Expression responses to the heat treatment in N. noltii were very different between the northern and southern population with a very weak response in the southern and a strong
response in the northern population ( Fig. 1). We further investigated genes responsible for the divergent expression in the northern population. Because no biological replication was available, we modeled the biological variation in response to heat for N. noltii via the biological variation between treatments of the southern population. Investigation of the strong northern response revealed differential expression of 369 genes between treatments with 28 genes up-regulated and 341 genes down-regulated upon SP600125 heat treatment (see Section 2. “Differential gene expression”"; workflow: Fig. S4; Table S2). The up-regulated set of genes in the northern CHIR-99021 ic50 population consisted of only 28 genes, none of which encoded an HSP gene or were enriched in any functional category
(Table S2). Conversely, the large set of 341 down-regulated genes in response to heat included enriched functions for cell wall modification, synthesis and degradation, hormone metabolism (brassinosteroids and gibberelins), protein synthesis and various functions combined under “misc” (Fig. 4). Although “stress” associated functions were not significantly enriched, various subcategories were present [Fig. 4; “stress.abiotic”: 1 gene (osmotin 34), “stress.abiotic.cold”: 2 genes, “stress.abiotic.drought/salt”: mafosfamide 4 genes, “stress.abiotic.heat”: 1 gene (heat-shock protein binding) “stress.abiotic.unspecified”: 4 genes] (Table S2). Shoots from both species displayed decreased shoot counts in response to heat stress (see Section 3.6 “Effects of the heat wave simulation on population performance”). We therefore investigated the role of HSP expression in both species, as HSPs are well known markers for heat stress. For each species, expression profiles for all 78 genes annotated with the functional term “stress.abiotic.heat” of all four libraries were compared with the constructed maximum and minimum expression profile of the respective
species via MDS analysis. These constructed maximum (minimum) expression profiles of HSP genes for each species were obtained by taking the maximum (minimum) expression value of each gene out of the four respective libraries. For N. noltii, none of the libraries grouped with the maximal expression profile (Fig. S6A). In contrast, heat-treated libraries of Z. marina showed a clear grouping with the constructed maximal expression profile, while control libraries were more similar to the minimum expression profile (Fig. S6B). This suggests that while HSPs were up-regulated under the simulated heat at 26 °C in Z. marina, no up-regulation of well-known members of the heat shock protein family occurred in N. noltii.