close

Вход

Забыли?

вход по аккаунту

?

Feature MultiplexingЧImproving the Efficiency of Microarray Devices.

код для вставкиСкачать
Zuschriften
common attribute shared by all of these devices is the
convention of distributing each sensor probe onto an exclusive array feature. This methodology imposes the requirement
that the arrays possess a number of features equal to the
number of probes, thus requiring array designers to investigate miniaturization to enable higher-density array construction.[6, 7] However, decreasing the feature size correspondingly reduces the number of sensing molecules per
feature, and the reduction of feature size and spacing is
fundamentally limited by the array technology (namely,
optimized feature spacing is at present on the order of
hundreds of microns).
A novel platform-independent probe-loading methodology has been developed that circumvents the need for the
continual shrinkage of sensor features. This methodology
increases the information density of the array not by
miniaturizing the features in the array, but rather by changing
the way probes are distributed within the array (see Figure 1).
DNA Sensing
DOI: 10.1002/ange.200502151
Feature Multiplexing—Improving the Efficiency
of Microarray Devices**
Matthew J. Schmid, Kalpana Manthiram,
Scott M. Grayson, James C. Willson, Jason E. Meiring,
Kathryn M. Bell, Andrew D. Ellington, and
C. Grant Willson*
Since the genomic sequencing of more than 100 organisms
first began, an almost overwhelming compendium of genetic
and proteomic targets has been identified. Researchers are
now beset with the task of characterizing these targets, as well
as identifying and characterizing their extant natural variants.
Combinatorial analytical techniques, such as microarray
analysis, have emerged as the preferred tools for performing
these characterizations and have enabled the recent commercialization of DNA[1–4] and proteomic[5] sensor arrays. A
[*] Dr. M. J. Schmid, K. Manthiram, Dr. S. M. Grayson, Dr. J. E. Meiring,
Prof. C. G. Willson
Department of Chemical Engineering
The University of Texas at Austin
Austin, TX 78712 (USA)
Fax: (+ 1) 512-471-7222
E-mail: willson@che.utexas.edu
J. C. Willson, K. M. Bell, Prof. A. D. Ellington, Prof. C. G. Willson
Department of Chemistry and Biochemistry
The University of Texas at Austin
Austin, TX 787412 (USA)
[**] We acknowledge Benjamen Rathsack and David Johnson for their
involvement in developing the shape-indexed MUFFINS platform.
We also thank the Welch Foundation, MURI, The Department of
Army Research (project DAAD 19-99-1-0207), and The Beckman
Foundation (project 26-7506-95) for financial support, as well as the
NSF/IGERT program (Cellular and Molecular Imaging for Diagnostics and Therapeutics) for academic support (M.J.S.).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
3416
Figure 1. Three-feature standard arrays (with one probe per feature)
can only detect three targets, whereas multiplexed arrays can detect six
target sequences.
This technique, called feature multiplexing, achieves
increases in density by incorporating multiple probes into
each feature using a unique encoding system (see Figure 1).
The use of this methodology allows a given number of
features (n) to be used to deploy a much greater quantity of
probes (up to 2n 2) when sensing for a single analyte. This
methodology can dramatically increase the capacity or
information density of an array because there is this
exponential relationship between the number of array
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. 2006, 118, 3416 –3419
Angewandte
Chemie
features and the number of probes that the array can
accommodate. A demonstration of feature multiplexing in
array-based diagnostics is presented herein.
To illustrate the principle of feature multiplexing, a small
single-nucleotide-polymorphism (SNP) detection array was
constructed and its performance verified in three separate
detection assays.[8] Due to the clinical importance of the p53
gene, the array was designed to screen for the presence of 29
potential cancer-causing missense SNPs within a short test
section of that gene.[9] The array contained a total of 31
different single-stranded oligonucleotide probes: 29 test
probes complementary to the SNPs for which the array had
been designed and two reference probes complementary to
the different nonoverlapping halves of the “wild-type” (or
reference) p53 gene sequence. These 29 test probes were
distributed into just five array features by using feature
multiplexing, thus constituting an approximate sixfold
increase in information density. The reference probes served
as controls in our assays; their function and their placement
within the array will be discussed in more detail later.
As for all SNP detection arrays, the probe design of a
feature-multiplexed array has an enormous impact on the
signal-to-noise characteristics of the array. An ideal probe
should form a stable double helix with its complement, and
only its complement under the hybridization conditions of the
detection assay. Thus, a probe set was designed such that a
single base-pair mismatch would be sufficient to disrupt
duplex formation. Likewise, all of the probes should produce
a signal of uniform intensity in response to the detection of
their complement. To that end, the probes in this array were
carefully designed by a custom software program to all have
nearly the same melting temperature (Tm = 30.8 8C, s =
1.0 8C), and hence the same duplex stability under the
optimized hybridization conditions (Table 1).
To distinguish the output of one probe from that of
another, each probe is indexed by assignation of a unique
distribution pattern within the multiplexed features.
Although there were many ways of choosing the distribution
patterns for these probes, the following procedure was
developed for ease of encoding and decoding. All the
probes were assigned a number (1–29), and that number
was converted into its five-bit binary expression (e.g.,
probe 21 = 10101). The pattern of “1”s in the five-bit binary
expression was used to determine the placement of each
probe into a set of multiplexed features (Table 2). For
example, probe 21 (or “10101”) was placed in features 1, 3,
and 5. Likewise, the signal response can be readily interpreted, as each positive response represents a “1” and each
Table 1: Probe sequences used in the detection assays and their
positions relative to the reference p53 gene sequence.[a, b]
Table 2: Binary loading patterns for each of the probes used in the array.
Probe
Probe sequence[b]
Tm [8C]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Ref 1
Ref 2
GTGAGTGTTTG
GGTGGGTGTT
GTGTGTGTTTG
GTGCATGTTTG
GGTGCCTGTT
GTGCTTGTTTG
GGTGCGTATTT
TGCGTCTTTG
TGCGTTTTTGT
TGCGTGATTG
CGTGGTTGTG
CGTGCTTGTG
TGTTAGTGCCT
CGTGTTGGTG
CGTGTTCGTG
CGTGTTTATGC
TGTTTCTGCCT
GTGTTTTTGCC
TGTTTGAGCCT
GTTTGGGCCT
TGTACCTGTCC
TGTTTGTCCCT
TTGTTCCTGTC
TTGTGACTGTC
TTGTGGCTGTC
TTGTGTCTGTC
TGTGCCAGTC
GTGCCGGTC
GTGCCCGTC
GTGCGTGTTT
GTGCCTGTCC
28.7
31.0
29.9
30.6
31.9
30.5
31.4
29.0
31.9
29.2
30.6
31.5
31.4
30.6
31.2
29.9
32.5
30.9
32.5
31.7
31.8
31.6
29.6
29.8
30.2
29.8
31.5
31.2
31.2
30.3
33.2
[a] The 5’-methacrylamide modifications on the probes are not shown;
mutant bases are printed in bold. [b] p53 reference gene sequence: 5’TGAGGTGCGTGTTTGTGCCTGTCCTGG-3’
Angew. Chem. 2006, 118, 3416 –3419
Probe
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Ref 1
Ref 2
00001
00010
00011
00100
00101
00110
00111
01000
01001
01010
01011
01100
01101
01110
01111
10000
10001
10010
10011
10100
10101
10110
10111
11000
11001
11010
11011
11100
11101
N.A.
N.A.
Sq
Feature type/shape[a]
Cir
Ch
H
Tra
X[b]
P[b]
T[c]
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
[a] Legend: Sq = square, Cir = circle, Ch = chevron, H = hexagon, Tra =
trapezoid, X = cross, P = pentagon, T = triangle. [b] Reference feature.
[c] Control feature. N.A. = not applicable.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.de
3417
Zuschriften
negative a “0” in the binary expression of the probe that
detected its complement (see Figure 2).
Figure 2. Each multiplexed feature is uniquely shaped and corresponds
to a bit in a binary expression. The binary expression (fluorescence = “1” and non-fluorescence = “0”) is converted into a number,
which identifies the probe that has detected its complement (in this
case, probe 21).
In addition to the five multiplexed features, our array also
contained three control features that were not multiplexed,
the two reference probes, and a control containing no DNA.
Because each reference probe was complementary to a
different half of the test sequence, their output could be
used to determine which portion contained a mutation (see
Figure 3), and therefore could corroborate the output of the
multiplexed features.
tures INdexed by Shape) platform[10] because it offers the
opportunity to readily incorporate multiple probes in a single
feature. The shape-encoded features were constructed by
contact lithography using a photopolymerizable liquid hydrogel prepolymer containing poly(ethylene glycol) diacrylate
(PEG-da), water, probe DNA, and trace amounts of a
photoinitiator. To covalently attach the probe sequences to
the hydrogel matrix, each of the probes used in this
demonstration contained a polymerizable 5’-acrydite modification.[10] Each batch of features was uniquely shaped, so as to
provide a means of rapidly identifying the function of each
feature within the randomly ordered array. The features can
be immobilized onto a support matrix, but for this demonstration the array features were in their loose form.
The detection assays were conducted by placing a set of
array features (the five multiplexed features and the three
control features) in a buffer solution containing fluorescently
labeled target DNA. The three assays consisted of two
sequences which contained a single base mutation and a third
sequence without a mutation, identical to the reference
sequence (Table 3). After the features had hybridized with
Table 3: Target sequences used in the detection assay.
Target
Sequence[a]
p53 “wild type”
Probe 7
Probe 21
3’-CCACGCACAAACACGGACAGG-5’-Cy3
3’-CCACGCATAAACACGGACAGG-5’-Cy3
3’-CCACGCACAAACATGGACAGG-5’-Cy3
[a] Mutant bases are printed in bold.
Figure 3. The output from the two reference features verifies which
subsection of the target contains a mutation.
The feature multiplexing methodology can be used in a
multitude of array platforms, including traditional Cartesianindexed arrays. However, this demonstration was performed
by using a randomly ordered/shape-indexed approach called
the MUFFINS (Mesoscale Unaddressed Functionalized Fea-
3418
www.angewandte.de
the target analytes, they were rinsed and imaged on a
fluorescence microscope. Bright-field and fluorescence
micrographs of the arrays are presented in Figure 4. In all
cases, the decoded output of each of the detection assays
agreed with the known identity of the target sequence.
Array designers have focused on decreasing the physical
size of the individual sensing features to meet the need for
increased information density. Arrays can hold twice as many
probes, a linear improvement, by halving the area of each of
the features. On the other hand, multiplexing enables
exponential increases in information density when sensing a
single analyte simply by distributing multiple probes in each
feature (see Figure 5). The above proof-of-concept study has
demonstrated that this method is capable of yielding a sixfold
increase in density with only five multiplexed features. Higher
compressions are possible, but require an increase in the
number of probes that must be placed at each feature. Feature
multiplexing is a very versatile technique, applicable to a
variety of sensing chemistries and platforms in which one is
attempting to identify a single unknown substance. Virtually
any microarray device in which multiple probes can be placed
in a single feature can benefit from this technique, thus
decreasing the need for further miniaturization and improving the time and cost efficiency of their manufacture.
Received: June 20, 2005
Published online: April 11, 2006
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. 2006, 118, 3416 –3419
Angewandte
Chemie
.
Keywords: DNA · hydrogels · multiplexing · sensors
[1] A. C. Pease, D. Solis, E. J. Sullivan, M. T. Cronin, C. P. Holmes,
S. A. Fodor, Proc. Natl. Acad. Sci. USA 1994, 91, 5022.
[2] M. J. OFDonnell-Maloney, D. P. Little, Genet. Anal. Tech. Appl.
1996, 13, 151.
[3] F. J. Steemers, J. A. Ferguson, D. R. Walt, Nat. Biotechnol. 2000,
18, 91.
[4] R. G. Sosnowski, E. Tu, W. F. Butler, J. P. OFConnell, M. J.
Heller, Proc. Natl. Acad. Sci. USA 1997, 94, 1119.
[5] H. Zhu, M. Bilgin, R. Bangham, D. Hall, A. Casamayor, P.
Bertone, N. Lan, R. Jansen, S. Bidlingmaier, T. Houfek, T.
Mitchell, P. Miller, R. A. Dean, M. Gerstein, M. Snyder, Science
2001, 293, 2101.
[6] M. O. Reese, R. M. van Dam, A. Scherer, S. R. Quake, Genome
Res. 2003, 13, 2348.
[7] R. J. Lipshutz, S. P. A. Fodor, T. R. Gingeras, D. J. Lockhart, Nat.
Genet. 1999, 21, 20.
[8] Full experimental details are available in the Supporting
Information.
[9] Y. Cho, S. Gorina, P. D. Jeffery, N. P. Pavletich, Science 1994, 265,
346.
[10] J. E. Meiring, M. J. Schmid, S. M. Grayson, B. M. Rathsack,
D. M. Johnson, R. Kirby, R. Kannappan, K. Manthiram, B. Hsia,
Z. L. Hogan, A. D. Ellington, M. V. Pishko, C. G. Willson, Chem.
Mater. 2004, 16, 5574.
Figure 4. Bright field (left) and fluorescence (right) micrographs of the
sensor response during the three detection assays. Below each set of
micrographs is a schematic representation that illustrates the decoding process.
Figure 5. The graph of two trend lines which illustrate the linear
improvements gained by decreasing the size of the features relative to
the exponential improvements that result from feature multiplexing.
Angew. Chem. 2006, 118, 3416 –3419
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.de
3419
Документ
Категория
Без категории
Просмотров
0
Размер файла
334 Кб
Теги
features, efficiency, microarrays, devices, multiplexingчimproving
1/--страниц
Пожаловаться на содержимое документа