Transcriptome analysis of the vernalization response in barley (Hordeum vulgare) seedlings

Abstract

Temperate cereals, such as wheat (Triticum spp.) and barley (Hordeum vulgare), respond to prolonged cold by becoming more tolerant of freezing (cold acclimation) and by becoming competent to flower (vernalization). These responses occur concomitantly during winter, but vernalization continues to influence development during spring. Previous studies identified VERNALIZATION1 (VRN1) as a master regulator of the vernalization response in cereals. The extent to which other genes contribute to this process is unclear. In this study the Barley1 Affymetrix chip was used to assay gene expression in barley seedlings during short or prolonged cold treatment. Gene expression was also assayed in the leaves of plants after prolonged cold treatment, in order to identify genes that show lasting responses to prolonged cold, which might contribute to vernalization-induced flowering. Many genes showed altered expression in response to short or prolonged cold treatment, but these responses differed markedly. A limited number of genes showed lasting responses to prolonged cold treatment. These include genes known to be regulated by vernalization, such as VRN1 and ODDSOC2, and also contigs encoding a calcium binding protein, 23-KD jasmonate induced proteins, an RNase S-like protein, a PR17d secretory protein and a serine acetyltransferase. Some contigs that were up-regulated by short term cold also showed lasting changes in expression after prolonged cold treatment. These include COLD REGULATED 14B (COR14B) and the barley homologue of WHEAT COLD SPECIFIC 19 (WSC19), which were expressed at elevated levels after prolonged cold. Conversely, two C-REPEAT BINDING FACTOR (CBF) genes showed reduced expression after prolonged cold. Overall, these data show that a limited number of barley genes exhibit lasting changes in expression after prolonged cold treatment, highlighting the central role of VRN1 in the vernalization response in cereals.

Figure 1. Overview of sampling methods for microarray analysis.

A) Barley seeds (cv. Sonja) were germinated and grown in darkness at either 20°C over 5 days (control) or 4°C over 49 days (prolonged cold). Seedlings were then shifted from the control treatment to 4°C for 24 hours (short term cold) or shifted from the prolonged cold treatment to 20°C for 24 hours (post cold). In all treatments the shoot apex remained at an early stage of vegetative development, but plants grown from seedlings that experienced prolonged cold flower rapidly when shifted to normal growth conditions, unlike control seedlings germinated at 20°C . B) To identify contigs that show a sustained response to prolonged cold, barley seeds were germinated in the dark at 4°C for 49 days and then transferred to growth in glasshouse conditions until they reached the three leaf stage (10 days after the end of cold treatment). Non-vernalized control plants were grown simultaneously under the same conditions and were sampled at the equivalent developmental leaf stage (14 days). Grey shading indicates low-temperature.

Figure 2. Changes in gene expression in seedlings after different low-temperature treatments.

A) The total numbers of contigs showing differential expression (>2 fold change, p<0.01) in the short term cold versus control (ST-C), the prolonged cold versus control (PLC-C) or the post cold versus control (POC-C) samples. B) The total numbers of genes up-regulated in the same comparisons. C) Genes down-regulated in the same comparisons. D) The total number of contigs up or down-regulated in the prolonged versus short term cold treatments. E) The number of contigs up or down-regulated in the post versus prolonged cold treatments. Dark shading indicates 8 fold change of greater, intermediate shading indicates 4–8 fold, light shading indicates 2–4 fold change.

Figure 3. Venn diagram showing the differentially expressed contigs across the different treatments.

A summary of the contigs that showed a two fold or greater change in transcript levels across the different treatments when compared to the control treatment (p<0.1). Shaded area indicates contigs that were significantly changed in the samples treated with prolonged cold and one day after prolonged cold treatment, relative to the control.

Figure 4. Principle component analysis of microarray data.

Principal component analysis was applied to differentially expressed contigs identified in the different seedling treatments (see methods). Closed triangles (▴) indicate samples from the no cold control treatment. Open triangles (Δ) represent the 1 day after prolonged cold treatment. Closed squares (▪) represent the short term cold treatment. Open squares (□) represent the prolonged cold treatment.

Figure 5. K-means cluster analysis of differentially expressed contigs.

The mean for each cluster is shown as black dots and lines and the gray lines represent the expression pattern of individual contigs. The three replicates for each treatment data are shown.

Figure 6. Examples of temperature responsive contigs.

The normalized expression levels for individual contigs (A–L), as assayed by Affymetrix Barley1 Chip (RMA normalization) for control seedlings, short term or prolonged cold treatment, and one day after prolonged cold (a). Also shown are expression levels for contigs in the fully expanded second leaf of plants after vernalization (10 days post-vernalization), versus the equivalent leaves from non-vernalized control plants (b).