Sweet potato cultivars Xushu18 and Taizhong 6 were grown in greenhouse as described previously [13]. For cold, drought and salt stress treatments, three-week-old Taizhong 6 plants were kept at 4°C, or treated with 25% PEG6000 or 250 mM NaCl. The fourth fully expanded leaves counted from the tops of plants were collected at 3 h, 6 h, 12 h, 24 h and 48 h after the treatments for RNA isolation. Ten plants were used for each treatment. For IbSINA5 expression analysis, fibrous roots, storage roots, stems, young leaves, mature leaves and senescent leaves of ten-week-old Taizhong 6 sweet potato plants were used. Total RNA was extracted as described previously [26].

qRT-PCR was carried out using IbSINA5 primers (forward: 5’-ATGTATAAAATGGAGATTGAAAGC-3’; reverse: 5’-TTAAGTGCTACAAATATTCG-3’) as described previously [26]. The expression level of IbCBF3, IbCOR413, IbCOR314 and IbCOR27 were examined with the house keeping gene IbTubulin as an internal control [13].

The encoding cDNA of IbSINA5 was cloned from sweet potato cultivar Taizhong 6 using its forward and reverse primers based on the gene information from the database [27]. The amplified cDNA fragment was cloned to the pEASY-T5 Zero Cloning Vector (Transgen, China) after sequence confirmation.

To detect the subcellular localization of IbSINA5 protein, IbSINA5 was fused in frame to the N-terminal of GFP in the pBI221:GFP vector. The resultant pBI221:IbSINA5-GFP and pBI221:GFP was transiently expressed in onion epidermal cells, respectively, as described previously [28]. The location of GFP and IbSINA5-GFP in the onion cells was observed with a confocal laser scanning microscopy (Leica TCS SP2).

To construct the plant expression vector pCAMBIA-2301s2:IbSINA5, the full length cDNA of IbSNIA5 was cloned into the modified pCAMBIA-1301s2 vector, driven by two copies of cauliflower mosaic virus (CaMV) 35S promoter. The resultant construct pCAMBIA-2301s2:IbSINA5 was introduced into sweet potato cv. Xushu18 via Agrobacterium-mediated transformation as described previously [29].

For the molecular confirmation of transgenic plants, PCR and qRT-PCR were carried out to verify the integration and expression of IbSNIA5. To assess the expression levels of IbSNIA5, IbCBF3 and IbCOR genes, total RNAs extracted from the mature leaves (the fourth leaves counted from the tops of plants) of four-week-old wild type and transgenic plants before and after cold treatment were used for qRT-PCR analyses. All gene specific primer sequences were the same as used by Jin et al. [13].

For cold resistance analysis, four-week-old wild type and transgenic sweet potato plants grown in soil were kept at 4°C for 48 h and recovery at 25°C for 24 h. The phenotypes of plants before and after the treatments were photographed and the contents of proline, MDA and hydrogen peroxide (H₂O₂) in the mature leaves were measured as described previously [30,31].

To generate the reporter construct 35S::REN-ProIbSIN5::LUC, the promoter region (2001 bp) of IbSNIA5 was inserted into pGreenII0800. To generate the effecter vector pGreenII62-SK-IbCBF3, IbCBF3 were cloned into pGreenII62-SK. Sweet potato leaf protoplast preparation and transformation were carried out as described by Wang et al. [32]. Transient dual-luciferase assays in sweet potato leaf protoplast were performed as described previously [33,34].

To analyze the transcription function of IbCBF3, the transcriptional factor effector pGreenII62-SK-IbCBF3 was transfected into sweet potato leaf protoplast. The transient expressions of IbCBF3, IbSNIA5 and IbCOR genes in sweet potato leaf mesophyll protoplasts were examine by qRT-PCR as described previously [35].

ANOVA (one-way) was used to generate every P value. The variability was indicated with the standard deviation (SD). *, ** and *** indicate p-value < 0.05, < 0.01 and < 0.001, respectively. Data are shown as mean ± standard deviation (SD) from three biological replicates each.

Ubiquitination mediated protein degradation at different plant growth stages and in response to various abiotic stresses has been well studied [36,37]. The Arabidopsis SINAT5 (AtSINAT5) functioned as a RING type E3 ubiquitin ligase to control lateral root growth by degrading NAC1, whereas SINA2, a TRAF-like SINA family protein lacking RING domain, positively regulated the resistance to drought stress [19,25]. To explore the biological function of SINA family proteins in tuberous crops, we blast searched the sweet potato genome database using Arabidopsis SINA2 (https://www.ipomoea-genome.org/genome_ jbrowse.html). A total number of seven SINA2 homologues were identified. Similar to the previously reported SINA and some other TRAF-like family proteins, they all contain a conserved C-terminal TRAF-like domain [14–16]. However, among them, same as the Arabidopsis SINA2 and animal’s TRAF1 proteins, IbSINA5 also lacks the RING domain (Fig. 1). Therefore, IbSINA5 does not have E3 ligase activity required for its target protein degradation.

To dissect the function of IbSNIA5 in the resistance to abiotic stresses in sweet potato, its expression levels in wild type Taizhong 6 plants were examined by qRT-PCR. IbSNIA5 was ubiquitously expressed in all the tested tissues and organs including fibrous roots, storage roots, stems, young leaves, mature leaves and senescent leaves, with a predominant expression in fibrous roots (Fig. 2A). Previously, we reported that SINA2 was strongly induced by ABA and drought [25]. We examined the expression levels of IbSINA5 in three-week-old Taizhong 6 plants after low temperature (4°C), 25% PEG6000 or 250 mM NaCl treatments by qRT-PCR. We found that IbSNIA5 was remarkably induced by cold, drought and salt stress, indicating its possible roles in plant response to various abiotic stresses (Figs. 2B–2D).

The subcellular localization of a certain protein may reflect its possible functions in the relevant biological processes it participates. To determine the subcellular localization of IbSNIA5, pBI221:IbSINA5-GFP and pBI221:GFP was transiently expressed in onion epidermal cells. IbSINA5-GFP fusion protein was mainly localized in nuclei and cytoplasm (Fig. 2E). Similar subcellular localization was also observed with SINA2-YFP fusion protein in Arabidopsis protoplasts [25]. The nuclei and cytoplasm localization of IbSINA5 implies that it may have a regulatory role in the expression of its upstream transcriptional factors in nuclei and its downstream functional genes in cytoplasm.

Base on our previous study that overexpression of SINA2 significantly augmented the resistance to drought stress in Arabidopsis plants, and the observation that IbSINA5 was dramatically induced by cold and drought stresses, we speculated that overexpression of IbSINA5 in sweet potato may also promote the cold resistance [25]. To verify this hypothesis, the plant expression construct pCAMBIA-2301s2:IbSINA5 containing the coding sequence of IbSINA5 was introduced into the genome of Xushu18 (Fig. 3A). A total number of 34 independent hygromycin resistant lines were obtained, and seven of them were randomly selected. PCR and qRT-PCR analyses confirmed the integration of IbSINA5 in the sweet potato genome, and the overexpression of IbSINA5 in all the tested transgenic lines (Figs. 3B and 3C).

To determine whether overexpression of IbSINA5 would improve the resistance of transgenic sweet potato plants to cold stress, we compared their growth phenotypes with wild type plants. Three transgenic lines with different IbSINA5 overexpression levels (L2, L5 and L7) were selected. Four-week-old wild type and transgenic plants grown in greenhouse were kept at 4°C for 48 h, then recovered at 25°C for 24 h. All transgenic plants showed less cold damage and successfully recovered after they were transferred back to room temperature, whereas wild type plants exhibited more severe cold damage and failed to recover (Fig. 4A). The damage extents were consistent with IbSINA5 overexpression levels, as indicated by the more severe damage in transgenic line L5, which showed relatively lower IbSINA5 overexpression level, than in transgenic lines L2 and L7, which showed relatively higher IbSINA5 overexpression levels (Fig. 4B).

It is well known that COR genes play a crucial role in plant response to cold stress [38]. We examined the expressions of IbCOR413, IbCOR314 and IbCOR27 in wild type and transgenic plants. As we have expected, the expression levels of IbCOR genes in transgenic plants were significantly higher than in wild type plants (Fig. 4C). Previous study has showed that overexpression of IbCBF3 conferred improved tolerance to low temperature and drought stress on transgenic sweet potato, leading to decreased accumulation of MDA and H₂O₂ [13]. Similar results were also observed in transgenic sweet potato plants overexpressing IbSINA5. Under normal growth condition, no significant difference was observed. However, under cold stress condition, a higher proline, and lower MDA and H₂O₂ content was detected in transgenic plants (Figs. 4D–4F). The increased IbCOR gene expression and antioxidant ability might help explain the improved cold resistance in IbSINA5 transgenic plants.

Previous work has demonstrated that IbCOR genes were up-regulated by IbCBF3 [13]. We performed PlantCARE analysis with the promoter sequence of IbSINA5, and a dehydration-responsive element-binding/C-repeat-binding factor (DREB/CBF) binding (DRE core) element was identified (Fig. 5). To explore whether IbSINA5 expression was regulated by CBFs in sweet potato, dual-luciferase assays were carried out. We found that IbCBF3 significantly activated the promoter of IbSINA5, implying that IbSINA5 expression could be regulated by IbCBF3 (Figs. 6A and 6B). We then transiently expressed IbCBF3 in the leaf protoplasts of wild type Xushu 18 plants. We observed that transient expression of IbCBF3 increased the transcription of IbSINA5, IbCOR413, IbCOR314 and IbCOR27 (Figs. 6C and 6D).

Taken together, our results suggests that overexpression of IbSINA5 up-regulates IbCOR gene expression and enhances cold resistance in transgenic sweet potato plants. IbCBF3 could be an efficient activator of IbSINA5 to regulate the expression of IbCOR genes, possible in a CBF-SINA-COR module.