The Identification of MATE Antisense Transcripts in Soybean Using Strand-Specific RNA-Seq Datasets

Abstract

Natural antisense transcripts (NATs) have been generally reported as negative regulators of their sense counterparts. Multidrug and toxic compound extrusion (MATE) proteins mediate the transport of various substrates. Although MATEs have been identified genome-wide in various plant species, their transcript regulators remain unclear. Here, using the publicly available strand-specific RNA-seq datasets of Glycine soja (wild soybean) which have the data from various tissues including developing pods, developing seeds, embryos, cotyledons and hypocotyls, roots, apical buds, stems, and flowers, we identified 35 antisense transcripts of MATEs from 28 gene loci after transcriptome assembly. Spearman correlation coefficients suggested the positive expression correlations of eight MATE antisense and sense transcript pairs. By aligning the identified transcripts with the reference genome of Glycine max (cultivated soybean), the MATE antisense and sense transcript pairs were identified. Using soybean C08 (Glycine max), in developing pods and seeds, the positive correlations between MATE antisense and sense transcript pairs were shown by RT-qPCR. These findings suggest that soybean antisense transcripts are not necessarily negative transcription regulators of their sense counterparts. This study enhances the existing knowledge on the transcription regulation of MATE transporters by uncovering the previously unknown MATE antisense transcripts and their potential synergetic effects on sense transcripts.

Keywords: MATE (multidrug and toxic compound extrusion) transporter; natural antisense transcript (NAT); sense transcript; soybean; strand-specific RNA-seq.

Figure 1

A heatmap of the predicted relative expression levels of MATE antisense transcripts identified from strand-specific RNA-seq datasets of pod (7 DAF), pod (14 DAF), pod (40 DAF), seed (14 DAF), seed (40 DAF), embryo, cotyledon and hypocotyl, root, apical bud, stem, and flower of soybean accession W05 (Table S1) [30,31].

Figure 2

Antisense MATE transcripts and their sense counterparts in soybean accession W05 with positively correlated expression levels identified based on the publicly available RNA-seq datasets. Left panel. (A) A heatmap of the expression levels of the antisense-sense transcript pairs (joined by a square bracket) that are predicted to be correlated, with the corresponding Spearman correlation coefficients. *Gsoja_MSTRG.47510 and Gsoja_MSTRG.47511 are both antisense transcripts corresponding to the sense transcript Glysoja.11G029339, with the Spearman correlation coefficients being 0.68 and 0.74 respectively; (B) the genomic locations of three of the antisense-sense transcript pairs having high Spearman correlation coefficients are showcased.

Figure 3

The expression levels of the antisense and their corresponding sense transcripts in the developing seed and pod were analyzed using RT-qPCR. The antisense and sense pairs are (A) MSTRG.9425 and Glyma03G0018400; (B) MSTRG.47510 and Glyma.11G026300; (C) MSTRG.45962 and Glyma.10G287600; (D) MSTRG.5534 and Glyma.02G097800; (E) MSTRG.84397 and Glyma.19G120900 (GmMATE4 [20]); (F) The color key of the bars shown in panel A–E. Three RT-qPCR reactions were done for each template and primer pair combination. One RT-qPCR was regarded as one technical replicate. Each bar represents the mean of three technical replicates ± standard deviation. The expression was normalized to the reference gene, VPS [38] using the 2^−ΔΔCT method. Similar expression trends were obtained from another biological replicate using a separate set of seed and pod from three other plants. Results of the biological replicates are shown in Supplementary Figure S1.