Global gene expression profile during low temperature in the brain of grass carp (Ctenopharyngodon idellus)

Grass carp is an important commercial fish widely cultured in China. Large range of temperature, in particular extremely low temperature, has dramatic effects on the aquaculture of this teleost. However, there is relatively little research on the molecular responses in the fish exposed to cold. Given the limited vision of approaches targeting individual genes, we investigated the transcriptome profiles of brain in response to cold in order to comprehensively characterize molecular mechanisms behind it. This study indicated that the estrogen signaling pathway was inhibited in brain when grass carp acclimated to low temperature, while terpenoid backbone biosynthesis pathway and steroid biosynthesis pathway were significantly activated. Such a result implied the crucial role of cholesterol in cold acclimation. Moreover, plenty of differentially expressed genes associated with spliceosomes were enriched during cooling process, which suggested alternative splicing may be involved in the regulation of biological process in acclimation to temperature changes. In researches on extremely low-temperature tolerance, we identified four genes (DUSP1, HSPA6, NR4A1 and GADD45B) associated with MAPK signaling pathway. The four genes, extensively up-regulated at 4°C and remained relatively low expression at moderate temperature, were closely related with extremely cold condition. Further examination of the candidate genes can provide insights into the mechanisms of grass carp to endure extremely low temperature in the winter.

126 annotated to all genes of the map was more than 20%, were selected for the enrichment analysis.
127 The statistical test for enrichment analysis was performed by R scripts executing Fisher's exact test 128 with the p-value ≤0.05. 129 2.5. Validation of expression profiles using RT-qPCR 130 Given the instability of beta-actin expression in cold condition, the gene RPL13A was took as 131 the internal control which steadily expressed during the cold adaptation [28]. Then 4 genes (5 132 transcripts) were selected and the length of PCR products was about 120-250bp (Primer list in 133 Table S4) and Tm ~55 0 C. We conducted the project that decreased the water temperature from 27 0 C to 4 0 C and 138 increased it back to 27 0 C to simulate the temperature changing in nature. The RNA-seq data sets at 139 five temperature points (27 0 C -12 0 C -4 0 C -12 0 C -27 0 C) which sequentially reached during process 140 were obtained and named as A, B, C, D, E. Each data set has 8.45±0.26M clean reads with 141 96.03±0.76% mapping rate (Table S1). Subsequently, we implemented the transcriptome assembly 142 using published gene structure annotation information of grass carp as a reference [29] and 143 generated a new annotation file containing 52,580 transcripts of 37,531 genes.

144
With the reference genome information, all assembled transcripts were classified into twelve 145 categories (Table 1). Apart from 32,811 reference transcripts, we found another 19,769 new In order to investigate the consecutive and phased changes of the transcriptome profiles with 152 the temperature fluctuating, we identified cumulative differentially expressed genes (cDEGs) / 153 transcripts (cDETs) and phased differentially expressed genes (pDEGs) / transcripts (pDETs).

154
On one hand, we sequentially compared test groups B, C, D, E with control A to identify cDEGs 155 and cDETs. As the temperature dropped from 27 0 C to 12 0 C , a total of 1,863 genes were differentially 156 expressed with 4,875 cDETs. When temperature was continuously dropped to 4 0 C, the number of 157 cDEGs and cDETs increased to 2,951 and 6,244. Conversely, when the water temperature was 158 raised from 4 0 C to 12 0 C, the number of cDEGs and cDETs was slightly reduced to 2,782 and 6,134.
159 In addition, as the water temperature rose to 27 0 C, we observed a sharp drop in the number of cDEGs 160 and cDETs to 183 and 1,965 ( Figure 1A).

161
On the other hand, we separately investigated the difference of transcriptome profiles between 162 every two adjacent temperature points (B vs. A, C vs. B, D vs. C, E vs. D) to find pDEGs and pDETs.
163 Interestingly, as shown in Figure 1B  In the first stage with temperature decreasing from 27 0 C to 12 0 C, overall 10 pathways were 177 significantly enriched including three for metabolism (glycine, serine and threonine metabolism, 178 steroid biosynthesis, terpenoid backbone biosynthesis), two for apoptosis (necroptosis, ferroptosis), 179 two for translation (RNA transport, mRNA surveillance pathway), one for transcription (Spliceosome), 180 one for protein metabolism (proteasome), and one for immune system (antigen processing and 181 presentation).

182
In the second stage of cooling with temperature down to 4 0 C, we found 9 conspicuous pathways: 183 estrogen signaling pathway, steroid biosynthesis, spliceosome, phagosome, antigen processing and 184 presentation, proximal tubule bicarbonate reclamation, and three pathways for protein metabolism 185 (proteasome, protein export, protein processing in endoplasmic reticulum).

186
After stayed one day in 4 0 C, the water was slowly heated to 12 0 C. Although the pathway of

210
The cold adaptation of fish was known as a continuous process which may last from days to 211 months with the transcriptome consecutively changing [30]. However, at very low temperature in 212 winter, many eurythermic fishes are able to enter dormancy, a dramatically distinct physiological 213 status[31]. The obviously different response to moderate cold condition and extremely low 214 temperature implies dissimilar mechanisms for cold adaptation. Grass carp enters dormancy when 215 the ambient temperature is below 10 0 C. Therefore, we assume that grass carp have different 216 mechanisms for cold acclimation during moderate cold and extremely cold periods. In our research, 217 we attempted to illuminate the cold adaptation mechanism which functions throughout low 218 temperature period and the low-temperature tolerance mechanism which activates during winter 219 dormancy.

272
Similarly, our results revealed many transcript isoforms were sensitive to cold and have distinct 273 expressional patterns from each other. Besides, we found the spliceosome significantly enriched 274 throughout whole low-temperature phase. Therefore, we speculated that this post-transcriptional 275 regulation was involved in the process of adaptation to cold for grass carp. While the cold adaptation of fish is a chronic and gradual process, extreme cold certainly bring 278 more stress to organisms than moderate cold conditions. Clearly, the fact that the capability to 279 survive from radically low temperature is crucial to fishes implies extremely low temperature 280 possibly triggers unique mechanisms. In order to reveal the mechanisms of low-temperature 281 tolerance in grass carp, we focused on the differentially expressed genes at 4 0 C compared to 12 0 C.
282 Surprisingly, the pDEGs (84) between 12 0 C and 4 0 C was dramatically less than those (1,863)  In general, MAPKs directly regulate the protein function through phosphorylation or indirectly 288 influence the biological process by signal transduction. This pathway has been extensively 289 concerned in freeze-resistance of plants [45,46], and has recently been enriched in the liver 291 signaling pathway is involved in the adaptation to low temperature in plants and animals, the genes 292 involved are different. Although the specific mechanism of MAPK signaling pathway in cold 293 acclimation is unclear, the correlation between relevant genes and cold endurance of organism 294 have been increasingly investigated. In present research, we focused on genes associated with 295 MAPK signaling pathway and analyzed their expression patterns in response to temperature 296 changes. Furthermore, we verified their expression profile by RT-qPCR.

297
Dual specificity phosphatase 1 (DUSP1) is a phosphatase which specifically phosphorylate 298 tyrosine and threonine, and its expression increased in our RNA-seq data at 4 0 C, consistent with 299 RT-qPCR reconfirmation. Knocking down DUSP1 can significantly increase the apoptotic rate of 300 zebrafish ZF4 cells in low temperature [48]. This gene was also found highly related with low-301 temperature-induced embryonic diapause in Blue-breasted quail [49].

302
Nuclear receptor subfamily 4 group A member 1 (NR4A1), a member of the steroid-thyroid 303 hormone-retinoid receptor superfamily, was confirmed to be upregulated in brown adipose tissue 304 (BAT) of mice exposed to cold [50]. Relatively few studies reported the effect of NR4A1 in the social 305 behavior in zebrafish [51,52], and no report has been publicated about the relationship between 306 NR4A1 and cold adaptation of fish. However, we found NR4A1 significantly up-regulated at 4 0 C in 307 grass carp brain, while its expression remained in low-level at higher temperatures. The notable 308 expression pattern of NR4A1 implied an important role in grass carp to tolerate the extremely low 309 temperature.

310
Heat shock 70kDa protein 6 gene (HSPA6) is the 6th member of heat shock protein family A 311 (Hsp70), a famous molecular chaperones family. The members of Hsp70 were often found active in 312 the studies of cold acclimation or cold shock [53,54]. Likewise, we found that HSPA6 expression 313 increased significantly as the water temperature decreased, and the two isoforms of HSPA6 314 presented different expression patterns. One isoform (HSPA6.1) was highly up-regulated at 4 0 C, 315 while the other isoform (HSPA6.2) incrementally expressed with temperature cooling from 27 0 C to 316 4 0 C. In consideration of the dissimilar sensitivity to temperature, we suppose that the two isoforms 317 of HSPA6 have different impacts on the cold acclimation of grass carp.

318
Growth arrest and DNA-damage-inducible (GADD45) is a group of genes including three 319 paralogs: GADD45A, GADD45B and GADD45G. Many studies have proven the involvement of 320 GADD45 in the regulation of growth and apoptosis. Although a recent study revealed that the mice 322 about the role of GADD45 in cold stress. According our results, both GADD45B and GADD45G 323 seemed to be involved in this process, while the expression of GADD45A was not sensitive to 324 temperature changes.

326
In the research, the response of grass carp to low temperature was divided into the cold 327 adaption and the cold dormancy. The former was a systematic process, involving the hormone