BADGE, Beads Array for the Detection of Gene Expression, a high-throughput diagnostic bioassay

Abstract

Several methods are presently available for gene expression analysis. However, few of them are suitable for detection of moderate numbers of genes in thousands of samples with high speed and low cost. There is great demand for such a method for use in diagnostics and screening. To address this need, we have developed an assay for gene expression analysis using microspheres and a fluidic instrument made by Luminex. The assay is named Beads Array for the Detection of Gene Expression (BADGE). BADGE can monitor up to 100 genes in a single reaction, and it takes only 1 h to hybridize and <20 sec to read the results of all 100 genes in a sample for the detection process. For the genes detected in five independent replicate experiments, the standard deviation was <35% of the mean. We have monitored multiple pathogenesis-related genes simultaneously in chemical-treated and control Arabidopsis samples employing the BADGE assay. The data were compared with those obtained from an established technology, Affymetrix GeneChip. The changes in expression profiles were very similar. Our study showed that the BADGE assay was capable of profiling expression of multiple genes at affordable cost and rapid speed.

Figure 1

Specificity of hybridization was determined in the hexaplexed assay (A) and the icosaplexed assay (B). Oligonucleotides representing 20 genes were coupled to 20 different color-coded microspheres. Various amounts of the complementary target oligonucleotides tagged with biotin were mixed in four tubes as indicated in Table 2. NTSR served as a negative control in the experiment. (A) Only the first six targets (see Table 2) and their corresponding capture-probe-coupled microspheres were added to the assay. (B) All 20 targets and their corresponding capture-probe-coupled microspheres were added to the assay. In both experiments A and B the background fluorescence of each set of microspheres was subtracted.

Figure 2

Specific hybridization in the presence of biotin-labeled wild-type Arabidopsis cRNA. The M13 capture probe was coupled to microspheres and the complementary sequence tagged with biotin was used as a target in this experiment. The M13 target oligonucleotide was added as 0.1, 1.0, 10.0, and 100.0 fmoles to the assay in the absence or presence of the indicated amount of biotin-labeled cRNA prepared from wild-type Arabidopsis leaves. In the presence of 10 μg of cRNA, the addition of 0.1, 1.0, 10.0, and 100.0 fmoles of target was approximately equal to a spiking ratio of 1:300,000, 1:30,000, 1:3,000, and 1:300, respectively.

Figure 3

Simultaneous detection of 20 genes from either BTH- or water-treated Arabidopsis samples as a function of time (1, 3, 5, 7, and 9 d). The left panels (A,C,E,G) are the results detected from BTH-treated samples, and the right panels (B,D,F,H) are the results detected from water-treated samples (control). The 20 genes were organized into four groups in this figure for presentation according to their properties and their expression levels affected by BTH treatment. The first group contained BTH-induced genes (A,B), the second group had unchanged PR genes (C,D), the third group included BTH-repressed genes (E,F), and the fourth group consisted of control genes (G,H).

Figure 4

Different amounts of cRNA (1.25, 2.5, 5.0, and 10 μg) prepared from Arabidopsis samples at 3 d after treatment with either BTH (A) or water (B) were used in this experiment to test if the signals were linear with the target amount. The results of the three PR genes from the first group classified in Figure 3 (PR5, PR1, and PRX1, induced by BTH) and the results from an internal control UBQ4 are presented in this figure. Linear signals were detected from 1.25 μg to 10 μg of biotin-labeled cRNA for the three PR genes in the sample that was treated with BTH (A). In contrast, the expression levels of PR5, PR1, and PRX1 in the control sample remained low (B). However, the expression signal of the internal control UBQ4 was always linearly related to the added target cRNA amount in both BTH-treated and control samples (A,B).

Figure 5

Biotinylated cRNA samples prepared from Arabidopsis leaves at 3 d after BTH treatment (B3D) were used for this experiment. The cRNA was hybridized to either microspheres (coupled with single-probe, 1P; three-probe, 3P; and six-probe, 6P; respectively) made by Luminex or Arabidopsis GeneChip made Affymetrix in separate experiments. The hybridization signals of the genes obtained from the two technologies were normalized to the expression level of UBQ4 as shown in A. The fold change in hybridization signal of the BTH-treated sample to the control sample is presented in B.