Transcriptional analysis of highly syntenic regions between Medicago truncatula and Glycine max using tiling microarrays

Abstract

Background: Legumes are the third largest family of flowering plants and are unique among crop species in their ability to fix atmospheric nitrogen. As a result of recent genome sequencing efforts, legumes are now one of a few plant families with extensive genomic and transcriptomic data available in multiple species. The unprecedented complexity and impending completeness of these data create opportunities for new approaches to discovery.

Results: We report here a transcriptional analysis in six different organ types of syntenic regions totaling approximately 1 Mb between the legume plants barrel medic (Medicago truncatula) and soybean (Glycine max) using oligonucleotide tiling microarrays. This analysis detected transcription of over 80% of the predicted genes in both species. We also identified 499 and 660 transcriptionally active regions from barrel medic and soybean, respectively, over half of which locate outside of the predicted exons. We used the tiling array data to detect differential gene expression in the six examined organ types and found several genes that are preferentially expressed in the nodule. Further investigation revealed that some collinear genes exhibit different expression patterns between the two species.

Conclusion: These results demonstrate the utility of genome tiling microarrays in generating transcriptomic data to complement computational annotation of the newly available legume genome sequences. The tiling microarray data was further used to quantify gene expression levels in multiple organ types of two related legume species. Further development of this method should provide a new approach to comparative genomics aimed at elucidating genome organization and transcriptional regulation.

Figure 1

Tiling microarray analysis of the 1 Mb syntenic regions. A representative Gene Browser window is shown in which predicted genes are aligned to the chromosomal coordinates. Arrows indicate the direction of transcription. The interrogating tiling probes are also aligned to the chromosome coordinates with the fluorescence intensity value depicted as a vertical bar in the six organ types. From top to bottom: nodule, root, stem, leaf, flower and seed.

Figure 2

Tiling microarray detection of the predicted genes in the 1 Mb region syntenic between barrel medic and soybean. (a) Pie charts showing the number and percentage of genes detected by tiling arrays in at least one of the six examined organ types. (b) Tiling array detection rates of predicted genes in the six organ types in barrel medic and soybean.

Figure 3

Analysis of the frequency of TARs in different organ types. (a) Percentage and number of TARs detected by tiling arrays in one, two, three, four, five and all six organ types in barrel medic and soybean. (b) Organ-specific number of TARs detected from only one organ type by tiling arrays in barrel medic and soybean.

Figure 4

Classification of TARs based on physical location relative to the predicted genes. (a) Pie charts showing percentage of all identified TARs in different genome components relative to the predicted gene structures in barrel medic and soybean. (b) Number of non-exonic TARs in different sub-genic regions in barrel medic and soybean.

Figure 5

Analysis of differentially expressed genes. Heat maps represent unsupervised clustering of differentially expressed genes in barrel medic and soybean. The red, yellow, and blue colors depict positive deviation, no deviation, and negative deviation of the transcription level, respectively.

Figure 6

Verification of tiling array detected differentially expressed genes. (a) RT-PCR analysis of the transcript abundance in six organ types for six selected soybean genes that are preferentially expressed in the nodule. Total RNA (5 μg) was reverse transcribed and 5% of the product used as template for PCR, which was carried out for 35 cycles. (b) Organ type-specific variation of the expression level of the gene Gm_121, as determined by median-polishing of the tiling array data. Dashed lines indicate the deviation value at p = 0.001.

Figure 7

Analysis of the transcription patterns of collinear genes. The collinear genes in both barrel medic and soybean are ordered by chromosome position. For each gene, the deviation of transcription level was calculated based on median polishing for the six organ types (see Materials and methods). The gene order was then plotted against the corresponding deviation value in each of the six organ types, which is color-coded.