3.1.2 ⎪ Expression of CBF1–3 genes and leaf transcriptome
In both ecotypes, the strongest CBF1–3 transcript expression was also seen in plants grown under HLC (Fig. 3). As was the case for photosynthetic capacity and leaf dry mass per area, CBF1 andCBF3 expression were also greater in SW compared to IT plants grown under HLC. In plants grown under LLC on the other hand, only IT moderately induced CBF1–3 but SW did not (Fig. 3).
Growth under HLC compared to LLW also resulted in sweeping differences in the leaf transcriptome in both ecotypes (Fig. 4). 1912 and 1415 genes were induced in HLC-grown IT and SW, respectively, with an adjustedP -value of less than 0.01 and a minimum fold-change of 2 (Table S1, S2). Similar numbers of genes were downregulated under HLC compared to LLW, i.e., 1671 and 1531 genes, for IT and SW, respectively (Tables S3, S4). For both ecotypes grown under each of the four conditions, all three biological replicates co-clustered (after hierarchical clustering) and the transcriptomic response of IT and SW in HLC conditions co-branched (Fig. 4a). This co-branching of HLC transcriptomes of the two ecotypes was due at least in part to large blocks of co-clustering genes either induced specifically or downregulated under HLC in both ecotypes (Fig. 4a, Tables S5–S7). Genes specifically induced under HLC were strongly enriched for a number of gene ontology (GO) categories, of which the most enriched unique categories were starch catabolism/starch metabolism (GO:0005983/GO:0005982), sucrose biosynthesis (GO:0005986), and cold acclimation (GO:0009631) (Table S8). GO analysis also revealed pathways repressed specifically in HLC; the three categories most strongly downregulated in HLC were water transport (GO:0006833), brassinosteroid metabolism (GO:0016131), and auxin polar transport (GO:0009926) (Table S9).
In addition to these shared responses, there were also substantial differences between IT and SW in how gene expression responded to HLC (Table S10-16). In SW compared to IT, growth under HLC conditions enriched more strongly for induction of genes in the overlapping GO categories of cold acclimation (GO:0009631), response to water deprivation (GO:0009414), and response to abscisic acid (GO:0009737) (Table S11). Conversely, the three most enriched unique GO categories among genes with significantly higher induction in IT compared to SW in HLC were protein refolding (GO:0042026), glucosinolate biosynthesis (GO:0019761), and phosphate starvation response (GO:0016036) (Table S14). Moreover, genes exhibiting greater down-regulation in HLC in IT compared to SW were enriched for cytokinin-activated signaling pathway (GO:0009736), auxin polar transport (GO:0009926), and response to auxin (GO:0009733) (Table S16).
Genes in the photosynthesis GO category (GO:0015979) also exhibited unique expression patterns in each ecotype under HLC (Fig. S2). With the exception of the light-stress-induced light-harvesting complexLHCB4.3 (AT2G40100), genes involved in light harvesting were downregulated in both ecotypes under HLC, but more so in IT relative to SW. Conversely, genes more strongly induced in IT compared to SW (Fig. S2) had in common that they are typically induced under abiotic and/or oxidative stress (see discussion), such as chloroplast glucose-6 phosphate/phosphate translocator GPT2 (AT1G61800), chloroplast envelope K+/H+ antiporterKEA2 (AT4G00630), phosphofructokinase (AT1G76550), cytosolic fumarase (AT5G76550), ferritin (AT2G30400/AT3G56090), and pyridoxal phosphate synthase protein (AT5G01410).
For both ecotypes, genes preferentially induced under HLC were enriched for those that had also been induced by CBF1–3 overexpression in the absence of either high light or cool temperatures (Park et al. 2018) (P -values of 10−19 and 10−18 for IT and SW, respectively; Fig. 4b, Table S17). Moreover, these genes preferentially induced under HLC were also enriched for genes expressed at lower levels in it:cbf123 and sw:cbf123 following sudden transfer from warm growth conditions to 4°C for 24 h (Park et al. 2018) (P -values of 10−28 and 10−38, for IT and SW, respectively; Table S18). Overall, while CBF1–3 target genes (Jiaet al. 2016; Park et al. 2018) were strongly induced in both ecotypes in HLC, these genes tended to be more strongly induced in SW compared to the IT ecotype in this condition. Examples for genes in this previously defined CBF1–3-regulated group that were more strongly induced in SW included the cold-acclimation-regulating protein kinaseCIPK25 (AT5G25110) and a suite of cold-induced dehydrin proteins,COR47 (AT1G20440), LTI30 (AT3G50970), and LTI29(AT1G20450) (Fig. 4b). In contrast, CBF1–3-regulated protein phosphatase EGR2 (AT5G27930), a repressor of growth, was more strongly induced in IT compared to SW under HLC (Fig. 4b).