High throughput gene expression measurement with real time PCR in a microfluidic dynamic array

Abstract

We describe a high throughput gene expression platform based on microfluidic dynamic arrays. This system allows 2,304 simultaneous real time PCR gene expression measurements in a single chip, while requiring less pipetting than is required to set up a 96 well plate. We show that one can measure the expression of 45 different genes in 18 tissues with replicates in a single chip. The data have excellent concordance with conventional real time PCR and the microfluidic dynamic arrays show better reproducibility than commercial DNA microarrays.

Figure 1. The 48.48 dynamic array chip.

A. Photograph of a 48.48 dynamic array chip showing the position of the sample inlets and the detector inlets in which the gene expression assay reagents are added. The check valves allow pressure to be applied and released. The accumulators provide reservoirs to hold the pressure and keep the valves closed during the reaction. The integrated Fluidic Circuit (IFC) is in the center of the chip. This is a network of fluid lines, NanoFlex™ valves and reaction chambers. The insert shows a blow-up of a portion of the IFC with one of the 2304 individual reaction chambers (RC) and the associated containment valves (CV) and interface valve (IV) There are two containment valves and one interface valve associated with each reaction chamber. B. A computer generated image (heat map) of a 48.48 dynamic array chip obtained after thermal cycling of the chip. Each of the squares represents 1 reaction chamber from the chip. The color indicates the C_T value according to the legend shown on the right. Black chambers indicate a C_T>40.

Figure 2. Comparisons of data.

The correlation coefficient r is shown above each plot. For details regarding these comparisons see Materials and Methods. A. Comparison of Duplicate samples from the GeneNote database for 45 genes and two tissues, heart and liver. B. Comparison of data from the 48.48 dynamic array obtained for preamplifid cDNA (PA cDNA) and data from the GeneNote database for the same genes and tissues. C. Comparison of the data from the 7900HT Sequence Detection System obtained with cDNA and data from the GeneNote database. D. Comparison of data from the 7900HT obtained with cDNA and the 48.48 dynamic array obtained with PA cDNA for the same genes and tissues. E. Comparison of the ΔΔC_T values for replicates on one of the 48.48 dynamic array chips used for gene expression analysis for all 44 genes and 18 tissues. F. Pair-wise comparison of the mean ΔΔC_T values from two chips for the same 44 genes and 18 tissues.

Figure 3. Estimated copy numbers.

Using the standard curve shown in Fig. S2 the actual number of copies in the reaction chambers was estimated from the mean C_T values of all the data obtained on the three chips used previously. The data is plotted as the log₁₀. Error bars were determined from the mean C_T values +/−one standard deviation. The raw data for the standard curve in Figure S2 is presented in Table S4.