A versatile assay for high-throughput gene expression profiling on universal array matrices

Abstract

We report a flexible, sensitive, and quantitative gene-expression profiling system for assaying more than 400 genes, with three probes per gene, for 96 samples in parallel. The cDNA-mediated annealing, selection, extension and ligation (DASL) assay targets specific transcripts, using oligonucleotides containing unique address sequences that can hybridize to universal arrays. Cell-specific gene expression profiles were obtained using this assay for hormone-treated cell lines and laser-capture microdissected cancer tissues. Gene expression profiles derived from this assay were consistent with those determined by qRT-PCR. The DASL assay has been automated for use with a bead-based 96-array matrix system. The combined high-throughput assay and readout system is accurate and efficient, and can cost-effectively profile the expression of hundreds of genes in thousands of samples.

Figure 1

DASL: a cDNA-based assay for RNA profiling. RNA is converted to cDNA, using biotinylated oligo-d(T)₁₈ and random hexamers, and immobilized to a streptavidin-coated solid support. Two oligonucleotides are designed to interrogate each target site on the cDNA. The upstream oligonucleotide consists of two parts, the gene-specific sequence and a universal PCR primer sequence (P1) at the 5′-end. The downstream oligonucleotide consists of three parts, the gene-specific sequence, a unique address sequence which is complementary to a capture sequence immobilized on the array, and a universal PCR primer sequence (P2′) at the 3′-end. A single address sequence is uniquely associated with a single target site. The upstream oligonucleotide hybridizes to the targeted cDNA site, extends and ligates to its corresponding downstream oligonucleotide to create a PCR template that can be amplified with universal PCR primers (P1 and P2). The PCR products, which are fluorescently labeled by incorporation of the 5′-labeled primer P1, are hybridized to capture sequences on the beads in the array, and fluorescence intensity is measured for each bead.

Figure 2

DASL assay performance, dynamic range and differential expression detection. Intensity data from six synthetic RNAs (cat, cre, gus, lacZ, luc, and bla) are shown as log intensity vs. log RNA input (molecules) in the DASL assay. The RNA input is given as molecules of spiked transcript per DASL reaction. The error bars represent the 95% confidence interval of intensity values for six replicate assays.

Figure 3

Correlation of DASL with qRT–PCR. The logarithmic fold difference in abundance in pairwise comparisons between A-20, EL-4, and a 1: 1 mix of the two was estimated for expressed genes in both the DASL assay (fold difference in array intensity, x-axis) and qRT–PCR (fold difference in abundance derived from crossover threshold, y-axis).

Figure 4

Cell-specific responses to androgen treatment. (A) Expression of 212 human prostate-related genes was monitored in prostate cell lines treated with androgen or vehicle. Data for each condition were plotted to illustrate the impact of treatment with androgen (y-axis) vs. vehicle (x-axis). (B) Agglomerative clustering using data from the 212 genes. For each cell line, three randomly selected biological replicates are presented for both androgen-treated (filled box) and untreated (open box) conditions. The distance between subclusters (y-axis, the height) measures the divergence of their expression patterns.

Figure 4

Cell-specific responses to androgen treatment. (A) Expression of 212 human prostate-related genes was monitored in prostate cell lines treated with androgen or vehicle. Data for each condition were plotted to illustrate the impact of treatment with androgen (y-axis) vs. vehicle (x-axis). (B) Agglomerative clustering using data from the 212 genes. For each cell line, three randomly selected biological replicates are presented for both androgen-treated (filled box) and untreated (open box) conditions. The distance between subclusters (y-axis, the height) measures the divergence of their expression patterns.

Figure 5

Reproducible gene-expression profiling of laser-captured samples. Total RNAs from laser-captured samples from different numbers of cells were assayed in triplicate. Cell number indicates the number of cells' worth of RNA input to each DASL assay. Correlation between individual replicates is shown by plotting the intensities for two replicates against the intensities of the third.

Figure 6

Agilent Bioanalyzer image of PC-3 RNA samples after incubation at 95°C for various periods of time. (L) RNA size ladder; (lanes 1–7) intact RNA; (lanes 2–6) 5, 10, 20, 40, and 60 min incubation at 95°C.