is more
hydrophilic and favors O/W emulsion. Conversely, surfactants with lower HLB values
favor w/o emulsions. Intermediate values of approximately 6-9 indicate weak surface
activity and thus the surfactant will not strongly solubilize in aqueous or hydrophobic
phases.
Large molecule surfactants or biopolymers (i.e., polysaccharides or proteins) can
also act to stabilize oil droplets, typically via steric hindrances, charged or hydrophobic
interactions (Dickinson, 1997, 1999). In the case of globular proteins, they are able to
unfold (loss of three-dimensional structure) at the interface so that the hydrophilic
residues orient towards the bulk aqueous phase while hydrophobic residues align within
the lipid core. Many proteins are known to form a thick, viscoelastic interfacial film,
which helps to prevent droplets from merging and coalescing (Georgieva, Schmitt, LealCalderon, & Langevin, 2009). Similarly, surface active polysaccharides function to
stabilize droplets via similar mechanisms, but may also have non adsorbed components
that act in the continuous phase to form a stabilizing network or by thickening the
18
continuous phase to prevent droplet coalescence (Matalanis, Jones, & McClements, 2011).
Generally speaking, surfactants are able to form smaller droplet sizes, but biopolymers
(when conditions are optimized) can demonstrate better long-term stability. In a study to
compare the effects of small and large molecule emulsifiers on β-carotene nanoemulsions,
Mao et al. (2009) demonstrated that larger protein emulsifiers resulted in an increased
droplet diameter but greater resistance to coalescence over time compared to those
formulated with smaller surfactants, Tween 20 and DML. This may be explained by the
strong protein interfacial layers that prevent aggregation, despite their inability to pack
closely to interface to form smaller droplets.
Sufficient quantity of emulsifier must be used to ensure that enough is present to
fully cover the surface of the particle (Walstra, 2003). However, when formulating
emulsions for food applications, all materials should be fit for consumption, generally
recognized as safe (GRAS) and used in appropriate amounts that fall within regulatory
limits. The US Food and Drug Administration has limitations on certain food additive
surfactant compounds. For example, polysorbate 80, used alone or in combination with
polysorbate 65, cannot exceed 0.1% in ice cream products (21 CFR 172.840).
1.5.3 Factors affecting stability
Overtime, emulsions and nanoemulsions break down, via e.g., flocculation,
gravitational separation, Ostwald ripening, and coalescence, (Figure 1-2) as the free
energy of the separated phases is greater than that of the dispersed system (McClements,
2012). Emulsions with smaller droplets tend to be more stable compared to those with
larger ones and are less susceptible to aggregation and gravitational separation (Ostertag
et al., 2012). However, the amount of emulsifier needed to stabilize larger emulsion
droplets is less due to the increased surface area-to-volume ratio observed for small
droplets.
19
Figure 1-2 Schematic of droplet behavior that leads to destabilization. Small, stable
droplets can undergo reversible creaming or flocculation, both of which can lead to
coalesence if the interfacial film is ruptured. Small droplets are especially susceptible to
Ostwald ripening and will diffuse through the continuous phase to become incorporated
into larger droplets. Both coalesced droplets and large droplets resulting from Ostwald
ripening can lead to phase separation.
The type of emulsifier has a major effect on the stability of emulsions and
nanoemulsions. For example, proteins can be highly susceptible to changes in pH or salt
concentration as this environmental change will alter the protein structure. Conversely,
modifications in protein structure could have positive effects on emulsion stability as
partial denaturation could lead to increased hydrophobicity (Kim, Cornec, & Narsimhan,
2005). The amount of emulsifier also plays a critical role, as an inadequate concentration
will be incapable of stabilizing and preventing droplet coalescence. Conversely, at
excessively high concentrations, excess emulsifier will exist in the aqueous phase, which
can have detrimental effects on droplet stability (e.g., depletion flocculation) or can have
an additional effect on chemical stability that is not accounted for. For example, excess
proteins and surfactants in the aqueous phase have been shown to scavenge free-radicals
in the system (Let, Jacobsen, Sørensen, & Meyer, 2007; Tong, Sasaki, Mcclements, &
Decker, 2000), which is not necessarily a negative effect. However, if the investigator is
unaware of the excess amount, he/she may inaccurately associate high oxidative stability
20
with the interface when the scavenging abilities of the nonadsorbed emulsifier may
actually be the cause.
The type of oil used in the dispersed phase also has an effect on emulsion stability.
For nanoemulsions, which are especially susceptible to Ostwald ripening, using a more
hydrophobic oil may be beneficial as the poor solubility with the aqueous phase will help
to retard the rate of mass transfer from smaller droplets to larger ones (McClements,
2012).
1.5.4 Fabrication and droplet size reduction
Emulsion processing techniques can be categorized as low or high-energy
methods. While low-energy methods provide some benefits—lower energy requirements
and reduced power costs—they may form emulsions that are more susceptible to
coalescence and require more surfactant. In a study comparing spontaneous
emulsification and microfluidization, (Yang, Marshall-Breton, Leser, Sher, &
McClements, 2012) demonstrated that the amount of surfactant needed to form droplets
for spontaneous emulsification was roughly ten-times greater than that required for
microfluidization. Thus, a major disadvantage with low-energy methods is its inability to
efficiently produce ultrafine emulsions. Strictly speaking, although spontaneous
emulsification can form droplets in the nanoscale region, these are classified as
microemulsions due to their thermodynamic stability and ability to develop structures
based on temperature and composition (Fryd & Mason, 2012). Nanoemulsions cannot
spontaneously self-assemble in this manner since they require a high-energy input. Phase
inversion is a low-energy method that can produce kinetically stable emulsions in the
nanoscale range. During this process, an aqueous phase is titrated into a surfactant-oil
phase while being constantly stirred. The resulting mixture is a W/O emulsion that
converts to an intermediate O/W/O stage before forming a final O/W emulsion (Ostertag
et al., 2012). Paralleling the disadvantage of spontaneous emulsification, phase inversion
also requires relatively high surfactant amounts to achieve smaller, stabilized droplets. If
greater surfactant amount is permitted, phase inversion offers a relatively simple, lowenergy process to produce ultra fine emulsions without specialized equipment.
21
Conversely, high-energy emulsification methods may be more feasible for the food
industry as they are typically easier to scale up (Dinglery & Gohla, 2008; Salvia-Trujillo,
Rojas-Graü, Soliva-Fortuny, & Martín-Belloso, 2013). High shear methods allow for a
lower surfactant-to-oil ratios compared to low-energy methods. This may be desirable as
there are limitations on the amount of emulsifier that can be added into food formulations.
High-pressure homogenization (HPH), one technique for producing
nanoemulsions, uses high pressures (100-2000 bar) to push a liquid sample through a
narrow gap into an interaction chamber (Mader, 2006). This induces shear stress to
disrupt droplets and reduce their size. Aside from shear, other important factors that
contribute to droplet rupture during HPH are cavitation, impact, turbulent flow, and
pressure gradient. The system is comprised of two major parts, the homogenization valve
and the high-pressure pump (Saravacos & Kostaropoulos, 2016). The valve includes a
plunger and a valve seat, between which forms a ring gap. In particular, HPH processing
may be the most feasible for the food industry due to the relative ease of equipment
operation and high throughput of the system. The particle size obtained after
homogenization is crucial since it affects the texture, appearance, bioavailability, and
stability of the product (Acosta, 2009; D Julian McClements, Decker, & Weiss, 2007;
Qian & McClements, 2011) . In general, higher pressures and greater number of passes
through the interaction chamber produce smaller, stabilized droplets (Becher, 1967).
However, due to physical limitations of equipment and time restraints, most protocols
incorporate 3-5 passes as any more does not significantly contribute to particle size
reduction (Tan & Nakajima, 2005a; Yuan, Gao, Zhao, & Mao, 2008). In some cases,
excessive passes can lead to overprocessing, causing particle sizes to increase due to
recoalescence (Jafari, Assadpoor, He, & Bhandari, 2008). Therefore, determining optimal
pressure and passes for industrial applications is important for efficiency and product
quality.
A better understanding of how equipment design and processing parameters
affects the fabrication of nanoemulsions is critical especially when the most efficient
surfactant cannot be used due to constraints on concentration or negative taste and
ingredient interactions within the food matrix. With regards to the equipment design, the
efficiency of homogenization is affected by the type of valve used, where valves that are
22
flat-seated with convoluted surfaces allow for efficient droplet breaking compared to
smooth surfaced sloped-seated valves (Saravacos & Kostaropoulos, 2016). The chamber
design also affects homogenization efficiency as demonstrated by Donsì, Sessa, & Ferrari
(2011) in their study on the effect of chamber geometry on particle size, mean droplet
size, and polydispersity index for nanoemulsions fabricated by HPH. Although the
chamber geometry was not shown to significantly impact the minimum mean droplet size
differences were observed for polydispersity index. Nonuniform chambers with
randomized turbulent and laminar regions did not allow for efficient disruption of the lipid
particles—thus causing an increased droplet size distribution, i.e. a greater polydispersity
index.
Microfluidization is another high-energy technique commonly used in the food
industry due to its ability to form droplets with narrow size distributions (Augustin &
Sanguansri, 2009) and large scale equipment availability. A higher number of passes
through the equipment result in smaller average droplets up to a certain maximum
saturation radius, above which additional shearing force at the same pressure will not
further reduce the particle size (Salvia-Trujillo et al., 2013). This may be due to
inefficiency in a single pass, where droplets located central experience a greater shear
flow than the particles closer to the channel wall (Meleson, Graves, & Mason, 2004). Lee
& Norton (2013) compared the performance of a microfluidizer with that of a highpressure valve homogenizer for producing food grade nanoemulsions. A broader droplet
size distribution was observed with high-pressure homogenization, however a
disadvantage of microfluidization is that agglomeration may occur above a given pressure
threshold ~40-60 MPa (Jafari, He, & Bhandari, 2007). However, it is difficult to
definitively compare these two systems since, as mentioned previously, other factors such
as geometry of interaction chamber and emulsifier type can noticeably affect the resulting
nanoemulsion.
Ultrasound technology has also been used to fabricate ultrafine emulsions. As
previously mentioned in the extraction section, this high-energy method induces
cavitation, which in turn can be manipulated by external pressure to induce emulsification
(Patist & Bates, 2008). This phenomena can thus be exploited to produce emulsions with
small particle sizes. Although ultrasound has been successfully used to produce
23
emulsions, there are limited published studies done for natural color applications. Cilek et
al. (2012) found that longer ultrasound times (up to 20 minutes) resulted in smaller
particle sizes, but did not significantly alter the color of encapsulated red phenolic powder
from sour cherry pomace. Research has been done on the use of ultrasound technology
for nanoemulsion fabrication (Kentish et al., 2008; Tang, Shridharan, & Sivakumar, 2013)
however, there is limited work done specifically on lycopene.
Other techniques for forming nanoemulsions may combine different types of
equipment in the procedure to minimize energy costs employed by methods solely relying
on high-energy processes. High-energy emulsification-evaporation techniques combine
the use of high shear techniques with rotary evaporation and have previously been used in
pharmaceutical and food applications. Tan & Nakajima (2005) used an emulsificationevaporation technique with hexane organic phase and β-carotene to produce
nanoemulsions. The process involved the use of a homogenizer to produce a course
emulsion, followed by microfluidization, and then rotary evaporation to remove the
hexane. A disadvantage to this method is the additional evaporation step and the
possibility of residual hexane in the formed food emulsion. Although the resulting
nanoemulsions were found to be structurally stable for up to 12 weeks, the β-carotene
content significantly decreased during the storage study (Tan & Nakajima, 2005b).
Similarly, Silva et al., (2011) used a combination of homogenization and rotary
evaporation that produced physically stable, but chemically unstable β-carotene
nanoemulsions that displayed degraded color with increasing storage time. In both cases,
increased shear rate of the homogenization step correlated with the increased β-carotene
degradation, possibly due to the increase in free radical generation or greater
solubilization of oxygen. Thus, the results from these studies suggest the processing
technique may have affected the stability of the nanoemulsions and that emulsion
composition and processing conditions should be considered when incorporating
carotenoids or other phytochemicals.
24
1.5.5 Stabilization of phytochemicals and potential enhanced bioavailibility
Extracted carotenoids, such as lycopene, are highly susceptible to oxidation and
isomerization by light and heat (John Shi, Dai, Kakuda, Mittal, & Xue, 2008). Emulsionbased delivery systems may alleviate this issue by stabilizing labile compounds into
dispersed oil droplets; thus protecting them from the action of pro-oxidants and allowing
further incorporation within the food system. In the case of carotenoids, emulsification
can improve the nutritional availability as compounds naturally compared to many
conventional food matrices, in which carotenoids would otherwise bind with protein or
exist in crystalline form, which can limit release and absorption (William et al., 1998).
Emulsion delivery systems have the potential to be used for various ingredient
technologies, including flavors, antioxidants, antimicrobials, and pigments. Among the
potential compounds of interest, β-carotene has been well studied and stabilized in oil-inwater nanoemulsions (Tan & Nakajima, 2005a, 2005b; Yuan et al., 2008). Other
carotenoids, such as astaxanthin (Affandi, Julianto, & Majeed, 2011; N Anarjan et al.,
2012; Navideh Anarjan, Mirhosseini, Baharin, & Tan, 2011) have also demonstrated
superior stability in nanoemulsions. Aside from investigating stability, assessing potential
bioactivity and health functionality of formulated nanoemulsions has been studied for
several different hydrophobic compounds. Some examples include improved
bioavailability, anti-inflammatory behavior, and anti-tumor effects of curcumin (Ahmed,
Li, McClements, & Xiao, 2012; Ganta, Devalapally, & Amiji, 2010; X. Wang et al.,
2008), improved bioavailability of lutein (Vishwanathan, Wilson, & Nicolosi, 2009),
sustained release of lycopene for UVA protection (Butnariu & Giuchici, 2011) .
Emulsions with small droplets, such as nanoemulsions, are of particular interest as
they have a greater relative interfacial area and have demonstrated the ability to improve
functionality and bioavailability of encapsulated compounds (Acosta, 2009; Cheong, Tan,
Man, & Misran, 2008; Ochekpe, Olorunfemi, & Ngwuluka, 2009).
Although emulsion delivery systems show promise, they should be further studied
for specific ingredients, food systems, and processing conditions to optimize the system
for the particular application. Improvements from previously encapsulated compounds
can also be made by further adjusting the size of the dispersed droplets, composition, and
25
properties of the interface while comparing this to encapsulation efficiency and
bioavailability (Qian & McClements, 2011). Since foods are complex matrices, there are
often difficulties with adding bioactive ingredients to new foods without compromising
quality, stability or sensory attributes (Patel & Velikov, 2011; Qian, Decker, Xiao, &
McClements, 2012; Sagalowicz & Leser, 2010) . In particular, adding compounds in the
soluble form may result in an altered flavor and color profile, while addition of insoluble
ingredients can change the texture and decrease the bioaccessibility of the compound.
Emulsions address these issues by retaining insoluble nature while in the food, to
minimize changes in flavor profile, yet assume a soluble behavior during digestion to
improve bioaccessibility (Patel & Velikov, 2011) .
1.6 Digestion, bioaccessibility, and bioavailability of carotenoids and
emulsion systems
In order to develop emulsion delivery systems that can enhance bioavailability of
phytochemicals, the process of digestion and absorption must be understood.
Bioaccessibility refers to the amount of compound that is released from the food matrix
and available for intestinal absorption (Heaney, 2001). In the case of carotenoids, this can
be estimated as the amount transferred to mixed micelles during digestion. Like
bioaccessibility, bioavailibility is dependent on gastrointestinal digestion but also includes
absorption, tissue accumulation, and potential bioactivity. Consequently, bioavailability
can be defined as the relative amount of compound that reaches circulation and can be
utilized (Wood, 2005). Typically, this is measured in humans as the level extracted from
serum or triglyceride rich lipoprotein response (Gann et al., 1999; Stahl & Sies, 1992). In
animal models, serum and other fluids can also be collected, but animal sacrifice and
subsequent tissue harvest provide critical information regarding carotenoid organ
accumulation.
26
1.5.4 Techniques for assessing in vitro bioaccessibility and bioavailability
Although human clinical trials and animal models provide more insight into how
compounds are metabolized and distributed in blood and tissues, they can be costly and
time consuming. In vitro digestion and cellular models allow for cost-effective
alternatives that do not have the same ethical restrictions that animal and human studies
would.
Static in vitro digestion models are widely used across various laboratory groups
and can be easily replicated as specialized equipment is not needed. However, a
drawback of simple static models is that they may not accurately mimic the complexities
of the human digestion process. Additionally, large variations can exist across different
lab groups even when using seemingly similar protocols. In order to address the
discrepancies, propositions have been made in order to develop a standardized basic
protocol that could be adapted and employed across several laboratory groups (Minekus et
al., 2014). Despite how meticulously a harmonized static model is developed, it is
unlikely that it can account for the nuances of the human gastrointestinal tract, e.g., the
complexity of adapting enzyme concentration or liquid volumes based off of digestion.
Advanced, dynamic digestion models may be better at simulating peristaltic movement,
changing pH, and automatically adjusting enzyme and salt concentrations. Although not
as widely used as static models, several dynamic models have been developed, including
the multi-compartmental TIM 1 (Blanquet et al., 2004; Déat et al., 2009) developed by
TNO (Netherlands), a bi-compartmental infant digestion model (Ménard et al., 2014)
developed by INRA (France) , a mono-compartmental Human Gastric Simulator (Kong &
Singh, 2010) developed at UC Davis (USA), and a mono-compartmental Gastric Digestor
Simulator (Kozu et al., 2014) developed at the University of Tsukuba (Japan). Currently,
a major downside of dynamic models is the relatively high cost and limited availability of
the specialized equipment. Regardless of the type of digestion model chosen (static or
dynamic), valuable, comparative data for digestibility and bioaccessibility can be attained,
assuming that: 1) the digestive conditions are appropriately selected and 2) proper controls
are incorporated into the study design.
While these developing digestion models involve computerized, mechanical
instrumentation, commonly used in vitro absorption models are biological. The
27
mechanism of cellular uptake varies depending on the nature of the compound (e.g., polar
or apolar) and can be passive or facilitated. For this reason, biological specimens are
helpful as they can mimic the behavior of intestinal epithelial cells. In the past, several
models, such as intestinal loops, vascularly perfused intestine, and everted sacs have been
used, but have limitations (Hidalgo, Raub, & Borchardt, 1989) and culture of actual
duodenal epithelia cells have been unsuccessful (Quaroni, Wands, Trelstad, & Isselbacher,
1979). Hidalgo, Raub, & Borchardt (1989) demonstrated that a human carcinoma cell
line, Caco-2 HTB-37, grows rapidly and differentiates into a squamous monolayer that,
when cultured appropriately, mimics the morphology of duodenal epithelial cells. Since
then, Caco-2 HTB-37 cells have been used by several groups to assess potential
bioavailability of various phytochemical compounds (Garrett, Failla, & Sarama, 1999;
Lipkie et al., 2014; Neilson, Song, Sapper, Bomser, & Ferruzzi, 2010; Redan et al., 2016;
Yee, 1997). Intracellular uptake, which refers to the amount of compound that
accumulates within the cells, can be studied using a single-compartment model (Garrett et
al., 1999; Lipkie et al., 2014). In this model, cells are grown on the bottom of the well and
the treatment media is gently added above (Figure 1-3). Following incubation of the cells
with the treatment, the cells can be washed, scraped and collected. Cellular transport or
chylomicron secretion, can be modeled using a two-compartment model (Failla,
Chitchumronchokchai, Ferruzzi, Goltz, & Campbell, 2014), in which the cells are grown
on a semi-permeable insert that is suspended above the bottom of the well (Figure 1-3).
Treatments and cell collection are handled analogously to the single-compartment model,
except that the amount of compound transported is collected from the bottom or
basolateral compartment while uptake can be considered the sum of what is in the cells
plus the concentration in the basolateral compartment.
28
Figure 1-3 Single- and two-compartment model for in vitro cellular uptake and
transport/chylomicron secretion, respectively. In the two-compartment model, the cells
are grown on an insert within the apical compartment. Carotenoids that are collected from
the basolateral compartment are assumed to be transported out of the enterocyte.
1.5.5 Carotenoid digestion and absorption
As lipophilic compounds, carotenoids are believed to undergo a gastrointestinal
fate that is analogous to other lipid compounds. Enzymes from the oral, gastric, and
intestinal phases act to break down macronutrients in the meal to allow release of
carotenoids and nutrients. Carotenoids are solubilized with oil droplets (if they were not
already incorporated within lipid droplets) in the stomach before entering the duodenum.
During intestinal digestion, bile salts displace proteins and other surface-active
compounds at the oil-water interface to form a mixed micelle. The mixed micelle, which
comprises of a lipid core (assuming the lipid content in the meal was sufficient), carries
hydrophobic carotenoids across the unstirred water layer to the enterocyte (Figure 1-4).
Intracellular carotenoid uptake is believed to be primarily a facilitated process, but may be
passive when pharmacological doses are present (Reboul & Emmanuelle, 2013). Several
membrane proteins, e.g., CD36, NPC1L1, have been identified as facilitators that
transport carotenoids into the enterocyte, but lycopene uptake is believed to be controlled
by a scavenger receptor protein SR-BI (Van Bennekum et al., 2005).
29
Figure 1-4 Schematic of lycopene absorption. Food and lipids are digested allowing for
release of mono- and diglycerols, free fatty acids, and lycopene. Bile salts, mono- and
diglycerides, and cholesterol stabilize mixed micelles, which can then transport lipophilic
compounds across the unstirred water layer to cross into the enterocyte passively
(diffusion) or actively through a membrane protein receptor (SR-BI).
1.6.3 Potential fate of emulsion delivery systems in the gastrointestinal tract
With the increased popularity of emulsion delivery research, there has been an
explosion of research that aims at understanding the potential biological fate of edible
delivery systems. Food-based delivery systems may be digested and absorbed similarly to
conventional food products, however, the biological fate in vivo will depend on the
structure of the delivery system, nature of the materials used, and the size and morphology
of the droplet or particle (Figure 1-5). With smaller droplet sizes, particularly with
nanoemulsions, questions have arised related to how these systems might be developed to
enhance stability, bioavailability, and if there are any associated safety risks. Most of the
safety concerns are typically associated with inorganic materials such as metals, which
have a higher risk for accumulation in the body. Additionally, since extremely small
particles (diameter ~10-30 nm), have shown the ability to cross the blood brain barrier
(Chau, Wu, & Yen, 2007), it is likely that highly stable and small structures could freely
30
cross membranes and potentially accumulate and lead to toxicity. The emulsions in this
dissertation research (Chapters 3-5) were formulated from biodegradable, food ingredients
and have droplet diameters >150 nm and have thus been assumed to be relatively safe to
handle.
Figure 1-5 Schematic of potential delivery system behavior and bioactive compound
uptake. A formulated delivery system may be resistant or partially resistant to digestion.
Complexes formed with biopolymers or stable delivery systems may be capable of
transporting bioactive compounds across the unstirred water layer without needing to be
reorganized into a mixed micelle. Uptake into the enterocyte is expected to follow similar
pathways as that of conventional carotenoid uptake (passive diffusion or facilitiated
difussion via a membrane protein). If particles, complexes, or delivery systems are small
enough, they may be capable of passing between the enterocytes (paracelluar).
1.7 Aims of the research
Considering the great potential that lycopene and other phytochemicals have as
functional ingredients, there is a need to develop efficient strategies for extraction and
encapsulation of high-value plant-based compounds. The literature shows various
extraction techniques have been explored for phytochemical extraction from plant
31
material. However, limited work exists specifically on lycopene. Considering that cislycopene has gained attention as a more bioavailable carotenoid compared to the all-transform (Cooperstone et al., 2015; Failla et al., 2008; Unlu, Bohn, Francis, Clinton, &
Schwartz, 2007), it is of interest to consider how processing conditions will affect isomer
recovery.
Emulsion-delivery systems have wide applicability to the food industry.
Consequently, it is important to gain a better understanding of how the chosen ingredients
(e.g., emulsifiers) affect stability (both physical and chemical) and if these systems truly
allow for enhanced performance and/or absorbability. Answering the latter question is
somewhat challenging as the human gastrointestinal tract is complex and difficult to
model. Additionally, there are several factors that simultaneously affect physical stability,
chemical retention, and bioaccessibility (e.g, surface area/droplet size, chemical
composition of emulsifier) that can be difficult to control across treatment types. A vast
majority of the in vitro digestion emulsion studies focus on digestibility, which is a crucial
aspect affecting potential biological fate. However, further work is needed to gain a
mechanistic understanding of how the interface affects bioaccessibility and bioavailability
of lipophilic bioactive compounds.
Thus, the aims of this research were to: 1) determine the effect of MAE conditions
on cis- and trans- lycopene recovery from tomato peels, a processing byproduct, 2)
determine the effect of dairy or plant protein on the physicochemical stability of lycopeneloaded emulsions, 3) investigate the effect of mixed protein interfaces on interfacial
rheology and emulsion stability, and 4) assess in vitro bioaccessibility and cellular uptake
of lycopene from emulsions.
32
1.8 References
21 CFR 172.840. (2016). Food additives permitted for direct addition to food for human
consumption. Retrieved July 4, 2017, from
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfCFR/CFRSearch.cfm?fr=172.8
40
Acosta, E. (2009). Bioavailability of nanoparticles in nutrient and nutraceutical delivery.
Current Opinion & Interface Science, 14, 3–15.
http://doi.org/10.1016/j.cocis.2008.01.002
Affandi, M. M. M., Julianto, T., & Majeed, A. (2011). Development and stability
evaluation of astaxanthin nanoemulsion. Asian J Pharm Clin Res, 4, 142–148.
Ahmed, K., Li, Y., McClements, D. J., & Xiao, H. (2012). Nanoemulsion- and emulsionbased delivery systems for curcumin: Encapsulation and release properties. Food
Chemistry, 132, 799–807. http://doi.org/10.1016/j.foodchem.2011.11.039
Alminger, M., Aura, A.-M., Bohn, T., Dufour, C., El, S. N., Gomes, A., … Santos, C. N.
(2014). In Vitro Models for Studying Secondary Plant Metabolite Digestion and
Bioaccessibility. Comprehensive Reviews in Food Science and Food Safety, 13(4),
413–436. http://doi.org/10.1111/1541-4337.12081
Anarjan, N., Mirhosseini, H., Baharin, B. S., & Tan, C. P. (2011). Effect of processing
conditions on physicochemical properties of sodium caseinate-stabilized astaxanthin
nanodispersions. LWT - Food Science and Technology, 44(7), 1658–1665.
http://doi.org/http://dx.doi.org/10.1016/j.lwt.2011.01.013
Anarjan, N., Tan, C. P., Nehdi, I. A., & Ling, T. C. (2012). Colloidal astaxanthin:
Preparation, characterization and bioavailability evaluation. Food Chemistry, 153(3),
1301–1309.
http://doi.org/http://dx.doi.org.ezproxy.lib.purdue.edu/10.1016/j.foodchem.2012.05.0
91
Anton, N., & Vandamme, T. F. (2011). Nano-emulsions and Micro-emulsions:
Clarifications of the Critical Differences. Pharmaceutical Research, 28, 978–985.
http://doi.org/10.1007/s11095-010-0309-1
33
Augustin, M. A., & Sanguansri, P. (2009). Chapter 5 Nanostructured Materials in the
Food Industry. In L. T. Steve (Ed.), Advances in Food and Nutrition Research (Vol.
Volume 58, pp. 183–213). Academic Press.
http://doi.org/http://dx.doi.org/10.1016/S1043-4526(09)58005-9
Barth, M. M., Zhou, C., Kute, K. M., & Rosenthal, G. A. (1995). Determination of
Optimum Conditions for Supercritical Fluid Extraction of Carotenoids from Carrot
(Daucus carota L.) Tissue. Journal of Agricultural and Food Chemistry, 43(11),
2876–2878. http://doi.org/10.1021/jf00059a019
Baysal, T., Ersus, S., & Starmans, D. A. J. (2000). Supercritical CO2 Extraction of βCarotene and Lycopene from Tomato Paste Waste. Journal of Agricultural and Food
Chemistry, 48(11), 5507–5511. http://doi.org/10.1021/jf000311t
Becher, P. (1967). Effect of preparation parameters on the initial size distribution function
in oil-in-water emulsions. Journal of Colloid and Interface Science, 24(1), 91–96.
http://doi.org/http://dx.doi.org/10.1016/0021-9797(67)90282-2
Blanquet, S., Zeijdner, E., Beyssac, E., Meunier, J.-P., Denis, S., Havenaar, R., & Alric,
M. (2004). A Dynamic Artificial Gastrointestinal System for Studying the Behavior
of Orally Administered Drug Dosage Forms Under Various Physiological
Conditions. Pharmaceutical Research, 21(4), 585–591. Retrieved from
https://www.researchgate.net/profile/Robert_Havenaar/publication/8565526_A_Dyn
amic_Artificial_Gastrointestinal_System_for_Studying_the_Behavior_of_Orally_Ad
ministered_Drug_Dosage_Forms_Under_Various_Physiological_Conditions/links/0
0b4951a8860eded01000000/
Butnariu, M. V, & Giuchici, C. V. (2011). The use of some nanoemulsions based on
aqueous propolis and lycopene extract in the skin’s protective mechanisms against
UVA radiation. Journal of Nanobiotechnology, 9(3), 1–9.
Calvo, M. M., García, M. L., & Selgas, M. D. (2008). Dry fermented sausages enriched
with lycopene from tomato peel. Meat Science, 80(2), 167–172.
34
Careri, M., Furlattini, L., Mangia, A., Musci, M., Anklam, E., Theobald, A., & von Holst,
C. (2001). Supercritical fluid extraction for liquid chromatographic determination of
carotenoids in Spirulina Pacifica algae: a chemometric approach. Journal of
Chromatography A, 912(1), 61–71. http://doi.org/http://dx.doi.org/10.1016/S00219673(01)00545-3
Castenmiller, J. J. M., & West, C. E. (1998). Bioavailibility and bioconversion of
carotenoids. Annual Review of Nutrition, 18(1), 19–38.
http://doi.org/doi:10.1146/annurev.nutr.18.1.19
Chan, C.-H., Yusoff, R., Ngoh, G.-C., & Kung, F. W.-L. (2011). Microwave-assisted
extractions of active ingredients from plants. Journal of Chromatography A,
1218(37), 6213–6225. http://doi.org/http://dx.doi.org/10.1016/j.chroma.2011.07.040
Chang, C., Tu, S., Ghosh, S., & Nickerson, M. T. (2015). Effect of pH on the interrelationships between the physicochemical, interfacial and emulsifying properties for
pea, soy, lentil and canola protein isolates. Food Research International, 77(3), 360–
367. http://doi.org/10.1016/j.foodres.2015.08.012
Chau, C.-F., Wu, S.-H., & Yen, G.-C. (2007). The development of regulations for food
nanotechnology. Trends in Food Science & Technology, 18(5), 269–280.
http://doi.org/http://dx.doi.org/10.1016/j.tifs.2007.01.007
Cheong, J. N., Tan, C. P., Man, Y. B. C., & Misran, M. (2008). α-Tocopherol
nanodispersions: Preparation, characterization and stability. Journal of Food
Engineering, 89(2). http://doi.org/http://dx.doi.org/10.1016/j.jfoodeng.2008.04.018
Chung, H.-Y., Rasmussen, H. M., & Johnson, E. J. (2004). Lutein bioavailability is higher
from lutein-enriched eggs than from supplements and spinach in men. The Journal of
Nutrition, 134(8), 1887–93. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/15284371
Cilek, B., Luca, A., Hasirci, V., Sahin, S., Sumnu, G., & Cilek, B. (2012).
Microencapsulation of phenolic compounds extracted from sour cherry pomace:
effect of formulation, ultrasonication time and core to coating ratio. European Food
Research and Technology, 235(4), 587–596. Retrieved from
http://search.proquest.com/docview/1223037250?accountid=13360
35
Cooperstone, J. L., Ralston, R. A., Riedl, K. M., Haufe, T. C., Schweiggert, R. M., King,
S. A., … Schwartz, S. J. (2015). Enhanced bioavailability of lycopene when
consumed as cis-isomers from tangerine compared to red tomato juice, a randomized,
cross-over clinical trial. Molecular Nutrition & Food Research, 59(4), 658–69.
http://doi.org/10.1002/mnfr.201400658
da Silva, R. P. F. F., Rocha-Santos, T. A. P., & Duarte, A. C. (2016). Supercritical fluid
extraction of bioactive compounds. TrAC Trends in Analytical Chemistry, 76, 40–51.
http://doi.org/10.1016/j.trac.2015.11.013
Dalgleish, D. G. (2006). Food emulsions—their structures and structure-forming
properties. Food Hydrocolloids, 20, 415–422.
http://doi.org/10.1016/j.foodhyd.2005.10.009
Dandekar, D. V, & Gaikar, V. G. (2002). Microwave assisted extraction of curcuminoids
from Curcuma longa. Separation Science and Technology, 37(11), 2669–2690.
http://doi.org/10.1081/SS-120004458
de la Fuente, J. C., Oyarzún, B., Quezada, N., & del Valle, J. M. (2006). Solubility of
carotenoid pigments (lycopene and astaxanthin) in supercritical carbon dioxide. Fluid
Phase Equilibria, 247(1–2), 90–95.
http://doi.org/http://dx.doi.org/10.1016/j.fluid.2006.05.031
Déat, E., Blanquet-Diot, S., Jarrige, J.-F., Denis, S., Beyssac, E., & Alric, M. (2009).
Combining the Dynamic TNO-Gastrointestinal Tract System with a Caco-2 Cell
Culture Model: Application to the Assessment of Lycopene and α-Tocopherol
Bioavailability from a Whole Food. Journal of Agricultural and Food Chemistry,
57(23), 11314–11320. http://doi.org/10.1021/jf902392a
Delazar, A., Nahar, L., Hamedyazdan, S., Sarker, S. D., Sarker, S. D., & Nahar, L. (2012).
Microwave-assisted extraction in natural products isolation. (J. M. Walker,
Ed.)Natural Products, Methods in Molecular Biology (3rd ed., Vol. 864). London,
UK: Springer Science+Business Media, LLC.
Dickinson, E. (1997). Properties of Emulsions Stabilized with Milk Proteins: Overview of
Some Recent Developments. Journal of Dairy Science, 80(10), 2607–2619.
http://doi.org/10.3168/jds.S0022-0302(97)76218-0
36
Dickinson, E. (1999). Adsorbed protein layers at fluid interfaces: interactions, structure
and surface rheology. Colloids and Surfaces B: Biointerfaces, 15(2), 161–176.
http://doi.org/10.1016/S0927-7765(99)00042-9
DiMascio, P., Devasagayam, T. P. A., Kaiser, S., & Sies, H. (1990). Carotenoids,
tocopherols and thiols as biological singlet molecular oxygen quenchers (Vol. 18).
http://doi.org/10.1042/bst0181054
Dinglery, A., & Gohla, S. (2008). Production of solid lipid nanoparticles (SLN): scaling
up feasibilities. Journal of Microencapsulation, 19(1), 11–16.
http://doi.org/10.1080/02652040010018056
Donsì, F., Sessa, M., & Ferrari, G. (2011). Effect of Emulsifier Type and Disruption
Chamber Geometry on the Fabrication of Food Nanoemulsions by High Pressure
Homogenization. Industrial & Engineering Chemistry Research, 51(22), 7606–7618.
http://doi.org/10.1021/ie2017898
El Darra, N., Grimi, N., Maroun, R., Louka, N., & Vorobiev, E. (2013). Pulsed electric
field, ultrasound, and thermal pretreatments for better phenolic extraction during red
fermentation. European Food Research and Technology, 236(1), 47–56.
http://doi.org/10.1007/s00217-012-1858-9
Failla, M. L., Chitchumronchokchai, C., Ferruzzi, M. G., Goltz, S. R., & Campbell, W. W.
(2014). Unsaturated fatty acids promote bioaccessibility and basolateral secretion of
carotenoids and α-tocopherol by Caco-2 cells. Food & Function, 5(6), 1101.
http://doi.org/10.1039/c3fo60599j
Failla, Mark, L., Chitchumroonchokchai, Chureeporn, Ishida, & Betty. (2008). In Vitro
Micellarization and Intestinal Cell Uptake of cis Isomers of. The Journal of Nutrition
Mar, 138(3), 482–486.
Fathi, M., Mozafari, M. R., & Mohebbi, M. (2012). Nanoencapsulation of food
ingredients using lipid based delivery systems. Trends in Food Science &
Technology, 23(1), 13–27. http://doi.org/http://dx.doi.org/10.1016/j.tifs.2011.08.003
Fellows, P. (2000). Processing using electric fields, high hydrostatic pressure, light or
ultrasound. In P. Fellows (Ed.), Food Processing Technology (Second, pp. 210–226).
Boca Raton, FL: CRC Press.
37
Frankel, E. N. (1993). In search of better methods to evaluate natural antioxidants and
oxidative stability in food lipids. Trends in Food Science and Technology, 4(7), 220–
225. http://doi.org/10.1016/0924-2244(93)90155-4
Fryd, M. M., & Mason, T. G. (2012). Advanced Nanoemulsions. Annual Review of
Physical Chemistry, 63(1), 439–518. http://doi.org/10.1146/annurev-physchem032210-103436
Gann, P. H., Ma, J., Giovannucci, E., Willett, W., Sacks, F. M., Hennekens, C. H., &
Stampfer, M. J. (1999). Lower prostate cancer risk in men with elevated plasma
lycopene levels: Results of a prospective analysis. Cancer Research, 59(1), 1225–
1230.
Ganta, S., Devalapally, H., & Amiji, M. (2010). Curcumin enhances oral bioavailability
and anti-tumor therapeutic efficacy of paclitaxel upon administration in
nanoemulsion formulation. Journal of Pharmaceutical Sciences, 99(11), 4630–4641.
http://doi.org/10.1002/jps.22157
Garrett, D. A., Failla, M. L., & Sarama, R. J. (1999). Development of an in Vitro
Digestion Method To Assess Carotenoid Bioavailability from Meals. Journal of
Agricultural and Food Chemistry, 47(10), 4301–4309.
http://doi.org/10.1021/jf9903298
Gärtner, C., Stahl, W., & Sies, H. (1997). Lycopene is more bioavailable from tomato
paste than from fresh tomatoes. The American Journal of Clinical Nutrition, 66(1),
116–122.
Georgieva, D., Schmitt, V. E., Leal-Calderon, F., & Langevin, D. (2009). On the Possible
Role of Surface Elasticity in Emulsion Stability. Langmuir, 25(10), 5565–5573.
http://doi.org/10.1021/la804240e
Hagiwara, A., Yoshino, H., Ichihara, T., Kawabe, M., Tamano, S., Aoki, H., … Shirai, T.
(2002). Prevention by natural food anthocyanins, purple sweet potato color and red
cabbage color, of 2-amino-1methyl-6-phenylimidazo[4,5-b]pyridine (PhIP)associated colorectal carcinogenesis in rats. Journal of Toxicological Sciences, 27(1),
57–68. http://doi.org/http://dx.doi.org/10.2131/jts.27.57
38
Hao, J., Han, W., Huang, S., Xue, B., & Deng, X. (2002). Microwave-assisted extraction
of artemisinin from Artemisia annua L. Separation and Purification Technology,
28(3), 191–196. http://doi.org/http://dx.doi.org/10.1016/S1383-5866(02)00043-6
Heaney, R. P. (2001). Factors influencing the measurement of bioavailability, taking
calcium as a model. In The Journal of nutrition (Vol. 131, p. 1344S–8S). American
Society for Nutrition. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/11285351
Hidalgo, I. J., Raub, T. J., & Borchardt, R. T. (1989). Characterization of the human colon
carcinoma cell line (Caco-2) as a model system for intestinal epithelial permeability.
Gastroenterology, 96(3), 736–749. http://doi.org/10.1016/0016-5085(89)90897-4
Hiranvarachat, B., Devahastin, S., Chiewchan, N., & Vijaya Raghavan, G. S. (2013).
Structural modification by different pretreatment methods to enhance microwaveassisted extraction of β-carotene from carrots. Journal of Food Engineering, 115(2),
190–197. http://doi.org/http://dx.doi.org/10.1016/j.jfoodeng.2012.10.012
Howitt, C. A., & Pogson, B. J. (2006). Carotenoid accumulation and function in seeds and
non-green tissues. Plant, Cell and Environment, 29(3), 435–445.
http://doi.org/10.1111/j.1365-3040.2005.01492.x
Jafari, S. M., Assadpoor, E., He, Y., & Bhandari, B. (2008). Re-coalescence of emulsion
droplets during high-energy emulsification. Food Hydrocolloids, 22(7), 1191–1202.
http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2007.09.006
Jafari, S. M., He, Y., & Bhandari, B. (2007). Production of sub-micron emulsions by
ultrasound and microfluidization techniques. Journal of Food Engineering, 82(4),
478–488. http://doi.org/http://dx.doi.org/10.1016/j.jfoodeng.2007.03.007
Jain, T., Jain, V., Pandey, R., Vyas, A., & Shukla, S. S. (2009). Microwave assisted
extraction for phytoconstituents- An overview. Asian Journal of Research Chemistry,
2(1), 19–25.
Jankowiak, L., Trifunovic, O., Boom, R. M., & Van Der Goot, A. J. (2014). The potential
of crude okara for isoflavone production. Journal of Food Engineering, 124, 166–
172. http://doi.org/10.1016/j.jfoodeng.2013.10.011
39
Jassie, L., Revesz, R., Kierstead, T., Hasty, E., & Matz, S. (1997). Microwave-assisted
solvent extraction. (H. M. S. Kingston & S. J. Haswell, Eds.)Microwave-Enhanced
Chemistry.
Jha, P., Flather, M., Lonn, E., Farkouh, M., & Yusuf, S. (1995). The Antioxidant Vitamins
and Cardiovascular Disease: A Critical Review of Epidemiologic and Clinical Trial
Data. Annals of Internal Medicine, 123(11), 860. http://doi.org/10.7326/0003-4819123-11-199512010-00009
Joslin, C. G., Gray, C. G., Goldman, S., Tomberli, B., & Li, W. (1996). Solubilities in
supercritical fluids from the virial equation of state. Molecular Physics, 89, 489–503.
Kalnin, D., Schafer, O., Amenitsch, H., & Ollivon, M. (2004). Fat Crystallization in
Emulsion: Influence of Emulsifier Concentration on Triacylglycerol Crystal Growth
and Polymorphism. Crystal Growth & Design, 4(6), 1283–1293.
http://doi.org/10.1021/cg030071k
Kaufmann, B., & Christen, P. (2002). Recent extraction techniques for natural products:
microwave-assisted extraction and pressurized solvent extraction. Phytochemical
Analysis, 13, 105–113. http://doi.org/10.1002/pca.631
Kentish, S., Wooster, T. J., Ashokkumar, M., Balachandran, S., Mawson, R., & Simons,
L. (2008). The use of ultrasonics for nanoemulsion preparation. Innovative Food
Science & Emerging Technologies, 9(2), 170–175.
http://doi.org/http://dx.doi.org/10.1016/j.ifset.2007.07.005
Kim, D. A., Cornec, M., & Narsimhan, G. (2005). Effect of thermal treatment on
interfacial properties of β-lactoglobulin. Journal of Colloid and Interface Science,
285, 100–109. http://doi.org/10.1016/j.jcis.2004.10.044
Knoblich, M., Anderson, B., & Latshaw, D. (2005). Analyses of tomato peel and seed
byproducts and their use as a source of carotenoids. Journal of the Science of Food
and Agriculture, 85(7), 1166–1170.
Kong, F., & Singh, R. P. (2010). A Human Gastric Simulator (HGS) to Study Food
Digestion in Human Stomach. Journal of Food Science, 75(9), E627–E635.
http://doi.org/10.1111/j.1750-3841.2010.01856.x
40
Kozu, H., Nakata, Y., Nakajima, M., Neves, M. A., Uemura, K., Sato, S., … Ichikawa, S.
(2014). Development of a Human Gastric Digestion Simulator Equipped with
Peristalsis Function for the Direct Observation and Analysis of the Food Digestion
Process. Food Science and Technology Research, 20(2), 225–233.
http://doi.org/10.3136/fstr.20.225
Labuza, T. P., & Dugan, L. R. (1971). Kinetics of lipid oxidation in foods. Critical
Reviews in Food Technology, 2(3), 355–405.
http://doi.org/10.1080/10408397109527127
Lampe, J. W., & Chang, J.-L. (2007). Interindividual differences in phytochemical
metabolism and disposition. Seminars in Cancer Biology, 17(5), 347–53.
http://doi.org/10.1016/j.semcancer.2007.05.003
Lee, L., & Norton, I. T. (2013). Comparing droplet breakup for a high-pressure valve
homogeniser and a microfluidizer for the potential production of food-grade
nanoemulsions. Journal of Food Engineering, 114(2), 158–163.
http://doi.org/http://dx.doi.org/10.1016/j.jfoodeng.2012.08.009
Let, M. B., Jacobsen, C., Sørensen, A.-D. M., & Meyer, A. S. (2007). Homogenization
Conditions Affect the Oxidative Stability of Fish Oil Enriched Milk Emulsions:
Lipid Oxidation. Journal of Agricultural and Food Chemistry, 55(5), 1773–1780.
http://doi.org/10.1021/jf062391s
Lipkie, T. E., Banavara, D., Shah, B., Morrow, A. L., McMahon, R. J., Jouni, Z. E., &
Ferruzzi, M. G. (2014). Caco-2 accumulation of lutein is greater from human milk
than from infant formula despite similar bioaccessibility. Molecular Nutrition &
Food Research, 58(10), 2014–2022. http://doi.org/10.1002/mnfr.201400126
Livny, O., Kaplan, I., Reifen, R., Polak-Charcon, S., Madar, Z., & Schwartz, B. (2002).
Lycopene inhibits proliferation and enhances gap-junction communication of KB-1
human oral tumor cells. The Journal of Nutrition, 132(12), 3754–3759.
41
Lu, Y., & Foo, L. Y. (2000). Antioxidant and radical scavenging activities of polyphenols
from apple pomace. Food Chemistry, 68, 81–85. Retrieved from
https://www.researchgate.net/profile/Yinrong_Lu/publication/222366229_Antioxida
nt_and_radical_scavenging_activities_of_polyphenols_from_apple_pomace__vegetables/links/541211130cf2fa878ad39653/Antioxidant-and-radical-scavengingactivities-of-polyphenols-f
Macías-Sánchez, M. D., Fernandez-Sevilla, J. M., Fernández, F. G. A., García, M. C. C.,
& Grima, E. M. (2010). Supercritical fluid extraction of carotenoids from
Scenedesmus almeriensis. Food Chemistry, 123(3), 928–935.
http://doi.org/http://dx.doi.org/10.1016/j.foodchem.2010.04.076
Mader, K. (2006). Solid lipid nanoparticles as drug carriers. In V. P. Torchilin (Ed.),
Nanoparticulaes as drug carriersr (pp. 187–212). London, UK: Imperial College
Press. Retrieved from
https://ebookcentral.proquest.com/lib/purdue/reader.action?docID=1681646&ppg=2
17
Mahdi Jafari, S., He, Y., & Bhandari, B. (2006). Nano-Emulsion Production by
Sonication and Microfluidization—A Comparison. International Journal of Food
Properties, 9(3), 475–485. http://doi.org/10.1080/10942910600596464
Mao, Li., Xu, D., Yang, J., Yuan, F., Gao, Y., & Zhao, J. (2009). Effects of Small and
Large Molecule Emulsifiers on the Characteristics of b-Carotene Nanoemulsions
Prepared by High Pressure Homogenization. Food Technology Biotechnology, 47(3),
336–342.
Matalanis, A., Jones, O. G., & McClements, D. J. (2011). Structured biopolymer-based
delivery systems for encapsulation, protection, and release of lipophilic compounds.
Food Hydrocolloids, 25(8), 1865–1880.
http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2011.04.014
McClements, D. J. (2012). Crystals and crystallization in oil-in-water emulsions:
Implications for emulsion-based delivery systems. Advances in Colloid and Interface
Science, 174, 1–30. http://doi.org/10.1016/j.cis.2012.03.002
42
McClements, D. J. (2012). Nanoemulsions versus microemulsions: terminology,
differences, and similarities. Soft Matter, 8(1), 1719–1729.
http://doi.org/0.1039/c2sm06903b
McClements, D. J., Decker, E. A., & Weiss, J. (2007). Emulsion-based delivery systems
for lipophillic bioactive components. Journal of Food Science, 72(8), R109–R124.
http://doi.org/10.1111/j.1750-3841.2007.00507.x
Meleson, K., Graves, S., & Mason, T. G. (2004). Formation of concentrated
nanoemulsions by extreme shear. Soft Materials, 2(2–3), 109–123.
http://doi.org/http://dx.doi.org/10.1081/SMTS-200056102
Ménard, O., Cattenoz, T., Guillemin, H., Souchon, I., Deglaire, A., Dupont, D., & Picque,
D. (2014). Validation of a new in vitro dynamic system to simulate infant digestion.
Food Chemistry, 145, 1039–1045. http://doi.org/10.1016/j.foodchem.2013.09.036
Minekus, M., Alminger, M., Alvito, P., Ballance, S., Bohn, T., Bourlieu, C., … Brodkorb,
A. (2014). A standardised static in vitro digestion method suitable for food – an
international consensus. Food Funct., 5(6), 1113–1124.
http://doi.org/10.1039/C3FO60702J
Narang, A. S., Delmarre, D., & Gao, D. (2007). Stable drug encapsulation in micelles and
microemulsions. International Journal of Pharmaceutics, 345(1–2), 9–25.
http://doi.org/http://dx.doi.org/10.1016/j.ijpharm.2007.08.057
Neas, E. D., & Collins, M. J. (1988). Introduction to Microwave Sample Preparation. In
H. M. Kingston & L. B. Jassie (Eds.), (pp. 7–32). Washington, D.C.: American
Chemical Society.
Nedovic, V., Kalusevic, A., Manojlovic, V., Levic, S., & Bugarski, B. (2011). An
overview of encapsulation technologies for food applications. Procedia Food
Science, 1(0), 1806–1815.
http://doi.org/http://dx.doi.org/10.1016/j.profoo.2011.09.265
Neilson, A. P., Song, B. J., Sapper, T. N., Bomser, J. A., & Ferruzzi, M. G. (2010). Tea
catechin auto-oxidation dimers are accumulated and retained by Caco-2 human
intestinal cells. Nutrition Research, 30(5), 327–340.
http://doi.org/10.1016/j.nutres.2010.05.006
43
Nobre, B., Marcelo, F., Passos, R., Beirão, L., Palavra, A., Gouveia, L., & Mendes, R.
(2006). Supercritical carbon dioxide extraction of astaxanthin and other carotenoids
from the microalga Haematococcus pluvialis. European Food Research and
Technology, 223(6), 787–790. http://doi.org/10.1007/s00217-006-0270-8
Ochekpe, N. A., Olorunfemi, P. O., & Ngwuluka, N. C. (2009). Nanotechnology and drug
delivery Part 2: Nanostructures for drug delivery. Tropical Journal of
Pharmaceutical Research, 8(3), 275–287.
Okigbo, R. N., Anuagasi, C. L., & Amadi, J. E. (2009). Advances in selected medicinal
and aromatic plants indigenous to Africa. Journal of Medicinal Plants Research,
3(2), 86–95. Retrieved from http://www.academicjournals.org/JMPR
Ostertag, F., Weiss, J., & McClements, D. J. (2012). Low-energy formation of edible
nanoemulsions: Factors influencing droplet size produced by emulsion phase
inversion. Journal of Colloid and Interface Science, 388(1), 95–102.
http://doi.org/http://dx.doi.org/10.1016/j.jcis.2012.07.089
Pan, J.-L., Wang, H.-M., Chen, C.-Y., & Chang, J.-S. (2012). Extraction of astaxanthin
from Haematococcus pluvialis by supercritical carbon dioxide fluid with ethanol
modifier. Engineering in Life Sciences, 12(6), 638–647.
http://doi.org/10.1002/elsc.201100157
Paré, J. R. J. (1994). Microwave extraction of volatile oils. .
Patel, A. R., & Velikov, K. P. (2011). Colloidal delivery systems in foods: A general
comparison with oral drug delivery. LWT - Food Science and Technology, 44(9),
1958–1964. http://doi.org/http://dx.doi.org/10.1016/j.lwt.2011.04.005
Patist, A., & Bates, D. (2008). Ultrasonic innovations in the food industry: From the
laboratory to commercial production. Innovative Food Science & Emerging
Technologies, 9(2), 147–154.
http://doi.org/http://dx.doi.org/10.1016/j.ifset.2007.07.004
44
Pingret, D., Fabiano-Tixier, A.-S., & Chemat, F. (2013). Ultrasound-assisted Extraction.
In M. A. Rostagno, J. M. Prado, J. H. Clark, & G. A. Kraus (Eds.), Natural Product
Extraction: Principles and Applications (pp. 89–112). Retrieved from
http://books.google.com/books?hl=en&lr=&id=GKqfELA7Nk8C&oi=fnd&pg=PA8
9&dq=ultrasound+color+extraction&ots=50URnU0n87&sig=Gw8726oat9IbIgY67a
GBtmJKsXg#v=onepage&q=ultrasound color extraction&f=false
Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2012). Physical and chemical
stability of β-carotene-enriched nanoemulsions: Influence of pH, ionic strength,
temperature, and emulsifier type. Food Chemistry, 132(3), 1221–1229.
http://doi.org/http://dx.doi.org/10.1016/j.foodchem.2011.11.091
Qian, C., & McClements, D. J. (2011). Formation of nanoemulsions stabilized by model
food-grade emulsifiers using high-pressure homogenization: Factors affecting
particle size. Food Hydrocolloids, 25(5), 1000–1008.
http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2010.09.017
Quaroni, A., Wands, J., Trelstad, R. L., & Isselbacher, K. J. (1979). Epitheliod cell
cultures from rat small intestine: characterization by morphologic and immunologic
criteria. Journal of Cell Biology, 80, 248–265. Retrieved from
http://jcb.rupress.org/content/jcb/80/2/248.full.pdf
Rao, J., & Mcclements, D. J. (2011). Formation of Flavor Oil Microemulsions,
Nanoemulsions and Emulsions: Influence of Composition and Preparation Method. J.
Agric. Food Chem, 59, 5026–5035. http://doi.org/10.1021/jf200094m
Rao, J., & McClements, D. J. (2012). Food-grade microemulsions and nanoemulsions:
Role of oil phase composition on formation and stability. Food Hydrocolloids, 29(2),
326–334. http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2012.04.008
Reboul, E., Borel, P., Mikail, C., Abou, L., Charbonnier, M., Caris-Veyrat, C., … Amiot,
M.-J. (2005). Enrichment of Tomato Paste with 6% Tomato Peel Increases Lycopene
and β-Carotene Bioavailability in Men. The Journal of Nutrition, 135(4), 790–794.
Retrieved from http://jn.nutrition.org/content/135/4/790.abstract
Reboul, E., & Emmanuelle. (2013). Absorption of Vitamin A and Carotenoids by the
Enterocyte: Focus on Transport Proteins. Nutrients, 5(9), 3563–3581.
http://doi.org/10.3390/nu5093563
45
Redan, B. W., Chegeni, M., Ferruzzi, M. G., Tresserra-Rimbau, A., Rimm, E. B., MedinaRemón, A., … Bai, H.-W. (2016). Differentiated Caco-2 cell monolayers exhibit
adaptation in the transport and metabolism of flavan-3-ols with chronic exposure to
both isolated flavan-3-ols and enriched extracts. Food Funct., 24, 639–647.
http://doi.org/10.1039/C6FO01289B
Richins, R. D., Hernandez, L., Dungan, B., Hambly, S., Holguin, F. O., & O’Connell, M.
A. (2010). A “green” extraction protocol to recover red pigments from hot Capsicum
fruit. HortScience, 45(7), 1084–1087.
Rodriguez-Amaya, D. B. (2001). A guide to carotenoid analysis in foods. ILSI press
Washington^ eD. C DC.
Roldán-Gutiérrez, J. M., & Dolores Luque de Castro, M. (2007). Lycopene: The need for
better methods for characterization and determination. TrAC Trends in Analytical
Chemistry, 26(2), 163–170.
http://doi.org/http://dx.doi.org/10.1016/j.trac.2006.11.013
Rozzi, N. L., Singh, R. K., Vierling, R. A., & Watkins, B. A. (2002). Supercritical Fluid
Extraction of Lycopene from Tomato Processing Byproducts. Journal of Agricultural
and Food Chemistry, 50(9), 2638–2643. http://doi.org/10.1021/jf011001t
Sagalowicz, L., & Leser, M. E. (2010). Delivery systems for liquid food products. Current
Opinion in Colloid & Interface Science, 15(1), 61–72.
Salvia-Trujillo, L., Rojas-Graü, M. A., Soliva-Fortuny, R., & Martín-Belloso, O. (2013).
Effect of processing parameters on physicochemical characteristics of microfluidized
lemongrass essential oil-alginate nanoemulsions. Food Hydrocolloids, 30(1), 401–
407. http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2012.07.004
Saravacos, G., & Kostaropoulos, A. E. (2016). Mechanical Processing Equipment. In G.
V. Barbosa-Canovas (Ed.), Handbook of Food Processing Equipment (2nd ed.). New
York, NY: Springer International Publishing. http://doi.org/10.1007/978-3-31925020-5_4
Sesso, H. D., Buring, J. E., Norkus, E. P., & Gaziano, J. M. (2004). Plasma lycopene,
other carotenoids, and retinol and the risk of cardiovascular disease in women. Am J
Clin Nutr, 79, 47–53. Retrieved from http://ajcn.nutrition.org/content/79/1/47.full.pdf
46
Sesso, H. D., Liu, S., Gaziano, J. M., & Buring, J. E. (2003). Dietary lycopene, tomatobased food products and cardiovascular disease in women. Journal of Nutrition,
133(7), 2336–2341.
Shahidi, F., & Han, X. (1993). Encapsulation of food ingredients. Critical Reviews in
Food Science and Nutrition, 33(6), 501–547.
http://doi.org/10.1080/10408399309527645
Shi, J., Dai, Y., Kakuda, Y., Mittal, G., & Xue, S. J. (2008). Effect of heating and
exposure to light on the stability of lycopene in tomato puree. Food Control, 19(5),
514–520.
Shi, J., & Xue, S. J. (2010). Supercritical-fluid extraction of lycopene from tomatoes. In S.
S. H. Rizvi (Ed.), Separation, extraction and concentration processes in the food,
beverage and nutraceuitcal industries (pp. 245–619). Philedelphia, PA: Woodhead
Publishing.
Silva, H. D., Cerqueira, M. A., Souza, B. W. S., Ribeiro, C., Avides, M. C., Quintas, M.
A. C., … Vicente, A. A. (2011). Nanoemulsions of β-carotene using a high-energy
emulsification–evaporation technique. Journal of Food Engineering, 102(2), 130–
135. http://doi.org/http://dx.doi.org/10.1016/j.jfoodeng.2010.08.005
Singh, R. P., & Heldman, D. R. (2009). Heat Transfer in Food Processing. In Introduction
to Food Engineering (4th ed., pp. 247–279). Burlington, MA: Academic Press.
Retrieved from https://app-knovelcom.ezproxy.lib.purdue.edu/web/view/pdf/show.v/rcid:kpIFEE0005/cid:kt00CBU4H
1/viewerType:pdf/root_slug:introduction-food-engineering/url_slug:heat-transfer-infood?cid=kt00CBU4H1&b-toc-cid=kpIFEE0005&b-toc-root-slug=introductionfood-en
Sopher, D. E., & David Sopher, B. E. (1964). Indigenous Uses of Turmeric (Curcuma
domestica) in Indigenous Uses of Turmeric (Curcuma domestica) in Asia and
Oceania. Anthropos, 59(1), 93–127. Retrieved from
http://www.jstor.org/stable/40456285
Spigno, G., & De Faveri, D. M. (2009). Microwave-assisted extraction of tea phenols: A
phenomenological study. Journal of Food Engineering, 93(2), 210–217.
http://doi.org/http://dx.doi.org/10.1016/j.jfoodeng.2009.01.006
47
Springob, K., & Kutchan, T. M. (2009). Introduction to the different classes of natural
products. In A. E. Osbourn & V. Lanzotti (Eds.), Plant-derived Natural Products:
synthesis, function, and application (pp. 3–50). New York, NY: Springer. Retrieved
from https://link-springer-com.ezproxy.lib.purdue.edu/content/pdf/10.1007%2F9780-387-85498-4.pdf
Stahl, W., & Sies, H. (1992). Uptake of lycopene and its geometrical isomers is greater
from heat-processed than from unprocessed tomato juice in humans. The Journal of
Nutrition, 122(11), 2161–6. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/1432255
Sun, Y., Liao, X., Wang, Z., Hu, X., & Chen, F. (2007). Optimization of microwaveassisted extraction of anthocyanins in red raspberries and identification of
anthocyanin of extracts using high-performance liquid chromatography-mass
spectrometry. European Food Research Technology, 225(3–4), 511–523.
http://doi.org/10.1007/s00217-006-0447-1
Surh, J., Decker, E., & McClements, D. (2006). Influence of pH and pectin type on
properties and stability of sodium-caseinate stabilized oil-in-water emulsions. Food
Hydrocolloids, 20(5), 607–618. http://doi.org/10.1016/j.foodhyd.2005.07.004
Surh, Y.-J. (2003). Cancer chemoprevention with dietary phytochemicals. Nature Reviews
Cancer, 3, 768–780. http://doi.org/10.1038/nrc1189
Tan, C. P., & Nakajima, M. (2005a). Effect of polyglycerol esters of fatty acids on
physicochemical properties and stability of β-carotene nanodispersions prepared by
emulsification/evaporation method. Journal of the Science of Food and Agriculture,
85, 121–126.
Tan, C. P., & Nakajima, M. (2005b). β-Carotene nanodispersions: preparation,
characterization and stability evaluation. Food Chemistry, 92(4), 661–671.
http://doi.org/http://dx.doi.org/10.1016/j.foodchem.2004.08.044
Tang, S. Y., Shridharan, P., & Sivakumar, M. (2013). Impact of process parameters in the
generation of novel aspirin nanoemulsions – Comparative studies between ultrasound
cavitation and microfluidizer. Ultrasonics Sonochemistry, 20(1), 485–497.
http://doi.org/http://dx.doi.org/10.1016/j.ultsonch.2012.04.005
48
Tong, L. M., Sasaki, S., Mcclements, D. J., & Decker, E. A. (2000). Mechanisms of the
Antioxidant Activity of a High Molecular Weight Fraction of Whey. Journal of
Agricultural and Food Chemistry, 48(5), 1473–1478.
http://doi.org/10.1021/jf991342v
Undevia, S. D., Gomez-Abuin, G., & Ratain, M. J. (2005). Pharmacokinetic variability of
anticancer agents. Nature Reviews Cancer, 5(6), 447–458.
http://doi.org/10.1038/nrc1629
Unlu, N. Z., Bohn, T., Francis, D., Clinton, S. K., & Schwartz, S. J. (2007). Carotenoid
Absorption in Humans Consuming Tomato Sauces Obtained from Tangerine or
High-β-Carotene Varieties of Tomatoes. Journal of Agricultural and Food
Chemistry, 55, 1597–1603. http://doi.org/10.1021/JF062337B
Vainioa, H., & Rautalahti, M. (1998). An International Evaluation of the Cancer
Preventive Potential of Carotenoids. Cancer Epidemiology, Biomarkers &
Prevention, 7, 725–728. Retrieved from
http://cebp.aacrjournals.org/content/cebp/7/8/725.full.pdf
Van Bennekum, A., Werder, M., Thuahnai, S. T., Han, C.-H., Duong, P., Williams, D. L.,
… Hauser, H. (2005). Class B Scavenger Receptor-Mediated Intestinal Absorption of
Dietary -Carotene and Cholesterol. Biochemistry, 44, 4517–4525.
http://doi.org/10.1021/bi0484320
Vaughn Katherine, L. S., Clausen Edgar, C., King Jerry, W., Howard Luke, R., & Julie, C.
D. (2008). Extraction conditions affecting supercritical fluid extraction (SFE) of
lycopene from watermelon. Bioresource Technology, 99(16), 7835–7841.
http://doi.org/http://dx.doi.org/10.1016/j.biortech.2008.01.082
Vilkhu, K., Mawson, R., Simons, L., & Bates, D. (2008). Applications and opportunities
for ultrasound assisted extraction in the food industry — A review. Innovative Food
Science & Emerging Technologies, 9(2), 161–169.
http://doi.org/http://dx.doi.org/10.1016/j.ifset.2007.04.014
Vishwanathan, R., Wilson, T. A., & Nicolosi, R. J. (2009). Bioavailability of a
nanoemulsion of lutein is greater than lutein supplement. Nano Biomed Eng, 1(1),
38–49. http://doi.org/10.5101/nbe.v1i1.p38-49
49
Walstra, P. (2003). Formation of emulsions and foams. In Physical Chemistry of Foods
(pp. 398–436). New York, NY: Marcel Dekker, Inc.
Walstra, P. (2008). Dispersed systems: basic considerations. In Fennema’s Food
Chemistry (4th ed., pp. 783–847). Boca Raton, FL: CRC Press.
Wandrey, C., Bartkowiak, A., & Harding, S. E. (2010). Materials for encapsulation.
Encapsulation Technologies for Active Food Ingredients and Food Processing.
http://doi.org/10.1007/978-1-4419-1008-0_3
Wang, L., Tabor, R., Eastoe, J., Li, X., Heenan, R. K., & Dong, J. (2009). Formation and
stability of nanoemulsions with mixed ionic-nonionic surfactants. Physical Chemistry
Chemical Physics, 11(1), 9772–9778. http://doi.org/10.1039/b912460h
Wang, L., & Weller, C. L. (2006). Recent advances in extraction of nutraceuticals from
plants. Trends in Food Science & Technology, 17(6), 300–312.
http://doi.org/http://dx.doi.org/10.1016/j.tifs.2005.12.004
Wang, X., Jiang, Y., Wang, Y.-W., Huang, M.-T., Ho, C.-T., & Huang, Q. (2008).
Enhancing anti-inflammation activity of curcumin through O/W nanoemulsions.
Food Chemistry, 108(2), 419–424.
http://doi.org/http://dx.doi.org/10.1016/j.foodchem.2007.10.086
Williams, A. W., Boileau, T. W. M., & Erdman, J. W. (1998). Factors influencing the
uptake and absorption of carotenoids. Experimental Biology and Medicine, 218(2),
106–108.
Wood, R. (2005). Bioavailability: definition, general aspects and fortification. In B.
Caballero, A. Prentice, & L. Allen (Eds.), Encyclopedia of human nutrition (2nd ed.).
Retrieved from
https://www.ars.usda.gov/research/publications/publication/?seqNo115=166852
Yang, Y., Marshall-Breton, C., Leser, M. E., Sher, A. A., & McClements, D. J. (2012).
Fabrication of ultrafine edible emulsions: Comparison of high-energy and lowenergy homogenization methods. Food Hydrocolloids, 29(2), 398–406.
http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2012.04.009
Yee, S. (1997). In Vitro Permeability Across Caco-2 Cells (Colonic) Can Predict In Vivo
(Small Intestinal) Absorption in Man—Fact or Myth. Pharmaceutical Research,
14(6), 763–766. http://doi.org/10.1023/A:1012102522787
50
Yuan, Y., Gao, Y., Zhao, J., & Mao, L. (2008). Characterization and stability evaluation
of β-carotene nanoemulsions prepared by high pressure homogenization under
various emulsifying conditions. Food Research International, 41(1), 61–68.
http://doi.org/http://dx.doi.org/10.1016/j.foodres.2007.09.006
Zou, T., Wang, D., Guo, H., Zhu, Y., Luo, X., Liu, F., & Ling, W. (2012). Optimization of
microwave-assisted extraction of anthocyanins from mulberry and identification of
anthocyanins in extract using HPLC-ESI-MS. Journal of Food Science, 77(1), C46c50. http://doi.org/10.1111/j.1750-3841.2011.02447.x
Zuidam, N. J., & Shimoni, E. (2010). Overview of Microencapsulates for Use in Food
Products or Processes and Methods to Make Them. In N. J. Zuidam & V. Nedvoic
(Eds.), Encapsulation Technologies for Active Food Ingredients and Food
Processing (pp. 3–29). New York, NY: Springer Science + Business Media, LLC.
51
CHAPTER 2 MICROWAVE-ASSISTED EXTRACTION OF
LYCOPENE IN TOMATO PEELS: EFFECT OF EXTRACTION
CONDITIONS ON ALL-TRANS AND CIS- ISOMER YIELDS
Kacie K.H.Y. Ho1, Mario G. Ferruzzi1, Andrea M.1 Liceaga, and M Fernanda San
Martín-González 1
1
Department of Food Science, Purdue University, 745 Agricultural Mall Dr, West
Lafayette, IN, 47906, USA
Reproduced with permission. Full citation:
Ho, K. K. H. Y., Ferruzzi, M. G., Liceaga, A. M., & San Martín-González, M. F. (2015).
Microwave-assisted extraction of lycopene in tomato peels: Effect of extraction conditions
on all-trans and cis-isomer yields. LWT-Food Science and Technology, 62(1), 160-168.
2.1 Introduction
The tomato industry is a multi-billion dollar market with the US being a top
producer of tomatoes for processed foods (Thornsbury, 2012). In 2009, production
exceeded 13 million tons (Economic Research Service, 2010), of which, 12% (the peel
portion) was considered waste despite having more lycopene than the pulp by weight (AlWandawi, Abdul-Rahman, & Al-Shaikhly, 1985; George, Kaur, Khurdiya, & Kapoor,
2004). Lycopene, C40H56, is the primary pigment responsible for the red hue in tomatoes,
watermelon, and blood oranges (Rodriguez-Amaya, 2001). As an acyclic, highly
conjugated isoprenoid, lycopene is the most potent singlet oxygen quencher of all
carotenoids (Di Mascio, Kaiser, & Sies, 1989). Consumption of lycopene from tomatoes
has been associated with protection against oxidative DNA damage and anticancer
properties (Agarwal & Rao, 2000), thus making it a compound of interest amongst
medical and nutrition researchers.
52
Aside from potential health benefits, lycopene offers an alternative to synthetic
food colorants. From a processing standpoint, extraction can be difficult as food grade
solvent choices are limited. However, isolating lycopene from tomato peels can reduce
the overall cost by adding value to an otherwise discarded waste product.
Lycopene is
insoluble in water and poorly soluble in organic solvents, which limits its removal from
raw plant material. However, extraction efficiency of carotenoids can be improved by
using solvent combinations to facilitate partitioning. Previous research indicated that
solvent systems containing hexane and ethyl acetate are the most efficient for carotenoid
extraction from tomato seeds and peels (Strati & Oreopoulou, 2011). Despite
improvements, the extraction procedure itself is time consuming and poses the risk of
degradation as samples are exposed to heat for extended periods of time. Due to this
limitation, pure lycopene is often expensive (Ascenso, Pinho, Eleutério, Praça, Bentley,
Oliveira, et al., 2013). Improvements in extraction efficiency or reduction in extraction
time may reduce the processing costs while producing a high value color.
In its natural form, lycopene is heat resistant and present in a thermodynamically
stable, all-trans, crystal within the chromoplasts of plant cells (Harris & Spurr, 1969).
Conventional extraction often requires heat to facilitate the migration of solvent to extract
pigment compounds. Although increased temperatures correspond with improved
solubility and organelle membrane disruption, heat exposure should be limited when
possible due to the thermolabile nature of carotenoids once they are in solvent (RodriguezAmaya, 2001). Although lycopene has been shown to be more stable in general against
isomerization and degradation compared to β-carotene (Nguyen & Schwartz, 1998)
previous studies have demonstrated that heat treatments, longer than 1 hour, favored the
trans-to-cis isomer conversion of lycopene while light irradiation induced cis-isomer
degradation over time in tomato products (Chen, Shi, Xue, & Ma, 2009; Shi, Dai, Kakuda,
Mittal, & Xue, 2008).
Microwave-assisted extraction (MAE) may provide a solution for this since this
technology induces rapid heating primarily within polar constituents due to dipole rotation
and ionic drifting (Neas & Collins, 1988). In theory, superheating of polar cellular
components will improve migration of lycopene into the extraction solvent, while the
short treatment times limit heat exposure of the nonpolar components. Previously, MAE
53
has been used to enhance extraction of catechins, anthocyanins and curcuminoids (Baiano,
Bevilacqua, Terracone, Contò, & Del Nobile, 2014; Dandekar & Gaikar, 2002; Zou,
Wang, Guo, Zhu, Luo, Liu, et al., 2012) among others has improved efficiency compared
to conventional extraction. Although MAE of various phytochemicals has been
investigated, limited research has been done on the effect of MAE on cis vs. trans isomer
yield. Thus, the objectives of this study were to 1) determine the optimal MAE conditions
for lycopene from tomato peels using response surface methodology and 2) evaluate the
effect of treatment on cis- and trans- lycopene yields.
2.2 Materials and methods
2.2.1 Reagents and standards
All-trans-lycopene standard and all reagents were purchased from Sigma
Chemical Co. (St. Louis, MO). Solvents were purchased from J.T. Baker (Phillipsburg,
NJ). Tomato peels were generously donated by a Red Gold Co. (Elwood, IN). To prevent
light-induced degradation of lycopene, all extractions were done in yellow light and
extraction solvents contained butylated hydroxytoluene (BHT) to limit oxidation
occurring during the centrifugation and handling of the extracts.
2.2.3 Raw materials and sample preparation
Tomato peels were obtained from a local processing facility as a byproduct of
tomato paste. During the tomato processing, caustic lye was used to remove peels.
Consequentially, received tomato peels were collected in bulk and neutralized with
hydrochloric acid until a pH of 7 was obtained. Excess moisture was removed by
squeezing peels with a cheesecloth prior to storage. All samples were flushed with
nitrogen and stored at -20°C until further processing.
Since smaller particle sizes better facilitate extraction, the peels were further
processed prior to microwave treatment. Frozen peels were ground using a spice grinder
until a particle size of < 0.5 cm was achieved. The moisture content of the ground peels
was analyzed with a MAX2000 Computrac Moisture Analyzer (Arizona Instruments,
Chandler, AZ USA). Ideally, the moisture content of each sample should be quantified,
54
however, due to the destructive nature of moisture analysis, the frozen supply of ground
tomato peels were sampled from ten different locations within the sample stock. The
mean value (70.345 + 1.405) was later used to calculate the extraction yield of lycopene
per weight of tomato peel on a dry weight basis. Although using the mean moisture
content is not the best way to express the data, the variability between sampled portions
was low (<2%).
Peels were not dried as the water present increased polarity, which could aid in
selective heating during microwave irradiation. Ground peels were stored in glass, screw
top bottles, flushed with nitrogen, and stored at -20°C until treated.
2.2.4 Experimental design
Response surface methodology (RSM) was used to determine the effect of
extraction parameters on lycopene yield. Initially, RSM was conducted to assess four
factors, solvent ratio (X1), solid-liquid ratio (X2), microwave power (X3), and energy
equivalents (X4), which were varied by adjusting treatment time, with a Box-Behnken
design comprising of 3 center points (Table 2-1). Treatment times were calculated based
off of the relationship between Power, energy, and time and are listed in Appendix A,
Table A-1. A secondary RSM was employed to investigate solvents containing a higher
ethyl acetate (EA) percentage. For this only two factors, solvent ratio (X1) and microwave
power (X2), were studied with a central composite design (CCD) with two center points
(Table 2-1). A second-order polynomial equation (Eq. 2.1) was used to express the
response yield of all-trans-lycopene and cis-lycopene (Yi) as a function of the
experimental factors (Xi) for each RSM design:
 = 0 + ∑=1   + ∑=1   2 + ∑=1    (2.1)
where b0 is a constant, bn, bmn, and bnm are the linear, quadratic, and interaction
coefficients, respectively. The multiple regression models were analyzed separately for
each Yi, such that one response was a function of four (low EA) or two (high EA)
independent variables. The model was predicted using regression analysis and analysis of
variance (ANOVA).
55
Table 2-1 Response surface methodology parameters
Low EA Experiments
High EA Experiments
Coded Value
Coded Value
Factor
-1
0
1
-1
0
1
1:0
1.5:0.5
1:1
2:8
1:9
0:1
1:20
2:20
4:20
N/A; Fixed at 1:20 g/mL
Power (W)
400
800
1600
400
Energy (kJ)
24
36
48
Solvent ratio
(mL Hexane : mL
Ethyl acetate)
Solid-liquid ratio
(g/mL)
800
1600
N/A; Fixed at 24 kJ
2.2.5 Microwave-assisted extraction of lycopene
Ground tomato peels were thawed to room temperature and weighed into teflonlined extraction vessels at 1, 2, or 4 g. Precisely 20 mL of corresponding solvent was
added with a magnetic stir bar prior to capping. A Mars Xpress microwave extraction
system (CEM Corp., Matthews, NC) was used at 400, 800, and 1600W at varying times to
achieve delivered energy equivalents of 24, 36, and 48 kJ. Within the closed microwave
system, 8 extraction vessels were arranged in a carousel following CEM Corp. protocol.
Although 8 vessels were irradiated, only three vessels were sampled and analyzed as the
triplicates per treatment.
Approximately 10 mL of saturated sodium chloride in water was added to the
treated samples to facilitate partitioning and to break emulsions formed at the interface.
This was then transferred to a 50 mL polypropylene tube and centrifuged in a 5804
centrifuge (Eppendorf, Hamburg, Germany) at 4,472g. The organic phase was collected
and centrifugation was repeated with additional solvent two more times to ensure
collection of all extracted lycopene. All organic phases were pooled, filtered through
anhydrous sodium sulfate to remove residual water and adjusted to 50 mL prior to drying
56
2 mL aliquots under nitrogen and freezer storage, at -20°C. Although direct injection
would be more efficient, hexane was removed to prevent solvent effects during analysis.
2.2.6 Conventional extraction of lycopene
Conditions used for the conventional extraction were selected to emulate the
optimum conditions determined by response surface methodology. Conventional
extraction was conducted with 1 g of ground tomato peels and 20 mL of a 1:1 (mL:mL)
mixture of hexane (1 mg mL-1 BHT)-ethyl acetate in a 50 mL polycenrifuge tube. The
tube was placed in a shaking water bath for 15 seconds at 45 °C, which falls within the
temperature ranges observed for the optimal MAE treatment. Since conventional solvent
extraction typically involves a longer heating time, another treatment was done following
the same conditions, except the heat treatment was extended to 30 minutes. The
conventional methods used for high EA treatment (0:1 solvent ratio, 1:20 solid-liquid
ratio, 400 W, 24 kJ equivalents for 1 minute) comparison followed the same protocol,
except 20 mL of ethyl acetate (1 mg mL-1 BHT) was used as the solvent and heated for 1
minute and 30 minutes. Following heat treatment, the extraction process was the same as
that done for MAE after microwave irradiation.
2.2.7 Quantification with high performance liquid chromatography (HPLC-DAD)
Carotenoid analysis was done using reversed phase HPLC-DAD based on the
method used by Kean, Hamaker, and Ferruzzi (2008) using an Agilent 1200 Series HPLC,
equipped with a diode array detector and a YMC Carotenoid S-3 C-30 column (2.0 × 150
mm, 3 µm particle size). A binary mobile phase of methanol with 2% aqueous
ammonium acetate (pH=4.5) and ethyl acetate was used at a flow rate of 0.37 mL/min
with a gradient as follows: 0% B (0 minutes), 80% B (6 minutes), 100% B (12 minutes), 0
% B (14 minutes). Precisely 10 µL of sample was injected and lycopene was quantified at
470 nm. Peak spectra were collected within the 200-600 nm range and analyzed with
Chemstation software (Agilent Technologies, Santa Clara, CA).
Chromatograms of all-trans-lycopene standard yielded a peak at a retention time
of ~11 minutes. Three cis-isomers of lycopene were separated and all-trans-lycopene and
57
isomers were identified by comparing retention times with carotenoid profiles of a test
salad containing known carotenoids (Goltz, Campbell, Chitchumroonchokchai, Failla, &
Ferruzzi, 2012) to rule out extraction of non-lycopene carotenoids. Since only translycopene is readily available as a standard, cis-isomers were collectively quantified from
the calibration curve of the all-trans-lycopene. The 5-cis-isomer, which was observed as a
pronounced shouldering peak off of trans-lycopene, was quantified along with the other
cis-isomers.
For calibration, a small amount of all-trans-lycopene standard was solubilized in
petroleum ether to make a stock lycopene solution with an absorbance ~0.8. The
absorbance of this solution and subsequent dilutions were read using a UV-Vis DU 800
spectrophotometer (Beckman and Coulter, Inc., Brea, CA) at 470 nm. The stock solution
was diluted to concentrations between 6.0-0.04 μM. The absorbance, done in triplicate,
was then used to calculate the concentration of the stock and six dilutions with a molar
extinction coefficient of 1.85 x 105 L*mol-1*cm-1. Each lycopene dilution was dried under
nitrogen and analyzed with HPLC-DAD to correlate peak area with lycopene
concentration. The coefficient of determination (R2) of the calibration curve was 0.999,
with a limit of detection (LOD) and limit of quantitation (LOQ) of 0.31 and 0.94 μM,
respectively. The LOD and LOQ were calculated based off of the standard deviation (SD)
of the intercept and slope, based on Validation of Analytical Procedures Methodology
Q2B ICHHT (2005).
2.2.8 Electron microscopy imaging
Transmission electron microscopy (TEM) was used to assess the effect of
treatment on tomato peel structure. A non-extracted ground tomato peel sample,
optimally treated samples (low EA) and 30-minute conventionally extracted samples were
imaged by the Purdue Life Science Microscopy Facility (West Lafayette, IN). Processed
tomato peels were received in acetone, transferred to fresh acetone containing 0.01 g mL-1
osmium tetroxide. After several rinses in fresh acetone they were embedded in EMbed812 resin. Thin sections were cut on a Reichert-Jung Ultracut E ultramicrotome and
stained with 0.02 g mL-1 uranyl acetate and lead citrate. Images were acquired on a FEI
Tecnai T20 electron microscope equipped with a LaB6 source and operating at 200 kV.
58
Since the ground tomato peels used for this experiment were previously processed,
fresh tomatoes were sampled from a local grocery store and used as a reference for tomato
structure. These were fixed in 0.025 g mL-1 glutaraldehyde in 100 moles mL-1 sodium
cacodylate buffer, post-fixed in buffered 0.01 g mL-1 osmium tetroxide containing 0.008
mg mL-1 potassium ferricyanide, dehydrated with a graded series of ethanol, and
embedded in EMbed-812 resin. Thin sections were cut stained, and visualized following
the same protocol as done for the ground tomato peels.
2.2.9 Statistical analysis
Statistical analysis was conducted with JMP version 10 (SAS Institute Inc. 2012
Cary, NC). Data was subjected to analysis of variance (ANOVA) where factors and
values were considered significant at P<0.05. Pairwise comparisons between control and
optimized extraction yields were conducted post-hoc using the Tukey-Kramer method
(α=0.05). Lycopene content was expressed as mg/100g dry weight and each data point is
represented by the mean values and SD of three independent extractions.
2.3 Results and discussion
2.3.1 Lycopene recovery of low EA extractions
In all MAE extractions, the primary compound was all-trans-lycopene (Figure 21). Only two parameter estimates, interaction effects solvent type*power and
power*energy were significant at the α=0.05 level, while the solid-liquid ratio did not
appear to significantly (P=0.330) affect the all-trans-lycopene extraction yield. Based on
the RSM the maximum predicted extraction yield of all-trans-lycopene was determined to
be 10.362 mg/100g (Figure 2-2) with a solvent ratio of 1:1 treated for 15 seconds (24 kJ
equivalents) at 1600W.
59
Figure 2-1 Representative chromatogram of carotenoid extract from MAE of tomato peels
at 470 nm. Suspected peak identies are as follow: (a) β-carotene, (b) cis-lycopene isomer,
(c) cis-lycopene isomer, (d) all-trans-lycopene, (e) 5-cis-lycopene.
Statistical analysis of cis-isomer extraction yield indicated that the solid-liquid
ratio and the interaction effect of solvent ratio*solid-liquid ratio were significant. This
suggests that cis-isomer yields are increased as the solid-liquid ratio decreases and the EA
proportion increases (Figure 2-3). In most cases (treatments 4 vs. 22, 3 vs. 23, 5 vs. 25,
and 8 vs. 24), solvent ratio with a higher proportion of EA was shown to increase the %
cis yield (Table 2-2). No parameters were found to be significant for affecting total
lycopene yield.
60
60
Figure 2-2 Response surface plots for all-trans-lycopene yield from low EA MAE with solvent ratio vs. power (top row) and energy
vs. power (bottom row) plotted. Power levels are fixed at (a) 24kJ, (b) 36 kJ, and (c) 48 kJ and solvent ratios are fixed at (d) 1:0
hexane:EA, (e) 1.5:0.5 hexane:EA, and (f) 1:1 mL hexane : mL EA solvent ratio. The maximum predicted extraction yield was (g)
10.362 mg/100g with a treatment comprising of: 1:1 mL hexane : mL EA solvent ratio, 1600 W, 24 kJ. Plotted response values
represent predicted values from the model.
61
Table 2-2 Cis-, trans-, and total lycopene yields from low EA MAE
Treatment
No.
Coded
Lycopene Yield mg/100 g
factors
(X1-X4)
Cis Isomers
% cis
Trans
% trans
Total
1
0+0-
1.429
+
0.04
15.339
7.727
+
0.251
82.943
9.316
+
0.289
2
0000
1.701
+
0.555
21.247
6.615
+
1.207
82.626
8.006
+
1.193
3
-00+
2.128
+
0.102
19.460
8.499
+
1.033
77.723
10.935
+
0.994
4
--00
0.537
+
0.363
8.894
5.501
+
1.08
91.106
6.038
+
1.399
5
-0-0
1.791
+
0.39
17.949
7.877
+
1.7
78.944
9.978
+
2.094
6
0000
1.838
+
0.428
20.696
6.73
+
0.668
75.780
8.881
+
1.088
7
00+-
1.675
+
0.328
20.269
6.278
+
0.533
75.968
8.264
+
0.359
8
-00-
1.878
+
0.223
22.128
6.295
+
0.568
74.172
8.487
+
0.774
9
0++0
1.27
+
0.026
16.920
6.079
+
1.451
80.989
7.506
+
1.479
10
0000
2.039
+
0.05
23.442
6.339
+
1.474
72.879
8.698
+
1.445
11
+0+0
2.38
+
0.031
19.883
9.279
+
0.864
77.519
11.97
+
0.898
12
0-0-
3.012
+
1.652
27.253
7.397
+
0.67
66.929
11.052
+
2.279
13
00-+
2.065
+
0.197
21.907
7.041
+
0.743
74.698
9.426
+
0.632
14
0+-0
1.178
+
0.177
18.134
5.16
+
0.468
79.433
6.496
+
0.537
15
0-0+
3.946
+
0.245
34.111
6.999
+
1.45
60.503
11.568
+
1.608
61
62
Table 2-2 continued
16
0-+0
1.676
+
1.747
22.881
5.043
+
0.546
68.846
7.325
+
2.211
17
0+0+
1.103
+
0.19
19.546
4.383
+
0.431
77.671
5.643
+
0.459
18
0--0
2.209
+
1.38
30.269
4.473
+
0.914
61.291
7.298
+
2.289
19
-0+0
1.295
+
0.672
24.160
3.752
+
0.693
70.000
5.36
+
1.357
20
-+00
1.192
+
0.037
21.214
4.268
+
0.198
75.957
5.619
+
0.222
21
00++
0.873
+
0.069
27.758
2.802
+
0.525
89.094
3.145
+
0.631
22
+-00
4.174
+
0.495
34.160
7.435
+
0.276
60.848
12.219
+
0.266
23
+00+
1.721
+
0.152
23.057
5.43
+
1.031
72.749
7.464
+
1.131
24
+00-
1.546
+
0.097
19.513
6.064
+
1.509
76.537
7.923
+
1.608
25
+0-0
1.677
+
0.506
20.137
6.344
+
0.84
76.177
8.328
+
1.155
26
++00
1.009
+
0.069
20.100
3.855
+
0.093
76.793
5.02
+
0.157
27
00--
1.391
+
0.454
26.225
3.594
+
0.097
67.760
5.304
+
0.451
* Lycopene yields represent means + SD (n=3)
62
63
Comparison between lycopene yields of the predicted optimized MAE treatment
(1:1 solvent ratio, 1:0 solid-liquid ratio, 1600 W, 24 kJ equivalents done for 15 seconds)
vs. conventional treatment conducted at the same time and temperature in a shaking water
bath indicated that MAE exhibited a significantly greater all-trans-lycopene yield
compared to the 15-second but not the 30-minute conventional extraction (Figure 2-4).
However, no differences were found between cis-isomer and total lycopene yield.
Figure 2-3 Response surface plot for cis-lycopene yield from low EA MAE. The
maximum cis-isomer extraction yield (a) was predicted to be 4.450 mg/100g with a
solvent ratio of 1:1 mL hexane: mL EA and a 1:20 solid-liquid ratio.
It should be pointed out that when the predicted optimal conditions were actually
tested, the all-trans-lycopene yield obtained (Figure 2-4) was less than the value predicted
by the model (Figure 2-2). Statistical analysis indicated that the model for all-translycopene had a significant lack of fit with P-value=0.0143 and a low coefficient of
determination (R2=0.58). Treatment 11 (1:1 solvent ratio, 2:20 solid–liquid ratio, 1600
W, 36 kJ equivalents with a 30 second treatment) had the greatest all-trans-lycopene yield
at 9.279 + 0.864 mg/100 g (Table 2-2). This may be due to the need for additional energy,
more than 24kJ, to extract lycopene at the given solvent and power level. A statistical
drawback of the Box-Behnken design is that over interpretation due to extrapolation
64
towards the corners of the response surface can occur. Thus, a second RSM experiment
(high EA) using a CCD was conducted focusing on only two factors.
20
18
Control 1 min
Control 30 min
MAE 1 min
Yield (mg/100g)
16
e
14
de
12
10
c
8
6
a
a
c
d
a
4
b
2
0
cis
trans
total
Figure 2-4 Comparison of control (conventional) methods vs. optimized low EA MAE.
The MAE conditions used (1:1 solvent ratio, 1:20 solid-liquid ratio, 1600 W, 24 kJ
equivalents for 15 seconds) were determined as optimal by RSM. Extraction yields of cis,
trans, and total lycopene are shown where same letters denote values that are not
significantly different at the α=0.05 level based on the Tukey Kramer method for pairwise
comparisons. Response values shown represent the mean + SD (n=3).
2.3.2 Lycopene recovery of high EA extractions
Since RSM demonstrated increasing all-trans-lycopene yields with increasing EA
concentrations (Figure 2-2), this second set of experiments employed solvent mixtures
with lower hexane-to-EA ratios (2:8, 1:9, 0:1) and fixed all treatments at 24 kJ equivalents
with 400 W (1 minute), 800 W (30 seconds), or 1600 W (15 seconds). Solid-liquid ratio
(1:20) was fixed because a limited supply of sample was available. Surface plots also
indicated that adjustments in power and energy could improve yields, however, the Mars
Xpress microwave extraction system has only three power settings, 400, 800, and 1600 W,
thus preventing the ability to increase or decrease power. Energy was also limited as
preliminary testing demonstrated that high-energy inputs caused solvent evaporation,
65
which would effectively shut down the system for safety reasons. Since a higher
proportion of EA increases the polarity of the solvent and the rate of heating, a fixed low
energy equivalent (24 kJ) was chosen for the high EA experiments.
ANOVA of high EA MAE indicated that there was a significant difference
amongst treatments(P=0.0164) for all-trans-lycopene extraction. In this case, only
solvent ratio was found to be a significant factor influencing the extraction yield of alltrans-lycopene. The model did not exhibit a significant lack of fit (P=0.1624), and
predicted a maximum yield of 13.872 mg/100g (Figure 2-5) for an extraction with ethyl
acetate at a power of 400 W, which for a 24 kJ equivalent had a treatment time of 1
minute. For cis-isomer extraction, no significant difference was found amongst
treatments. However, the % cis of the extracts was greatest for treatments 1 and 9, which
were the only ones using a solvent ratio of 2:8 (Table 2-3).
66
Table 2-3 Cis-, trans-, and total lycopene yields from high EA MAE
Lycopene Yield mg/100 g
Treatment
Coded factors
No.
(X1-X4)
1
--
3.909
+
0.243
43.991
4.633
+
1.944
52.138
8.886
+
1.817
2
00
3.857
+
2.336
39.490
5.562
+
2.956
56.947
9.767
+
0.621
3
+-
3.44
+
0.636
19.799
12.195
+
0.884
70.187
17.375
+
0.253
4
-0
3.409
+
0.436
35.378
5.891
+
0.41
61.135
9.636
+
0.851
5
0-
3.288
+
0.898
25.980
9.028
+
0.178
71.334
12.656
+
0.867
6
+0
2.624
+
0.295
17.344
12.175
+
1.611
80.475
15.129
+
1.892
7
0+
2.263
+
0.366
22.493
7.454
+
0.346
74.088
10.061
+
0.379
8
00
2.824
+
0.998
30.799
6.002
+
0.857
65.460
9.169
+
0.205
9
-+
3.753
+
0.082
42.287
4.788
+
0.461
53.949
8.875
+
0.42
10
++
3.577
+
0.806
24.470
10.711
+
0.671
73.273
14.618
+
1.471
Cis Isomers
% cis
Trans
% trans
Total
* Lycopene yields represent means + SD (n=3)
66
67
Figure 2-5 Response surface plot for all-trans-lycopene yield from high EA MAE. The
maximum all-trans-extraction yield (a) was predicted to be 13.872 mg/100g with a full
EA solvent and when treated at 400 W. Solvent ratio significantly affected the extraction
yield (P=0.004) while power did not (P=0.210). Plotted response values indicate mean +
SD (n=3).
68
The actual optimal all-trans-lycopene yield was determined as 13.592 mg/100 g
(Figure 2-6), which was statistically greater than the 1 minute (P=0.0006) and 30-minute
conventional extraction (P<0.0001). Similarly, the total lycopene yield was significantly
greater for the optimized MAE treatment compared to the 30-minute control (P<0.0001)
and conventional control (P=0.006). Significant differences were not observed amongst
treatments for cis-isomer yield (P>0.05). However, the proportion of cis-isomers to
extracted trans-lycopene is dramatically higher for the 30-minute control compared to the
1-minute control and the optimized MAE treatment. The relative increase in cis-isomers
may be due to the longer treatment time.
20
18
Control 1 min
Control 30 min
MAE 1 min
Yield (mg/100g)
16
d
14
12
8
4
e
b
10
6
g
f
a
a
a
c
2
0
cis
trans
total
Figure 2-6 Comparison of control (conventional) methods vs. the high EA MAE
treatment with the highest all-trans-lycopene yield. The MAE conditions used (0:1
solvent ratio, 1:20 solid-liquid ratio, 400 W, 24 kJ equivalents for 60 seconds) were
determined as optimal by RSM. Extraction yields of cis, trans, and total lycopene are
shown where same letters denote values that are not significantly different at the α=0.05
level based on the Tukey Kramer method for pairwise comparisons. Response values
shown represent the mean + SD (n=3).
The literature reports different lycopene recoveries depending on the extraction
method and type of raw material. Enzyme assisted extraction, was found to be extremely
efficient with 440 mg/100g of lycopene from tomatoes (Lavecchia & Zuorro, 2008),
although the process can be costly. It should also be noted that whole tomatoes may
69
contain more lycopene since they have not been previously processed. Studies done on
tomato peels reported yields ranging from 0.639-73.40 mg/100 g (Kaur, Wani, Oberoi, &
Sogi, 2008; Knoblich, Anderson, & Latshaw, 2005; Shi, Yi, Xue, Jiang, Ma, & Li, 2009).
Kaur, Wani, Oberoi, and Sogi (2008) found that a maximum recovery of 1.98 mg
lycopene/100 g was attainable when tomato skins (0.15 mm particle size) were
conventionally extracted with hexane:acetone:alcohol (2:1:1 mL:mL:mL) with 0.5 mg
mL-1 BHT at a 1:30 solid-liquid ratio (w/v), at 50°C for 8 minutes four times.
Specifically, lycopene yield increased as a function of extraction number (repeated on
one sample) and decreasing particle size. The authors hypothesized that the extractions
conducted at 50°C allowed for better breakdown of chromoplast membranes compared to
cooler conditions, yet did not induce significant degradation compared to treatments done
at 60°C. Shi, Yi, Xue, Jiang, Ma, and Li (2009) determined a higher total lycopene
content in dried tomato skins at ~13.0 mg/100g when extracted with hexane overnight at
45°C. The authors compared this to supercritical fluid extraction and achieved a
maximum recovery of 73.3% when ethanol and olive oil were used as modifiers at 75°C
and 35 MPa, which is less efficient compared to the results in this study. Although the
results presented in this study from MAE are likely an improvement over conventional
solvent extraction and over supercritical fluid extraction, lycopene yields were still
significantly lower than expected. Lower yields in this study may be partially due to
previous processing (i.e. hot break) that peels underwent, which lowered the amount of
extractable lycopene (Kaur, Wani, Oberoi, & Sogi, 2008) or due to differences in tomato
variety (George, Kaur, Khurdiya, & Kapoor, 2004).
Tomato peels following treatments (MAE and conventional) were still visibly
orange, suggesting that lycopene remained in the peel as a non-extractable fraction.
Calvo, Dado, and Santa-Maria (2007) encountered a similar result when they heated
freeze-dried tomato peels in ethanol or ethyl acetate at temperatures ranging from 25-60°
C. Ethanol was found to have a greater lycopene extraction yield, possibly due to its
ability to better penetrate the peels compared to ethyl acetate, however, residual pigment
remained in the sample following treatment. Attempts were made to remove all the
apparent color from the tomato peels (data not shown), however none of the procedures
were able to completely remove pigments. A well established extraction procedure
70
involving sonication of chloroform-soaked peels (Jun, 2006; Naviglio, Caruso, Iannece,
Aragòn, & Santini, 2008; Rozzi, Singh, Vierling, & Watkins, 2002) was tested, however,
the treated peels exhibited little noticeable color following treatment and yielded low
lycopene concentrations. Modifications were also made to the extraction procedure
following methods done previously by Goltz, Campbell, Chitchumroonchokchai, Failla,
and Ferruzzi (2012) by testing different solvent types and ratios, further reducing particle
size with a high shear mixer, and using caustic pretreatments (done at various times and
temperatures) with 0.4 g mL-1 potassium hydroxide in methanol. However, the latter
resulted in the pigment loss into the aqueous phase, due to either degradation of the
carotenoids (exhibited by lack of color or the development of a dark color). To determine
absolute lycopene content in tomato peels, future studies require treatments that can
effectively disrupt or degrade the physical cell structure barriers without affecting
embedded lycopene.
2.3.3 Transmission electron microscopy of tomato peels
TEM micrographs were unable to display cell ultrastructure of byproduct tomato
peels, possibly due to the extent of processing the peels underwent prior to receiving.
However, the ultrastructure with lycopene bodies can be seen in the fresh tomato peel
samples as the spherical electron dense (dark) regions (Figure 7a).
71
a
b
Lycopene
bodies
c
Limited fissures
d
Fissures
Fissures
Figure 2-7 TEM images of a fresh tomato peel with no extraction (a), byproduct tomato
peel with no extraction (b), byproduct tomato peel subjected to control extraction for 30
minutes (c), and byproduct tomato peel subjected to MAE (1:1 solvent ratio, 1:20 solidliquid ratio, 1600 W, 24 kJ, for 15 seconds) (d). Visibly more holes and fissures are
present in extracted samples, thus suggesting that MAE, and to some extent conventional
extraction, cause structural disruption. Scale bars indicate 1 µm.
The non-treated byproduct tomato peel (Figure 7b) appeared to be significantly
less damaged compared to treated peels (Figure 7c & 7d). In particular, the MAE (1:1
solvent ratio, 1:20 solid-liquid ratio, 1600 W, 24 kJ, for 15 seconds) treated samples
exhibited significantly more structural disruption as fissures and gray wholes appeared to
be more prevalent. This suggests that the MAE was better able to disrupt cellular
structure to reduce physical extraction barriers.
2.4 Conclusions
Optimization data indicated that solvent ratio and microwave power in relation to
energy equivalents significantly affected the all-trans-lycopene extraction yield. Cisisomer extraction was primarily affected by the solvent ratio and solid-liquid ratio. The
maximum all-trans-lycopene yield of ~13 mg/100 g was obtained with ethyl acetate at
400 W, with a 24 kJ equivalent (1 minute). Significantly more all-trans-lycopene was
72
extracted with ethyl acetate via MAE compared to a 1-minute and 30-minute
conventional treatment. TEM suggested that selective, physical disruption occurs in the
tomato peels during MAE. All-trans-lycopene has been of interest for food and
pharmaceutical industries since it is the most stable isomer. Additionally, all-translycopene exhibits greater color intensity compared to cis-isomers due to a hypsochromic
shift and smaller extinction coefficient of the latter (Schieber & Carle, 2005). However,
interest in cis-lycopene is growing as there is some evidence indicating that these isomers
are more bioavailable compared to the all-trans form (Boileau, Boileau, & Erdman,
2002). Although certain limitations to MAE exist (i.e. consumer preference against
solvent use and challenges with scaling up) the findings of this study offer applicable
information that could steer other extraction techniques towards cis or trans-isomer
recovery, depending on the application.
2.5 Acknowledgements
This material is based upon work supported by the National Science Foundation
Graduate Research Fellowship under Grant No. DGE-1333468 and the Purdue
Department of Food Science Industry Fellows Program. The authors thank Red Gold (IN,
USA) for providing tomato peels, Mohneet Ahuja from the Purdue University Statistical
Consulting Service for advice on statistical design and Laurie Mueller from the Purdue
Life Science Microscopy Facility for her expertise with TEM.
73
2.6 References
Agarwal, S., & Rao, A. V. (2000). Tomato lycopene and its role in human health and
chronic disease. Canadian Medical Association, 163(6), 739-744.
Al-Wandawi, H., Abdul-Rahman, M., & Al-Shaikhly, K. (1985). Tomato processing
wastes as essential raw materials source. Journal of Agricultural and Food
Chemistry, 33(5), 804-807.
Ascenso, A., Pinho, S., Eleutério, C., Praça, F. G., Bentley, M. V. L. B., Oliveira, H.,
Santos, C., Silva, O., & Simões, S. (2013). Lycopene from Tomatoes: Vesicular
Nanocarrier Formulations for Dermal Delivery. Journal of Agricultural and Food
Chemistry, 61(30), 7284-7293.
Baiano, A., Bevilacqua, L., Terracone, C., Contò, F., & Del Nobile, M. A. (2014). Single
and interactive effects of process variables on microwave-assisted and
conventional extractions of antioxidants from vegetable solid wastes. Journal of
Food Engineering, 120(0), 135-145.
Boileau, T. W.-M., Boileau, A. C., & Erdman, J. W. (2002). Bioavailability of all-trans
and cis-isomers of lycopene. Experimental Biology and Medicine, 227(10), 914919.
Calvo, M. M., Dado, D., & Santa-Maria, G. (2007). Influence of extraction with ethanol
or ethyl acetate on the yield of lycopene, β-carotene, phytoene, and phytofluene
from tomato peel powder. European Food Research Technology, 224(1), 567571.
Chen, J., Shi, J., Xue, S. J., & Ma, Y. (2009). Comparison of lycopene stability in waterand oil-based food model systems under thermal- and light-irradiation treatments.
LWT - Food Science and Technology, 42(3), 740-747.
Dandekar, D. V., & Gaikar, V. G. (2002). Microwave assisted extraction of curcuminoids
from Curcuma longa. Separation Science and Technology, 37(11), 2669-2690.
Di Mascio, P., Kaiser, S., & Sies, H. (1989). Lycopene as the most efficient biological
carotenoid singlet oxygen quencher. Archives of Biochemistry and Biophysics,
274(2), 532-538.
74
Economic Research Service, U. S. D. A. (2010). U.S. Tomato Statistics (92010). In).
Online.
George, B., Kaur, C., Khurdiya, D. S., & Kapoor, H. C. (2004). Antioxidants in tomato
(Lycopersium esculentum) as a function of genotype. Food Chemistry, 84(1), 4551.
Goltz, S. R., Campbell, W. W., Chitchumroonchokchai, C., Failla, M. L., & Ferruzzi, M.
G. (2012). Meal triacylglycerol profile modulates postprandial absorption of
carotenoids in humans. Mol. Nutr. Food Res., 56, 866-877.
Harris, W. M., & Spurr, A. R. (1969). Chromoplasts of Tomato Fruits. II. The Red
Tomato. American Journal of Botany, 56(4), 380-389.
ICHHT, G. (2005). Validation of analytical procedures: text and methodology Q2 (R1).
IFPMA: Geneva.
Jun, X. (2006). Application of high hydrostatic pressure processing of food to extracting
lycopene from tomato paste waste. High Pressure Research, 26(1), 33-41.
Kaur, D., Wani, A. A., Oberoi, D. P. S., & Sogi, D. S. (2008). Effect of extraction
conditions on lycopene extractions from tomato processing waste skin using
response surface methodology. Food Chemistry, 108(2), 711-718.
Kean, E. G., Hamaker, B. R., & Ferruzzi, M. G. (2008). Carotenoid Bioaccessibility from
Whole Grain and Degermed Maize Meal Products. Journal of Agricultural and
Food Chemistry, 56(21), 9918-9926.
Knoblich, M., Anderson, B., & Latshaw, D. (2005). Analyses of tomato peel and seed
byproducts and their use as a source of carotenoids. Journal of the Science of
Food and Agriculture, 85(7), 1166-1170.
Lavecchia, R., & Zuorro, A. (2008). Improved lycopene extraction from tomato peels
using cell-wall degrading enzymes. European Food Research and Technology,
228(1), 153-158.
Naviglio, D., Caruso, T., Iannece, P., Aragòn, A., & Santini, A. (2008). Characterization
of High Purity Lycopene from Tomato Wastes Using a New Pressurized
Extraction Approach. Journal of Agricultural and Food Chemistry, 56(15), 62276231.
75
Neas, E. D., & Collins, M. J. (1988). Introduction to Microwave Sample Preparation. In
H. M. Kingston & L. B. Jassie (Eds.), (pp. 7-32). Washington, D.C.: American
Chemical Society.
Nguyen, M. L., & Schwartz, S. J. (1998). Lycopene stability during food processing.
Experimental Biology and Medicine, 218(2), 101-105.
Rodriguez-Amaya, D. B. (2001). A guide to carotenoid analysis in foods: ILSI press
Washington^ eD. C DC.
Rozzi, N. L., Singh, R. K., Vierling, R. A., & Watkins, B. A. (2002). Supercritical Fluid
Extraction of Lycopene from Tomato Processing Byproducts. Journal of
Agricultural and Food Chemistry, 50(9), 2638-2643.
Schieber, A., & Carle, R. (2005). Occurrence of carotenoid cis-isomers in food:
Technological, analytical, and nutritional implications. Trends in Food Science &
Technology, 16(9), 416-422.
Shi, J., Dai, Y., Kakuda, Y., Mittal, G., & Xue, S. J. (2008). Effect of heating and
exposure to light on the stability of lycopene in tomato puree. Food Control,
19(5), 514-520.
Shi, J., Yi, C., Xue, S. J., Jiang, Y., Ma, Y., & Li, D. (2009). Effects of modifiers on the
profile of lycopene extracted from tomato skins by supercritical CO2. Journal of
Food Engineering, 93(4), 431-436.
Strati, I. F., & Oreopoulou, V. (2011). Process optimisation for recovery of carotenoids
from tomato waste. Food Chemistry, 129(3), 747-752.
Thornsbury, S. (2012). Tomatoes. In Vegetables & Pulses, vol. 2012): USDA Economic
Research Service.
Zou, T., Wang, D., Guo, H., Zhu, Y., Luo, X., Liu, F., & Ling, W. (2012). Optimization
of microwave-assisted extraction of anthocyanins from mulberry and
identification of anthocyanins in extract using HPLC-ESI-MS. Journal of Food
Science, 77(1), C46-c50.
76
CHAPTER 3 PHYSICOCHEMICAL STABILITY OF LYCOPENELOADED EMULSIONS STABILIZED BY PLANT OR DAIRY
PROTEINS
Kacie K.H.Y. Hoab, Karin Schroëna , M. Fernanda Martín-Gonzálezb, Claire C. BertonCarabina
a
Food Process Engineering Group, Department of Agrotechnology and Food Sciences,
Wageningen University, Bornse Weilanden 9, 6708 WG Wageningen, The Netherlands
b
Department of Food Science, College of Agriculture, Purdue University, 745
Agriculture Mall Drive, West Lafayette, IN 47907, USA
Reproduced with permission. Full citation:
Ho, K. K.H.Y., Schroën, K., San Martín-González, M. F., & Berton-Carabin, C. C.
(2016). Physicochemical stability of lycopene-loaded emulsions stabilized by plant or
dairy proteins. Food Structure. 12: 34-42
3.1 Introduction
Recently, there has been growing interest in enriching food products with
bioactive ingredients (e.g., flavors, vitamins, antioxidants or phytochemicals) to produce
a desired functionality. Lycopene is the most potent singlet oxygen quencher amongst
carotenoids (Di Mascio, Kaiser, & Sies, 1989; Rao, Waseem, & Agarwal, 1998) that
could be used as a naturally derived antioxidant or as a health-promoting ingredient.
However, lycopene is largely insoluble in water and chemically labile. Therefore,
encapsulation strategies should be considered, such as using emulsion-based delivery
systems.
Dairy proteins have been extensively used for food applications, and in particular
to stabilize the interface in oil-in-water (O/W) emulsions. Compared to other emulsifiers
(e.g., surfactants or modified starch), dairy proteins, such as whey protein isolate (WPI)
77
and sodium caseinate (SC), can improve the physical and chemical stability of
carotenoid-loaded emulsions (Mao et al., 2009; Mao, Yang, Yuan, & Gao, 2010). The
high colloidal stability is attributed to the ability of dairy proteins to form thick and
sterically-stabilized interfacial layers (Dickinson, 2001). In emulsion stability, the
interfacial protein layer plays a critical role in the physical stabilization process (Wilde,
2000). Amongst dairy proteins, whey proteins (mostly represented by the globular protein
β-lactoglobulin) have a rigid structure, which is known to lead to different interfacial
organization compared to SC (primarily β-casein), which has a flexible structure
(Dickinson, 2013) and in turn may lead to different effects on the physical and perhaps
chemical stability of emulsions. Besides, Cornacchia & Roos (2011) found that the
different protein chemistries of WPI and SC affected β-carotene retention in O/W
emulsions, with the latter protein providing a better oxidative barrier. Dairy protein
emulsifiers have also proved to promote the bioavailability of bioactives: Interfacial WPI
combined with Tween 20 or sucrose laurate demonstrated improved cellular uptake of
lycopene and astaxanthin, compared to Tween 20 alone, from formulated emulsions in
colon carcinoma cells (lines HT-29 and Caco-2) in vitro (Ribeiro et al., 2006). Although
the mechanism of enhanced bioavailability was not elucidated, the authors alluded to
potential interactions between the carotenoids and β-lactoglobulin as a possible
explanation.
The drawback of using dairy proteins for producing functional food emulsions is
their low sustainability and impact on the environment (VandeHaar & St-Pierre, 2006;
Erb et al., 2016). Plant proteins represent a large and relatively underutilized resource that
is more sustainable and requires less energy for production compared to their animalderived counterpart (de Boer, Helms, & Aiking, 2006; O’Kane, Vereijken, Happe,
Gruppen, & J S Van Boekel, 2004). Recent reviews (Shi & Dumont, 2014; Song, Tang,
Wang, & Wang, 2011) have also highlighted functional properties of different biobased
films from plant proteins as the utilization of such renewable proteins has gained
popularity. Despite the growing interest for plant-derived proteins as emulsifiers (Chihi,
Mession, Sok, & Saurel, 2016), the link with stabilization of bioactive components in
O/W delivery systems is hardly ever made. Many plant proteins, including soy protein
isolate (SPI) and pea protein isolate (PPI) have been reported as promising functional
78
emulsifiers (Aoki, Taneyama, & Inami, 1980; Bengoechea, Cordobés, & Guerrero, 2006;
Lam & Nickerson, 2013; Pelgrom, Berghout, Van Der Goot, Boom, & Schutyser, 2014;
Phoon, San Martin-Gonzalez, & Narsimhan, 2014), yet it is still arguable whether they
perform as well as dairy proteins, or even outperform them (Chove, Grandison, & Lewis,
2001). SPI and PPI are both from commonly consumed plant sources and exhibit good
emulsifying properties as they have been shown to form stable O/W droplets that were
not drastically bigger compared to β-lactoglobulin-stabilized droplets (Benjamin, Silcock,
Beauchamp, Buettner, & Everett, 2014). Interfacial properties of SPI and PPI have also
been studied and demonstrate potential to physically stabilize O/W emulsions by forming
strong viscoelastic films (Chang et al. 2015). Despite the numerous studies characterizing
soy and pea protein functionality, limited work (Fernandez-Avila, Arranz, Guri, Trujillo,
& Corredig, 2016; Tapal & Tiku, 2012) has been conducted specifically on SPI,
consisting primarily of globular proteins glycinin and conglycinin (Chronakis, 1996), and
PPI, consisting primarily of legumin and vicilin/convicilin (O’Kane et al., 2004), for
improving bioactive delivery. Tapal & Tiku (2012) conducted research on curcumin and
SPI complexation and found that >80% of the bioactive was retained during simulated
gastric conditions. Fernandez-Avila et al. (2016) also found promising results for plant
protein (SPI and PPI)-stabilized emulsions, as conjugated linoleic acid (CLA) delivery
was enhanced compared against non-emulsified CLA for both proteins in a Caco-2 cell
model. Despite these promising first results, it is still unknown whether plant proteins
could be a valuable alternative to dairy proteins for the production of functional
emulsions loaded with bioactives, such as lycopene. In fact, direct comparisons between
plant and dairy proteins and the link between interfacial properties and bioactive
encapsulation have hardly been touched upon.
For the design of emulsion-based encapsulation systems, we believe it is necessary to
connect the physicochemical stability of emulsions with the structural organization of the
oil-water interface. Consequently, the aims of this study were to determine the effect of
interfacial dairy or plant protein on the: 1) physical stability (particle size and zeta
potential) and 2) chemical stability (lycopene retention) of emulsions, and 3) interfacial
organization (adsorption kinetics and dilatational rheology). Ultimately, we have
79
attempted to relate these findings and provide guidelines for the design of sustainable
protein-stabilized emulsion-based delivery systems.
3.2 Materials and methods
3.2.1 Materials
Rapeseed oil and tomato paste for lycopene extraction were purchased from local
supermarkets (Wageningen, Netherlands). MP Alumina N-Super I (MP Biomedicals,
France) was mixed with rapeseed oil overnight as previously described (Berton, Genot, &
Ropers, 2011) to strip the oil of tocopherols and surface-active impurities. All-translycopene standard, all solvents (analytical grade) and other reagents were purchased from
Sigma Aldrich (Zwijndrecht, the Netherlands). Proteins were generously donated by the
suppliers as follows: 97.5% purity WPI (BIPRO, Davisco, Switzerland), 80% purity SC
(Sodium Caseinate S, DMV International, Amersfoort, Netherlands), and 90% purity SPI
(soy protein isolate SUPRO EX 37, Solae Europe SA, Switzerland) and 80-90% PPI (pea
protein isolate NUTRALYS F85, Roquette, France). Ultrapure water (Millipore Milli-Q
water purification system) was used for all experiments.
3.2.2 Preparation of lycopene oil stock
Approximately 250 g of tomato paste were combined with 10 g of celite, 10 g of
sodium bicarbonate, and 500 mL of an extraction solvent (1:1 v/v hexane (0.1% butylated
hydroxytoluene w/v) – ethyl acetate). The mixture was held under a stream of nitrogen
and in an ice-bath while stirring at 250 rpm with an overhead IKA mixer for 1.5 hours.
The mixture was then vacuum filtered with No. 1 filter paper (Whatman, United
Kingdom) to separate solids from liquids, transferred to a separatory funnel, and washed
with a saturated solution of sodium chloride in water. The lower aqueous phase was
drained and the upper hexane layer was collected, flushed with nitrogen and rotary
evaporated almost to dryness. Stripped oil (~80 g) was added to solubilize the lycopene
crystals prior to transferring to a borosilicate screw top bottle. The resulting lycopene-inoil mixture was held under a stream of nitrogen to remove residual solvent until constant
weight was achieved. This entire process was repeated 10 times and individual batches of
80
lycopene oil were pooled, prior to aliquoting into 35 mL batches, flushing with nitrogen,
and storing at -20˚C. The lycopene content of the stock oil was determined after dilution
in hexane spectrophotometrically at 471 nm, analyzed using high-performance liquid
chromatography (Kean, Hamaker, & Ferruzzi, 2008), and then compared against an alltrans-lycopene standard to identify cis- and trans- isomers (Ho, Ferruzzi, Liceaga, & San
Martín-González, 2015). The resulting stock oil had a total lycopene concentration of
0.236 mg/g of oil and consisted primarily of all-trans-lycopene (~90%).
3.2.3 Preparation of the aqueous phase
WPI and SC were added to 0.01 M phosphate buffer (pH=7) and stirred with a
magnetic stir bar overnight at room temperature at 100 rpm prior to emulsification the
following morning. SPI and PPI both contained a non-soluble fraction and thus required
additional pre-treatment prior to use in emulsification. SPI or PPI were combined with
0.01 M phosphate buffer (pH=7) and stirred for 48 hours at 200 rpm at 4˚C. The resulting
mixtures were centrifuged at 10,000 x g for 10 min at 20˚C. The supernatant was
collected and centrifuged again under the same conditions. The resulting supernatant,
containing the soluble protein fraction, was carefully collected and stored at 4˚C prior to
use. The soluble protein concentration was estimated following a standard protocol for
BCA Protein Assay (Thermoscientific, 2015). The day prior to emulsification, soluble
plant protein solutions were diluted with 0.01 M phosphate buffer (pH=7) to obtain 5 or 7
g/L of protein for SPI and PPI, respectively. The diluted solutions were stirred with a
magnetic stir bar overnight at room temperature at 100 rpm prior to emulsification the
next morning.
3.2.4 Preparation of lycopene-loaded emulsions
Preliminary experiments were conducted to determine the optimal quantity of
protein to use that would allow for small, physically stable droplets while limiting
(<30%) excess emulsifier in the aqueous phase by following an adapted protocol (Berton,
Genot, et al., 2011). The aqueous phase of emulsions made with varying concentrations
(5-20 g/L) of protein was collected after centrifugation at 1840 x g for 1.5 hours. The
amount of protein in the aqueous phase was then estimated as previously described
81
(Thermoscientific, 2015) at 562 nm using a DU 720 UV-Vis spectrophotometer
(Beckman Coulter, Woerden, Netherlands). Selected concentrations of proteins for
emulsions were determined to be 5 g/L for WPI, SC, and SPI and 7 g/L for PPI as these
allowed for a small droplet size (0.1-0.2 μm) while limiting the excess protein to <30% of
soluble protein (Appendix B, Figure B-1).
Aliquots of lycopene stock oil were removed from freezer storage and placed in
an ultrasonic water bath for 30 min to solubilize lycopene crystals in the oil. A coarse
emulsion was prepared by mixing the lycopene oil stock (10% wt) with aqueous protein
solution (90% wt) via an Ultra Turrax at 11,000 rpm for 30 seconds. The coarse emulsion
was then immediately passed through a high pressure M-110Y Microfluidizer
(Microfluidics, Massachusetts, USA) for five times at 800 bar. The freshly prepared
emulsions were flushed with nitrogen and stored in borosilicate screw top vials at 4˚C, in
the dark. The resulting emulsions were sampled and measured (for physical stability) and
aliquoted and stored (for chemical stability) at 0, 3, 7, and 14 days. Aliquots for lycopene
quantification were stored in glass vials, flushed with nitrogen, and stored at -20˚C until
tested.
3.2.5 Particle size measurment
Emulsion droplet size was measured using a static light scattering instrument
(Mastersizer 2000, Malvern Instruments Ltd.; Worcestershire, UK). Non-diluted
emulsion samples were directly added to an attached Hydro SM small volume sampling
unit for measurement. In order to assess if samples flocculated, 1 mL of emulsion was
added to 4 mL of 10% sodium dodecyl sulfate (SDS) solution in water, vortexed, and
then the droplet size was measured again.
All samples were measured within an obscuration range of 12-16%. Particle size
of emulsion droplets is reported as the volume weighted mean (d4,3) and represents the
average of three independent emulsion measurements, each of which were the average of
three measurements.
82
3.2.6 Zeta potential measurement
The zeta potential of emulsions was determined by measuring the electrophoretic
mobility of droplets via laser Doppler velocimetry using a Zetasizer Nano ZS (Malvern
Instruments Ltd.; Worcestershire, UK). Measurements were conducted with a backscatter
detection angle of 173˚C and calculated following the Smoluchowski model with
refractive indices of 1.330 and 1.475 for water and canola oil, respectively. Samples were
diluted with ultrapure water to 1.25% (v/v) and measured after 2 minutes of equilibration
at 25°C with 3 measurements per sample. The zeta potential values were expressed as the
average from three independent samples.
3.2.7 Lycopene extraction and quantification in emulsions
Lycopene was extracted from emulsion samples using a method previously
described (Ax, Mayer-Miebach, Link, Schuchmann, & Schubert, 2003) with
modifications. Precisely 3 mL of ethanol, 1 mL of saturated sodium chloride in water,
and 4 mL of solvent (0.1% BHT in hexane w/v) were added to 1 mL of emulsion sample.
The samples were then vortexed and flushed with nitrogen prior to sonication for 5
minutes. Following this, a Pasteur pipette was used to carefully collect the upper hexane
phase. Extraction with additional solvent was repeated until the hexane phase was
colorless (4 repetitions). Extracts were diluted with hexane to achieve absorbance values
between 0.1 – 0.8 and measured with a UV-VIS spectrophotometer at 471 nm. The total
lycopene content was calculated using a molar extinction coefficient of 1.85 x 105 M-1
cm-1, which was calculated as described previously (Britton, Liaaen-Jensen, & Pfander,
2004). The chemical stability of lycopene was expressed as the relative retention of
lycopene (Crelative) and the absolute lycopene content. The relative lycopene retention and
the encapsulation efficiency (EE) are defined as follows (Eq. 1 and 2):
Crelative (%) = (Ct / C0)* 100
(1)
EE = (C0 / Ci) * 100
(2)
Where Ct is the lycopene content (mg/100 g of emulsion) in the lycopene at time t and C0
is the lycopene present in the emulsion on day 0 of storage. Ci represents the amount of
83
lycopene initially added to 100 g of emulsion. The absolute lycopene content was
determined as the lycopene content (mg lycopene/100 g of emulsion) measured at each
time point. Lycopene stability was determined in triplicate from three independent
emulsions.
3.2.8 Adsorption kinetics of interfacial films
The interfacial tension at the interface between stripped oil and aqueous protein
solutions was measured using an automated drop tensiometer (Teclis, Longessaigne,
France). Preliminary experiments (data not shown) indicated that there was no observable
change in the adsorption kinetics of whey proteins when lycopene was present in the oil,
compared to pure stripped oil (for a lycopene-to-whey protein ratio similar to that in
emulsion systems). Therefore, stripped canola oil was used as the oil phase for this series
of experiments. It was used to fill a 0.5 mL glass syringe, connected to a 16-gauge
stainless steel needle to form a model oil droplet (surface area of 60 mm2). The
continuous phase was protein solutions (0.1 g/L) in 10 mM phosphate buffer (pH 7) in a
40 x 23.6 x 15 mm glass cuvette (Hellma Analytics, Jena, Germany). Protein adsorption
kinetics was measured during 2-hour runs to ensure equilibrium and was run in, at least,
duplicate to ensure repeatability. Interfacial tension was determined by fitting the
experimental data to the Young-Laplace equation. Following each experiment, needles
and syringes were cleaned with a 1% detergent solution (Hellmanex, Hellma Analytics,
Jena , Germany) using an ultrasonic bath. Prior to use, needles and syringes were rinsed
with ethanol and copious amounts of ultrapure water.
3.2.9 Interfacial rheology of interfacial films
Following the 2-hour equilibration period used to allow for protein adsorption at
the oil-water interface, oscillation cycles were applied to the model drop to investigate
the viscoelastic response of the protein interfacial film to dilatational deformation. The
drop was subjected to amplitude sweeps (2-35%) under a constant frequency of 0.01 Hz.
The dilatational elastic modulus (E’d) and the dilatational viscous modulus (E’’d) were
determined from the intensity and phase of the first harmonic of a Fourier transform of
the oscillating surface tension signal, and are defined as follows (Eq. 3.3, 3.4):
84
E’d = ∆γ (A0/∆A) cosδ
(3.3)
E’’d = ∆γ (A0/∆A) sinδ
(3.4)
Where ∆γ is the change in surface tension, A0 is the initial drop surface, ∆A is the change
in drop surface during the oscillations and δ is the phase shift.
The loss tangent (tan θ) was calculated by the following equation:
tan θ = E”d / E’d
(3.5)
2.2.5. Statistical analysis
All emulsions were prepared in triplicate with physical and chemical stability
measurements reported as the mean and SD of all measurements per emulsion type.
Statistical analysis was conducted with JMP version 11 (SAS Institute Inc.; Cary NC,
USA). Data were subjected to one-way analysis of variance (ANOVA) with α=0.05. The
Tukey-Kramer method was conducted post-hoc for mean comparisons (α=0.05).
3.3 Results and discussion
3.3.1 Physical stability of lycopene-loaded emulsions
All emulsions exhibited small droplet size (d4,3: 0.2 – 0.5 μm) between 0 and 7
days of storage (Figure 3-1) with span values between 2.17-3.16 (Appendix B, Figure B2). SC-, WPI-, and PPI-stabilized emulsions remained physically stable with a similar
droplet size at day 14 compared to day 0 (Figure 3-2), in contrast to SPI-stabilized
emulsions, which exhibited significantly larger d4,3 value at day 14.
85
3
WPI
SC
SPI
PPI
a
Particle Size (um)
2.5
2
1.5
1
0.5
a
a
a
0
0
3
7
Time (Days)
14
Figure 3-1 Particle size (d4,3; left y-axis) of lycopene-loaded emulsions over time.
Response values shown represent the mean + SD (n=3), with letters denoting samples
that are significantly different at a given storage time (α=0.05)
In order to understand what caused this, all emulsion samples were measured with and
without SDS to check for flocculation. The SC and PPI-stabilized emulsions exhibited
similar particle size distributions at day 0 and day 14, with and without SDS (Figure 3-2),
indicating they were not subjected to flocculation. Conversely, the SPI-stabilized
emulsion showed particle size distributions that exhibited a left-shift when diluted in SDS
solution, both at t = 0 and 14 days indicating that some flocculation occurred. Yet, after
treatment with SDS the particle size distribution of SPI-stabilized emulsions was similar
at day 0 and day 14, indicating that the emulsion was stable to coalescence. The particle
size distribution of the WPI-stabilized emulsion shifted to higher values after 14 days
compared to the initial measurement, which remained unchanged after SDS treatment,
indicating that coalescence occurred to a limited extent.
86
A
Day 0
Day 0 w/ SDS
Day 14
Day 14 w/ SDS
B
8
Volume (%)
Volume (%)
8
6
4
2
6
4
2
0
0.01
1
0
100
0.01
1
Size (µm)
100
0.01
1
Size (µm)
100
Size (µm)
D 8
6
6
Volume (%)
Volume (%)
C 8
4
4
2
2
0
0
0.01
0.1
1
10
Size (µm)
100
Figure 3-2 Comparison of particle size distribution of lycopene-loaded emulsions
stabilized with WPI (A), SC (B), SPI (C), and PPI (D) at day 0 and day 14, with and
without 1% SDS. Identical distributions with and without SDS dilution suggest that
flocculation did not occur in such samples. When Day 0 and Day 14 distributions are
identical the emulsions are stable.
All samples exhibited negative initial zeta potentials between -45 and -60 mV,
which did not change over the course of 14 days (Appendix B, Figure B.3). Large
negative zeta potential values were expected as emulsions were prepared at a pH above
the isoelectric point of all proteins tested. Although zeta potential can give an indication
of electrostatic stabilization, proteins are mostly known for the formation of thick,
viscoelastic layers at the oil-water interface that are directly linked to their efficiency at
87
preventing emulsion droplet coalescence (Dickinson, Owusu, Tan, & Williams, 1993), as
will be discussed in more detail in the interfacial rheology section.
3.3.2 Encapsulation stability of lycopene-loaded emulsions
All emulsions contained around 1.4 mg lycopene/100 g emulsion directly after
their preparation, and by the end of storage after 14 days they all had a relative lycopene
retention >65% (Figure 3-3) corresponding to >0.8 mg/100g emulsion.
WPI
Relative Lycopene Retention (%)
100
90
80
70
60
50
40
30
20
10
0
ab
a
b
SC
ab
SPI
b ab
b
ab
a
3
PPI
7
Time (days)
ab
ab
a
14
Figure 3-3 Relative retention of lycopene, as a function of time for lycopene-loaded
emulsions. Response values shown represent the mean + SD (n=3), with same letters
denoting values that are not significantly different (α=0.05).
The highest relative lycopene retention amongst emulsions was with SC at ~87%, closely
followed by PPI, with a retention of ~81%. Both values were significantly higher than
found for the WPI- and SPI-stabilized emulsions. SC has been reported to better protect
emulsions against lipid oxidation compared to WPI, and also better than SPI (Hu,
McClements, & Decker, 2003), which is in accordance with our findings. The relatively
low stability obtained with WPI compared to the work of Hu may be the result of the
difference in pH that was applied, 3.0 versus 7.0 used for this study: isoelectric points are
88
~5.1 for WPI (Alting, Hamer, de Kruif, & Visschers, 2000) and ~5.6 for SPI and PPI
(Chove et al., 2001; Liu, Elmer, Low, & Nickerson, 2010).
3.3.3 Adsorption kinetics
Interfacial tension at the oil-water interface with proteins initially dissolved in the
aqueous phase was determined and expressed as a function of time (log scale) as shown
in Figure 3-4. In the absence of protein, the stripped oil-water interface exhibited a
constant interfacial tension at ~36 mN/m (data not shown) and was in accordance with
values previously obtained in our laboratory for stripped vegetable oil, whereas a
decrease in interfacial tension over time was observed when proteins were present. SC,
SPI, and PPI led to roughly similar equilibrium interfacial tensions of approximately 15.8
mN/m, 15.6 mN/m, and 15.9 mN/m, respectively, by the end of the two hour run while
WPI led to a higher value at roughly 18.3 mN/m, indicating that it is less surface active in
comparison to the other proteins.
Figure 3-4 Adsorption kinetics of WPI (A), SC (B), SPI (C), and PPI (D) at the O/W
interface as a function of time (log scale). The slope of the line correlates with the rate of
adsorption to the interface. The dashed line represents the interfacial tension of the
stripped O/W interface in the absence of protein at ~36 mN/m.
SC appeared to have the fastest rate of adsorption, followed by the plant
proteins—PPI being faster than SPI—with WPI exhibiting the slowest rate of adsorption
89
at the oil-water interface. SC adsorbs quickly to the interface due to a relatively higher
amount of nonpolar groups compared to proteins such as WPI (Dickinson, 2011; Nakai &
Li-Chan, 1988). SC differs from WPI, SPI, and PPI in its structure; specifically β-casein
consists of flexible, random coil proteins with little secondary structure due to the number
and distribution of prolyl residues, and to a lack of covalent intramolecular bonding
(Dickinson, 2001), which makes caseins flexible, amphiphilic proteins. Conversely,
disulfide bridges and cysteine residues in β-lactoglobulin, the main component of WPI,
stabilize the protein’s globular tertiary structure (McClements, Monahan, & Kinsella,
1993), which makes the molecule considerably less flexible; this affects the structure of
the formed interfacial films, which is investigated in more detail in the next section.
3.3.4 Interfacial rheology
Coalescence can happen if a hole is created in the interfacial film that separates
two colliding droplets. Such a rupture can be seen as a dilatational deformation, thus we
tested the dilatational properties of protein-stabilized interfaces (Bos & van Vliet, 2001;
Murray, 2011). With the exception of WPI, the elastic and viscous moduli of the protein
layers did not have a large dependence on the applied deformation (Figure 3-5), implying
that the measurements were conducted within the linear viscoelastic regime. Compared to
all other samples, the SC layer exhibited substantially lower elastic moduli (Figure 3-5A),
and thus higher loss tangent (Figure 3-5B), while the elastic moduli for WPI, SPI, and
PPI all appear to be substantially higher (>15 mN/m) than their corresponding loss
moduli. This indicates that the SC layer exhibited more viscous behaviour compared to
the other protein layers, which is likely due to the random coil and lack of secondary
structure characteristic of SC (Dickinson, 1992). Our findings are in agreement with other
studies in which SC was also reported to form viscous layers at the oil-water interface
(Erni, Windhab, & Fischer, 2011) due to loose packing and weak interactions between
interfacial casein proteins (Dickinson, 2001). A viscous interface, which is
characteristically less dense and compact compared to an elastic one, is formed with SC
primarily due to its flexibility as a protein, but also due to its hydrophobicity as SC
90
preferentially orients along the oil phase as opposed to building adsorbed layers at the oilwater interface (Maldonado-Valderrama et al., 2005).
Compared to SC-based interfaces, WPI-based ones exhibited a more elastic
behaviour, which can be attributed to strong intermolecular interactions and a high twodimensional packing efficiency at the interface (Dickinson, 2001). SPI- and PPI-based
layers exhibited loss tangents more similar to that of the WPI-based layer, which was
expected since plant proteins are globular (Boye et al., 2010) and known to produce an
interconnected, viscoelastic monolayer at the oil-water interface (Chang et al., 2015).
Figure 3-5 Elastic (filled shapes) and loss (open shapes) moduli (A) and loss tangent (B)
of proteins at deformations between 0.03-0.35. Higher loss tangent values indicate a more
viscous response, while lower values indicate a more elastic behavior. Response values
shown represent the mean + SD (n=3). Statistical differences amongst protein films are
shown (B) with same letters denoting values that are not significantly different (α=0.05).
3.3.5 Comparison and design considerations for protein-stabilized emulsions
All our emulsions had similar and small droplet size, therefore, effects of
interfacial area, that are reported to potentially influence chemical stability (Lethuaut,
Métro, & Genot, 2002) or not (Berton‐ Carabin, Ropers, & Genot, 2014; Hu,
McClements, & Decker, 2003; Osborn & Akoh, 2004) can rather safely be disregarded in
the interpretation of the results. Besides, we designed our emulsions in such a way that
the fraction and concentration of non-adsorbed proteins was low, so that the contribution
91
of this non-adsorbed fraction to their physicochemical stability was presumably limited
(Berton et al., 2011; Faraji et al., 2004).
Most probably, the protein properties and the resulting interfacial layers affect
lycopene stability. Steric forces influence emulsion physical stability, particularly for SCstabilized emulsions, as electrostatic forces are expected to play a lesser role in
stabilization for flexible proteins (Dickinson, 2010), while for the other less flexible
proteins, thicker layers are expected to stabilize the interfaces. Hu et al. (2003) discussed
the amino acid composition of SC, which contains relatively high amounts of
antioxidative tyrosine, proline, and methionine, as a potential explanation for improved
oxidative stability of emulsions stabilized with SC compared to SPI and WPI, although
they express that this relationship is not clear. In another study, high-pressure processing
at 1379 bar vs. 345 bar was reported to induce a tighter packing in the cross-linked
interfacial layer of SC-stabilized emulsions, which was related to a higher oxidative
stability (Phoon, Paul, Burgner, Fernanda San Martin-Gonzalez, & Narsimhan, 2014).
Other studies have reported that increasing processing temperature of protein-stabilized
emulsions results in further unfolding of proteins and potential alteration of conformation
(Let, Jacobsen, Sørensen, & Meyer, 2007). In particular, whey proteins have been
reported to exhibit antioxidant properties post-homogenization due to the unfolding and
exposure of sulfhydryl groups, which can either repel (Min Hu, D. Julian McClements, &
Decker, 2003) or scavenge free radicals (Let et al., 2007; Tong, Sasaki, Mcclements, &
Decker, 2000).
From the above it is clear that interfacial properties are related to the
physicochemical stability of an emulsion, which is mostly linked to providing a denser
barrier against oxidizing agents and coalescence (Georgieva, Schmitt, Leal-Calderon, &
Langevin, 2009), however it is difficult to find clear experimental evidence for this. As
discussed previously, elastic interfaces are the result of an interconnected protein
network. The gel-like viscoelastic interface observed in this study amongst WPI, SPI, and
PPI-stabilized emulsions would be expected to form a rigid layer, which in theory could
better physically stabilize the system and limit contact between the lipid phase and
oxidizing agents. However, globular proteins may exhibit localized empty patches due to
92
depletion (Bos & van Vliet, 2001), which potentially has detrimental consequences for
lycopene stability.
Despite the mechanical and structural properties of the interface, chemical
properties, such as oxygen permeability through a protein layer, should also be taken into
consideration. β-casein films at the air-water interface were found to have a higher
oxygen permeability compared to that of β-lactoglobulin (Toikkanen et al., 2014), while
β-casein-stabilized emulsions have been found to exhibit better oxidative stability (based
on oxygen uptake and formation of conjugated dienes, hexane, and propanal) in various
conditions compared to β-lactoglobulin-stabilized emulsions (Berton, Ropers, Bertrand,
Viau, & Genot, 2012; Berton, Ropers, Viau, & Genot, 2011), and this is most probably
caused by the fact that caseins are better at scavenging free radicals (Clausen, Skibsted, &
Stagsted, 2009) and binding iron (Faraji, Mcclements, & Decker, 2004; Sugiarto, Ye,
Taylor, Singh, & Singh, 2010) compared to whey proteins.
Yet, protein flexibility and interfacial elasticity alone cannot be used to simply explain
the stability of lycopene-loaded emulsions. It is likely that chemical properties of the
proteins aided in lycopene stability, although future work could be done to directly assess
this. Especially pea protein is of great interest; given its relatively high stability and
encapsulation capacity, it is expected to serve as a genuine alternative for animal-based
proteins in emulsion formulations.
3.4 Conclusions
This work systematically investigated the physical and chemical stability of
lycopene-loaded emulsions prepared using various proteins as emulsifiers. Especially
emulsions stabilized with casein and pea protein exhibited both high chemical
(encapsulation % > 80%) and physical stability (no change in particle size) after 14 days.
Interestingly, no correlation could be found between the elasticity of the protein layers at
model oil-water interfaces, and the physicochemical stability of the corresponding
emulsions. This is most probably due to the fact that adsorbed casein molecules induced
strong steric repulsion, resulting in an additional emulsion stabilization effect, and
lycopene protection effects due to the protein ability to chelate metals ions and scavenge
free radicals.
93
Performance of each protein could be ranked for each property measured,
however, it is perhaps more valuable to consider the collective characteristics for each of
the protein-stabilized emulsions. Although SC appeared to perform optimally, PPI was a
strong plant contender and demonstrated comparably good properties as it stabilized
emulsions against flocculation and coalescence, exhibited relatively rapid protein
adsorption, and stabilized lycopene to a similar extent as SC. Overall, SC and PPI both
exhibited relatively good physical and chemical stabilization for lycopene-loaded
emulsions, while SPI and WPI exhibited better stabilization for either physical or
chemical stabilization, rather than both (Table 3.1).
This research demonstrates that selected plant proteins can perform well compared to
dairy proteins for lycopene encapsulation and have potential as dairy alternatives for
chemical protection against oxidation in colloidal systems.
Table 3-1 Summary comparison of physical and chemical properties lycopene-loaded
emulsions stabilized with WPI, SC, SPI, or PPI. Proteins that strongly demonstrated
relatively high (++++) values for a given characteristic are compared against those with
intermediate (+++ or ++) and lower (+) values.
Small
Physical
Fast
Highly Elastic
Lycopene
Droplet Size
Stability
Adsorption
Interface
Retention
WPI
++
++
+
++++
+
SC
+++
++++
++++
+
++++
SPI
+
+
++
++++
++
PPI
++
+++
+++
++
+++
3.5 Acknowledgements
This research was based upon work supported by the National Science
Foundation Graduate Research Fellowship (Grant No. DGE-1333468) through the
Graduate Research Opportunities World Wide (GROW) program in support with the
Netherlands Organisation for Scientific Research (NWO), and the Wageningen
94
University Graduate School (VLAG). Laboratory work was conducted at Wageningen
University in the Food Process Engineering Group. The authors would like to thank the
laboratory technicians of the Food Process Engineering Group at Wageningen University,
including Jos Sewalt, Martin de Wit, and Jarno Gieteling for assistance with experimental
set up, and Maurice Strubel for his HPLC expertise.
95
3.6 References
Alting, A. C., Hamer, R. J., de Kruif, C. G., & Visschers, R. W. (2000). Formation of
Disulfide Bonds in Acid-Induced Gels of Preheated Whey Protein Isolate. Journal of
Agricultural and Food Chemistry, 48(10), 5001–5007.
http://doi.org/10.1021/jf000474h
Aoki, H., Taneyama, O., & Inami, M. (1980). Emulsifyng properties of soy protein:
Characteristics of 7S and IIS proteins. Journal of Food Science, 45(3), 534–538.
http://doi.org/10.1111/j.1365-2621.1980.tb04095.x
Ax, K., Mayer-Miebach, E., Link, B., Schuchmann, H., & Schubert, H. (2003). Stability
of lycopene in oil-in-water emulsions. Engineering in Life Sciences, 3(4), 199–201.
http://doi.org/10.1002/elsc.200390028
Bengoechea, C., Cordobés, F., & Guerrero, A. (2006). Rheology and microstructure of
gluten and soya-based o/w emulsions. Rheologica Acta, 46(1), 13–21.
http://doi.org/10.1007/s00397-006-0102-6
Benjamin, O., Silcock, P., Beauchamp, J., Buettner, A., & Everett, D. W. (2014).
Emulsifying Properties of Legume Proteins Compared to β-Lactoglobulin and
Tween 20 and the Volatile Release from Oil-in-Water Emulsions. Journal of Food
Science, 79(10), E2014–E2022. http://doi.org/10.1111/1750-3841.12593
Berton‐ Carabin, C. C., Ropers, M., & Genot, C. (2014). Lipid Oxidation in Oil‐ in‐
Water Emulsions: Involvement of the Interfacial Layer. Comprehensive Reviews in
Food Science and Food Safety, 13(5), 945–977.
Berton, C., Genot, C., & Ropers, M.-H. (2011). Quantification of unadsorbed protein and
surfactant emulsifiers in oil-in-water emulsions. Journal of Colloid and Interface
Science, 354(2), 739–748. http://doi.org/http://dx.doi.org/10.1016/j.jcis.2010.11.055
Berton, C., Ropers, M.-H., Bertrand, D., Viau, M., & Genot, C. (2012). Oxidative
stability of oil-in-water emulsions stabilised with protein or surfactant emulsifiers in
various oxidation conditions. Food Chemistry, 131(4), 1360–1369.
http://doi.org/10.1016/j.foodchem.2011.09.137
96
Berton, C., Ropers, M.-H., Viau, M., & Genot, C. (2011). Contribution of the Interfacial
Layer to the Protection of Emulsified Lipids against Oxidation. Journal of
Agricultural and Food Chemistry, 59(9), 5052–5061.
http://doi.org/10.1021/jf200086n
Bos, M. A., & van Vliet, T. (2001). Interfacial rheological properties of adsorbed protein
layers and surfactants: a review. Advances in Colloid and Interface Science, 91(3),
437–471. http://doi.org/10.1016/S0001-8686(00)00077-4
Boye, J. I., Aksay, S., Roufik, S., Ribéreau, S., Mondor, M., Farnworth, E., &
Rajamohamed, S. H. (2010). Comparison of the functional properties of pea,
chickpea and lentil protein concentrates processed using ultrafiltration and
isoelectric precipitation techniques. Food Research International, 43(2), 537–546.
http://doi.org/10.1016/j.foodres.2009.07.021
Britton, G., Liaaen-Jensen, S., & Pfander, H. (2004). Carotenoid Handbook Info.pdf.
Basel, Switzerland: Birkhauser Verlag.
Chang, C., Tu, S., Ghosh, S., & Nickerson, M. T. (2015). Effect of pH on the interrelationships between the physicochemical, interfacial and emulsifying properties
for pea, soy, lentil and canola protein isolates. Food Research International, 77(3),
360–367. http://doi.org/10.1016/j.foodres.2015.08.012
Chihi, M.-L., Mession, J., Sok, N., & Saurel, R. (2016). Heat-Induced Soluble Protein
Aggregates from Mixed Pea Globulins and β-Lactoglobulin. Journal of Agricultural
and Food Chemistry, 64(13), 2780–2791. http://doi.org/10.1021/acs.jafc.6b00087
Chove, B. E., Grandison, A. S., & Lewis, M. J. (2001). Emulsifying properties of soy
protein isolate fractions obtained by isoelectric precipitation. Journal of the Science
of Food and Agriculture, 81(8), 759–763. http://doi.org/10.1002/jsfa.877
Chronakis, I. S. (1996). Network formation and viscoelastic properties of commercial soy
protein dispersions: effect of heat treatment, pH and calcium ions. Food Research
International, 29(2), 123–134. http://doi.org/10.1016/0963-9969(96)00018-X
Clausen, M. R., Skibsted, L. H., & Stagsted, J. (2009). Characterization of Major Radical
Scavenger Species in Bovine Milk through Size Exclusion Chromatography and
Functional Assays. Journal of Agricultural and Food Chemistry, 57(7), 2912–2919.
http://doi.org/10.1021/jf803449t
97
Cornacchia, L., & Roos, Y. H. (2011). Stability of β-Carotene in Protein-Stabilized Oilin-Water Delivery Systems. Journal of Agricultural and Food Chemistry, 59(13),
7013–7020. http://doi.org/10.1021/jf200841k
de Boer, J., Helms, M., & Aiking, H. (2006). Protein consumption and sustainability:
Diet diversity in EU-15. Ecological Economics, 59(3), 267–274.
http://doi.org/10.1016/j.ecolecon.2005.10.011
Di Mascio, P., Kaiser, S., & Sies, H. (1989). Lycopene as the most efficient biological
carotenoid singlet oxygen quencher. Archives of Biochemistry and Biophysics,
274(2), 532–538. http://doi.org/http://dx.doi.org/10.1016/0003-9861(89)90467-0
Dickinson, E. (1992). Faraday research article. Structure and composition of adsorbed
protein layers and the relationship to emulsion stability. Journal of the Chemical
Society, Faraday Transactions, 88(20), 2973. http://doi.org/10.1039/ft9928802973
Dickinson, E. (2001). Milk protein interfacial layers and the relationship to emulsion
stability and rheology. Colloids and Surfaces B: Biointerfaces, 20(3), 197–210.
http://doi.org/10.1016/S0927-7765(00)00204-6
Dickinson, E. (2010). Flocculation of protein-stabilized oil-in-water emulsions. Colloids
and Surfaces. B, Biointerfaces, 81(1), 130–40.
http://doi.org/10.1016/j.colsurfb.2010.06.033
Dickinson, E. (2011). Mixed biopolymers at interfaces: Competitive adsorption and
multilayer structures. Food Hydrocolloids, 25(8), 1966–1983.
http://doi.org/10.1016/j.foodhyd.2010.12.001
Dickinson, E. (2013). Stabilising emulsion-based colloidal structures with mixed food
ingredients. Journal of the Science of Food and Agriculture, 93(4), 710–21.
http://doi.org/10.1002/jsfa.6013
Dickinson, E., Owusu, R. K., Tan, S., & Williams, A. (1993). Oil-soluble Surfactants
Have Little Effect on Competitive Adsorption of β-Lactalbumin and βLactoglobulin in Emulsions. Journal of Food Science, 58(2), 295–298.
http://doi.org/10.1111/j.1365-2621.1993.tb04259.x
98
Erb K.H., Lauk C., Kastner T., Mayer A., Theurl M.C., Haberl H. (2016). Exploring the
biophysical option space for feeding the world without deforestation. Nature
Communications, In press.
Erni, P., Windhab, E. J., & Fischer, P. (2011). Emulsion drops with complex interfaces:
Globular versus flexible proteins. Macromolecular Materials and Engineering,
296(3-4), 249–262. http://doi.org/10.1002/mame.201000290
Faraji, H., Mcclements, D. J., & Decker, E. A. (2004). Role of Continuous Phase Protein
on the Oxidative Stability of Fish Oil-in-Water Emulsions. Journal of Agricultural
and Food Chemistry, 52(14), 4558–4564. http://doi.org/10.1021/jf035346i
Fernandez-Avila, C., Arranz, E., Guri, A., Trujillo, A. J., & Corredig, M. (2016).
Vegetable protein isolate-stabilized emulsions for enhanced delivery of conjugated
linoleic acid in Caco-2 cells. Food Hydrocolloids, 55, 144–154.
http://doi.org/10.1016/j.foodhyd.2015.10.015
Georgieva, D., Schmitt, V. E., Leal-Calderon, F., & Langevin, D. (2009). On the Possible
Role of Surface Elasticity in Emulsion Stability. Langmuir, 25(10), 5565–5573.
http://doi.org/10.1021/la804240e
Ho, K. K. H. Y., Ferruzzi, M. G., Liceaga, A. M., & San Martín-González, M. F. (2015).
Microwave-assisted extraction of lycopene in tomato peels: Effect of extraction
conditions on all-trans and cis-isomer yields. LWT - Food Science and Technology,
62(1), 160–168. http://doi.org/10.1016/j.lwt.2014.12.061
Hu, M., McClements, D. J., & Decker, E. A. (2003). Lipid oxidation in corn oil-in-water
emulsions stabilized by casein, whey protein isolate, and soy protein isolate. Journal
of Agricultural and Food Chemistry, 51(6), 1696–1700.
http://doi.org/10.1021/jf020952j
Kean, E. G., Hamaker, B. R., & Ferruzzi, M. G. (2008). Carotenoid Bioaccessibility from
Whole Grain and Degermed Maize Meal Products. Journal of Agricultural and
Food Chemistry, 56(21), 9918–9926. http://doi.org/10.1021/jf8018613
Lam, R. S. H., & Nickerson, M. T. (2013). Food proteins: a review on their emulsifying
properties using a structure-function approach. Food Chemistry, 141(2), 975–84.
http://doi.org/10.1016/j.foodchem.2013.04.038
99
Let, M. B., Jacobsen, C., Sørensen, A.D. M., & Meyer, A. S. (2007). Homogenization
Conditions Affect the Oxidative Stability of Fish Oil Enriched Milk Emulsions:
Lipid Oxidation. Journal of Agricultural and Food Chemistry, 55(5), 1773–1780.
http://doi.org/10.1021/jf062391s
Lethuaut, L., Métro, F., & Genot, C. (2002). Effect of droplet size on lipid oxidation rates
of oil-in-water emulsions stabilized by protein. Journal of the American Oil
Chemists’ Society, 79(5), 425–430. http://doi.org/10.1007/s11746-002-0500-z
Liu, S., Elmer, C., Low, N. H., & Nickerson, M. T. (2010). Effect of pH on the functional
behaviour of pea protein isolate–gum Arabic complexes. Food Research
International, 43(2), 489–495. http://doi.org/10.1016/j.foodres.2009.07.022
Maldonado-Valderrama, J., Fainerman, V. B., Gálvez-Ruiz, M. J., Martín-Rodriguez, A.,
Cabrerizo-Vílchez, M. A., & Miller, R. (2005). Dilatational Rheology of β-Casein
Adsorbed Layers at Liquid-Fluid Interfaces. Journal of Physical Chemistry, 109(37),
17608–17616. http://doi.org/10.1021/jp050927r
Mao, Li., Xu, D., Yang, J., Yuan, F., Gao, Y., & Zhao, J. (2009). Effects of Small and
Large Molecule Emulsifiers on the Characteristics of β-Carotene Nanoemulsions
Prepared by High Pressure Homogenization. Food Technology Biotechnology,
47(3), 336–342.
Mao, Li., Yang, J., Yuan, F., & Gao, Y. (2010). Effects of homogenization models and
emulsifiers on the physicochemical properties of β-carotene nanoemulsions. Journal
of Dispersion Science and Technology, 31(7).
http://doi.org/10.1080/01932690903224482
McClements, D. J., Monahan, F. J., & Kinsella, J. E. (1993). Disulfide Bond Formation
Affects Stability of Whey Protein Isolate Emulsions. Journal of Food Science,
58(5), 1036–1039. http://doi.org/10.1111/j.1365-2621.1993.tb06106.x
Min Hu, D. Julian McClements, A., & Decker, E. A. (2003). Impact of Whey Protein
Emulsifiers on the Oxidative Stability of Salmon Oil-in-Water Emulsions. Journal
of Agricultural and Food Chemistry, 51(5), 1235–1439.
Murray, B. S. (2011). Rheological properties of protein films. Current Opinion in Colloid
& Interface Science, 16(1), 27–35. http://doi.org/10.1016/j.cocis.2010.06.005
100
Nakai, S., & Li-Chan, E. (1988). Hydrophobic interactions in food systems. CRC Press.
O’Kane, F. E., Vereijken, J. M., Happe, R. P., Gruppen, H., & J S Van Boekel, M. A.
(2004). Heat-Induced Gelation of Pea Legumin: Comparison with Soybean
Glycinin. Journal of Agricultural and Food Chemistry, 52(16), 5071–5078.
http://doi.org/10.1021/jf035215h
Osborn, H. T., & Akoh, C. C. (2004). Effect of emulsifier type, droplet size, and oil
concentration on lipid oxidation in structured lipid-based oil-in-water emulsions.
Food Chemistry, 84(3), 451–456. http://doi.org/10.1016/S0308-8146(03)00270-X
Pelgrom, P. J. M., Berghout, J. A. M., Van Der Goot, J., Boom, R. M., & Schutyser, M.
A. I. (2014). Preparation of functional lupine protein fractions by dry separation.
LWT - Food Science and Technology, 59, 680–688.
http://doi.org/10.1016/j.lwt.2014.06.007
Phoon, P. Y., Paul, L. N., Burgner, J. W., Fernanda San Martin-Gonzalez, M., &
Narsimhan, G. (2014). Effect of Cross-Linking of Interfacial Sodium Caseinate by
Natural Processing on the Oxidative Stability of Oil-in-Water (O/W) Emulsions.
Journal of Agricultural and Food Chemistry, 62(13), 2822–2829.
http://doi.org/10.1021/jf403285z
Phoon, P. Y., San Martin-Gonzalez, M. F., & Narsimhan, G. (2014). Effect of hydrolysis
of soy β-conglycinin on the oxidative stability of O/W emulsions. Food
Hydrocolloids, 35, 429–443. http://doi.org/10.1016/j.foodhyd.2013.06.024
Rao, A. V, Waseem, Z., & Agarwal, S. (1998). Lycopene content of tomatoes and tomato
products and their contribution to dietary lycopene. Food Research International,
31(10), 737–741. http://doi.org/http://dx.doi.org/10.1016/S0963-9969(99)00053-8
Ribeiro, H. S., Guerrero, J. M. M., Briviba, K., Rechkemmer, G., Schuchmann, H. P., &
Schubert, H. (2006). Cellular Uptake of Carotenoid-Loaded Oil-in-Water Emulsions
in Colon Carcinoma Cells in Vitro. Journal of Agricultural and Food Chemistry,
54(25), 9366–9369. http://doi.org/10.1021/jf062409z
Shi, W., & Dumont, M.J. (2014). Review: bio-based films from zein, keratin, pea, and
rapeseed protein feedstocks. Journal of Material Science, 49(5), 1915–1930.
http://doi.org/10.1007/s10853-013-7933-1
101
Song, F., Tang, D.L., Wang, X.L., & Wang, Y.-Z. (2011). Biodegradable Soy Protein
Isolate-Based Materials: A Review. Biomacromolecules, 12(10), 3369–3380.
http://doi.org/10.1021/bm200904x
Sugiarto, M., Ye, A., Taylor, M. W., Singh, H., & Singh, H. (2010). Milk protein-iron
complexes: Inhibition of lipid oxidation in an emulsion. Dairy Science &
Technology, 90(1), 87–98. http://doi.org/10.1051/dst/2009053
Tapal, A., & Tiku, P. K. (2012). Complexation of curcumin with soy protein isolate and
its implications on solubility and stability of curcumin. Food Chemistry, 130, 960–
965. http://doi.org/10.1016/j.foodchem.2011.08.025
Thermoscientific. (2015). Pierce BCA Protein Assay Kit. Retrieved June 14, 2016, from
https://tools.thermofisher.com/content/sfs/manuals/MAN0011430_Pierce_BCA_Pro
tein_Asy_UG.pdf
Toikkanen, O., Lä Hteenmä Ki, M., Moisio, T., Forssell, P., Partanen, R., & Murtomä, L.
(2014). Study of Oxygen Transfer across Milk Proteins at an Air−Water Interface
with Scanning Electrochemical Microscopy. Journal of Agricultural and Food
Chemistry, 62(10), 2284–2288. http://doi.org/10.1021/jf5008715
Tong, L. M., Sasaki, S., Mcclements, D. J., & Decker, E. A. (2000). Mechanisms of the
Antioxidant Activity of a High Molecular Weight Fraction of Whey. Journal of
Agricultural and Food Chemistry, 48(5), 1473–1478.
http://doi.org/10.1021/jf991342v
VandeHaar M.J., St-Pierre N. (2006). Major advances in nutrition: relevance to the
sustainability of the dairy industry. Journal of Dairy Science, 89(4),1280-1291.
Wilde, P.J. (2000). Interfaces: their role in foam and emulsion behaviour. Current
Opinion in Colloid & Interface Science, 5(3), 176–181.
http://doi.org/10.1016/S1359-0294(00)00056-X
102
CHAPTER 4 SYNERGISTIC AND ANTAGONISTIC EFFECTS OF
PLANT AND DAIRY PROTEIN BLENDS ON THE
PHYSICOCHEMICAL STABILITY OF LYCOPENE-LOADED
EMULSIONS
Kacie K.H.Y. Hoab, Karin Schroëna , M. Fernanda Martín-Gonzálezb, Claire C. BertonCarabina
a
Food Process Engineering Group, Department of Agrotechnology and Food Sciences,
Wageningen University, Bornse Weilanden 9, 6708 WG Wageningen, The Netherlands
b
Department of Food Science, College of Agriculture, Purdue University, 745
Agriculture Mall Drive, West Lafayette, IN 47907, USA
In preparation for review with Food Hydrocolloids
4.1 Introduction
Oil-in-water (O/W) emulsions have been largely studied as delivery systems for
lipophilic bioactive compounds, such as carotenoids (Cornacchia & Roos, 2011; Qian,
Decker, Xiao, & McClements, 2013; Salvia-Trujillo, Qian, Martín-Belloso, &
McClements, 2013). Lycopene, an unsaturated pigment from tomatoes, is the most
hydrophobic and most potent singlet oxygen quencher of all carotenoids (Di Mascio,
Kaiser, & Sies, 1989; Rao, Waseem, & Agarwal, 1998). This makes it a particularly
interesting, yet challenging compound to disperse, stabilize, and ultimately deliver using
aqueous-based systems.
Dairy proteins, such as sodium caseinate and β-lactoglobulin, are commonly used
to stabilize O/W emulsions (Dalgleish, 1997; Dickinson, 1994, 2001; Hailing & Walstra,
1981; McClements, 2004). Recently, interest has been rising in using non-animal derived
proteins, such as plant proteins, which may constitute a more sustainable alternative.
103
From a sustainability perspective, plant proteins (e.g., soy and pea proteins) require less
resources for production than dairy proteins and are claimed to exhibit fair emulsion
stabilization properties (Benjamin, Silcock, Beauchamp, Buettner, & Everett, 2014; Liu
& Tang, 2014). We previously reported on the physicochemical stability of lycopeneloaded emulsions stabilized by dairy or plant protein (Ho, Schroën, San Martín-González,
& Berton-Carabin, 2016). We observed that sodium caseinate (SC) and, to a lesser extent,
pea protein isolate (PPI) provided high physical and chemical stability to lycopeneloaded emulsions while whey protein isolate (WPI) and soy protein isolate (SPI)
performed less well, for at least one of both aspects.
The question we then posed ourselves is if blending these proteins would lead to
better results. Substantial work has already been done on the behavior of different dairy
protein blends (e.g. WPI-SC) at the oil-water interface (Britten & Giroux, 1991;
Dalgleish, Goff, & Luan, 2002; Seta, Baldino, Gabriele, Lupi, & de Cindio, 2012; Ye,
2008). Britten & Giroux (1991) observed preferential interfacial adsorption of SC from
WPI-SC blends in O/W emulsions with 30% (w/w) oil phase. This preferential adsorption
is due to the ability of SC to rapidly cover the interface (Seta, Baldino, Gabriele, Lupi, &
Cindio, 2014). Dickinson (2011) has reviewed competitive adsorption in mixed
biopolymer systems, which has been a topic of interest for many years for dairy protein
blends and protein-polysaccharide blends. In mixed systems, SC is known to dominate
the interface and dictate the interfacial behavior compared to less hydrophobic and nonflexible proteins, such as WPI.
Only a few studies have investigated the effect of blends containing dairy and plant
proteins. Aoki, Shirase, Kato, & Watanabe (1984) found that SC rapidly and
preferentially adsorbs to the interface in emulsions when used together with soy protein
isolate. Other studies have indicated that blends containing SC and pea proteins can
enhance emulsion physical stability compared to SPI or PPI alone due to a synergistic
interaction between the flexible SC molecules and plant proteins (Ji et al., 2015) or their
aggregates (Yerramilli, Longmore, & Ghosh, 2017). In particular, globular proteins, e.g.,
β-lactoglobulin, have relatively high packing densities and the ability to form strong
protein-protein interactions (Dickinson, 2001), which could be beneficial. Considering
this, we selected four proteins to blend: PPI, SPI, WPI and SC; the last one being an
104
industrially proven standard. Our study aimed at assessing the effect of different
interfacial dairy-plant protein blends on: 1) the physical stability (droplet size and surface
charge) and 2) chemical stability (lycopene retention) of emulsions, and 3) the interfacial
properties (adsorption kinetics and dilatational rheology). We ultimately attempted using
these findings to propose hypotheses regarding protein blend behavior at the interface.
4.2 Materials and methods
4.2.1 Materials
All-trans-lycopene standard, all analytical grade solvents and reagents were
purchased from Sigma Aldrich (Zwijndrecht, the Netherlands). Water was purified with
an EMD Millipore Milli-Q water purification system and had a resistivity of 18.2
MΩ⋅cm. Proteins were generously donated as follows: WPI, 97.5% purity (BiPRO,
Davisco, Switzerland), SC, 80% purity (Sodium caseinate S, DMV International,
Amersfoort, the Netherlands), SPI, 90% purity (SUPRO EX 37, Solae Europe SA,
Switzerland), and PPI, 80-90% purity (NUTRALYS F85, Roquette, France). More
information on the solubility and properties in aqueous solution of the different proteins
can be found in our previous work (Ho et al., 2016). Tomato paste, for lycopene
extraction, and canola oil were purchased from local retailers. Canola oil was stripped of
tocopherols and surface-active impurities via mixing with MP Alumina N-Super (MP
Biomedicals, France) as described previously (Berton, Genot, & Ropers, 2011).
4.2.2 Lycopene extraction and oil phase preparation
Lycopene was extracted as previously described (Ho et al., 2016). Briefly, ~250 g
of tomato paste, ~10 g of celite, ~10 g of sodium bicarbonate, and 500 mL of a 1:1
hexane (0.1% w/v butylated hydroxytoluene)-ethyl acetate mixture were stirred on ice
under a stream of nitrogen, at 250 rpm with an overhead IKA mixer for 1.5 hours.
Following vacuum filtration, the liquid portion was washed with saturated sodium
chloride solution before draining the bottom, aqueous layer from a separatory funnel. The
organic portion of the liquid extract was collected, flushed with nitrogen and evaporated
using a rotative agitator. Approximately 80 g of stripped canola oil were added to the dry
105
lycopene extract, transferred to a borosilicate bottle, and held under a stream of nitrogen
to ensure evaporation of residual solvent. This process was repeated several times and all
resulting lycopene-oil batches were pooled prior to storage at -20 °C. Lycopene content
was determined by measuring the absorbance of the extract at 471 nm using a UV-vis
spectrophotometer [molar extinction coefficient of 1.85 x 105 M-1cm-1 (Britton, LiaaenJensen, & Pfander, 2004)]. Lycopene concentration in the stock oil was 0.24 mg/g oil.
Cis- and trans- lycopene isomers were separated with high-performance liquid
chromatography using a YMC Carotenoid S-3 C30 column (YMC America, Inc.,
Allentown, PA, USA) as previously described by Kean, Hamaker, & Ferruzzi (2008).
Isomers were compared against an all-trans-lycopene standard (Ho, Ferruzzi, Liceaga, &
San Martín-González, 2015) to determine the stock as ~90% all-trans-lycopene.
4.2.3 Aqueous phase preparation
]Dairy proteins (WPI, SC) were solubilized in 0.01 M phosphate buffer, pH 7.0,
overnight at room temperature using a magnetic stirrer (100 rpm). Plant proteins were
only partially soluble (SPI~30%, PPI~25%) in buffer, thus buffer-protein isolate mixtures
were subjected to centrifugation (10,000 x g ,10 minutes, 4 °C) after overnight stirring to
remove the insoluble fraction, which negatively affected emulsion stability (Ho et al.,
2016). The protein content of the plant protein solutions was estimated using a standard
BCA protocol (Smith et al., 1985) and then diluted with buffer to 5 and 7 g/L for SPI and
PPI, respectively. Protein concentrations were selected such that at similar emulsion
droplet sizes the amount of non-adsorbed protein in the aqueous phase was minimized
following previous findings (Ho et al., 2016), for which used an adapted protocol from
Berton et al. (2011) was used. When needed, protein solutions were combined as 1:1 or
3:1 w/w ratios and gently stirred (100 rpm) for 1 hour prior to emulsification.
4.2.4 Emulsion preparation
Aliquots of frozen lycopene stock oil were thawed in an ultrasonic bath for 30
minutes, which allowed for lycopene crystal solubilisation in the oil. The aqueous phase
and the oil phase (10% w/w) were combined and mixed with a rotor-stator homogenizer
(UltraTurrax, IKA-Werke GmbH & Co., Staufen, Germany) at 11,000 rpm for 30 s to
106
form a coarse emulsion, which was immediately passed through a high pressure M-110Y
Microfluidizer (Microfluidics, Massachusetts, USA) at 800 bar, five times. The resulting
fine emulsion was flushed with nitrogen and stored in borosilicate vials in the dark at 4
°C. Aliquots were taken from the emulsions at 0-14 days for physical characterization
and lycopene stability. Those used for lycopene quantification were stored in the dark in
glass vials, flushed with nitrogen, and held at -20 °C until the samples were extracted and
analyzed.
4.2.5 Droplet size measurement
Emulsion droplet size was determined using a static light scattering instrument
(Mastersizer 2000, Malvern Instruments Ltd., Worcestershire, UK). Droplet size was
reported as the volume-weighted mean (d3,2) and was measured within an obscuration
range of 12-16%. The refractive indices applied for the continuous and dispersed phases
were 1.330 and 1.475, respectively. For emulsions that were suspect of flocculation,
aliquots were diluted (1:4 w/w) in a 10% sodium dodecylsulfate (SDS) solution prior to
measurement to determine the size of the individual droplets.
4.2.6 Zeta potential measurement
Emulsions were diluted with ultrapure water to a dispersed phase fraction of
0.122% (v/v) and analyzed at 25 °C following a 2 min equilibration period. The
electrophoretic mobility of droplets was measured via laser Doppler velocimetry using a
Zetasizer Nano ZS (Malvern Instruments Ltd, Worcestershire, UK) with a backscatter
angle of 173°. Zeta potential was calculated by the Malvern software using the
Smoluchowski model (refractive indices of the continuous and dispersed phases 1.330
and 1.475, respectively). Zeta potential values were expressed as the mean of three
independent samples, of which each was measured in triplicate.
4.2.7 Determination of lycopene retention in emulsions
Lycopene was extracted from emulsions using a method developed by Ax,
Mayer-Miebach, Link, Schuchmann, & Schubert (2003) with modifications (Ho et al.,
2016). Emulsion aliquots (1 mL) were combined with 4 mL of hexane containing 0.1%
BHT (w/v), 3 mL of ethanol, and 1 mL of saturated sodium chloride in water. The
107
mixture was vortexed, held under a stream of nitrogen, and then placed into an ultrasonic
bath for 5 minutes. The organic solvent layer was carefully removed with a Pasteur
pipette while the remaining liquid was re-extracted with fresh solvent 3 additional times.
The extracts from each sample were pooled and then diluted appropriately to achieve an
absorbance value between 0.1-0.8. Lycopene was quantified as described in Chapter 3
and was expressed as a relative retention value (%), which is defined as follows (Eq.
(4.1)):
Lycopene retention (%) = (Ct/C0) x 100
(4.1)
Where C0 is the initial lycopene content (mg/100 g of emulsion) and Ct is the lycopene
content at time t. Encapsulation efficiency was defined as C0, compared against Ci, the
amount of lycopene added (Eq. (2)):
Encapsulation efficiency (%) = (C0/Ci) x 100
(4.2)
To prevent carotenoid degradation during analysis, samples were held on ice, extracted in
the dark, and flushed with nitrogen to limit exposure to heat, light, and oxygen.
4.2.8 Adsorption kinetics of interfacial films
An automated drop tensiometer (Teclis, Longessaigne, France) was used to
measure the interfacial tension at the oil-water interface. The experimental set-up
consisted of a model oil drop (stripped oil) in a glass cuvette (Hellma Analytics, Jena,
Germany) containing an aqueous protein blend (0.1 g/L) in phosphate buffer (10 mM, pH
7.0). Lycopene was not included in the oil drop because preliminary experiments
indicated that added lycopene did not have any effect on adsorption kinetics compared to
lycopene-free stripped oil, which is explained by its high lipophilicity that keeps the
component away from the interface (data not shown). Each model oil drop (60 mm2
surface area) was formed with a 0.5 mL glass syringe, which was connected to a 16gauge stainless steel needle. Interfacial tension, calculated via the Young-Laplace
108
equation, was recorded for 2 hours. Between sample runs, cuvettes, needles, and syringes
were profusely cleaned with ethanol, 1% detergent solution (Hellmanex, Hellma
Analytics, Jena, Germany), and ultrapure water.
4.2.9 Interfacial rheology of interfacial films
For each individual protein and protein-mixture, dilatational deformation sweeps
were performed with the drop tensiometer following the 2-hour equilibration period
described in the previous section. The model drop was subjected to amplitude sweeps in
which the drop was compressed and expanded to 2-35% of its original area (60 mm2)
with a constant frequency (0.01 Hz). The dilatational elastic (E’d) and loss (E”d) moduli
were calculated as follows (Eqs. (3.3) and (3.4)):
E’d = Δγ (A0/ΔA) cosδ
(4.3)
E”d= Δγ (A0/ΔA) sinδ
(4.4)
where Δγ is the change in interfacial tension, A0 is the initial drop surface area, ΔA is the
change in drop surface area during the oscillations, and δ is the phase shift.
The loss tangent (tan θ) was calculated and expressed as a function of the applied
deformation ((A-A0)/A0) (Eq. 4.5):
tan θ = E”d / E’d
(4.5)
A limitation of using the first harmonic Fourier moduli is that any nonlinearities present
in the raw signal are disregarded (Ewoldt, Hosoi, & McKinley, 2007; Van Kempen,
Schols, Van Der Linden, & Sagis, 2013). Lissajous-Bowditch plots can thus be used to
analyze complex interfaces that experience non-linear interfacial tension response upon
compression and expansion. For this, the surface pressure is plotted as a function of
deformation. The resulting plot shape can be related to film rheology (Figure 4-1), but
more importantly, asymmetries in shape can be indicative of a complex interface (Sagis
& Scholten, 2014) with e.g., distinct protein domains.
109
Figure 4-1 Examples of Lissajous-Bowditch curves depicting viscous (a), viscoelastic
(b), elastic (c), and non-linear viscoelastic (d) interfaces. Figure adapted from Deshpande
(2010) and Sagis & Scholten (2014).
4.2.10 Statistical analysis
Values obtained for physical and chemical stability represent the mean and SD
from three independently prepared emulsions. JMP version 11 (SAS Institute Inc., Cary
NC, USA) was used to statistically analyze data with one-way ANOVA and to compare
means post-hoc with the Tukey-Kramer method. In all statistical analyses, significance
was established with α=0.05.
4.3 Results and discussion
4.3.1 Physical stability of lycopene-loaded emulsions
All fresh emulsions showed fairly similar droplet diameters (d3,2), between 0.180.21 μm. During the 14-day storage period, most emulsions were macroscopically stable
and did not show creaming or phase separation (Figure 4-2), which was confirmed by
static light scattering measurements (Figure 4-3A, B, C, D, G). The emulsions stabilized
with PPI-WPI blends exhibited very good physical stability, as the droplet size
distribution remained unchanged over time and regardless of dilution with SDS (Figure
4-3B, D), which indicates that the samples were not susceptible to coalescence or
flocculation. The emulsion stabilized with the SPI-PPI blend (Figure 4-3G) also showed
good physical stability, with the exception of a slight shift towards a right skewed
distribution at day 14 without SDS dilution, indicating minor and reversible flocculation.
110
Emulsions stabilized with the 1:1 and 3:1 SPI-WPI blends (Figure 4-3A, C) were
probably slightly flocculated, as indicated by a minor droplet population visible around 1
– 10 μm, that disappeared when the samples were diluted in SDS solution.
Figure 4-2 Visual appearance of lycopene-loaded emulsions stabilized by WPI (a), SC
(b), SPI (c), PPI (d) or protein blends, 1:1 SPI-WPI (e), 3:1 SPI-WPI (f), 1:1 PPI-WPI
(g), 3:1 PPI-WPI (h), 1:1 SPI-SC (i), 1:1 PPI-SC (j), 1:1 SPI-PPI (k) on day 14 of storage.
SC-blend samples (i, j) exhibit a lighter color compared to the other emulsions and an
orange creamed layer (highlighted in the dashed line box).
Figure 4-3 Comparison of droplet size distributions of lycopene-loaded emulsions
stabilized with 1:1 SPI-WPI (A), 1:1 PPI-WPI (B), 3:1 SPI-WPI (C), 3:1 PPI-WPI (D),
1:1 SPI-SC (E), 1:1 PPI-SC (F), and 1:1 SPI-PPI (G) at day 0 and 14 with and without
1% SDS. Identical distributions with and without SDS dilution suggest that flocculation
111
did not occur. When Day 0 and Day 14 distributions are identical the emulsions are
physically stable.
Quite different behavior was observed for the emulsions stabilized by SC-blends.
These samples showed macroscopic phase separation, with the formation of a creamed
layer, over 14 days of incubation (Figure 4-2). The emulsion stabilized with the 1:1 PPISC blend was the most unstable and exhibited a dramatic increase in droplet diameter
from ~0.18 to 1.18 μm by day 14 (Figure 4-4). Although the droplet size for the emulsion
stabilized with the SPI-SC blend did not appear to grow as dramatically (Appendix A,
Table A-2), the span was significantly increased by day 14 (Appendix A, Table A-3).
These effects were to some extent caused by flocculation, as can be deduced from the
distributions obtained after dilution with SDS that have lower average droplet size, but
coalescence did take place since the droplet size distribution no longer coincides with the
one measured initially (Figure 4-3E, F). The dramatic changes in droplet sizes for SCblends imply that there could be antagonistic effects on physical stability when these
plant proteins are blended with SC.
10
*
1:1 SPI-WPI
Droplet size (μm)
1:1 PPI-WPI
1:1 SPI-SC
1:1 PPI-SC
1
3:1 SPI-WPI
3:1 PPI-WPI
1:1 SPI-PPI
0
0
3
7
14
Time (Days)
Figure 4-4 Droplet size, d3,2 of lycopene-loaded emulsions over time. Response values
shown represent the mean + SD (n=3). At day 14, an asterisk (*) denotes a value that is
112
significantly (p > 0.05) different. Statistical differences for values at days all other time
points (days 0, 3, and 7) are listed in Appendix A,Table A-2
In contrast, Yerramilli et al. (2017) observed an improvement in emulsion
stability when using blends of PPI and SC. An increase in surface hydrophobicity of PPISC blends post homogenization was noted and it was hypothesized that the SC molecules
were able to surround excess PPI to limit depletion flocculation and to better solubilize
PPI in the aqueous phase (Yerramilli et al., 2017). Similarly, Ji et al. (2015) observed a
synergistic effect for SC and SPI which resulted in stable emulsions, while the individual
proteins did not. The observed differences could be due to the concentrations that are
used: <1% (w/w) to limit non-adsorbed protein in the current work, while both Ji et al.
(2015) and Yerramilli et al. (2017) worked with higher protein concentrations (2% and 510%, respectively) and likely had larger proportions of non-adsorbed protein in the
aqueous phase. In our study, we used the soluble fraction because it seemed to positively
impact emulsions (Ho et al., 2016), while Yerramilli et al. (2017) used prehomogenization to break up aggregates to enhance dispersibility. Both protein
concentration and solubility/presence of insoluble fraction may thus have influenced
emulsion stability.
The zeta potential of different emulsions stabilized by protein blends was and remained
largely negative over 14 days for all samples (Figure 4-5; pH 7 > IEP proteins), which is
in line with our previous results for the individual WPI, SC, SPI, and PPI (Ho et al.,
2016). It is important to note that the electrostatic charge of emulsion droplets is not
necessarily indicative of the emulsions’ physical stability; for protein-stabilized
emulsions, the formation of thick, viscoelastic interfacial films is expected to be a major
113
factor in that respect, which is further investigated in section 3.3.
Figure 4-5 Initial zeta potential of lycopene-loaded emulsions fabricated with proteins
and protein blends. The results previously obtained for emulsions stabilized with the
individual proteins are displayed as reference (within the gray dashed-line box), while the
results obtained for emulsions stabilized with protein blends are shown to the right. Data
shown represent the mean + SD (n=3), with same superscript letters denoting values that
are not significantly different (p > 0.05).
4.3.2 Chemical stability of lycopene in emulsions
Lycopene retention was expressed as a percentage since the initial encapsulation
efficiencies (Appendix A, Table A-4) varied slightly amongst samples. Encapsulation
efficiencies for all emulsions were between ~47-63%, with the efficiencies of the
emulsions stabilized with the blends being statistically similar to that of at least one of the
emulsions stabilized with the individual protein components. This suggests that blending
plant and dairy protein does not significantly change the initial encapsulation efficiency
compared to using one protein alone.
Due to the physical destabilization of the emulsions stabilized with SC-based
blends, their lycopene retention was lower than that of all other blends at day 14 as
reported in the supplementary information (Appendix A, Table A-5). All other emulsions
exhibited at least 60% retention of lycopene after 14 days of storage, with globular
114
protein blend-based emulsions retaining >80% (Figure 4-6).
120
b
Lycopene Retention (%)
100
80
bc
abc
ac
b
c
abc
bc
a
60
40
20
0
WPI
SC
SPI
PPI
1:1 SPI- 1:1 PPI- 3:1 SPI- 3:1 PPI- 1:1 SPIWPI
WPI
WPI
WPI
PPI
Figure 4-6 Percent lycopene retention after 14 days of storage. The results previously
obtained for emulsions stabilized with the individual proteins are displayed as reference
(within the gray dashed-line box), while the results obtained for emulsions stabilized with
protein blends are shown to the right. Values shown represent the mean + SD (n=3), with
same superscript letters denoting values that are not significantly different (p > 0.05).
Overall, emulsions stabilized by WPI in combination with a plant protein (in
particular, the 1:1 blends SPI-WPI and PPI-WPI), showed a higher lycopene retention
over time compared to emulsions stabilized with the individual protein counterparts.
When both plant proteins were used as a blend (1:1 SPI-PPI), the lycopene retention was
one of the highest observed. These results suggest that blending plant protein with either
WPI or another globular plant protein can improve the chemical stability of lycopeneloaded emulsions compared to using WPI alone. Additionally, compared to SC, the best
performing individual protein, blends containing plant proteins performed statistically
similarly or better with regards to chemical stability.
Although SC has been shown to outperform WPI and plant proteins with regards
to the chemical stabilization of lipophilic ingredients in emulsions (Ho et al., 2016; Hu,
McClements, & Decker, 2003), some of the protein blend-based emulsions presented
here had lycopene retentions that were similar to SC, which suggests that plant protein
blends could be used as an alternative. This can be related to many proteins, including
115
WPI, SC, SPI, and PPI, having antioxidant properties (Han & Baik, 2008; Mcgookin &
Augustin, 1991; Peña-Ramos & Xiong, 2003; Peng, Xiong, & Kong, 2009), e.g. exposure
of free-radical scavenging residues were reported for β-lactoglobulin (Elias, Kellerby, &
Decker, 2008). However, it is also important to note that proteins in the aqueous phase
may have influenced the lycopene stability through their antioxidant action, in spite of the
limit set to <30% (Berton‐ Carabin, Ropers, & Genot, 2014; Elias, McClements, &
Decker, 2005; Faraji, Mcclements, & Decker, 2004; Min Hu, D. Julian McClements, &
Decker, 2003).
4.3.3 Adsorption kinetics
Figure 4-7A distinguishes the commonly described steps for protein adsorption:
(i) an initial induction period, corresponding to diffusion of proteins to the interface; (ii) a
steep decline in the interfacial tension, corresponding to the interface becoming filled;
and (iii) a slow decline in interfacial tension, corresponding to conformational changes of
the adsorbed layer (Beverung, Radke, & Blanch, 1999). For all tested individual proteins
and protein blends, equilibrium interfacial tensions at the stripped oil-water interface
were ~15-16 mN/m and were not largely different except for WPI, which was slightly
less surface-active with an equilibrium value of ~18 mN/m.
WPI-based blends appeared to reach the equilibrium interfacial tension faster than
WPI alone, and also systematically showed a lower first measurement point, indicative of
fast (sub-second) adsorption of material (Figure 4-7B, C). Although plant proteins, in
particular PPI, appeared to have a faster rate of adsorption than that of WPI, the WPIbased blends exhibited an even more immediate drop in interfacial tension. Although
speculative, it is possible that interactions taking place in the aqueous phase induce some
conformational changes that favor early adsorption of protein material at the oil-water
interface.
116
Figure 4-7 Adsorption kinetics of the three stages of adsorption, typically expected for
the dynamic interfacial tension response of proteins (A) is shown as an example adapted
from Beverung, Radke, & Blanch (1999) and the observed adsorption kinetics for SPI
and WPI (B), PPI and WPI (C), and SC and plant protein blends (D) at the O-W
interface. The interfacial tension as function of time for 3:1 SPI-WPI (a), 1:1 SPI-WPI
(b), 3:1 PPI-WPI (c), 1:1 PPI-WPI (d), 1:1 SPI-PPI (e), 1:1 SPI-SC (f), and 1:1 PPI-SC
(g) are plotted as data points connected by solid lines. Dashed lines correspond to
individual proteins (WPI, SPI, PPI, and SC) and the interfacial tension of oil-water
interface is represented by a blue dashed horizontal line for reference.
The initial adsorption behavior of SC-based blends was similar as found for SC,
which was less fast compared to only PPI or SPI (Figure 4-7D). This could be indicative
of SC preferentially adsorbing and dominating the interface (in the early stages of
adsorption). Britten & Giroux (1991) observed that interfacial tension of casein-whey
composite blends was dominated by casein due to its preferential adsorption at the
interface. Seta et al. (2014) studied β-lactoglobulin and β-casein mixed layer films at the
oil-water interface and observed that adsorption kinetics of mixtures fell between those of
the individual proteins. In both studies, it was noted that there appeared to be a greater
influence of casein, and this could be due to the flexible nature of the random coil protein
that allows SC to rapidly unfold and adsorb at the interface (Mackie, Gunning, Wilde, &
117
Morris, 2000; Mitchell, 1986). For the 1:1 SPI-PPI blend, the adsorption behavior was
close to both SPI and PPI (Figure 4-7D), so it is difficult to conclude how the blend
behavior is related to its individual counterparts.
4.3.4 Interfacial rheology
The rheological properties of the interfacial protein-based films can provide
insight into emulsion stability to coalescence, as the film rupture between two emulsion
droplets can be seen as dilatational deformation (Bos & van Vliet, 2001; Murray, 2011).
The elastic and loss moduli of all samples did not appear to have a major dependence on
the applied deformation (Figure 4-8A, B), which suggests that the measurements were
mostly conducted in the linear viscoelastic regime, as was the case for the individual
proteins (WPI, SC, SPI, PPI; Ho et al., 2016). Of all the proteins measured, SC clearly
showed the lowest Ed’ and highest loss tangent. The random coil structure and flexible
behavior of SC contributes to its ability to sterically stabilize droplets (Nylander, 1998)
but limit its ability to pack tightly at the interface (Dickinson, 2001) with adjacent SC
proteins being unable to form strong intermolecular protein interactions and interfacial
network. Except for the 1:1 PPI-SC mixture, all blends showed tangents in between those
of the individual constituent proteins and relatively elastic behavior compared to SC
(Figure 4-8C), which suggests that the interface elasticity is determined by contributions
from both proteins, and is most likely due to the ability of globular proteins to form
viscoelastic films via intermolecular interactions (Chang, Tu, Ghosh, & Nickerson, 2015;
Dickinson, 2001).
118
Figure 4-8 Comparison of (A) WPI blends and (B) SC blends + SPI-PPI blend elastic
moduli, E’d (a) and loss moduli, E’’d (b) for protein films at the oil–water interface; and
(C) loss tangent (E’’d/E’d) at 0.15 deformation. Results for individual proteins are shown
as: (A, B) solid lines (E’d), dashed lines (E’’d) and (C) black bars for reference. Data
shown represent mean + SD (n=3). Statistical differences amongst protein films are
shown (C) with same letters denoting values that are not significantly different (α = 0.05).
When plotting the data as Lissajous-Bowditch curves, more information regarding
the interfacial behavior can be obtained (Sagis & Fischer, 2014). At low deformation of
0.1 the curves are symmetric (Figure 4-9). The film made of the 1:1 SPI-SC blend
exhibited a curve mostly similar to that of the individual SC film; the very low slope
indicative of a low viscoelastic modulus as a result of SC flexibility and ability to
structurally re-adjust during the compression cycle. In spite of the similarity of the
curves, the loss tangent suggests that both proteins are present in the interface (Figure 48C). For previous β-casein-β-lactoglobulin blend studies it was suggested that formation
of a primary casein layer took place at the interface with β-lactoglobulin weakly
interacting as a sort of loose secondary layer (Dickinson, Rolfe, & Dalgleish, 1990),
which could also be an explanation for the behavior of our 1:1 SPI-SC blend. The film
made with 1:1 PPI-SC showed notably similar behavior to that of the PPI film (Figure 49). PPI has a faster rate of adsorption than SPI but is slower than SC (Figure 4-7D), and
119
since displacement of SC is not likely, PPI could be filling vacant patches at the oil-water
interface that were not readily occupied by SC.
Figure 4-9 Lissajous-Bowditch plots of interfacial films prepared with proteins and
protein blends at the oil-water interface (amplitude = 0.1). Surface pressure (π) is plotted
against the applied deformation
When the droplet was subjected to 35% compression and expansion of its area,
most samples showed some asymmetries: in extension (i.e., top part of the curve, from
left to right), the slope of the curve is decreasing with increasing amplitude; whereas in
compression (i.e., bottom part of the curve, from right to left) it is increasing with
increasing amplitude. Such a shape points at strain softening and strain hardening
(Berton-Carabin, Schröder, Rovalino-Cordova, Schroën, & Sagis, 2016; Van Kempen et
al., 2013; Wan, Yang, & Sagis, 2016), which was particularly marked for the 1:1 PPI-SC
blend (Figure 4-10), and may indicate that distinct domains were present (Rühs,
Scheuble, Windhab, & Fischer, 2013) that get disconnected in extension, and jam in
120
compression. For SPI-WPI blends, the ratio influenced the profile of the resulting
Lissajous-Bowditch plots: at 1:1 ratio (Figure 4-10), the shape of the plot resembled that
of WPI only whereas at 3:1 ratio, it looked more like that of SPI only, which points out
that the initial amounts are of importance for the interfacial film formation and properties.
Figure 4-10 Lissajous-Bowditch plots of interfacial films prepared with proteins and
protein blends at the oil-water interface, at deformations (∆A/A0 (-)) 0.03 (green), 0.05
(blue), 0.1 (yellow), 0.15 (orange), 0.25 (gray), 0.3 (light blue). Surface pressure (π) is
plotted against the applied deformation.
4.3.5 Comparison of protein-protein blends and potential droplet stabilization
mechanisms
Based on our findings, we have hypothesized potential organization schemas to
explain the stability or instability of protein blend-stabilized emulsions (Figure 4-11). All
emulsions had (initially) relatively similar d3,2 values. Emulsions stabilized with WPIbased blends exhibited enhanced physical and chemical stability compared to emulsions
stabilized with individual WPI or plant protein. We hypothesize that the proteins in the
WPI-blends co-adsorbed, forming a thick, viscoelastic protein layer. From an oxidation
perspective, this co-adsorbed film could provide a protective effect through, e.g.,
121
enhanced metal or radical scavenging amino acids. Alternatively, excess protein in the
aqueous phase could contribute similarly as e.g. WPI has metal chelating and free
radicals scavenging properties (Tong, Sasaki, McClements, & Decker, 2000).
Figure 4-11 Schematic of possible scenarios for protein blend behavior at the oil-water
interface.
The most remarkable feature of the emulsions stabilized with SC-based blends
was their rapid physical destabilization (>7 days), which is in stark contrast to the good
physical stability of emulsions stabilized with SC only. There is thus an antagonistic
effect, and while SC is expected to reach the interface first (Figure 4-7D) and possibly
exclude other proteins from the interface, the presence of non-adsorbed protein could
cause depletion flocculation. Although depletion flocculation is considered a reversible
process, in unstable systems it can lead to coalescence and oiling off (Dickinson, 2001),
which we observed as visible phase separation in the emulsions stabilized with SCblends. In our study, the emulsion protein concentration was selected to limit nonadsorbed protein post homogenization, which makes it questionable whether sufficient
122
protein was present in the continuous phase to induce depletion flocculation.
Alternatively, it is presumable that SC rapidly formed an interfacial film around droplets
during homogenization, which could have prevented further adsorption of other proteins,
as previously observed by others (Aoki et al., 1984). Yet, the SC amount available for
interface stabilization was 2-fold lower than in the emulsions stabilized by SC only; this
could imply that in the SC-based blends, the SC amount was insufficient for long-term
droplet stabilization and was not compensated by the plant protein. A last possible
explanation relates to the asymmetry in the Lissajous plots observed for the PPI-SC
blend, suggesting that distinct SC protein domains may exist at the interface, which could
compromise film integrity and mechanical strength, due to weak packing ability and poor
affinity for neighbouring proteins (Dickinson, 2001).
4.4 Conclusions
Our findings indicate that plant-dairy protein blends have effects that exceed
those of individual proteins. Despite SC providing the best physicochemical stability of
lycopene-loaded emulsions amongst individual proteins, emulsions stabilized with blends
of SC and SPI or PPI were physically unstable, revealing an antagonistic effect on
physical stability. Emulsions stabilized with WPI-based blends exhibited better lycopene
retention compared to emulsions stabilized with individual plant proteins or WPI and
lycopene retention at day 14 was the highest amongst protein blend-based emulsions.
For each blend, both proteins were introduced simultaneously to the interface, which
consequently affected the ultimate rheological behavior. Proteins with the fastest
adsorption rate were expected to dominate the interface, as proteins can only be displaced
in very specific cases (flexible for globular proteins). Overall, blending plant proteins
seemed to produce synergistic (with WPI) or antagonistic (with SC) effects on the
physicochemical stability of lycopene-loaded emulsions. Although the underlying
mechanisms could not be identified completely, it is clear that plant proteins that are
currently underutilized are genuine alternatives to replace or partially replace dairy
proteins for stabilization of lycopene and other lipophilic bioactives in colloidal systems.
123
4.5 Acknowledgements
This research was based upon work supported by the National Science
Foundation Graduate Research Fellowship (Grant No. DGE-1333468) through the
Graduate Research Opportunities World Wide (GROW) program in support with the
Netherlands Organisation for Scientific Research (NWO), and the Wageningen
University Graduate School (VLAG). The authors would like to thank the laboratory
technicians of the Food Process Engineering Group at Wageningen University, including
Jos Sewalt and Martin de Wit, for assistance with experimental set up, and Maurice
Strubel for his HPLC expertise, and Ph.D. student Anja Schröder for her assistance with
the initial analysis of the Lissajous-Bowditch plots.
124
4.6 References
Aoki, H., Shirase, Y., Kato, J., & Watanabe, Y. (1984). Emulsion Stabilizing Properties
of Soy Protein Isolates Mixed with Sodium Caseinates. Journal of Food Science,
49(1), 212–216. http://doi.org/10.1111/j.1365-2621.1984.tb13710.x
Ax, K., Mayer-Miebach, E., Link, B., Schuchmann, H., & Schubert, H. (2003). Stability
of lycopene in oil-in-water emulsions. Engineering in Life Sciences, 3(4), 199–201.
http://doi.org/10.1002/elsc.200390028
Benjamin, O., Silcock, P., Beauchamp, J., Buettner, A., & Everett, D. W. (2014).
Emulsifying Properties of Legume Proteins Compared to β-Lactoglobulin and
Tween 20 and the Volatile Release from Oil-in-Water Emulsions. Journal of Food
Science, 79(10), E2014–E2022. http://doi.org/10.1111/1750-3841.12593
Berton-Carabin, C. C., Schröder, A., Rovalino-Cordova, A., Schroën, K., & Sagis, L. M.
C. (2016). Protein and lipid oxidation affect the viscoelasticity of whey protein
layers at the oil–water interface. European Journal of Lipid Science and
Technology, 118(11), 1630–1643.
Berton‐ Carabin, C. C., Ropers, M., & Genot, C. (2014). Lipid Oxidation in Oil‐ in‐
Water Emulsions: Involvement of the Interfacial Layer. Comprehensive Reviews in
Food Science and Food Safety, 13(5), 945–977.
Berton, C., Genot, C., & Ropers, M.-H. (2011). Quantification of unadsorbed protein and
surfactant emulsifiers in oil-in-water emulsions. Journal of Colloid and Interface
Science, 354(2), 739–748. http://doi.org/http://dx.doi.org/10.1016/j.jcis.2010.11.055
Beverung, C. J., Radke, C. J., & Blanch, H. W. (1999). Protein adsorption at the oil/water
interface: characterization of adsorption kinetics by dynamic interfacial tension
measurements. Biophysical Chemistry, 81(1), 59–80. http://doi.org/10.1016/S03014622(99)00082-4
Bos, M. A., & van Vliet, T. (2001). Interfacial rheological properties of adsorbed protein
layers and surfactants: a review. Advances in Colloid and Interface Science, 91(3),
437–471. http://doi.org/10.1016/S0001-8686(00)00077-4
Britten, M., & Giroux, H. J. (1991). Emulsifying Properties of Whey Protein and Casein
Composite Blends. Journal of Dairy Science, 74(10), 3318–3325.
125
http://doi.org/10.3168/jds.S0022-0302(91)78519-6
Britton, G., Liaaen-Jensen, S., & Pfander, H. (2004). Carotenoid Handbook Info.pdf.
Basel, Switzerland: Birkhauser Verlag.
Chang, C., Tu, S., Ghosh, S., & Nickerson, M. T. (2015). Effect of pH on the interrelationships between the physicochemical, interfacial and emulsifying properties
for pea, soy, lentil and canola protein isolates. Food Research International, 77(3),
360–367. http://doi.org/10.1016/j.foodres.2015.08.012
Cornacchia, L., & Roos, Y. H. (2011). Stability of β-Carotene in Protein-Stabilized Oilin-Water Delivery Systems. Journal of Agricultural and Food Chemistry, 59(13),
7013–7020. http://doi.org/10.1021/jf200841k
Dalgleish, D. G. (1997). Adsorption of protein and the stability of emulsions. Trends in
Food Science & Technology, 8(1), 1–6. http://doi.org/10.1016/S09242244(97)01001-7
Dalgleish, D. G., Goff, H. D., & Luan, B. (2002). Exchange reactions between whey
proteins and caseins in heated soya oil-in-water emulsion systems — behavior of
individual proteins. Food Hydrocolloids, 16(4), 295–302.
http://doi.org/10.1016/S0268-005X(01)00102-3
Di Mascio, P., Kaiser, S., & Sies, H. (1989). Lycopene as the most efficient biological
carotenoid singlet oxygen quencher. Archives of Biochemistry and Biophysics,
274(2), 532–538. http://doi.org/http://dx.doi.org/10.1016/0003-9861(89)90467-0
Dickinson, E. (1994). Protein-stabilized emulsions. Journal of Food Engineering, 22(1–
4), 59–74. http://doi.org/10.1016/0260-8774(94)90025-6
Dickinson, E. (2001). Milk protein interfacial layers and the relationship to emulsion
stability and rheology. Colloids and Surfaces B: Biointerfaces, 20(3), 197–210.
http://doi.org/10.1016/S0927-7765(00)00204-6
Dickinson, E. (2011). Mixed biopolymers at interfaces: Competitive adsorption and
multilayer structures. Food Hydrocolloids, 25(8), 1966–1983.
http://doi.org/10.1016/j.foodhyd.2010.12.001
Dickinson, E., Rolfe, S. E., & Dalgleish, D. G. (1990). Surface shear viscometry as a
probe of protein-protein interactions in mixed milk protein films adsorbed at the oilwater interface. International Journal of Biological Macromolecules, 12(3), 189–
126
194. http://doi.org/10.1016/0141-8130(90)90031-5
Elias, R. J., Kellerby, S. S., & Decker, E. A. (2008). Antioxidant Activity of Proteins and
Peptides. Critical Reviews in Food Science and Nutrition, 48(5), 430–441.
http://doi.org/10.1080/10408390701425615
Elias, R. J., McClements, D. J., & Decker, E. A. (2005). Antioxidant Activity of
Cysteine, Tryptophan, and Methionine Residues in Continuous Phase βLactoglobulin in Oil-in-Water Emulsions. Journal of Agricultural and Food
Chemistry, 53(26), 10248–10253. http://doi.org/10.1021/jf0521698
Ewoldt, R. H., Hosoi, A. E., & McKinley, G. H. (2007). New measures for characterizing
nonlinear viscoelasticity in large amplitude oscillatory shear. Journal of Rheology,
52(6), 1427–1458. http://doi.org/10.1122/1.2970095
Faraji, H., Mcclements, D. J., & Decker, E. A. (2004). Role of Continuous Phase Protein
on the Oxidative Stability of Fish Oil-in-Water Emulsions. Journal of Agricultural
and Food Chemistry, 52(14), 4558–4564. http://doi.org/10.1021/jf035346i
Hailing, P. J., & Walstra, P. (1981). C R C Critical Reviews in Food Science and
Nutrition Protein‐ stabilized foams and emulsions PROTEIN-STABILIZED
FOAMS AND EMULSIONS. Critical Reviews in Food Science and Nutrition,
15(2), 155–203. http://doi.org/10.1080/10408398109527315
Han, H., & Baik, B.-K. (2008). Antioxidant activity and phenolic content of lentils ( Lens
culinaris ), chickpeas ( Cicer arietinum L.), peas ( Pisum sativum L.) and soybeans (
Glycine max ), and their quantitative changes during processing. International
Journal of Food Science & Technology, 43(11), 1971–1978.
http://doi.org/10.1111/j.1365-2621.2008.01800.x
Ho, K. K. H. Y., Ferruzzi, M. G., Liceaga, A. M., & San Martín-González, M. F. (2015).
Microwave-assisted extraction of lycopene in tomato peels: Effect of extraction
conditions on all-trans and cis-isomer yields. LWT - Food Science and Technology,
62(1), 160–168. http://doi.org/http://dx.doi.org/10.1016/j.lwt.2014.12.061
Ho, K. K. H. Y., Schroën, K., San Martín-González, M. F., & Berton-Carabin, C. C.
(2016). Physicochemical stability of lycopene-loaded emulsions stabilized by plant
or dairy proteins. Food Structure.
http://doi.org/http://dx.doi.org/10.1016/j.foostr.2016.12.001
127
Hu, M., McClements, D. J., & Decker, E. A. (2003). Lipid Oxidation in Corn Oil-inWater Emulsions Stabilized by Casein , Whey Protein Isolate , and Soy Protein
Isolate. Journal of Agricultural and Food Chemistry, 51, 1696–1700.
Ji, J., Zhang, J., Chen, J., Wang, Y., Dong, N., Hu, C., … Wu, C. (2015). Preparation and
stabilization of emulsions stabilized by mixed sodium caseinate and soy protein
isolate. Food Hydrocolloids, 51, 156–165.
http://doi.org/10.1016/j.foodhyd.2015.05.013
Kean, E. G., Hamaker, B. R., & Ferruzzi, M. G. (2008). Carotenoid Bioaccessibility from
Whole Grain and Degermed Maize Meal Products. Journal of Agricultural and
Food Chemistry, 56(21), 9918–9926. http://doi.org/10.1021/jf8018613
Liu, F., & Tang, C.-H. (2014). Emulsifying Properties of Soy Protein Nanoparticles:
Influence of the Protein Concentration and/or Emulsification Process. Journal of
Agricultural and Food Chemistry, 62(12), 2644–2654.
http://doi.org/10.1021/jf405348k
Mackie, A. R., Gunning, A. P., Wilde, P. J., & Morris, V. J. (2000). Orogenic
Displacement of Protein from the Oil/Water Interface. Langmuir, 16(5), 2242–2247.
http://doi.org/10.1021/la990711e
McClements, D. J. (2004). Protein-stabilized emulsions. Current Opinion in Colloid &
Interface Science, 9(5), 305–313. http://doi.org/10.1016/j.cocis.2004.09.003
Mcgookin, B. J., & Augustin, M.-A. (1991). Antioxidant activity of casein and Maillard
reaction products from casein-sugar mixtures. Journal of Dairy Research, 58, 313–
320. http://doi.org/10.1017/S0022029900029885
Min Hu, D. Julian McClements, A., & Decker, E. A. (2003). Impact of Whey Protein
Emulsifiers on the Oxidative Stability of Salmon Oil-in-Water Emulsions. Journal
of Agricultural and Food Chemistry, 51(5), 1235–1439.
Mitchell, J. R. (1986). Foaming and emulsifying properties of proteins. Developments in
Food Proteins, 1986(4), 291–338.
Murray, B. S. (2011). Rheological properties of protein films. Current Opinion in Colloid
& Interface Science, 16(1), 27–35. http://doi.org/10.1016/j.cocis.2010.06.005
Nylander, T. (1998). Protein-lipid interactions. In Studies in Interface Science (Vol. 7, pp.
128
385–431). http://doi.org/10.1016/S1383-7303(98)80057-4
Peña-Ramos, E. A., & Xiong, Y. L. (2003). Whey and soy protein hydrolysates inhibit
lipid oxidation in cooked pork patties. Meat Science, 64(3), 259–263.
http://doi.org/10.1016/S0309-1740(02)00187-0
Peng, X., Xiong, Y. L., & Kong, B. (2009). Antioxidant activity of peptide fractions from
whey protein hydrolysates as measured by electron spin resonance. Food Chemistry,
113(1), 196–201. http://doi.org/10.1016/j.foodchem.2008.07.068
Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2013). Impact of lipid
nanoparticle physical state on particle aggregation and β-carotene degradation:
Potential limitations of solid lipid nanoparticles. Food Research International,
52(1), 342–349. http://doi.org/http://dx.doi.org/10.1016/j.foodres.2013.03.035
Rao, A. V, Waseem, Z., & Agarwal, S. (1998). Lycopene content of tomatoes and tomato
products and their contribution to dietary lycopene. Food Research International,
31(10), 737–741. http://doi.org/http://dx.doi.org/10.1016/S0963-9969(99)00053-8
Sagis, L. M. C., & Fischer, P. (2014). Nonlinear rheology of complex fluid–fluid
interfaces. Current Opinion in Colloid & Interface Science, 19(6), 520–529.
http://doi.org/10.1016/j.cocis.2014.09.003
Sagis, L. M. C., & Scholten, E. (2014). Complex interfaces in food: Structure and
mechanical properties. Trends in Food Science and Technology, 37(1), 59–71.
http://doi.org/10.1016/j.tifs.2014.02.009
Salvia-Trujillo, L., Qian, C., Martín-Belloso, O., & McClements, D. J. (2013). Influence
of particle size on lipid digestion and β-carotene bioaccessibility in emulsions and
nanoemulsions. Food Chemistry, 141(2), 1472–1480.
http://doi.org/http://dx.doi.org/10.1016/j.foodchem.2013.03.050
Seta, L., Baldino, N., Gabriele, D., Lupi, F. R., & Cindio, B. de. (2014). Rheology and
adsorption behaviour of β-casein and β-lactoglobulin mixed layers at the sunflower
oil/water interface. Colloids and Surfaces A: Physicochemical and Engineering
Aspects, 441(0), 669–677.
http://doi.org/http://dx.doi.org/10.1016/j.colsurfa.2013.10.041
129
Seta, L., Baldino, N., Gabriele, D., Lupi, F. R., & de Cindio, B. (2012). The effect of
surfactant type on the rheology of ovalbumin layers at the air/water and oil/water
interfaces. Food Hydrocolloids, 29(2), 247–257.
http://doi.org/http://dx.doi.org/10.1016/j.foodhyd.2012.03.012
Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano,
M. D., … Klenk, D. C. (1985). Measurement of protein using bicinchoninic acid.
Analytical Biochemistry, 150(1), 76–85. http://doi.org/10.1016/00032697(85)90442-7
Tong, L. M., Sasaki, S., Mcclements, D. J., & Decker, E. A. (2000). Mechanisms of the
Antioxidant Activity of a High Molecular Weight Fraction of Whey. Journal of
Agricultural and Food Chemistry, 48(5), 1473–1478.
http://doi.org/10.1021/jf991342v
Van Kempen, S. E. H. J., Schols, H. A., Van Der Linden, E., & Sagis, L. M. C. (2013).
Non-linear surface dilatational rheology as a tool for understanding microstructures
of air/water interfaces stabilized by oligofructose fatty acid esters. Soft Matter,
9(40), 9579e9592. http://doi.org/10.1039/c3sm51770e
Wan, Z., Yang, X., & Sagis, L. M. C. (2016). Contribution of Long Fibrils and Peptides
to Surface and Foaming Behavior of Soy Protein Fibril System. Langmuir.
http://doi.org/10.1021/acs.langmuir.6b01511
Ye, A. (2008). Interfacial composition and stability of emulsions made with mixtures of
commercial sodium caseinate and whey protein concentrate. Food Chemistry,
110(4), 946–952. http://doi.org/10.1016/j.foodchem.2008.02.091
Yerramilli, M., Longmore, N., & Ghosh, S. (2017). Improved stabilization of
nanoemulsions by partial replacement of sodium caseinate with pea protein isolate.
Food Hydrocolloids, 64, 99–111. http://doi.org/10.1016/j.foodhyd.2016.10.027
130
CHAPTER 5 BIOACCESSIBILITY AND CELLULAR UPTAKE OF
WHEY AND PEA PROTEIN LYCOPENE-LOADED EMULSIONS
5.1 Introduction
Protein-stabilized emulsion systems have been studied as a technology to
potentially enhance stability and bioavailability of carotenoids (Cornacchia & Roos,
2011; Qian, Decker, Xiao, & McClements, 2012). Although efforts have been made to
understand the process of digestion and cellular uptake of poorly absorbed compounds
from emulsion systems (Lu et al., 2016; Ribeiro et al., 2006; Yi, Lam, Yokoyama, Cheng,
& Zhong, 2014), limited work has been done that compares the potential biological fate
with a conventional food source and couples an in vitro digestion with cellular uptake in
Caco-2 cells. Previously, Ribeiro et al. (2006) investigated the effect of interfacial
composition on cellular uptake of lycopene from non-digested emulsions. While their
findings have applicability for parenteral nutrition, the effect of interfacial composition
on digestion, bioaccessibility and consequential absorption is still unclear. Studies have
indicated that there is a clear correlation with emulsion droplet size and bioavailability
(Desai, Labhasetwar, Walter, Levy, & Amidon, 1997; Lu et al., 2016). Thus, in order to
test the effect of two different interfacial proteins, we chose to compare whey protein
isolate, a well-studied dairy emulsifier and pea protein isolate, a plant protein that has
gained popularity as an emulsifier. Since these two proteins have very different
properties, efforts were taken to attain comparable droplet size. Thus, the objectives of
this study were to compare the effect of interfacial protein on lycopene: 1) micellarization
efficiency and 2) cellular uptake in Caco-2 cells.
5.2 Materials and Methods
5.2.1 Materials
All-trans-lycopene standard (>90% purity), USP lycopene standard, enzymes,
reagents and solvents (analytical grade) were obtained from Sigma Aldrich (St. Louis,
131
MO) unless stated otherwise. Canola oil was purchased from a local grocery store (West
Lafayette, IN). Double distilled water was used for droplet size measurement and for
buffer preparation.
5.2.2 Preparation of emulsion aqueous phase
Selected protein concentrations were based off of preliminary experiments and
previous work (Ho, Schroën, San Martín-González, & Berton-Carabin, 2016), which
aimed at ensuring small, stable droplets, while limiting the amount of excess protein in
the aqueous phase. Since Caco-2 cell uptake is reportedly size dependent (Desai et al.,
1997), emulsifier concentration was adjusted to ensure relatively similar droplet sizes for
all studied emulsions. Protein solutions were prepared as described previously (Ho et al.,
2016). Briefly, whey protein isolate (0.5% w/w) was combined with 0.01 M phosphate
buffer (pH=7.0) and allowed to hydrate overnight with light agitation to reduce foam
formation. Unlike the whey protein isolate, which was completely water soluble, the pea
protein isolate was only 25% soluble. Thus, the insoluble fraction was first removed by
hydrating the pea protein for 24 hours at 4°C at 600 rpm using a magnetic stir plate. The
solution was centrifuged twice at 10,000 x g for 10 minutes to separate the insoluble
proteins. The resulting soluble pea protein solution was then diluted with phosphate
buffer (pH=7.0) to attain a final concentration of 0.7% w/w.
5.2.3 Preparation of emulsion oil phase
Lycopene was removed from a USP standard using a known volume of hexane.
An aliquot was then taken, diluted with hexane, and absorbance read at 471 nm using a
UV-VIS DU 800 spectrophotometer (Beckman and Coulter, Inc., Brea, CA). The
absorbance, measured in triplicate, was then used to calculate the lycopene concentration
using a molar extinction coefficient of 1.85 x 105 L mol-1 cm-1 (G Britton, Liaaen-Jensen,
& Pfander, 2004). The hexane-lycopene stock was then combined with canola oil, and
held under a stream of nitrogen until constant weight to remove the hexane. The
resulting stock oil had a lycopene concentration of ~0.5 mg/g canola oil. HPLC-DAD
analysis indicated that the stock contained ~73% all-trans-lycopene and ~27% cislycopene.
132
5.2.4 Emulsion fabrication
Aqueous phase (either whey or soluble pea protein solution) was combined with
lycopene-containing oil to form a 10% w/w oil-in-water emulsion. The two phases were
mixed with a Polytron PT 2100 high shear mixer (Kinematica, Inc., Bohemia, NY) at
1,100 rpm to form a coarse emulsion, which was immediately passed through a
NanoDeBee high-pressure homogenizer (BEE International, South Easton, MA) at
20,000 psi for 5 passes. Finished emulsions were stored in amber glass containers,
flushed with nitrogen, and held at 4°C to delay microbial growth.
5.2.5 Droplet size and surface charge measurement
Emulsion droplet size and zeta potential was measured using a ZS Zetasizer Nano
ZS (Malvern Instruments, Ltd., Worchestershire, UK). For size measurement, samples
were diluted to 0.5% (v/v) with double distilled water to prevent multiple scattering. In
order to distinguish flocculated from individual droplets, emulsions were measured with
and without the addition of an aqueous 10% sodium dodecyl sulfate (SDS) solution. All
samples were measured following a 2 minute temperature equilibration at 25°C with the
average size expressed as the hydrodynamic diameter (Z-average) of three runs per
measurement. For zeta potential measurement, samples were diluted to 1.25% (v/v) with
double distilled water and measured following a 2 minute equilibration period at 25°C
with 3 measurements per sample. A back-scatter detection angle of 173°, refractive
indices of 1.330 and 1.475 for water and canola oil, respectively, were used with the
Smoluchowski model to calculate the resulting values. Both the droplet size and zeta
potential values were expressed as an average from three independent samples.
5.2.6 In vitro digestion of emulsions
Precisely 8 g of emulsion or lycopene oil with buffer (10% oil) were subjected to
digestion without the addition of canola oil. Tomato paste (2 g) was combined with 0.8 g
of canola oil to match the oil content in the other samples. Lycopene and oil content used
for all treatments was kept consistent because co-consumption of lipids can enhance
carotenoid uptake (Brown et al., 2004; Unlu, Bohn, Clinton, & Schwartz, 2005). Samples
133
were subjected to a simulated oral, gastric and intestinal digestion as previously described
by Garrett, Failla, & Sarama (1999) and as adapted by Kean, Hamaker, & Ferruzzi
(2008). Following the intestinal phase, samples were collected (digestive fractions) and
the remaining centrifuged in polycarbonate tubes at 10,000 x g for 1 hr using a Sorvall
RC 6 Plus centrifuge (Thermofisher Scientific, Waltham, MA). The supernatant was
filtered through a 0.22 μm cellulose acetate filter to isolate the aqueous micellar fraction.
Digestive and aqueous micellar fractions were collected, flushed with nitrogen and frozen
at -80°C until extraction and analysis.
5.2.7 Cell culture and treatments
In vitro bioavailibility was assessed as cellular uptake of lycopene using Caco-2
HTB-37 cells (ATCC, Manassas, VA) with a single compartment model. Cell
maintenance followed protocols described previously (Lipkie et al., 2014; Neilson, Song,
Sapper, Bomser, & Ferruzzi, 2010). Briefly, cells were cultured in vented flasks,
incubated at 37°C (5% CO2), and passaged at 80% confluency.
Cells (passages 26-28) were seeded at a density of 12.8 x 106 cells/well onto 6well polystyrene plates (Corning Inc., Corning, NY) and were fed with fresh media
(DMEM with 10% FBS, 1% HEPES, 1% penicillin/streptomycin, 1% nonessential amino
acids, and 0.1% gentamicin v/v) every other day and 24 hr prior to treatment. Cellular
uptake studies were conducted 10-12 days post-confluency. Cells were washed with
warm 0.1% fatty acid-free albumin in PBS (Redan et al., 2016) and washed twice with
warm PBS before being treated with aqueous micellar fractions diluted with warm
DMEM (1:3 v/v) following in vitro digestion. Cells were washed with warm 5 mM
sodium taurocholate in PBS to remove surface carotenoids (Ferruzzi, Lumpkin,
Schwartz, & Failla, 2006) then collected in cold PBS following 0, 3, and 6 hours of
treatment. Preliminary work (data not shown) showed that the 1:3 (v/v) dilution of
digested emulsions did not have a negative effect on monolayer integrity for Caco-2 cells
grown on transwells. However, following 6 hr of treatment, cells appeared to detach
from wells for lycopene oil treatments. This is likely due to proteases in the aqueous
fraction targeting the cellular protein since the digested meal for lycopene oil contained
no added protein. Thus, lycopene uptake at 0, 3, and 6 hr (Appendix B, Figure B-5) is
134
reported for all samples, except for the lycopene oil. In order to prevent cell detachment,
a second cell experiment was conducted by using a treatment media with a 1:4 v/v
dilution of thawed (stored at -80°C for <1 week) aqueous micellar fraction in warm
DMEM. For this set of treatments, uptake was measured after 3 hr of treatment. Protein
content of untreated cells was quantified via BCA Assay (QuantiPro BCA Assay Kit,
Sigma-Aldrich, St. Louis, MO).
5.2.8 Lycopene extraction and quantification via HPLC
All samples were extracted using a method previously described (Ax, MayerMiebach, Link, Schuchmann, & Schubert, 2003) and modified (Ho et al., 2016) with
minor adjustments. Briefly, 5 g of digested or aqueous fractions were combined with 4
mL of hexane (0.1% BHT w/v), 3 mL of ethanol, and 1 mL of aqueous saturated sodium
chloride. Samples were vortexed for 5 seconds and the upper hexane layer was collected.
This process was repeated with additional hexane until there was no visible color left (3-4
repetitions). All hexane layers per sample were pooled and dried under a stream of
nitrogen and held at -80°C before resolubilization and quantification via HPLC-DAD.
Undigested samples (described in Section 5.2.6) were extracted using an adapted protocol
to ensure lycopene release from tomato paste solids. Emulsion, lycopene oil, or tomato
paste was combined with 10 mL of hexane, 3 mL of ethanol, 1 mL of aqueous saturated
sodium chloride, and 1 g of celite. The mixtures were agitated using a high shear mixer
at 11,000 rpm for 20 seconds. Samples were then centrifuged at 2,500 x g for 2 min.
The upper, hexane layer was collected carefully with a Pasteur pipette and additional
hexane was added prior to vortexing (5 s) and centrifugation two more times. All hexane
layers were pooled per sample and dried under a stream of nitrogen prior to
resolubilization and HPLC-DAD analysis. Samples collected post cell treatments were
extracted and stored in the same manner as the digesta and aqueous fractions, except they
were subjected to ultrasonication at 17 Watts for 5 seconds (to lyse the cells) using a
Branson Sonifier (Branson Ultrasonics, Danbury, CT) prior to solvent addition. To limit
degradation and isomerization of carotenoid during extraction and analysis, precautions
were taken to limit heat, light, and oxygen exposure by holding samples on ice,
conducting work in subdued and yellow lights, and flushing samples with nitrogen gas.
135
Separation and quantification of all-trans-lycopene was conducted using an
Agilent 1200 Series HPLC equipped with a diode array detector (Agilent Technologies,
Santa Clara, CA). In addition, cis-isomers were tentatively identified as done previously
(Ho, Ferruzzi, Liceaga, & San Martín-González, 2015). In vitro bioaccessibility
following digestion was expressed as the micellarization efficiency (Eq. 5.1)
Micellarization efficiency = CAQ/CDG * 100 (5.1)
where CAQ and CDG represent the lycopene content in the aqueous micellar fraction and
the digesta fraction, respectively. In order to normalize for the lycopene content in the
media, cellular uptake was expressed as the % uptake (Eq. 2)
Lycopene % uptake = Ccells/CAQ *100
(5.2)
where Ccells and CAQ represent the lycopene content extracted from the cells and the
aqueous micellar fraction, respectively.
5.2.9 Statistical analysis
Data were statistically analyzed using JMP version 12 (SAS Institute, Cary, NC).
The Shapiro-Wilk test was used to determine normality of data prior to analysis. Data
that was not normally distributed was first subjected to a Box-Cox transformation prior to
further statistical analysis. Data were subjected to analysis of variance (ANOVA) with
α=0.05. The Tukey-Kramer method was conducted post-hoc for means comparison.
5.3 Results and discussion
5.3.1 Emulsion characterization
Droplet size was 214.3 + 0.7 and 440.5 + 15.5 nm for whey and pea-stabilized
emulsions, respectively. However, addition of 10% SDS resulted in a narrower droplet
size distribution for pea protein-stabilized emulsions (Figure 5-1), indicating that the
droplets existed in a flocculated state. Non-flocculated droplet sizes for both whey and
136
pea protein-stabilized emulsions were more similar in Z-average at 207.9 + 1.1 and 247.3
+ 0.4 nm, respectively. Zeta potential of the whey and pea-stabilized emulsions were 57.66 + 0.3 and -43.40 + 0.8 mV, respectively. Our previous work (Ho et al., 2016)
conducted on whey and pea protein-stabilized emulsions indicated similar zeta potential
values.
14
12
Whey
Volume (%)
10
Pea
8
Whey w/ SDS
6
PPI w/ SDS
4
2
0
0.1
10
Diameter (nm)
1000
Figure 5-1 Droplet size distributions of lycopene-loaded whey and pea protein emulsions
(solid lines) and diluted with 10% SDS (dashed lines) to estimate the size of nonflocculated droplets.
5.3.2 Lycopene content and digestive stability
All samples, emulsions (whey and pea protein), lycopene oil, and tomato paste
contained similar quantities of lycopene prior to simulated in vitro digestion.
Chromatograms indicated that the primary carotenoid in all samples was all-translycopene although a few smaller cis-lycopene peaks were observed (Figure 5-2).
137
Figure 5-2 Representative chromatograms of lycopene extracted from protein-stabilized
emulsions (whey or pea), lycopene oil, and tomato paste. Peaks for cis-isomers (a) and
all-trans-lycopene (b) are shown for all samples. The 5-cis-isomer can be observed as a
right shoulder off of the all-trans-lycopene peak.
Cis-lycopene was not significantly different amongst any samples, while alltrans-lycopene from the emulsions were not significantly different from the lycopene oil
nor the tomato paste controls (Figure 5-3A). Post digestion, the cis-lycopene content for
the tomato paste decreased and was significantly lower than that of the digested emulsion
systems and lycopene oil. Interestingly, trans-lycopene did not appear to decrease much
after digestion when stabilized in whey or pea protein emulsions (Figure 5-3B). These
values were significantly higher than that of the remaining lycopene in the lycopene oil
and the tomato paste after digestion. This indicates that all-trans-lycopene was more
stable from the emulsion systems compared to a non-emulsified control (lycopene oil)
and conventional food source (tomato paste).
138
B
A
Lycopene (nmol)
100
80
ab
60
40
a
a
a
ab
a
b
a
20
0
Cis
Trans
100
90
80
70
60
50
40
30
20
10
0
Whey
a a
Pea
b
LO
a a
a
c
TP
b
Cis
Trans
Figure 5-3 Lycopene content of emulsions (whey or pea protein-stabilized), lycopene oil
(LO), and tomato paste (TP) before (A) and after (B) digestion. Response values shown
represent the mean + SD (n=3), with same letters denoting values that are not
significantly different (α=0.05).
5.3.3 In vitro bioaccessibility of lycopene
Micellarization efficiency, which is the ratio of the lycopene in the aqueous
micellar fraction compared against the lycopene in the digesta fraction, was significantly
lower for pea protein emulsions compared to the non-emulsified lycopene oil (Figure 54). Overall, the controls (lycopene oil and tomato paste) had the highest relative
lycopene concentrations in the aqueous micellar fraction. Although these emulsion
systems were designed with the expectation of enhancing lycopene bioaccessibility,
interfacial stabilization with whey and pea proteins may have negatively impacted the
micellarization efficiency by limiting the release of lycopene from lipid droplets. The
purpose of digestion is to break down food so that compounds within the matrix become
accessible for absorption. In the case of the emulsions, the presence of proteases and
lipases are imperative for hydrolysis and consequently affect bioaccessibility via
digestion. Thus, to better understand the effects of the protein interfaces on
bioaccessibility, it is helpful to consider the digestibility of whey and pea proteins. In
139
bulk systems, Boirie et al. (1997) classified whey as a protein that is quickly digested
within the stomach while casein is slowly digestible due to coagulation. Postprandial
leucine oxidation was greater when whey was consumed compared to casein, despite
them having matched leucine content. Overduin, Guerin-Deremaux, Wils, & Lambers
(2015) compared the same pea protein isolate (NUTRALYS) and whey protein isolate
(BiPro) used in our study, but in relation to satiety and found that the pea protein acted as
an intermediate digestible protein compared to sodium caseinate, which maintained its
protein network, and whey protein isolate, which was quickly dissolved during digestion.
The authors hypothesized that the slower digestion of pea protein and casein allowed for
delayed bioavailability of protein in the small intestines compared to whey, which was
rapidly digested and oxidized in the stomach. In our study, differences in lycopene
oxidation were not observed between emulsion samples as the cis- and trans-lycopene
digesta quantities were not statistically different (Figure 5-3B). However, we cannot
assume that protein bulk behavior will be the same as behavior at the oil-water interface.
Digestion of adsorbed whey protein has been reported to be quicker and more extensive
compared to non-adsorbed proteins, possibly due to favorable conformational unfolding
at the interface that allows for better pepsin access (Macierzanka, Sancho, Mills, Rigby,
& Mackie, 2008; Sarkar, Goh, Singh, & Singh, 2009). Despite the rate and overall extent
of pepsinolysis being greater for interfacial proteins, SDS-PAGE indicated that a fraction
of adsorbed β-lactoglobulin likely remained partially folded and unavailable for
hydrolysis (Macierzanka et al., 2008). In our study, the low micellarization efficiencies
of the whey and pea protein emulsions may be a result of partially limited lycopene
release due to incomplete interfacial protein digestion. Yet, perhaps a more feasible
explanation relates to a potential complexation between hydrophobic pockets of
displaced/free protein and lycopene which could, in turn, have limited the
bioaccessibility. During the duodenal phase of digestion, interfacial proteins can be
almost completely displaced by bile salts (Maldonado-Valderrama et al., 2008) after,
which they can be digested by trypsin and chymotrypsin. However, if displaced proteins
were able to form complexes with lycopene, digestion could potentially be limited,
depending on the 3D structure and access to hydrolyzing enzymes. Pea protein is known
to exist in an aggregated or flocculated state (Shao & Tang, 2016) and although the pea
140
protein emulsions in this study exhibited reversible flocculation (Figure 5-1), it is not
clear how they behaved during this digestion process. One possibility is that pea proteinlycopene aggregates formed after bile salt displacement, may have been able to resist
complete digestion, and were too large to pass through the 0.22 μm filter used to separate
the aqueous micellar fraction, which would result in a lower micellarization efficiency.
Micellarization Efficiency (%)
90
a
80
70
ab
60
Whey
50
Pea
40
30
a
ab
LO
TP
20
10
ab
ab
b
b
0
Cis
Trans
Figure 5-4 Micellarization efficiency of lycopene from emulsions (whey or pea proteinstabilized), lycopene oil (LO), or tomato paste (TP). Response values shown represent
the mean + SD (n=3), with same letters denoting values that are not significantly different
(α=0.05).
The micellarizaiton efficiency of cis-lycopene appeared to be higher compared to
that of all-trans-lycopene (Figure 5-4). Other groups have found cis-lycopene to be more
bioaccessible. In their study on carotenoid bioaccessibility and cellular uptake from gac
fruit, Failla, Chitchumroonchokchai, & Ishida (2008) reported a higher micellarization
efficiency and cellular uptake for cis-isomers compared to trans-lycopene. Considering
that all-trans-lycopene is the dominant carotenoid before and after digestion in our
samples (Figure 5-3), the higher cis-lycopene micellarization is possibly due to a
preferential incorporation of cis- versus trans-isomers into mixed micelles. This may be
due to cis-isomers having bends in their chemical configuration, which decreases the
space they occupy and limits crystal formation (Boileau, Boileau, & Erdman, 2002;
George Britton, 1995) possibly enhancing their solubility.
141
5.3.4 Cellular uptake of lycopene
Lycopene uptake was not correlated with micellarization efficiency.
Interestingly, pea protein emulsions exhibited significantly greater cis and trans-lycopene
uptake efficiencies compared to all other samples (Figure 5-5A). Aside from the pea
emulsions, lycopene uptake (both cis- and trans-) was greatest for whey emulsions
followed by tomato paste and lycopene oil, although no significant differences were
observed amongst these three samples. Also, despite having lower micellarizaiton
efficiencies compared to the controls (Figure 5-4), both whey and pea protein emulsions
had absolute lycopene contents greater than lycopene oil and tomato paste (Figure 5-5B).
Similar to the trend observed for micellarization efficiency, cis-lycopene uptake appeared
to be greater than that of all-trans-lycopene. This parallels previous findings, which
observed greater cis-lycopene uptake in Caco-2 cells (Failla et al., 2008).
30
b
Lycopene uptake (%)
25
20
b
15
10
5
a
a
a
a
aa
Lycopene (picomole/μg protein)
B
A
0.08
a
Whey
0.06
Pea
ab
LO
a a
0.04
TP
b
b
a
0.02
a
0
0
Cis
Trans
Cis
Trans
Figure 5-5 Cellular uptake efficiency (A) and absolute accumulation (B) of cis- and
trans-lycopene after 3 hours of treatment with emulsions (whey or pea proteinstabilized), lycopene oil (LO), or tomato paste (TP). Response values shown represent
the mean + standard deviation (n=3), with same letters denoting values that are not
significantly different (α=0.05).
Carotenoid absorption is a complex process. The literature has shown evidence
supporting passive diffusion (Quick & Ong, 1990) and facilitated diffusion (Van
142
Bennekum et al., 2005), with the latter being applicable for non-pharmacological
carotenoid doses from food (Reboul & Emmanuelle, 2013). Lipkie et al. (2014) observed
that lutein uptake efficiency in Caco-2 cells was dose-dependent for human milk, but not
for formula. While the dose-independence of formula could be due to passive diffusion,
potentiators in human milk may have enhanced lutein uptake and suggests that
differences in food composition could affect carotenoid uptake efficiency. For lycopene
specifically, scavenger receptor protein SR-BI is believed to be the main membrane
protein responsible for uptake into the enterocyte (Van Bennekum et al., 2005). Yet,
studies that have utilized nanoemulsions or nanoparticle technology have suggested that
other mechanisms, such as endocytosis, passive diffusion, and paracellular transport (Lu
et al., 2016; McClements & Xiao, 2012) play a major role for cellular uptake when
encapsulation or delivery systems are utilized.
In this study, we did not test potential mechanisms for lycopene uptake, however
our findings suggest that differences in lycopene uptake efficiency occurred depending
on the food composition (i.e., emulsion vs. non-emulsified vs. tomato paste). Despite pea
protein emulsions having the lowest micellarization efficiency and the lowest lycopene
concentration in the aqueous fraction, they exhibited significantly greater relative cellular
uptake compared to all other samples (Figure 5-5). However, the absolute cis-lycopene
accumulation values (Figure 5-5B) for whey protein emulsions was the highest, although
not significantly greater than that of pea protein emulsions. The greater efficiency of
lycopene uptake with pea protein suggests that factors aside from lycopene concentration
affected lycopene accumulation in Caco-2 cells although it is not clear if this is via
facilitated action or via pea protein-related potentiators.
The aqueous fractions of digested whey and pea protein emulsions may have
comprised of not only mixed micelles, but also bound lycopene, which could be
associated with hydrophobic portions of non-adsorbed proteins (Lipkie et al., 2014;
Reboul & Borel, 2011). Whey protein has been classified as a lipocalin as it is able to
interact with hydrophobic compounds, such as retinol and β-carotene (Dufour & Haertle,
1991), and possibly enhance carotenoid delivery (Said, Ong, & Shingleton, 1989).
Lactolycopene, lycopene embedded within a β-lactoglobulin matrix, was shown to have
in vivo bioavailability in humans comparable to tomato paste, potentially due to the
143
reduced lycopene crystal size and/or the presence of the protein network (Richelle et al.,
2002). The effect of protein-carotenoid interactions on bioavailability may have been
limiting if the protein complexes were too large to be included in the aqueous fraction (as
discussed in Section 5.3.3). However, if a certain portion of protein-carotenoid
complexes exist in the aqueous micellar fraction, these could potentially have improved
mobility into the enterocyte. An analogous example of this is retinol binding protein,
which acts on retinol to 1) protect it from oxidation, 2) transport it from storage sites to
tissue, and 3) mediate uptake in cells via specific protein receptor (Blomhoff, Green,
Berg, & Norum, 1990). Although differing by ~70% in protein residues, β-lactoglobulin
reportedly behaves similarly as a lipocalin to retinol binding protein due to similar
tertiary structure (Godovac-Zimmermann, Conti, Liberatori, & Braunitzer, 1985). Mensi
et al. (2014) found that inhibition of SR-BI protein in Caco-2 cells limited β-carotene
uptake from mixed micelles, but not from β-lactoglobulin complexes. This suggests that
carotenoid-loaded mixed micelles are dependent on passive diffusion, while proteins in
the emulsion systems may have enhanced uptake via unspecified mechanisms. However,
further work would be needed to determine if the whey and pea proteins in this study
behaved as lipocalins for lycopene.
5.4 Conclusions
Overall significant differences in lycopene micellarization efficiency and cellular
uptake were observed amongst the emulsion samples (whey and pea protein-stabilized)
and when compared against the controls (lycopene oil and tomato paste). These findings
suggest that lycopene bioavailability is not solely dependent on micellarization
efficiency. Other factors, such as digestibility and interfacial protein behavior during
gastric and intestinal phases, may have had an effect on lycopene transfer.
Mechanistically, it is unclear how lycopene from protein-stabilized emulsions enters the
enterocyte, although some sort of facilitated protein process is potentially involved in the
presence of added protein. Overall the findings of this study furnish several opportunities
for further investigation. Aside from the unanswered questions, the results did indicate
that the type of interfacial protein will affect the in vitro bioaccessibility of lycopene and
that protein-stabilized emulsions can potentially enhance bioavailability of lipophilic
144
compounds compared to a non-emulsified (lycopene oil) and a conventional food control
(tomato paste).
5.5 Acknowledgements
The authors would like to thank Dr. Jozef Kokini for allowing us to use his
Zetasizer to characterize our emulsions and Dr. Kee-Hong Kim for allowing us to store
our samples in his -80°C freezer. In vitro digestions and cell work was conducted at the
North Carolina State University Plants for Human Health Institute, while emulsion
fabrication and characterization and carotenoid extraction and analysis was done at
Purdue University in the Department of Food Science.
145
5.6 References
Ax, K., Mayer-Miebach, E., Link, B., Schuchmann, H., & Schubert, H. (2003). Stability
of lycopene in oil-in-water emulsions. Engineering in Life Sciences, 3(4), 199–201.
http://doi.org/10.1002/elsc.200390028
Blomhoff, R., Green, M. H., Berg, T., & Norum, K. R. (1990). Transport and storage of
vitamin A. Science, 250, 399–404.
Boileau, T. W.-M., Boileau, A. C., & Erdman, J. W. (2002). Bioavailability of all-trans
and cis-isomers of lycopene. Experimental Biology and Medicine, 227(10), 914–
919.
Boirie, Y., Dangin, M., Gachon, P., Vasson, M. P., Maubois, J. L., & Beaufrère, B.
(1997). Slow and fast dietary proteins differently modulate postprandial protein
accretion. Proceedings of the National Academy of Sciences of the United States of
America, 94(26), 14930–5. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/9405716
Britton, G. (1995). Structure and properties of carotenoids in relation to function. The
FASEB Journal, 9(15), 1551–1558.
Britton, G., Liaaen-Jensen, S., & Pfander, H. (2004). Carotenoid Handbook Info.pdf.
Basel, Switzerland: Birkhauser Verlag.
Brown, M. J., Ferruzzi, M. G., Nguyen, M. L., Cooper, D. A., Eldridge, A. L., Schwartz,
S. J., & White, W. S. (2004). Carotenoid bioavailability is higher from salads
ingested with full-fat than with fat-reduced salad dressings as measured with
electrochemical detection. American Journal of Clinical Nutrition, 80(2), 396–403.
http://doi.org/80/2/396 [pii]
Cornacchia, L., & Roos, Y. H. (2011). Stability of β-Carotene in Protein-Stabilized Oilin-Water Delivery Systems. Journal of Agricultural and Food Chemistry, 59(13),
7013–7020. http://doi.org/10.1021/jf200841k
Desai, M. P., Labhasetwar, V., Walter, E., Levy, R. J., & Amidon, G. L. (1997). The
mechanism of uptake of biodegradable microparticles in Caco-2 cells is size
dependent. Pharmaceutical Research, 14(11), 1568–1573.
http://doi.org/10.1023/A:1012126301290
146
Dufour, E., & Haertle, T. (1991). Binding of retinoids and β-carotene to β-lactoglobulin.
Influence of protein modifications. Biochimica et Biophvsica Acta, 1079(3), 316–
320. Retrieved from http://ac.elscdn.com.ezproxy.lib.purdue.edu/016748389190075B/1-s2.0-016748389190075Bmain.pdf?_tid=cec41e2a-5cfc-11e7-ad6c00000aacb360&acdnat=1498762833_ae04321badcc2ff9decb197ef3f666e2
Failla, M. L., Chitchumroonchokchai, C., & Ishida, B. K. (2008). In Vitro Micellarization
and Intestinal Cell Uptake of cis Isomers of Lycopene Exceed Those of All-trans
Lycopene. The Journal of Nutrition, 138(3), 482–486. Retrieved from
http://jn.nutrition.org/content/138/3/482.abstract
Ferruzzi, M. G., Lumpkin, J. L., Schwartz, S. J., & Failla, M. (2006). Digestive Stability,
Micellarization, and Uptake of β-Carotene Isomers by Caco-2 Human Intestinal
Cells. Journal of Agricultural and Food Chemistry, 54(7), 2780–2785.
http://doi.org/10.1021/jf0530603
Garrett, D. A., Failla, M. L., & Sarama, R. J. (1999). Development of an in Vitro
Digestion Method To Assess Carotenoid Bioavailability from Meals. Journal of
Agricultural and Food Chemistry, 47(10), 4301–4309.
http://doi.org/10.1021/jf9903298
Godovac-Zimmermann, J., Conti, A., Liberatori, J., & Braunitzer, G. (1985). Homology
between the Primary Structures of ß-Lactoglobulins and Human Retinol-Binding
Protein: Evidence for a Similar Biological Function? Biol. Chem. Hoppe-Seyler,
366, 431–434. Retrieved from https://www-degruytercom.ezproxy.lib.purdue.edu/downloadpdf/j/bchm3.1985.366.issue1/bchm3.1985.366.1.431/bchm3.1985.366.1.431.pdf
Ho, K. K. H. Y., Ferruzzi, M. G., Liceaga, A. M., & San Martín-González, M. F. (2015).
Microwave-assisted extraction of lycopene in tomato peels: Effect of extraction
conditions on all-trans and cis-isomer yields. LWT - Food Science and Technology,
62(1), 160–168. http://doi.org/http://dx.doi.org/10.1016/j.lwt.2014.12.061
147
Ho, K. K. H. Y., Schroën, K., San Martín-González, M. F., & Berton-Carabin, C. C.
(2016). Physicochemical stability of lycopene-loaded emulsions stabilized by plant
or dairy proteins. Food Structure, 12, 34–42.
http://doi.org/http://dx.doi.org/10.1016/j.foostr.2016.12.001
Kean, E. G., Hamaker, B. R., & Ferruzzi, M. G. (2008). Carotenoid Bioaccessibility from
Whole Grain and Degermed Maize Meal Products. Journal of Agricultural and
Food Chemistry, 56(21), 9918–9926. http://doi.org/10.1021/jf8018613
Lipkie, T. E., Banavara, D., Shah, B., Morrow, A. L., McMahon, R. J., Jouni, Z. E., &
Ferruzzi, M. G. (2014). Caco-2 accumulation of lutein is greater from human milk
than from infant formula despite similar bioaccessibility. Molecular Nutrition &
Food Research, 58(10), 2014–2022. http://doi.org/10.1002/mnfr.201400126
Lu, W., Kelly, A. L., Maguire, P., Zhang, H., Stanton, C., & Miao, S. (2016). Correlation
of Emulsion Structure with Cellular Uptake Behavior of Encapsulated Bioactive
Nutrients: Influence of Droplet Size and Interfacial Structure. Journal of
Agricultural and Food Chemistry, 64(45), 8659–8666.
http://doi.org/10.1021/acs.jafc.6b04136
Macierzanka, A., Sancho, A. I., Mills, E. N. C., Rigby, N. M., & Mackie, A. R. (2008).
Emulsification alters simulated gastrointestinal proteolysis of β-casein and βlactoglobulin. Soft Matter, 5, 538–550. http://doi.org/10.1039/b811233a
Maldonado-Valderrama, J., Woodward, N. C., Gunning, A. P., Ridout, M. J., Husband, F.
A., Mackie, A. R., … Wilde, P. J. (2008). Interfacial Characterization of βLactoglobulin Networks: Displacement by Bile Salts. Langmuir, 24(13), 6759–6767.
http://doi.org/10.1021/la800551u
McClements, D. J., & Xiao, H. (2012). Potential biological fate of ingested
nanoemulsions: influence of particle characteristics. Food & Function, 3(3), 202–
220. http://doi.org/10.1039/C1FO10193E
Mensi, A., Borel, P., Goncalves, A., Nowicki, M., Gleize, B., Roi, S., … Reboul, E.
(2014). β-Lactoglobulin as a Vector for β-Carotene Food Fortification. Journal of
Agricultural and Food Chemistry, 62(25), 5916–5924.
http://doi.org/10.1021/jf501683s
148
Neilson, A. P., Song, B. J., Sapper, T. N., Bomser, J. A., & Ferruzzi, M. G. (2010). Tea
catechin auto-oxidation dimers are accumulated and retained by Caco-2 human
intestinal cells. Nutrition Research, 30(5), 327–340.
http://doi.org/10.1016/j.nutres.2010.05.006
Overduin, J., Guerin-Deremaux, L., Wils, D., & Lambers, T. T. (2015). NUTRALYS pea
protein: characterization of in vitro gastric digestion and in vivo gastrointestinal
peptide responses relevant to satiety. Food & Nutrition Research, 59(1), 25622.
http://doi.org/10.3402/fnr.v59.25622
Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2012). Physical and chemical
stability of β-carotene-enriched nanoemulsions: Influence of pH, ionic strength,
temperature, and emulsifier type. Food Chemistry, 132(3), 1221–1229.
http://doi.org/http://dx.doi.org/10.1016/j.foodchem.2011.11.091
Quick, T. C., & Ong, D. E. (1990). Vitamin A metabolism in the human intestinal Caco-2
cell line. Biochemistry, 29(50), 11116–11123. http://doi.org/10.1021/bi00502a015
Reboul, E., & Borel, P. (2011). Proteins involved in uptake, intracellular transport and
basolateral secretion of fat-soluble vitamins and carotenoids by mammalian
enterocytes. Progress in Lipid Research, 50(4), 388–402.
http://doi.org/10.1016/j.plipres.2011.07.001
Reboul, E., & Emmanuelle. (2013). Absorption of Vitamin A and Carotenoids by the
Enterocyte: Focus on Transport Proteins. Nutrients, 5(9), 3563–3581.
http://doi.org/10.3390/nu5093563
Redan, B. W., Chegeni, M., Ferruzzi, M. G., Tresserra-Rimbau, A., Rimm, E. B.,
Medina-Remón, A., … Bai, H.-W. (2016). Differentiated Caco-2 cell monolayers
exhibit adaptation in the transport and metabolism of flavan-3-ols with chronic
exposure to both isolated flavan-3-ols and enriched extracts. Food Funct., 24, 639–
647. http://doi.org/10.1039/C6FO01289B
Ribeiro, H. S., Guerrero, J. M. M., Briviba, K., Rechkemmer, G., Schuchmann, H. P., &
Schubert, H. (2006). Cellular Uptake of Carotenoid-Loaded Oil-in-Water Emulsions
in Colon Carcinoma Cells in Vitro. Journal of Agricultural and Food Chemistry,
54(25), 9366–9369. http://doi.org/10.1021/jf062409z
149
Richelle, M., Bortlik, K., Liardet, S., Hager, C., Lambelet, P., Baur, M., … Offord, E. A.
(2002). A food-based formulation provides lycopene with the same bioavailability to
humans as that from tomato paste. The Journal of Nutrition, 132(3), 404–8.
Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11880563
Said, H. M., Ong, D. E., & Shingleton, J. L. (1989). Intestinal uptake of retinol:
enhancement by bovine milk fl-lactoglobulin13. American Journal of Clinical
Nutrition, 49, 690–694. Retrieved from
https://www.researchgate.net/profile/David_Ong2/publication/20225748_Intestinal_
uptake_of_retinol_Enhancement_by_bovine_milk_lactoglobulin/links/0fcfd5096fe21b90d4000000.pdf
Sarkar, A., Goh, K. K. T., Singh, R. P., & Singh, H. (2009). Behaviour of an oil-in-water
emulsion stabilized by β-lactoglobulin in an in vitro gastric model. Food
Hydrocolloids, 23, 1563–1569. http://doi.org/10.1016/j.foodhyd.2008.10.014
Shao, Y., & Tang, C.-H. (2016). Gel-like pea protein Pickering emulsions at pH3.0 as a
potential intestine-targeted and sustained-release delivery system for β-carotene.
Food Research International, 79, 64–72.
http://doi.org/10.1016/j.foodres.2015.11.025
Unlu, N. Z., Bohn, T., Clinton, S. K., & Schwartz, S. J. (2005). Carotenoid absorption
from salad and salsa by humans is enhanced by the addition of avocado or avocado
oil. The Journal of Nutrition, 135(3), 431–436. http://doi.org/135/3/431 [pii]
Van Bennekum, A., Werder, M., Thuahnai, S. T., Han, C.-H., Duong, P., Williams, D. L.,
… Hauser, H. (2005). Class B Scavenger Receptor-Mediated Intestinal Absorption
of Dietary -Carotene and Cholesterol. Biochemistry, 44, 4517–4525.
http://doi.org/10.1021/bi0484320
Yi, J., Lam, T. I., Yokoyama, W., Cheng, L. W., & Zhong, F. (2014). Cellular Uptake of
β-Carotene from Protein Stabilized Solid Lipid Nanoparticles Prepared by
Homogenization–Evaporation Method. Journal of Agricultural and Food Chemistry,
62(5), 1096–1104. http://doi.org/10.1021/jf404073c
150
CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS
6.1 Summary and overall conclusions
Extraction and encapsulation strategies have been investigated for various
phytochemicals. While lycopene recovery from tomato peels had been investigated,
limited work had been done with microwave-assisted extraction, likely due to the
challenging and otherwise incompatible solubility with popular dielectric solvents.
Knowledge gaps also existed with regards to encapsulation and delivery of carotenoids in
emulsion-based systems. Although β-carotene emulsions are well studied, lycopene
delivery is relatively less explored. On a broader sense, the connection between
interfacial rheology and bioactive encapsulation is not well studied. Overall, the aims of
this dissertation research were achieved and the findings contributed insight to the less
understood corners of this research area.
In Chpater 2, microwave-assisted extraction (MAE) was a more efficient process
for recovering lycopene from tomato peels compared to a conventional solvent
extraction. Electron microscopy images suggested that MAE disrupted plant matter to a
greater extent compared to conventional extraction. Overall, the solvent ratio and
microwave power (in relation to energy equivalents) significantly affected all-translycopene yields while the cis-isomer recovery was primarily affected by the solid-liquid
ratio and solvent ratio. Generally speaking, these findings collectively showed the
potential for MAE as an efficient technology to recover high-value compounds from a
processing byproduct.
In Chapter 3, a comparison of the physical and chemical stability of dairy (whey
protein isolate and sodium caseinate) and plant (soy and pea protein isolate) indicated
that casein-stabilized emulsions were superior (but not significantly better than plant
proteins) in regards to lycopene retention compared to all other protein-stabilized
emulsions. While these findings indicated that plant proteins, particularly pea proteins,
could be used as an alternative to animal-derived proteins, this also indicated that there is
room for improvement for plant-based approaches.
151
Chapter 4 indicated that synergistic effects of physicochemical stability were
observed when the globular proteins were blended (dairy with plant protein or plant with
plant protein). Interfacial rheology suggested that complex interfacial structures existed
for some blends, while others appeared to perform similarly to one protein component.
Taken together, Chapters 3 and 4 demonstrate the potential for plant-derived proteins to
be used as emulsifiers for lycopene delivery.
The Chapter 5 findings were interesting in that lycopene micellarization
efficiency was not a predictor of relative cellular uptake. Overall, pea protein-stabilized
emulsions exhibited a significantly higher lycopene uptake efficiency in Caco-2 cells
compared to all other samples (whey protein emulsions, lycopene oil, and tomato paste).
Similar to findings elsewhere (Failla, Chitchumroonchokchai, & Ishida, 2008), cislycopene appeared to be more bioaccessible and more efficient at transferring into the
cell. Complementing the findings from Chapters 3 and 4, the digestive stability findings
suggested that the whey and pea protein emulsions were more effective at maintaining
cis- and trans-lycopene levels compared to the controls.
6.2 Future directions
One of the biggest challenges with implementing MAE at the industrial
scale is the scale-up due to technical design challenges (e.g., non-uniform heating,
operator safety, and the need for specialized equipment. However, the high capital costs
of equipment investment may be balanced if the targeted extractable compounds can be
used as high-value ingredients. Additionally, evidence suggests that optimizing factors,
such as balancing the sample thickness and applied frequency, can achieve deeper
microwave penetration depths to make the process more efficient (Ciriminna et al.,
2016). Aside from manufacturing challenges, there are inherent challenges with
extracting lycopene and other carotenoids. Due to the extreme hydrophobicity of
lycopene, MAE presents some challenges as an extraction method. Although the findings
from Chapter 2 demonstrate that there is a significant decrease in processing time and an
added effect of breaking down cellular matter and barriers to better facilitate extraction,
choices of soluble solvents are limited. The solvents used in Chapter 2, hexane and ethyl
acetate, can be used for specific food applications with certain constraints on residual
152
concentrations (21CFR173.228; 21CFR173.270). Other compounds, such as polyphenols
and essential oils, have been extracted using a solvent-free microwave technique (Sahin
et al., 2017; Wang et al., 2006). Thus, a valuable avenue for future work on carotenoid
MAE could focus on limiting the use of toxic solvents while optimizing yields.
The results from Chapters 3 and 4 are promising as the food industry could
benefit from having more alternatives to dairy protein ingredients. From a sustainability
perspective, it is unclear how significant the benefit of partially reducing dairy protein
usage in protein blends could be. However, the synergistic enhancement of protein
functionality when whey protein was combined with soy or dairy protein suggests that
protein-protein interactions could be forming and consequently affecting emulsion
performance. Future work could explore these potential interactions. Deeper
characterization of the interfacial proteins could convey information relating to the
surface hydrophobicity of adsorbed proteins, potential oxygen permeability of films, or
the antioxidant effects of the proteins and protein blends in the presence of lycopene.
The findings from Chapter 5 were undoubtedly interesting. Although hypotheses
can be drawn regarding lycopene uptake, a governing mechanism is currently unclear and
should be further explored. Coupling the in vitro digestion with particle size analysis,
microscopy, and zeta potential measurements would provide information about the
interfacial, size, and morphological changes of the emulsion droplets and could be used
to assess changes after each step of digestion or just prior to the hypothetical intestinal
absorption. Additionally, membrane protein function could be blocked to determine if
differences in cellular uptake are observed for passive or facilitated diffusion.
Considering that M-cells, a drastically less prevalent cell type in the small intestine, are
more efficient at absorbing compounds compared to epithelial cells, it may also be
interesting to consider if the interfacial protein has an effect on lycopene or carotenoid
uptake in this cell type. In the future, animal models could also be utilized, but care
should be taken when interpreting the data, particularly with rodent models, which have
higher proportions of M-cells compared to humans (Ermund, Gustafsson, Hansson, Keita,
& Konjufca, 2013).
153
6.3 References
21CFR173.228. (2016). Ethyl acetate. Retrieved from
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=173.22
8
21CFR173.270. (2016). Hexane. Retrieved from
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=173.27
0
Ciriminna, R., Carnaroglio, D., Delisi, R., Arvati, S., Tamburino, A., & Pagliaro, M.
(2016). Industrial Feasibility of Natural Products Extraction with Microwave
Technology. Sustainable Chemistry, 3, 549–555.
http://doi.org/10.1002/slct.201600075
Ermund, A., Gustafsson, J. K., Hansson, G. C., Keita, Å. V., & Konjufca, V. (2013).
Mucus Properties and Goblet Cell Quantification in Mouse, Rat and Human Ileal
Peyer’s Patches. PLoS ONE, 8(12), e83688.
http://doi.org/10.1371/journal.pone.0083688
Failla, M. L., Chitchumroonchokchai, C., & Ishida, B. K. (2008). In Vitro Micellarization
and Intestinal Cell Uptake of cis Isomers of Lycopene Exceed Those of All-trans
Lycopene. The Journal of Nutrition, 138(3), 482–486. Retrieved from
http://jn.nutrition.org/content/138/3/482.abstract
Sahin, S., Samli, R., Tan, A. S. B., Barba, F. J., Chemat, F., Cravotto, G., & Lorenzo, J.
M. (2017). Solvent-Free Microwave-Assisted Extraction of Polyphenols from Olive
Tree Leaves: Antioxidant and Antimicrobial Properties. Molecules, 22(7), 1056.
http://doi.org/10.3390/molecules22071056
Wang, Z., Ding, L., Li, T., Zhou, X., Wang, L., Zhang, H., … He, H. (2006). Improved
solvent-free microwave extraction of essential oil from dried Cuminum cyminum L.
and Zanthoxylum bungeanum Maxim. Journal of Chromatography A, 1102(1–2),
11–17. http://doi.org/http://dx.doi.org/10.1016/j.chroma.2005.10.032
154
APPENDIX A SUPPLEMENTARY TABLES
Table A- 1 Treatment times (seconds) for MAE of tomato peels based off of power and energy equivalents
Energy
equivalents
(kJ)
Power (W)
400
800
1600
24
60 s
30 s
15 s
36
90 s
45 s
22.5 s*
48
120 s
60 s
30 s
Treatment times were calculated (Power (W)= Energy (J) /Time (s)) based off of the specified power and energy equivalents listed in
Table 2-1.
*Treatment time was rounded to 23 seconds since MAE system could not accommodate for time increments < 1 s.
154
155
Table A- 2 Particle size, d3,2 of lycopene-loaded emulsions over time.
Sample
d3,2 (µm)
Day 0
Day 3
Day 7
Day 14
1:1 SPI-WPI
0.21 +
0.01a
0.21 + 0.01ab
0.22 +
0.01ab
0.25 +
0.01a
1:1 PPI-WPI
0.20 + 0.00ab
0.20 + 0.00ab
0.20 +
0.00bc
0.21 +
0.01a
1:1 SPI-SC
0.19 +
0.01bc
0.21 + 0.02ab
0.21 + 0.02abc
0.29 +
0.05a
1:1 PPI-SC
0.18 +
0.00c
0.19 +
0.01b
0.19 + 0.02abc
1.17 +
0.97b
3:1 SPI-WPI
0.21 +
0.00a
0.23 +
0.01a
0.24 +
0.01a
0.23 +
0.01a
3:1 PPI-WPI
0.18 +
0.00c
0.20 +
0.02b
0.19 +
0.01c
0.20 +
0.01a
0.20 + 0.01ab
0.20 +
0.01b
0.21 +
0.01bc
0.21 +
0.00a
1:1 SPI-PPI
155
156
Table A- 3 Span of lycopene-loaded emulsions overtime
Sample
Span
Day 0
Day 3
Day 7
Day 14
1:1 SPI-WPI
2.94 +
0.21b
2.71 + 0.27cd
2.44 +
0.21bcd
2.50 + 0.22a
1:1 PPI-WPI
2.21 + 0.00cd
2.25 + 0.03de
2.26 +
0.02cd
2.17 + 0.08b
1:1 SPI-SC
3.16 + 0.14ab
3.00 + 0.18bc
3.42 +
0.11ab
11.56 + 5.78b
1:1 PPI-SC
2.45 +
0.08c
2.43 + 0.02de
3.12 +
1.03abc
3.37 + 0.77b
3:1 SPI-WPI
3.08 + 0.16ab
3.34 + 0.07ab
3.40 +
0.35ab
2.30 + 0.31b
3:1 PPI-WPI
2.22 + 0.09cd
2.10 +
0.12e
2.17 +
0.04cd
2.12 + 0.04b
1:1 SPI-PPI
2.20 + 0.04cd
2.16 + 0.07de
2.14 +
1.51d
2.16 + 1.66b
156
157
Table A- 4 Encapsulation efficiency (%) of lycopene-loaded emulsions fabricated with protein and protein blends.
Sample
Encapsulation
efficiency (%)
WPI
58.28
+
0.02 abc
SC
60.29
+
0.01 ab
SPI
54.71
+
0.04 abcd
PPI
63.73
+
0.02 a
1:1 SPI-WPI
61.23
+
0.09 ab
1:1 PPI-WPI
58.24
+
0.03 abc
3:1 SPI-WPI
53.03
+
0.03 d
3:1 PPI-WPI
54.75
+
0.03 abcd
1:1 SPI-SC
53.03
+
0.01 bcd
1:1 PPI-SC
57.05
+
0.01bcd
1:1 SPI-PPI
47.91
+
0.01 cd
157
158
Table A- 5 Lycopene retention in emulsions stabilized with SC-blends
Sample
Lycopene Retention (%)
Day 3
Day 7
Day 14
1:1 SPISC
82.46 +
16.55
Not measurable
Not measurable
73.80 +
2.92
Not measurable
Not measurable
1:1 PPISC
158
159
APPENDIX B SUPPLEMENTARY FIGURES
Figure B- 1 Determination of optimal protein concentration. Particle size (left y-axis)
and correlating percent of excess protein (right y-axis) versus protein concentration added
to the emulsion for WPI (A), SC (B), SPI (C), and PPI (D). Dashed line denotes the
selected protein concentration.
160
Figure B- 2 Span of lycopene-loaded emulsions over time. Response values shown
represent the mean + SD (n=3), with same letters denoting values that are not
significantly different (α=0.05).
WPI
Zeta Potential (mV)
0
SC
SPI
0
PPI
Time (Days)
7
3
14
-10
-20
-30
-40
-50
-60
c
ab
a
b
b
a
a
a
a
b
a
a
b
a
a
a
-70
Figure B- 3 Initial zeta potential of lycopene-loaded emulsions fabricated with proteins
and protein blends. Response values shown represent the mean + SD (n=3), with same
letters denoting values that are not significantly different (α=0.05).
161
Encapsulation Efficiency (%)
70
60
ab
ab
WPI
SC
b
a
50
40
30
20
10
0
SPI
PPI
Figure B- 4 Encapsulation efficiency of lycopene in protein stabilized emulsions at t=14
days. Response values shown represent the mean + SD (n=3), with same letters denoting
values that are not significantly different (α=0.05).
A
Lycopene uptake (%)
a
4
a
2
6
Lycopene uptake (%)
B
6
Whey
4
Pea
a
2
b
b
c
0
0
2
Time (hr)
4
6
a
b
b
c
0
0
2
TP
4
6
Time (hr)
Figure B- 5 Cis- (A) and all-trans-lycopene (B) uptake from emulsions and tomato paste
overtime. Response values shown represent the mean + SD (n=4), with same letters
denoting values that are not significantly different (α=0.05).
162
APPENDIX C PUBLISHED ABSTRACTS
Ho, K. K. H. Y., Ferruzzi, M. G., Liceaga, A. M., & San Martín-González, M. F. (2015).
Microwave-assisted extraction of lycopene in tomato peels: Effect of extraction
conditions on all-trans and cis-isomer yields. LWT-Food Science and Technology, 62(1),
160-168.
Lycopene is the primary carotenoid in tomato peels, a processing byproduct, and
can be used as a natural color or bioactive ingredient. Unfortunately, extractions are
inefficient as lycopene is extremely nonpolar and susceptible to degradation. As a rapid
technique, microwave-assisted extraction (MAE) potentially offers efficient lycopene
recovery. Thus, the objectives of this research were to: 1) optimize MAE of lycopene
from tomato peels and 2) evaluate the effect of treatment on all-trans and isomer yields.
Response surface methodology (RSM) was employed to optimize lycopene extraction
with solvent ratio solid-liquid ratios, microwave power, and delivered as factors. High
performance liquid chromatography with a diode array detector (HPLC-DAD) was used
for isomer separation and quantification. Optimum MAE conditions were determined as:
0:10 solvent ratio at 400 W with a yield of 13.592 mg/100 g of extracted all-translycopene. RSM suggested that ethyl acetate was a better MAE solvent for lycopene
recovery as compared to hexane, which overall extracted less lycopene. HPLC-DAD
indicated that MAE significantly improved all-trans and total lycopene yields, while
conventional extraction demonstrated higher proportions of cis-isomer yields.
Additionally, electron micrographs showed that significant structural disruption occurred
in MAE-treated samples, possibly allowing for the improved lycopene extraction.
KEYWORDS: all-trans-lycopene, cis-isomers, microwave-assisted extraction, response
surface methodology
163
Ho, K. K.H.Y., Schroën, K., San Martín-González, M. F., & Berton-Carabin, C. C.
(2016). Physicochemical stability of lycopene-loaded emulsions stabilized by plant or
dairy proteins. Food Structure. 12: 34-42
Lycopene is a lipophilic bioactive compound that has many health benefits but
can be challenging to deliver in vivo. To mediate this, delivery strategies should be
developed, and protein-stabilized oil-in-water (O/W) emulsions have been suggested to
improve the physicochemical stability, bioaccessibility and bioavailability of lycopene. In
this research different proteins were compared to determine their impact on the physical
stability (droplet size, charge, interfacial rheology) and lycopene retention in canola O/W
emulsions. Two were of dairy (whey protein isolate, sodium caseinate) and two of plant
(soy and pea protein isolate) origin; plant proteins being of interest due to their wider
availability, reduced cost, and lower impact on the environment compared to dairy
proteins.
Particle size distribution for sodium caseinate and pea protein-stabilized
emulsions remained unchanged after 14 days of refrigerated storage, while whey and soy
protein isolate-stabilized emulsions became unstable. The droplet charge was largely
negative (~ -45 – -60 mV) for all emulsions and the lycopene concentration in plant
164
protein-stabilized emulsions at 14 days of storage was similar to that in sodium caseinatestabilized emulsions, but significantly higher than that in whey protein-stabilized
emulsions. While sodium caseinate formed relatively viscous films at the oil-water
interface, the other proteins showed more viscoelastic behaviour. In spite of this
difference, both the caseinate and pea protein stabilized emulsions were promising
delivery vehicles. This also indicates that plant-derived proteins can be feasible
alternatives to dairy emulsifiers.
KEYWORDS: Emulsions, plant proteins, dairy proteins, lycopene encapsulation,
physicochemical stability, interfacial rheology
165
VITA
Kacie K.H.Y. Ho
EDUCATION
•
Ph.D., Food Science
Purdue University, West Lafayette, IN, USA
Aug. 2012-Aug. 2017
Major Professor: Dr. M. Fernanda San Martín-Gonzalez
Dissertation Research: Microwave-assisted extraction of tomato peels and
physicochemical stability, in vitro bioaccessibility, and cellular uptake of lycopeneloaded emulsions
•
Bachelor of Science, Food Science & Human Nutrition
University of Hawaii at Manoa, Honolulu, HI
Magna Cum Laude
Aug. 2008-May 2012
•
Study Abroad, Study Abroad Program
International College of Seville, Seville, Spain
Sept. 2011-Dec. 2011
AWARDED FELLOWSHIPS
National Science Foundation Graduate Research Fellowship
Purdue University Doctoral Fellowship
Purdue Food Science Industry Fellows Program
Awarded: Sept. 2013
Awarded: Aug. 2012
Awarded: Aug. 2012
SCHOLARLSHIPS AND GRANTS
Graduate Research Opportunities Worldwide Grant
Phi Tau Sigma Student Achievement Scholarship
Purdue Graduate Student Government Travel Grant
Awarded: Apr. 2015
Awarded: Feb. 2015
Awarded: Jun. 2014, Apr. 2017
TEACHING HONORS
•
Graduate Instructional Development Certificate
Center for Instructional Excellence
Awarded: Spring 2017
166
•
Graduate Teaching Award
Teaching Academy, Center for Instructional Excellence
•
B.J. Liska Outstanding Teaching Assistant Award
Purdue Department of Food Science
Awarded: Spring 2017
Awarded: Fall 2016
ADDITIONAL RESEARCH EXPERIENCE
•
Visiting Research Scientist
Aug. 2015-Apr. 2016
Wageningen University, Wageningen, The Netherlands
Supervisors: Dr. Claire C. Berton-Carabin and Dr. Karin Schroën
Project: Effect of interfacial protein composition on physicochemical stability of
lycopene-loaded emulsions
Supervised MSc thesis research for student Jettie Faber
Project: Stability of lycopene-loaded solid lipid nanoparticles
•
Research Assistant for Food Processing Laboratory,
University of Hawaii, Honolulu, HI, USA
Supervisor: Dr. Soojin Jun
May 2011-May 2012
Tasks: Prepared and executed experiments related to combination ohmic-microwave
heating, FTIR spectroscopy for coffee authenticity, and used AutoCAD software to
design figures for COMSOL simulations
EMPLOYMENT
•
Global Food Research Intern
Cargill, Minneapolis, MN, USA
•
Agricultural Research Service Intern
Jun. 2013-Aug. 2010
Nutrient Data Lab Beltsville, US Department of Agriculture, MD, USA
•
Product Development Intern
Hawaiian Host, Inc., Honolulu, HI, USA
May 2013 -Aug. 2013
May 2009-Aug. 2009
TEACHING EXPERIENCE
•
Guest Lecturer for Food Analysis Lecture (FS 467)
Purdue University
Instructor: Dr. Suzanne Nielsen
•
Guest Lecturer for Introduction to Food Processing Lecture (FS 162) Mar. 29, 2017
Mar. 27 & 29 2017
167
Purdue University
Instructor: Simran Kaur
•
Guest Lecturer for Food Processing II Lecture (FS 442)
Purdue University
Instructor: Dr. M. Fernanda San Martín-Gonzalez
•
Teaching Assistant
Food Processing II Laboratory (FS 447), Purdue University
Section 001: 16 Students; Section 002: 12 Students
Instructor: Dr. M. Fernanda San Martín-Gonzalez
•
Teaching Assistant
Sept. 2011-Dec. 2011
Cultural Aspects of Food Habits (FSHN 476), University of Hawaii
Discussion-based, writing intensive course: 8 Students
Instructor: Dr. Wayne Iwaoka
Oct. 4 & 27, 2016
Aug. 2016- Dec. 2016
PRESENTATIONS
•
Ho, Kacie, Schroën, Karin, San Martín-Gonzalez, M. Fernanda, Berton-Carabin
Claire C. 2017. Plant proteins can partly replace dairy proteins in lycopene-loaded
emulsions to enhance physiochemical stability. Oral presentation for the 2017
American Oil Chemists’ Society. Orlando, FL, USA. May 2, 2017.
•
Ho, Kacie, San Martín-Gonzalez, M. Fernanda, Schroën, Karin, Berton-Carabin,
Claire C. 2016. Stability of lycopene-loaded emulsions: Effect of dairy and plant
proteins at the interface. Oral presentation for the 2016 Food Colloids Conference.
Hotel de Wageningsche Berg, Wageningen, Netherlands. April 11, 2016.
•
Ho, Kacie, San Martín-Gonzalez, M. Fernanda, Schroën, Karin, Berton-Carabin,
Claire C. 2015. Stability of lycopene-loaded emulsions: Effect of interface structure
and lipid physical state. Poster presentation for the 2015 Food Process Engineering
Group Day. Wageningen University, Wageningen, Netherlands. October 22, 2015.
Poster was awarded 1st Place for Best Overall Presentation (judged by students and
staff) and Favorite Poster (selected by B.Sc. and M.Sc. Students) out of 11 posters.
•
Ho, Kacie, Ferruzzi, Mario G., Liceaga, Andrea M., San Martín-Gonzalez, M.
Fernanda. Microwave-assisted extraction of all-trans-lycopene and effect of
extraction conditions on isomerization. Poster presentation for the 2014 Journey
Through Science Day with PepsiCo and the Science Alliance. New York Academy of
Sciences, New York, NY, USA. December 8, 2014.
Poster received one of three Excellence in Research Awards out of 50 invited
participants.
•
Ho, Kacie, Ferruzzi, Mario G., Liceaga, Andrea M., San Martín-Gonzalez, M.
Fernanda. Optimization of Microwave-Assisted Extraction of All-trans-lycopene:
168
Effect of Extraction Conditions on Carotenoid Profile. Poster presentation for the
2014 Institute of Food Technologists Annual Meeting and Food Expo. Morial
Convention Center, New Orleans, LA, USA. June 21 & 22, 2014.
Poster was awarded 2nd place in the Food Engineering Division out of over 200
accepted abstracts.
SERVICE TO PROFESSION AND UNIVERSITY
•
Higher Education Review Board Reviewer
Institute of Food Technologists (IFT)
•
President* and President-Elect
Phi Tau Sigma Honor Society (Purdue Chapter)
•
Academic & Professional Development Committee Chair
Purdue Graduate Student Government
•
Treasurer
Purdue Food Science Graduate Student Association
Aug. 2013-Aug. 2016
Aug. 2014-2015*, Aug 2013-2014
Aug. 2014-2015
an. 2013-Aug. 2014
OUTREACH
•
Food Science Demonstration Volunteer Apr. 13, 2013, Apr. 18, 2015, Apr. 8, 2016
Purdue Spring Fest
•
Assisting Food Science Workshop Facilitator
4-H Academy at Purdue University
Jun. 10, 2015
•
Research Poster Symposium Judge
Purdue Nanotechnology Student Advisory Council
Aug. 6, 2014
ENGAGEMENT
•
Chair* and Volunteer
Fall 2012, Spring 2013, Fall 2014, Spring 2015*
Next Generation Scholars, Purdue University
•
Product Development Researcher
Kulanui Brand, Univeristy of Hawaii
Jun. 2009-May 2012
DIVERSITY AND INCLUSION
•
Safe Zone Trainee
Purdue University LGBTQ Center
Apr. 20, 2017
169
•
Panelist, Body Image Discussion Panel, Ideal Asian Body Campaign Apr. 13, 2017
Purdue Asian American & Asian Resource and Cultural Center
CERTIFICATIONS AND INDUSTRY-RELATED COURSES
•
•
•
Better Process Control School, Purdue University
Sensory Evaluation Techniques, Purdue University
Aseptic Packaging and Processing, Purdue University
May 8-11, 2017
May 4-5, 2015
Spring 2015
PEER-REVIEWED PUBLICATIONS
•
Ho, Kacie K.H.Y., Schroën, Karin, San Martín-Gonzalez, M. Fernanda, BertonCarabin, Claire C. 2016. Physicochemical stability of lycopene-loaded emulsions
stabilized by plant or dairy proteins. Food Structure.
http://dx.doi.org/10.1016/j.foostr.2016.12.001
•
Ho, Kacie K.H.Y., Ferruzzi, Mario G., Liceaga, Andrea M., San Martin-Gonzalez, M.
Fernanda. 2015. Microwave-assisted extraction of lycopene in tomato peels: effect of
extraction conditions on all-trans and cis-isomer yields. LWT-Food Science and
Technology, 62(1), 160-168. http://dx.doi.org/10.1016/j.lwt.2014.12.061
MEMBERSHIPS
Institute of Food Technologists
Phi Tau Sigma Honorary
American Oil Chemists’ Society
Документ
Категория
Без категории
Просмотров
0
Размер файла
3 732 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа