*2.3. Identification of Genes Associated with HT Adaptation in Kenshin*

To identify HT adaptation-related genes, we adopted several selection criteria based on the results shown in Table 2. The selection criteria were: (1) expression levels in Kenshin at 22 ◦C over 2-fold higher than those in Chiifu (we hypothesized that genes essential for long-term adaptation to HT would exhibit high basal levels of expression); and (2) expression levels under both warming and minor-warming conditions at least 2-fold higher in Kenshin than in Chiifu. Sixty-four genes were identified (Table 3), including 16 genes that were upregulated in both lines under minor-warming and warming conditions (asterisks in Table 3): 10 HSR genes, 2 TF (transcription factor) genes, 1 SF (splicing factor) gene, and 3 other genes. The two transcription factor genes were *BrPIF6* (Bra007660; phytochrome-interacting factor 3-like 2; PIL2/PIF6) and *Bra006853* (MYB-like transcription factor family protein). Genes in other categories included *Bra037453* (disease resistance protein (CC-NBS-LRR class) family) for "response to stress", *BrGLN1.3* (Bra021276; glutamine synthetase 1.3)

for "ligase activity", and *BrROT3* (Bra011678; cytochrome P450 superfamily protein (ROT3)) for "lipid biosynthesis process". These 16 genes might play important roles in HT adaptation in *B. rapa*. We subjected three of these genes to further analysis: *BrHSFA2*, *BrHSP18.2s*, and *BrSMP1* (Bra023741). and *BrSMP1* (Bra023741). *2.4. Comparison of HT-Related Gene Expression between B. rapa and Arabidopsis* 

*Int. J. Mol. Sci.* **2018**, *19*, x FOR PEER REVIEW 13 of 22

synthetase 1.3) for "ligase activity", and *BrROT3* (Bra011678; cytochrome P450 superfamily protein

adaptation in *B. rapa*. We subjected three of these genes to further analysis: *BrHSFA2*, *BrHSP18.2s*,

#### *2.4. Comparison of HT-Related Gene Expression between B. rapa and Arabidopsis* Genes involved in thermotolerance (basal thermotolerance, acquired thermotolerance, and warming tolerance) and the associated marker genes are well known in the model plant *Arabidopsis*.

Genes involved in thermotolerance (basal thermotolerance, acquired thermotolerance, and warming tolerance) and the associated marker genes are well known in the model plant *Arabidopsis*. To determine whether the same set of genes functions in *B. rapa*, we compared the expression patterns of these genes and other *HSP* genes with our microarray data (Table 4). Warming genes (*PHYB*, *HSP70*, and *PIF4*) identified in *Arabidopsis* were highly expressed in all *B. rapa* samples, with no notable increase in expression upon warming treatment, implying that genes responsible for long-term HT adaptation in *B. rapa* are different from *Arabidopsis* warming genes. In other cases, we assumed differences among samples, such as two contrasting lines in *B. rapa* vs. an ecotype of *Arabidopsis*. *BrPIF4* (Bra000283) expression appeared to be somewhat related to HT adaptation in *B. rapa*. Two Chinese cabbage genes homologous to acquired thermotolerance-related genes in *Arabidopsis*, *BrROF2* and *BrHSFA2*, appear to be critical for warming adaptation in *B. rapa*. The expression levels of several *HSP* genes were also consistent with warming treatment, pointing to their possible involvement in adaptation to HT. To determine whether the same set of genes functions in *B. rapa*, we compared the expression patterns of these genes and other *HSP* genes with our microarray data (Table 4). Warming genes (*PHYB*, *HSP70*, and *PIF4*) identified in *Arabidopsis* were highly expressed in all *B. rapa* samples, with no notable increase in expression upon warming treatment, implying that genes responsible for long-term HT adaptation in *B. rapa* are different from *Arabidopsis* warming genes. In other cases, we assumed differences among samples, such as two contrasting lines in *B. rapa* vs. an ecotype of *Arabidopsis*. *BrPIF4* (Bra000283) expression appeared to be somewhat related to HT adaptation in *B. rapa*. Two Chinese cabbage genes homologous to acquired thermotolerance-related genes in *Arabidopsis*, *BrROF2* and *BrHSFA2*, appear to be critical for warming adaptation in *B. rapa*. The expression levels of several *HSP* genes were also consistent with warming treatment, pointing to their possible involvement in adaptation to HT. *2.5. Confirmation of Microarray Data via qRT-PCR* 

#### *2.5. Confirmation of Microarray Data via qRT-PCR* To confirm the expression levels of the genes detected by microarray analysis, we performed

To confirm the expression levels of the genes detected by microarray analysis, we performed RT-PCR analysis of several selected genes (Figure 5). Although RT-PCR appears to be less sensitive than microarray analysis, RT-PCR results are often used to support microarray data. The expression levels of most of these genes increased upon warming treatment (28 ◦C), with maximum levels detected at 45 ◦C. These genes included three heat-acclimation-related genes (*BrHSFA2*, *BrHSFB2A*, and *BrROF2*), various *HSP* genes (especially *sHSP*s), peroxidase family genes, and others. Several genes showed high basal expression levels that further increased upon warming conditions: *BrHSP98.7*, *BrHSP70*, *BrHSP21*, three *BrHSP20L*s (Bra30910, Bra01883, Bra01884), and *BrMPSR1*. As expected, all of these genes showed higher basal expression levels in Kenshin than in Chiifu. RT-PCR analysis of several selected genes (Figure 5). Although RT-PCR appears to be less sensitive than microarray analysis, RT-PCR results are often used to support microarray data. The expression levels of most of these genes increased upon warming treatment (28 °C), with maximum levels detected at 45 °C. These genes included three heat-acclimation-related genes (*BrHSFA2*, *BrHSFB2A*, and *BrROF2*), various *HSP* genes (especially *sHSP*s), peroxidase family genes, and others. Several genes showed high basal expression levels that further increased upon warming conditions: *BrHSP98.7*, *BrHSP70*, *BrHSP21*, three *BrHSP20L*s (Bra30910, Bra01883, Bra01884), and *BrMPSR1*. As expected, all of these genes showed higher basal expression levels in Kenshin than in Chiifu.

**Figure 5.** RT-PCR analysis of selected genes identified by microarray analysis. The expression levels of these genes obtained by microarray analysis are summarized in Table S16. **Figure 5.** RT-PCR analysis of selected genes identified by microarray analysis. The expression levels of these genes obtained by microarray analysis are summarized in Table S16.

#### *2.6. Expression of BrHSFA2 and BrHSP18.2 2.6. Expression of BrHSFA2 and BrHSP18.2*

Based on our data (Tables 3 and 4) and previous reports [22–24], we selected *BrHSFA2* and *BrHSP18.2A-C* for further analysis of their possible involvement in response to warming and HT conditions. To examine whether *BrHSFA2* undergoes alternative splicing upon HT exposure, as does Based on our data (Tables 3 and 4) and previous reports [22–24], we selected *BrHSFA2* and *BrHSP18.2A-C* for further analysis of their possible involvement in response to warming and HT conditions. To examine whether *BrHSFA2* undergoes alternative splicing upon HT exposure, as does *Arabidopsis HSFA2*, and whether the intron sequences of this gene are the same in Kenshin and Chiifu, we cloned and sequenced at least 10 clones of the *BrHSFA2* intron region including part of

*BrHSFA2* mRNA.

Exon 1 and Exon 2 from both lines. These 548 bp fragments, including the 337 bp intron sequence, were 100% identical between Chiifu and Kenshin (NCBI accession MH310901, MH310902). We then compared *BrHSFA2* with the homologous sequence from *Arabidopsis* to investigate whether *BrHSFA2* also undergoes alternative splicing (Figure S1). As shown in Figure S1B, *BrHSFA2* might contain a mini-exon with a TAG stop codon, which produces a truncated version of the BrHSFA2 polypeptide via alternative splicing. The truncated version of BrHSFA2 has different C-terminal amino acids from *Arabidopsis* HSFA2 (Figure S1C). Exon 1 and Exon 2 from both lines. These 548 bp fragments, including the 337 bp intron sequence, were 100% identical between Chiifu and Kenshin (NCBI accession MH310901, MH310902). We then compared *BrHSFA2* with the homologous sequence from *Arabidopsis* to investigate whether *BrHSFA2* also undergoes alternative splicing (Figure S1). As shown in Figure S1B, *BrHSFA2* might contain a mini-exon with a TAG stop codon, which produces a truncated version of the BrHSFA2 polypeptide via alternative splicing. The truncated version of BrHSFA2 has different C-terminal amino acids from *Arabidopsis* HSFA2 (Figure S1C).

*Int. J. Mol. Sci.* **2018**, *19*, x FOR PEER REVIEW 14 of 22

Chiifu, we cloned and sequenced at least 10 clones of the *BrHSFA2* intron region including part of

To confirm that alternative splicing occurs in *BrHSFA2*, we carried out qRT-PCR with a reverse primer consisting of possible mini-exon-derived mRNA (Table S1; Figure 6). The levels of an alternatively spliced form of the transcript were higher in Kenshin than in Chiifu upon warming conditions (Figure 6D), while the total transcript levels (full-length + alternatively spliced form) were higher in Chiifu (Figure 6C). The level of full-length mRNA was higher in Kenshin than in Chiifu (Figure 6B). These results indicate that alternative splicing occurs in the intron of *BrHSFA2*, that this process facilitates the expression of full-length *BrHSFA2* as in *Arabidopsis*, and that the levels of the alternatively spliced form of this gene are higher in Kenshin (adapted to HT) than in Chiifu. To examine any association of the alternative splicing of *BrHSFA2* with *BrSMP1*, encoding a spliceosome component and a candidate gene involved in HT adaptation in Kenshin (Table 3), we examined the expression of *BrSMP1* under the same conditions (Figure S3). The expression pattern of *BrSMP1* upon warming and HS was proportional to *BrHSFA2* expression, suggesting the possible involvement of BrSMP1 in alternative splicing of *BrHSFA2* upon HT treatment. To confirm that alternative splicing occurs in *BrHSFA2*, we carried out qRT-PCR with a reverse primer consisting of possible mini-exon-derived mRNA (Table S1; Figure 6). The levels of an alternatively spliced form of the transcript were higher in Kenshin than in Chiifu upon warming conditions (Figure 6D), while the total transcript levels (full-length + alternatively spliced form) were higher in Chiifu (Figure 6C). The level of full-length mRNA was higher in Kenshin than in Chiifu (Figure 6B). These results indicate that alternative splicing occurs in the intron of *BrHSFA2*, that this process facilitates the expression of full-length *BrHSFA2* as in *Arabidopsis*, and that the levels of the alternatively spliced form of this gene are higher in Kenshin (adapted to HT) than in Chiifu. To examine any association of the alternative splicing of *BrHSFA2* with *BrSMP1*, encoding a spliceosome component and a candidate gene involved in HT adaptation in Kenshin (Table 3), we examined the expression of *BrSMP1* under the same conditions (Figure S3). The expression pattern of *BrSMP1* upon warming and HS was proportional to *BrHSFA2* expression, suggesting the possible involvement of BrSMP1 in alternative splicing of *BrHSFA2* upon HT treatment.

**Figure 6.** *BrHSFA2* expression in Chiifu and Kenshin during warming and heat shock treatments. **q**RT-PCR was performed with primer sets described in Table S1 and data analysis was carried out using qPCR value of three replicates. (**A**) Genomic organization of *BrHSFA2* and possible mRNAs with primer positions indicated; (**B**) Full-length *BrHSFA2* mRNA levels; (**C**) *BrHSFA2* mRNA containing both full-length and truncated (alternatively spliced) forms; (**D**) Truncated form of **Figure 6.** *BrHSFA2* expression in Chiifu and Kenshin during warming and heat shock treatments. qRT-PCR was performed with primer sets described in Table S1 and data analysis was carried out using qPCR value of three replicates. (**A**) Genomic organization of *BrHSFA2* and possible mRNAs with primer positions indicated; (**B**) Full-length *BrHSFA2* mRNA levels; (**C**) *BrHSFA2* mRNA containing both full-length and truncated (alternatively spliced) forms; (**D**) Truncated form of *BrHSFA2* mRNA.

HSFA2 is responsible for maintaining HS memory up to two days in *Arabidopsis* by maintaining histone methylation, thereby enabling the quick induction of HSR genes upon recurring HS [23,24]. HSFA2 has the most pronounced effect on *Arabidopsis HSP18.2* (*Hsp18.1-CI*/AT5G59720) [23]. *B. rapa* HSFA2 is responsible for maintaining HS memory up to two days in *Arabidopsis* by maintaining histone methylation, thereby enabling the quick induction of HSR genes upon recurring HS [23,24]. HSFA2 has the most pronounced effect on *Arabidopsis HSP18.2* (*Hsp18.1-CI*/ AT5G59720) [23]. *B. rapa* possesses three homologs corresponding to *AtHSP18.2*, *BrHSP18.2A*

(Bra002539), *BrHSP18.2B* (Bra020295), and *BrHSP18.2C* (Bra006697), as listed in order from the highest to lowest identity with *AtHSP18.2*. These three genes encode highly identical polypeptides (93–96% identity) (Figure S2), but we successfully generated primer sets to distinguish each gene (Table S1). A gradual increase in temperature (by 5 ◦C every 2 h [HS]) strongly upregulated all three genes at 37 ◦C, with the greatest increase observed for *BrHSP18.2A* and *BrHSP18.2B* in Kenshin (Table 5). However, under warming conditions, compared with 27 ◦C during HS treatment, tremendous increases were observed in the expression of *BrHSP18.2A* and *BrHSP18.2B* in Kenshin, but there was a several-fold higher increase in *BrHSP18.2C* expression in Chiifu than in Kenshin. These results appear to reflect the different responses of *B. rapa* to HT acclimation compared with *Arabidopsis*, as well as differences between Kenshin and Chiifu.

**Table 5.** Expression of *BrHSP18.2* family genes in Chiifu and Kenshin under various temperature conditions. Expression level (fold change) was calculated based on qRT-PCR values of three replicates using *BrACT2* as a standard. Heat shock treatment was performed by increasing the temperature 5 ◦C every 2 h.

