|
|
|
|
Pharmaceutical
Discovery, Oct 1, 2005
By Patrizia
Di Pietro , Enrico Alessi , Floriana San Biagio ,
Luigi La Magna , Gaetano Panvini , Gianfilippo
Scicolone , Salvatore Oliveri , Salvo Coffa
|
|
|
|
| Oct
1, 2005 |
| By:
Chris
Heid |
| Pharmaceutical
Discovery |
|
Fluidigm has developed a system for real-time quantitative PCR (RT
QPCR), comprised of instrumentation and dynamic arrays—nanofluidic chips
for combining any N samples and and M assays. After a dynamic array is
loaded with cDNA samples and sets of primers and FRET probes,
instrumentation automatically combines samples and assays into all
possible pairings within discrete 10 nL reaction chambers. Our Dynamic
Array Reader continuously monitors reactions. In this note, we describe
experiments demonstrating that dynamic arrays yield reproducibility and
dynamic range of detection equivalent to conventional platforms while
offering orders of magnitude higher throughput than 96-well plates.
Introduction
Drug discovery research often requires conventional microarrays to
compare expression patterns for thousands of genes from a single sample.
Because microarray analysis lacks sufficient precision and dynamic range,
RT QPCR is used to validate expression changes for a small subset of
genes. RT QPCR provides greater sensitivity, dynamic range, and precision,
and also enables the analysis of a larger sample set, thus improving
statistical significance. However, its implementation on microplate
platforms is time consuming, logistically challenging, and expensive. In
addition, limitations in the amount of available sample may preclude the
analysis of many genes.
We have developed an RT QPCR system that enables the simultaneous
analysis of significantly more genes and samples while streamlining the
process. The system provides an almost 100-fold increase in throughput
over 96-well systems and tremendous savings in time, labor, and running
cost. These logistical advantages will allow researchers to economically
and quickly conduct larger, more meaningful gene expression validation
studies.
Figure 1. 48.48 R1 dynamic array.
Samples and reagents are loaded into the microtiter wells on
either side of the carrier. The integrated fluidic circuit in the
center contains more than 15,000 engineered features, including
reaction chambers, vias, valves, and channels. Each chip yields
2,304 unique data points.
|
This system includes a dynamic array for gene expression, chip-loading
instrumentation, and the Dynamic Array Reader. A dynamic array consists of
a carrier—having well spacing that conforms to microplate
standards so that samples and reagents can be loaded with pipettes or
dispensing robots—and an integrated fluidic circuit,
incorporating a dense network of valves, vias, chambers, and channels to
route fluids and partition reactions (Figure 1). The chip-loading
instrument facilitates the on-chip combination of samples and assay
components into all possible pairings within discrete 10-nL reaction
chambers. The Dynamic Array Reader performs thermal cycling for all
chambers simultaneously and collects real-time images of reactions
throughout the run. The resulting data are analyzed and cycle threshold
numbers (Cts) are calculated. First-generation dynamic arrays allow for a
total of 2,304 experiments. The 48.48 R1 version of the chip accepts any
48 cDNA samples and any 48 TaqMan® assays to create every pair-wise
combination (48 x 48). Next-generation dynamic arrays, available in 2006,
will provide a total of 9,216 experiments. Thus, dynamic arrays provide
both high throughput and unlimited assay choices.
Experimental Conditions
We designed an experiment to measure the quantitative power of dynamic
arrays when challenged with varying concentrations of a target sequence.
The study was performed using version 12.12 R16, that is, a dynamic array
configured to accept 12 samples and 12 assays, producing 144 pair-wise
combinations with 16 replicates for a total of 2,304 reactions.
The study was implemented as follows: the chip was loaded with at least
5 µL of serial five-fold dilutions of Random-primed BD™ qPCR Human
Reference cDNA (BD Biosciences) and at least 5 µL of a TaqMan® assay for
TIMP-3 (F primer: 5'-CTACCTGCCTTGCTTTGTGA-3'; R primer: 5'-
ACCGAAATTGGAGAGCATGT-3' ; probe: 5'-6-FAM/CCAAGAACGAGTGTCTCTGGACCG/3BHQ2-3',
Integrated DNA Technologies). Final primer and probe concentrations were
900 nM and 250 nM, respectively. Final cDNA quantities in each reaction
chamber ranged from 4.5 pg to 7.2 fg. A no-template sample and a genomic
DNA sample were also loaded as controls.
Figure 2. 10x magnification of the
48.48 R1 dynamic array. The sample and TaqMan® Universal PCR
Master Mix mixture are loaded into the 9-nL chamber (blue) and the
primers and probes are loaded into the 1-nL chamber (orange).
|
Our chip-loading instrument was used to
drive samples and assays from the carrier wells into the reaction chambers
of the integrated fluidic circuit. The sample included a 1:1 mixture of
sample and TaqMan® Universal PCR Master Mix (Applied Biosystems), and the
assay consisted of a 10x solution of primers and a TaqMan probe. Each
reaction chamber holds approximately 9 nL from the sample well and
approximately 1 nL from the assay well (Figure 2). The Dynamic Array
Reader was used to accomplish thermal cycling, imaging, and data
collection. Thermal cycling conditions were 10 min at 95 ºC, followed by
40 cycles of 95 ºC for 10 s and 60 ºC for 1 min. Our proprietary
software was used for quantitative analysis.
Results
Table 1. Standard deviations and CVs
(TIMP) on the 12.12 R16 dynamic array. The cDNA concentration and
amount in the reaction chambers for each dilution of the series
are shown. Mean Ct, standard deviation, and percentage CV are also
shown for each dilution. Sixteen replicate reactions were run for
all dilutions; however, for the least-concentrated sample, only
positive reactions were considered in calculating the mean Ct,
standard deviation, and percentage CV.
|
Table 1 shows the cDNA concentrations,
Cts, and the corresponding SDs and CVs. Replicate reactions show Cts for
each dilution with a standard deviation below 0.3 cycles, except for the
7.2-fg dilution. Seven of the 16 reactions for the 7.2-fg dilution were
negative while the 9 positive reactions had a Ct of about 26 and a higher
standard deviation than the more concentrated samples. These results are
explained by stochastic effects seen when the average copies per reaction
chamber approaches 1 copy; in fact, further experiments (see our following
application note on absolute quantification) verify that, on average, 16
fg cDNA contains one TIMP-3 copy. Furthermore, a single target copy is
detectable after about 26 cycles of PCR, 10 or more cycles earlier than
most other RT QPCR systems. This increased sensitivity is explained
primarily by the concentration difference seen between a 10-nL PCR
reaction and a 20-µL PCR reaction (2× = 2000, x = ~11).
Figure 3 shows amplification plots from the TIMP-3 dilution series
experiment, which demonstrate that RT QPCR on dynamic arrays work
consistently and correlate with input target sequence amounts.
Conclusions
Figure 3. Amplification plots from
the TIMP-3 dilution series on the 12.12 R16 dynamic array. In each
panel, 16 replicate RT QPCR reactions are plotted (FAM/ROX versus
cycle). The Cts were determined at 5 standard deviations above
background (shown by the black horizontal line). The gray areas
illustrate diffusion of the 6-FAM probe from the 1-nL volume into
the 9-nL volume. Studies have shown that allowing additional time
for diffusion prior to PCR does not affect Ct (data not shown).
|
Our Dynamic Array System gives RT QPCR
performance on par with industry leading systems. Our direct comparisons
with the ABI PRISM® 7900 Sequence Detection System have demonstrated
equivalent precision and discrimination (data not shown). However, our
system provides experiment throughput and sample conservation that is
vastly superior to alternatives. Thus, pharmaceutical researchers will be
able to measure far more genes per run and to utilize precious samples
more efficiently. Gene expression analysis is the first among a suite of
applications that we are commercializing for use with this system. Future
applications will include SNP genotyping, nanofluidic immunoassays, rare
mutation detection, exact quantification by PCR, and others.
1 The FID Crystallizer,
introduced as part of the TOPAZ™ System for protein crystallization.
|
|
|
|
