вход по аккаунту


Spatially Encoded Single-Bead Biginelli Synthesis in a Microstructured Silicon Array.

код для вставкиСкачать
Microscale Synthesis
DOI: 10.1002/anie.200503041
Spatially Encoded Single-Bead Biginelli Synthesis
in a Microstructured Silicon Array**
G. Alexander Groß,* Gnther Mayer, Jens Albert,
Daniel Riester, Jens Osterodt, Hanns Wurziger, and
Andreas Schober*
Microreaction technology has become a field of wide
scientific and industrial interest as it allows an increased
overall efficiency through savings in labor, materials, time,
and costs. The advantages of miniaturization have been
discussed in the context of the “lab-on-a-chip” concept.[1, 2]
For combinatorial chemistry, microfluidic approaches have to
cope with many reagent solutions with different fluid properties, vapor pressures, wettabilities, and other factors.[3] Despite
the many technical problems to be solved, the use of
microfluidic devices seems promising.[4] Owing to the distinct
architecture of such systems, the identity of products manufactured in a continuous process is always secured. Solidphase synthesis is used for the preparation of libraries with
[*] Dr. A. Schober
Center for Micro- and Nanotechnologies
Technical University of Ilmenau
98 693 Ilmenau (Germany)
Fax: (+ 49) 367-769-3499
Dr. G. A. Groß
Department of Physical Chemistry and Microreaction Technology
Technical University of Ilmenau
Weimarerstrasse 32, 98 693 Ilmenau (Germany)
Fax: (+ 49) 367-769-3179
Dr. G. Mayer, J. Albert
Microsystems Division
Institute for Physical High Technology e.V. Jena
Albert-Einstein-Strasse 9, 07 745 Jena (Germany)
Dr. D. Riester
Institute of Microbiology and Genetics
University of G?ttingen
Grisebachstrasse 8, 37 077 G?ttingen (Germany)
Dr. J. Osterodt, Dr. H. Wurziger
Merck KGaA
Frankfurterstrasse 250, 64 298 Darmstadt (Germany)
[**] The authors thank Dr. P. Raddatz and Dr. I. Lues for the pleasant
working conditions. The technical assistance of J?rg Heisiep and
Stephan Niedziella with microanalysis is gratefully acknowledged,
as well as the assistance and engineering support by Michael
Schmelz and GEnther Brenner. We particularly thank our colleagues
at the IPHT Jena for the microstructuring support and Dr. Andreas
Schwienhorst (University of G?ttingen) for helpful discussions.
Funding was provided by BMBF grant 0311766 with contributions
from Merck KgaA Darmstadt (BMBF grant 0311765) and partial
support by a grant of the ThEringer Ministerium fEr Kultur (TKM)
and the European Community (B 678-03001).
Supporting information for this article is available on the WWW
under or from the author.
large numbers of compounds. The bead-based “split-andrecombine” process is technically less demanding and very
efficient in terms of the number of products versus the
number of synthetic steps.[5, 6] However, a major disadvantage
is inherent in this process: the information on the reaction
sequence of each individual bead is lost. The necessary
product information has to be retrieved afterwards by
encoding or decoding strategies, which is a serial and timeconsuming process. This problem can be overcome by using a
spatially resolved approach. In this case, the prepared
compounds are encoded by their position. Most of these
approaches are restricted to chemical protocols similar to
those used in peptide chemistry.[7, 8] Microstructured silicon
nanotiter plates[2c] charged with a single bead per compound
allow the application of nearly all solid-phase chemistry
protocols, resulting in a bead array with immobilized substances with known processing history for each compound.[7–9]
The miniaturization of solid-phase chemistry to the
nanovolume scale is difficult and requires much effort to
solve the technical difficulties. Evaporation of solvents must
be avoided, even with long reaction times. This is exacerbated
even more if the necessary reaction temperatures increase the
vapor pressures of the solvents used. Cross-contamination
must also be avoided during filling, sealing, and reaction steps.
This is particularly important for the condensed 200-nLvolume reactor arrays with 12 9 12 wells on a 16 9 16-mm2sized chip. Such a format equates to 4700 wells on the MTP
footprint. It is amenable to selection, excision, and manipulation of individual beads and solutions by simple handling
steps in an automated setup.
Herein we describe the first miniaturized parallel organic
synthesis in nanoliter volumes on single beads at 100 8C for
14 h as a “proof-of-concept experiment”. Evaporation and
“cross-talk” was excluded by using a combination of a
pressurized sealing device (at the end a vice) and a specially
designed chip module with two covering stamps. Only silicon
and PTFE come into contact with chemicals. The brittle
silicon nanochips with their integrated reactor wells withstand
a sealing pressure of 3000 N because the lateral gasket
deformation was hindered by the setup used (Figure 1).[10] In
this way each single reaction well was sealed reliably even at
temperatures up to 100 8C.
The nanochips are made up from three microstructured
silicon wafers, which are stacked and secured by PDMS
(polydimethylsiloxane) bonding (Figure 2). The chip stack is
flanked by the two gasket layers 1 and 3 (Figure 2, types b–d)
at the top and bottom. The silicon layer 2 inbetween contains
the actual reaction wells with microsieves. These sieves allow
the removal of reaction or washing solutions under vacuum[2c]). Figure 2 shows different chip designs that have been
optimized with respect to tightness and reusability. This
means that there will be no permanent connection between
the chip, the gasket, and the cover, even if the vice force of
3000 N acts on the 2.56-cm2 chip. One major problem had to
be solved. In filled chambers of types a and b (Figure 2) all
inner surfaces are wetted up to the upper rim owing to the
surface tension of the solvent. If the gasket comes into contact
with the surface of the liquid, capillary forces cause contamination of all wells. In types c and d, additional inner rims
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3102 –3106
mitotic kinesin inhibitors, antibacterial and fungicidal compounds, and others.[12]
Nine copies of a DHPM library with 16 members was
synthesized copies on a single chip by following an adapted
macroscopic protocol.[11] The recommended solvent mixture
1,4-dioxane/iPrOH was changed to triglyme/iPrOH to obtain
a higher vapor pressure and boiling point. Appropriate N-3aryl-3-oxopropanamide beads were prepared before by
addition of aryl lithium enolates to immobilized 4-nitrophenyl
carbamates according to a literature protocol.[13] During
transfer of these beads into the chip wells, swollen beads
were sorted according to their horizontal diameter (350 10 mm) by using the image processor of the x,y,z manipulation
and pipetting robot (Figure 3). The loaded beads were placed
Figure 1. Synthesis module composed of the chip carrier with gaskets
and the closing vice. a) Chip carrier modules with PTFE gaskets.
b) Closed vice with the assembled chip module inside. c) SEM image
of a PTFE gasket molded by the chip after 24 h. d) Profile across the
used PTFE gasket ( 75 mm deep).
Figure 3. Bead handling. a) Partially swollen bead fixed through a
vacuum needle during the transfer and sorting process. b) Top view
into a chamber with one bead placed on the sieve bottom (there are
four reflections of the bead at the angular silicon walls).
in columns, and the necessary reagent solutions
(0.5 mmol L 1) were then added in lines to the chip to yield
the pattern outlined in Figure 4.
Scheme 1. The Biginelli reaction. TFA = trifluoroacetic acid.
Figure 2. SEM images of cross-sections of different types of reaction
chips. a) Microstructured 14 G 14 chip. b), c), d) Three layer sandwich
chip assemblies, PDMS bonded. Type d is equipped with an internal
rim. Layer captions: 1 and 3 are gasket-bearing layers, 2 contains the
reaction chambers.
prevent the wetting of the gasket-bearing layer. Type d in
Figure 2 best fulfils all the criteria.
The Biginelli reaction was chosen to test the parallel
synthesis on the nanovolume scale. This reaction requires
temperatures of near 100 8C for more than 12 h.[11] 4,6-Diaryl
1,3-dihydropyrimidin-2(1H)-ones (DHPMs) 4 are formed in a
three-component condensation of immobilized N-3-aryl-3oxopropanamides 1, aldehydes 3, and thiourea (2)
(Scheme 1). DHPMs are regarded as privileged structures
for drug research. They have been described as calciumchannel modulators, aa1-adrenergic-receptor antagonists,
Angew. Chem. Int. Ed. 2006, 45, 3102 –3106
A fused silica capillary connected to a syringe pump was
used for pipetting. The chip inside the holding device was
mounted onto the bottom gasket stamp and adjusted next to a
96-well storage plate at the pipetting robot. Reagent solutions
(150 mL) were aspirated from this plate and dispensed in 150nl portions into the target wells on the silicon chip by contact
pipetting. The wetting properties of the surfaces and the
geometry of the chip wells aid their precise filling into the
middle chip (see Figure 5).[14]
After dosing, the capillary was removed with only a very
small droplet remaining at the tip. This was removed by
rinsing the syringe system twice. The liquid transfer required
approximately 35 min, after which the cover stamp was
adjusted, fixed, and pressed together within the vice. The
sealed “synthesis module” was then placed in an incubator set
to 100 8C for 14 h. Afterwards, the setup was allowed to cool
to ambient temperature. To determine the tightness of the
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
vice force acting on the stack during the high-temperature
For reference purposes, 100 mg of the appropriate N-3aryl-3-oxopropionaminde beads were converted into 16
DHPM derivatives in standard glassware. Single-bead products of chip-processed and macroscopic-batch beads were
analyzed and compared. Therefore chip-processed beads of
four different products were chosen from three different
positions each. Analyzed chip positions marked in Figure 4
belong to the chromatograms shown in Figure 6.
Figure 4. Spatially encoded library inside the silicon chip and building
blocks used. A library of 16 DHPM derivatives was prepared in nine
copies inside a single silicon chip. Aldehydes (R2 = 1–4) were used in
columns, polymer-bound b-ketoamides (R1 = a–d) were used in rows.
The spatial position encoding is etched into the silicon surface at the
top and left border of the chip (A1–L12). HPLC analysis of marked
product positions is shown in Figure 6.
Figure 5. Contact pipetting of reagent solution into a silicon reaction
chamber. a) Fused silica capillary placed inside the reaction chamber.
b) Chamber with fused silica capillary after a dosage of 150 nL.
system, it was weighed before and after the experiment.
On average, a weight loss of less than 2 % was observed.
The chemical conversion was done by heating and
subsequent cooling of the reagent solutions. Due to
solvent evaporation and condensation, the surfaces of
the gasket-bearing lids became moist. The dismantling
Figure 6. Comparison of product chromatograms.
and removal of the cover lid led to cross-contamination.
When the vice is released, capillary forces move the
To cleave the products from the beads, single beads were
reagent solution into the opening slit between the lid
treated in HPLC vials for 1 h with 30 mL of the cleavage
and the silicon chip, thus contaminating the entire chip. This,
solution (DCE/TFA 1:1). The solvents were carefully evapohowever, is of no influence when the immobilized starting
rated from the cleavage products, and the resulting residues
materials in all wells are consumed. The subsequent washing
were dissolved at least for one hour in acetonitrile/water
steps remove excess reagents.
(10 mL; 95:5). These solutions were analyzed by HPLC (LCThe beads were washed five times with triglyme/iPrOH
Packings Instruments) in the presence of the beads. The
(3:2), 1,2-dichloroethane (DCE), and finally with iPrOH by
chromatograms of the macroscopic batch and chip synthesis
using the vacuum station next to the pipettor. Deliberate
showed no significant differences (Figures 6 and 7). As a spot
overfilling (300 mL) ensured the cleaning of all possibly
check, the collected material from 10 beads was analyzed by
contaminated surfaces and beads. In the last step, the beads
LC–MS in more detail. Masses of precursors and main
were dried under vacuum for 10 min. The PDMS-bonded chip
products were as expected. However, the precursor/product
stack withstands all experiments. This is due to the very small
ratio varies depending on the dissolving time after cleavage.
contact area between PDMS and solvents as well as the high
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3102 –3106
demand synthesis” of potential drug derivatives upstream to
screening procedures. The technological gap between highthroughput synthesis (HTS) and high-content screening
procedures can be overcome. Nanotiter plates on common
HTS equipment allows the production of enough material on
a single bead for several biological tests. The technology is
currently under development for automatic, large-scale
parallel combinatorial synthesis.
Received: August 26, 2005
Revised: December 19, 2005
Published online: March 30, 2006
Keywords: combinatorial chemistry · microreactors · nanochip ·
solid-phase synthesis · synthesis miniaturization
Figure 7. Chromatogram comparison of identical single-bead products
(arranged in ascending order): 1. batch-scale reference; 2. chip position A-5; 3. chip position E-5; 4. chip position I-5.
This indicates a rebinding process of the product to the beads
(i.e. Figure 6, chromatograms C-7 and K-7). The intensity of
the measured chromatograms of identical substances varies
by up to 50 % (peak area). This is a reflection of differences in
the sizes of the selected beads, dosing errors during liquid
handling, before and during injection into the HPLC system,
or a varying solubility of the received products. A comparison
of chromatograms of products from neighboring chip wells
indicates no cross-contamination. We conclude that each
individual reaction well of the silicon chip of type D
(Figure 2) was sealed properly during reaction.
We demonstrated the feasibility of a novel approach for
miniaturized parallel solid-phase synthesis of a library in 200nl volumes. Individual, spatially segregated synthesis beads
were addressed by a combination of silicon wafer technology
and system engineering. Hereby the problems of evaporation,
leakage, and cross-contamination in the small volume could
be solved by a specially designed chip and covering system.
The approach proved to be versatile for a solid-phase
Bigginelli synthesis at 100 8C for 14 h and should be feasible
for chemical protocols under less-demanding conditions.
As shown in the proof-of-concept experiment, standard
pipettor techniques can be used. Solvents with high vapor
pressures can be used (up to pressures of 150 mbar inside
the chamber). In this way individual reaction vessel could be
closed to allow the use of common organic solvents at
temperatures up to 120 8C. Generic solid-phase protocols can
be used in this setup.
The approach seems to be a promising alternative to
miniaturization for chemical synthesis, particularly for solidphase synthesis. The bead loading, pipetting, and washing
steps were carried out by one robotic system, thus the
complete process is suitable for automation. Our single-bead
synthesis approach could potentially be used for the “on
Angew. Chem. Int. Ed. 2006, 45, 3102 –3106
[1] a) D. J. Harrison, K. Fluri, K. Seiler, Z. Fan, C. S. Effenhauser,
A. Manz, Science 1993, 261, 895 – 897; b) M. A. Burns, Science
2002, 296, 1818 – 1819; c) “Microsystems for independent parallel chemical and biological processing”: A. Schober, A. Schwienhorst, J. M. KIhler, M. Fuchs, R. GJnther, M. ThJrk, Microsystems Technologies, (Eds.: H. Reichel, A. Heuberger), VDEVerlag, Berlin, 1994, pp. 381 – 390; d) for a recent example, see:
B. Zheng, R. F. Ismagilov, Angew. Chem. 2005, 117, 2576 – 2579;
Angew. Chem. Int. Ed. 2005, 44, 2520 – 2523.
[2] a) Microreactors (Eds.: W. Ehrfeld, V. Hessel, H. LIwe), WileyVCH, Weinheim, 2000, pp. 154 – 162; b) K. JLhnisch, V. Hessel,
H. LIwe, M. Baerns, Angew. Chem. 2004, 116, 410 – 451; Angew.
Chem. Int. Ed. 2004, 43, 406 – 446; c) “Nanotiterplates for
screening and synthesis”: G. Mayer, K. Wohlfahrt, A. Schober,
J. M. KIhler in Microsystem Technology: A Powerful Tool for
Biomolecular Studies in Biomethods (Ed.: H. P. Saluz), BirkhLuser, Basel, 1999, pp. 75 – 128; d) N. Schwesinger, O. Marufke, F. Qiao, R. Devant, H. Wurziger, Process MiniaturizationIMRET 2: 2nd International Conference on Microreaction
Technology. Topical Conference Reprints, 1998, New Orleans,
p. 124; e) A. Schober, G. Schlingloff, G. Mayer, A. Groß, J.
Albert, T. Henkel, H. Wurziger, Microsyst. Technol. 2004, 10,
281 – 292; f) E. Gottwald, A. Schober, Eur. Pharm. Rev. 2004, 4,
35 – 44.
[3] For recent examples of combinatorial microfluidics, see the
special issue: QSAR Comb. Sci. 2005, 24, 699 – 767.
[4] a) P. Watts, S. J. Haswell, Chem. Soc. Rev. 2005, 34, 235 – 246;
b) for instance see T. Kawaguchi, H. Miyata, K. Ataka, K. Mae,
J. Yoshida, Angew. Chem. 2005, 117, 2465 – 2468; Angew. Chem.
Int. Ed. 2005, 44, 2413 – 2416.
[5] a) A. Furka, F. SebestyPn, M. Asgedom, G. DibQ, Int. J. Pept.
Protein Res. 1991, 37, 487 – 493; b) M. A. Gallop, R. W. Barrett,
W. J. Dower, S. P. A. Fodor, E. M. Gordon, J. Med. Chem. 1994,
37, 1233 – 1251; c) E. M. Gordon, R. W. Barret, W. J. Dower,
S. P. A. Fodor, M. A. Gallop, J. Med. Chem. 1994, 37, 1385 – 1401.
[6] J. R. Schullek, J. H. Butler, Z. -J. Ni, D. Chen, Z. Yuan, Anal.
Biochem. 1997, 246, 20 – 29.
[7] S. P. A. Fodor, J. L. Read, M. C. Pirrung, L. Stryer, A. T. Lu, D.
Solas, Science 1991, 251, 767 – 773.
[8] M. Meldal, C. B. Holm, G. Bojeson, M. H. Jacobsen, A. Holm,
Int. J. Pept. Protein Res. 1993, 41, 250 – 260.
[9] R. Frank, Tetrahedron 1992, 48, 9217 – 9232.
[10] Sealing conditions were determined experimentally (data not
shown); for details, see patent: B. Diefenbach, H. Deppe, H.
Wurziger, A. Gross, G. Schlingloff, A. Schober, D. Tomandl
(Merck), 2000, WO 2001089680 [Chem. Abstr. 2001, 868299].
[11] G. A. Groß, H. Wurziger, A. Schober, J. Comb. Chem. 2006, 8,
153 – 155.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[12] For examples of solution-phase, polymer-supported, fluorous,
and soluble-polymer-supported DHPM syntheses, see: C. O.
Kappe, QSAR Comb. Sci. 2003, 22, 630 – 645 and references
therein; for a discussion on soluble- versus insoluble-polymerassisted synthesis, see: J. J. V. Eynde, O. Watte, Arkivoc 2003, 4,
93 – 101.
[13] G. A. Groß, H. Deppe, A. Schober, Tetrahedron Lett. 2003, 44,
3939 – 3942.
[14] An accuracy of 3 % was determined by optical measurement of
the filling level.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3102 –3106
Без категории
Размер файла
309 Кб
biginelli, synthesis, encoded, array, single, silicon, bead, spatially, microstructure
Пожаловаться на содержимое документа