Abstract Rice is susceptible to cold temperatures, especially during the seedling stage. Despite extensive research into the cold tolerance mechanisms of rice, the number of cloned genes remains limited. Plant subtilisin-like proteases (SUBs or SBTs) are protein-hydrolyzing enzymes which play important roles in various aspects of plant growth as well as the plant response to biotic and abiotic stress. The rice SUB gene family consists of 62 members, but it is unknown whether they are involved in the response to cold stress. In this study, we observed that a loss-of-function SUB4 mutant exhibited enhanced cold tolerance at the seedling stage. The sub4 mutant seedlings exhibited improved survival rates and related physiological parameters, including relative electrolyte conductivity, chlorophyll content, malondialdehyde content, and antioxidant enzyme activity. Transcriptomic analysis revealed that differentially expressed genes responsive to cold stress in the sub4 mutants were primarily associated with metabolism and signal transduction. Notably, the majority of cold-responsive genes were associated with cell wall functions, including those related to cell wall organization, chitin catabolic processes, and oxidoreductases. Our findings suggest that SUB4 negatively regulates the cold response in rice seedlings, possibly by modifying the properties of the cell wall. Subject terms: Plant biotechnology, Plant stress responses Introduction Rice (Oryza sativa L.), a staple food across the globe, is notably susceptible to cold temperatures throughout its growth and development, particularly during seed germination, seedling establishment, and pollen development at the booting stage^[42]1–[43]4. Early-sown rice seedlings must frequently contend with chilling (0 to 15 °C) stress during late spring, resulting in chlorosis, necrosis, or even death^[44]5,[45]6. Chilling stress and injury result in severe yield losses and present a significant global challenge to rice production^[46]7. Therefore, unraveling the mechanisms underlying cold tolerance has become a long-term objective of both foundational researchers and breeding practitioners in order to develop rice varieties with enhanced cold resistance. The plant response to cold is a complex multi-level process which involves low-temperature perception, a variety of signal transduction pathways, the expression of cold response genes, and changes in physiological metabolism. It is widely believed that plants perceive cold signals through alterations in plasma membrane (PM) fluidity. Cold-triggered reductions in PM fluidity trigger alterations in cellular Ca^2+ concentrations. As a secondary messenger, Ca^2+ regulates downstream genes by mediating signal transduction cascades^[47]8. The PM-localized COLD1 has been cloned and found to be a regulator of G-protein signaling, working to activate Ca^2+ channels involved in sensing cold^[48]9. The ICE-CBF-COR signal transduction pathway is activated following the perception of cold signals^[49]10,[50]11. As part of this pathway, the key transcription factor CBF (C-repeat Binding Factor) plays a crucial role in the plant response to cold temperatures^[51]12,[52]13. CBF expression is regulated by various factors, such as ICE1 (inducer of CBF expression 1), OST1 (open stomata 1), MAPKs (mitogen-activated protein kinases), ICE2, and CAMTA3 (calmodulin-binding transcription activator 3)^[53]14–[54]19. Downstream of this pathway, the products of COR (cold-responsive) genes are essential metabolites necessary for counteracting cold stress^[55]20. These include enzymes, molecular chaperones, and cold-responsive proteins involved in processes such as respiration, sugar metabolism, lipid metabolism, and antioxidant metabolism^[56]21–[57]23. Although extensive research has been conducted on the cold tolerance mechanisms of rice, only a handful of relevant genes have been cloned. As a result, the intricate molecular underpinnings of rice cold tolerance have not been fully elucidated, and a comprehensive cold signal response pathway has yet to be established in rice. Apart from COLD1, other cold-tolerance genes that have been cloned and functionally characterized in rice include qLTG3-1, qCTS-9, GSTZ2, LTG1, CTB4a, and LTT7^[58]7,[59]24–[60]28. These genes confer cold tolerance at various stages, including germination, seedling development, and booting. Serine proteases represent a major category of protein-hydrolyzing enzymes, distinguished from other proteases by using serine as the active site^[61]29. The majority of proteases found in plants are catalytic serine peptidases. Among these, the most prominent are those derived from bacterial subtilisin, i.e., the subtilisin-like protease (SUB or SBT) family^[62]30–[63]32. SUBs contain multiple conserved domains, endowing them with various functions during all stages of the plant life cycle. Specifically, plant SUBs have been reported to participate in an array of developmental processes such as embryogenesis, seedling establishment, vascular development, panicle development, and fruit ripening; biological functions such as cell division, tissue differentiation, programmed cell death, organ abscission, and senescence; and defensive roles against diverse pathogens^[64]31,[65]33–[66]38. In addition, SUBs appear to also play a role in the plant response to abiotic stress. For example, overexpressing SBT3.8 enhances osmotic tolerance in transgenic Arabidopsis thaliana plants, resulting in increased fresh weight and improved lateral root development^[67]39. In maize, four SUB genes were found to be differentially regulated under drought conditions^[68]40. GhSBT27A-silenced cotton plants are more sensitive to polyethylene glycol (PEG) stress and exhibit reduced drought tolerance^[69]41. Although these findings suggest that SUBs are important mediators of the plant abiotic stress response, the regulatory mechanisms underlying this connection remain unknown. Prior studies have reported that the rice SUB gene family contains 62 members, although none have been functionally characterized to date^[70]42,[71]43. Furthermore, whether or how these SUB genes participate in the rice cold stress response remains an open question. In this study, we identified a sub4 mutant exhibiting increased resistance to cold stress, suggesting that the SUB4 gene potentially plays a negative regulatory role within the cold response pathway. To investigate the molecular mechanism underlying the enhanced cold tolerance of the mutant seedlings, we conducted a comparative transcriptomic analysis between the wild type seedlings and the sub4 mutants under cold stress conditions. Finally, RNA-Seq identified cold-responsive genes associated with cell wall structure, remodeling, and redox homeostasis. These results suggest that the modification of cell wall properties could be a pivotal factor contributing to the enhanced cold tolerance observed in the sub4 mutant seedlings. Results The sub4 mutants exhibited enhanced cold tolerance at the seedling stage We first investigated the phenotypic responses of the sub4 mutants to cold stress. Both the wild type (WT) ZH11 and the sub4 mutant seedlings were exposed to 4 °C for 24–72 h. Seedling morphology and survival rates were observed during the subsequent recovery period. As shown in Fig. [72]1A,B, no significant phenotypic differences were observed between the sub4 and ZH11 seedlings after 24 h of exposure to 4 °C, with 70% and 60% of seedlings surviving, respectively. After 48 h, obvious phenotypic differences emerged, with the sub4 seedlings exhibiting a significantly higher survival rate than ZH11 seedlings. After 72 h, 100% of ZH11 seedlings died while the sub4 seedlings exhibited a 30% survival rate. Overall, the sub4 mutant seedlings exhibited enhanced cold tolerance compared to WT ZH11 seedlings. Fig. 1. [73]Fig. 1 [74]Open in a new tab Mutation of SUB4 increased the cold tolerance of rice seedlings. (A) Phenotypes of rice seedlings following recovery after exposure to 4 °C for 0, 24, 48, and 72 h. Scale bar = 5 cm. (B) Corresponding survival rates of seedlings from (A). A t-test was used to analyze statistical differences between means. **p < 0.01. (C) Identification of a 567th C deletion in the SUB4 gene derived from the sub4 mutant rice seedlings. (D) Predicted three-dimensional structure of the SUB4 protein and a schematic representation of the SUB4 amino acid sequence in wild type ZH11 and the sub4 mutant rice seedlings. Through gene sequencing and sequence alignment, we identified a 567th C deletion in the SUB4 gene coding sequence derived from the sub4 mutant seedlings (Fig. [75]1C). Utilizing the InterPro and SWISS-MODEL databases, we predicted the protein domains and three-dimensional structure of SUB4 (Fig. [76]1D). Deletion of the 567th C resulted in premature termination at 201 aa, the loss of the Subtilase domain, and the disruption of the intact protein’s structure. These alterations likely impair the subtilase functionality of the SUB4 protein, potentially contributing to the cold tolerance observed in the sub4 mutant seedlings. Physiological characterization of the cold-tolerant sub4 mutant Relative electrolyte conductivity (REC), chlorophyll content, malondialdehyde (MDA) content, and antioxidant activity serve as indicators of the degree of injury resulting from cold stress. We measured these physiological parameters before and after subjecting both WT ZH11 and the sub4 mutant rice seedlings to 4 °C for 48 h. In agreement with the phenotypic results, the sub4 mutant seedlings exhibited increased cold tolerance compared to ZH11 according to these parameters. Under unstressed conditions, both the sub4 and ZH11 seedlings exhibited approximately 20% REC (Fig. [77]2A). However, after 48 h of cold treatment, REC increased significantly to 85% in ZH11 seedlings but only to 65% in the sub4 seedlings, which was significantly lower than that of ZH11 (p < 0.01). Cold treatment resulted in increased MDA content in both rice lines (Fig. [78]2B). However, the MDA content in the mutant seedlings was significantly (p < 0.01) lower than in the WT seedlings following cold stress. The observed changes in REC and MDA suggest that the cells of the sub4 mutants experience less membrane damage in response to cold stress, thereby exhibiting enhanced cold tolerance. Fig. 2. [79]Fig. 2 [80]Open in a new tab Physiological indicators of cold stress in the sub4 mutant and wild type ZH11 seedlings following 48 h of 4 °C. (A) Relative electrolyte conductivity (REC). (B) Malondialdehyde (MDA) content. (C) Chlorophyll content. (D) Peroxidase (POD) activity. (E) Superoxide dismutase (SOD) activity. (F) Catalase (CAT) activity. (G) H[2]O[2] content. A t-test was used to analyze statistical differences between means. **p < 0.01. (H) 3,3ʹ-diaminobenzidine (DAB) staining of leaves from sub4 mutant and wild type ZH11 seedlings following 48 h of 4 °C. Bar: 1 μm. Under unstressed conditions, no significant differences in chlorophyll content were observed between the mutant and ZH11 seedlings (Fig. [81]2C). However, after 48 h of cold treatment, the chlorophyll content decreased in both rice lines. ZH11 experienced an approximately 2.5 mg/g reduction in chlorophyll content, while sub4 exhibited a decrease of approximately 2 mg/g. Because the sub4 mutants exhibited a higher chlorophyll content than WT seedlings, the mutants may have a higher photosynthetic output under cold stress. Peroxidase (POD), superoxide dismutase (SOD), and catalase (CAT) are antioxidant enzymes responsible for the removal of reactive oxygen species (ROS) in plant cells. Overall, the activities of POD and SOD exhibited similar trends in response to cold stress (Fig. [82]2D,E). Specifically, cold treatment increased the activities of both POD and SOD. However, the sub4 mutants exhibited significantly higher POD and SOD activities than ZH11 seedlings (p < 0.01). Notably, CAT activity exhibited the opposite trend (Fig. [83]2F). Under unstressed conditions, CAT activity was comparatively higher. However, after 48 h of cold stress, both the sub4 and ZH11 seedlings exhibited significantly decreased CAT activity. Similarly, the CAT activity was significantly higher in mutant seedlings than in WT seedlings (p < 0.01). The higher antioxidant activities observed in the cold-stressed mutant seedlings allows them to more effectively detoxify ROS, thereby contributing to enhanced cold tolerance. To further assess the reactive oxygen species (ROS) levels in the wild type (WT) and sub4 mutants, we conducted 3,3ʹ-diaminobenzidine (DAB) staining and quantified hydrogen peroxide (H[2]O[2]) content. Notably, the accumulation of H[2]O[2] in sub4 plants following cold treatment was significantly reduced compared to the WT (Fig. [84]2G). Consistent with this observation, under cold treatment, the DAB staining signal in sub4 mutant seedlings was weaker than that in the WT, indicating a substantial decrease in ROS content induced by low temperature in the mutants (Fig. [85]2H). Collectively, these results suggest that the accumulation of H[2]O[2] induced by cold stress is significantly associated with the cold tolerance of the sub4 mutants. Expression patterns of SUB4 and other cold-responsive genes Previous studies have found SUB4 to be expressed in various rice tissues. However, this gene exhibits particularly high expression in young leaves, spikes, anthers, and callus (according to data from EMBL-EBI). In this study, we evaluated the expression levels of SUB4 in seedling shoots and roots, as well as in roots, stems, leaves, and spikes sampled during the heading stage. Consistent with previously-published data, SUB4 exhibited the highest expression in seedling shoots, followed by spikes at the heading stage and seedling roots (Fig. [86]3A). These results suggest that the SUB4 gene may play metabolic roles during seedling growth and spike development. Fig. 3. [87]Fig. 3 [88]Open in a new tab Expression patterns of SUB4 and other cold-responsive genes. (A) Expression patterns of SUB4 in different rice tissues. S seedling stage, H heading stage. (B) Expression pattern of SUB4 respond to cold stress. Rice leaves of 2-week-old seedling at different time points after cold stress were used. (C) Expression patterns of cold stress-related genes in the sub4 mutant and wild type ZH11 seedlings. *p < 0.05, **p < 0.01. As shown in Fig. [89]3B, we performed quantitative PCR (qPCR) validation of SUB4 gene expression, using the leaves of two-week-old seedling harvest at different times after cold. The results revealed that SUB4 gene expression increased within the first 48 h, peaking (threefold increase) between 6 and 12 h. This suggests that SUB4 is involved in the early response to cold stress. Subsequently, we evaluated the expression of cold stress marker genes in the sub4 mutants, including OsCBF1, OsCBF2, and OsCBF3 associated with the CBF pathway, as well as OsPYL9, OsNAC5, OsABF2, OsP5CS, and OsCPK13 associated with the ABA signaling pathway (Fig. [90]3C). Genes associated with the CBF pathway exhibited universal cold-induced expression, particularly OsCBF1 and OsCBF3, whose expression levels increased dozens-fold following exposure to cold. In addition, their expression levels were significantly higher in mutant seedlings. In contrast, the expression patterns of genes associated with the ABA pathway, particularly OsABF2, OsP5CS, and OsCPK13, exhibited only slight differences. Specifically, these genes exhibited modest upregulation in response to cold, and their expression levels were comparable in both rice lines. These findings suggest that the SUB4 mutation potentially influences the expression of CBF pathway-related genes, resulting in increased cold tolerance in the sub4 mutants. Transcriptomic analysis in response to cold stress involved in SUB4 To elucidate the SUB4-specific mechanisms underlying the enhanced cold tolerance of the sub4 mutant seedlings, we conducted a comparative transcriptomic analysis between the sub4 mutant (Mu) and WT ZH11 (WT) seedlings. Given that SUB4 is involved in the early response to cold stress, we exposed rice seedlings to 4 °C for 0, 6, and 24 h. Each RNA-seq sample was prepared with three biological replicates, resulting in a total of 18 samples. Approximately 6.8 Gb of clean data was obtained for each sample. To validate the reliability of the transcriptomic data from each sample, we first conducted a principal component analysis (PCA) based on the FPKM values of all identified genes (Fig. [91]4A). Overall, the three biological replicates of each sample clustered together, and samples subjected to different durations of cold treatment were clearly separated. The Mu_6h and WT_6h comparison exhibited the most significant difference, suggesting that major transcriptomic changes occurred at the 6 h time point. Fig. 4. [92]Fig. 4 [93]Open in a new tab Transcriptomic analysis of the sub4 mutant and wild type ZH11 seedlings subjected to cold stress. (A) Principal component analysis (PCA) of gene expression levels in all 18 samples. (B–D) Volcano plots representing changes in gene expression in the ZH11 seedlings and the sub4 mutants at different time points (0, 6, or 24 h) under cold stress. The numbers of upregulated and downregulated genes are indicated in the plots. (E) Venn diagram illustrating the differentially expressed genes (DEGs) in the ZH11 seedlings and the sub4 mutants under cold stress at different time points. Mu, sub4 mutant. WT, ZH11. To explain their varied responses to cold, we compared differentially expressed genes (DEGs) identified in mutant and WT seedlings. DEGs were determined according to the following criteria: fold change (FC) ≥ 2 and false discovery rate (FDR) < 0.05. Before cold treatment, a comparison between WT_0h and Mu_0h revealed a total of 268 DEGs, consisting of 176 upregulated and 92 downregulated genes (Fig. [94]4B). Under cold stress, a total of 584 DEGs were identified (Fig. [95]4E), including 164 DEGs after 24 h of cold treatment, 94 of which were upregulated and 70 of which were downregulated (Fig. [96]4D). Notably, the greatest number (515) of DEGs between WT and Mu occurred after 6 h of cold treatment, including 299 upregulated genes and 216 downregulated genes (Fig. [97]4C). Additionally, 95 DEGs were common to all samples after 6 h and 24 h of cold treatment (Fig. [98]4E). Functional annotation and enrichment analysis of cold-responsive DEGs Next, we functionally annotated the 584 cold-responsive DEGs, including those induced after both 6 h and 24 h of cold treatment. According to the Gene Ontology (GO) enrichment analysis, these DEGs were most enriched in the “aminoglycan metabolic process”, “chitin metabolic process”, and “oxidation reduction” Biological Processes (BP); the “iron ion binding”, “tetrapyrrole binding”, and “chitinase activity” Molecular Functions (MF); and the “extracellular region” Cellular Component (CC) (Fig. [99]5A). Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis indicated that these genes participate in various metabolic pathways, including “starch and sucrose metabolism”, “phenylpropanoid biosynthesis”, and “amino sugar and nucleotide sugar metabolism” (Fig. [100]5B). Finally, these cold-responsive DEGs were also found to be associated plant hormone signal transduction and the MAPK signaling pathway. Fig. 5. [101]Fig. 5 [102]Open in a new tab Functional annotation of cold-responsive differentially expressed genes (DEGs). (A) GO enrichment analysis of DEGs between the ZH11 seedlings and the sub4 mutants under cold stress. The − Log[10] (p-value) represents the significance of GO enrichment in the Biological Process (BP), Molecular Function (MF), and Cellular Component (CC) categories. (B) KEGG pathway analysis of DEGs between the ZH11 seedlings and the sub4 mutants under cold stress. Differential expression of genes encoding cell wall-related proteins may confer cold tolerance to the sub4 mutants A close examination of the GO enrichment analysis revealed that the majority of cold-responsive DEGs were allocated to the BP category and were related to cell wall functions. These genes are characterized by their subcellular localization to either the cell wall or the extracellular space. Furthermore, the only significantly enriched term in the Cellular Component (CC) category was found to be “extracellular region” (Fig. [103]5A). These findings strongly suggested that the cold-responsive DEGs identified in the sub4 mutant are closely associated with cell wall functions. We further classified these DEGs into several groups according to their functions, including cell wall organization, chitin catabolic processes, oxidoreductases, and others (Fig. [104]6). Among them, 12 genes are involved in cell wall organization, including those encoding enzymes such as expansin, xyloglucan endotransglucosylase, pectin methylesterase, and cellulose synthase. Notably, five out of six chitinases were upregulated in the mutant seedlings following cold stress. Although chitinases are known for their anti-pathogen activities in rice, our results suggest that these enzymes may also play a potential role in cold tolerance. Additionally, DEGs encoding oxidoreductases included 6 peroxidases and ascorbate oxidase 2, indicating that there may be some variation in the ROS scavenging capabilities between the sub4 and ZH11. The germin-like family of glycoproteins is associated with the plant cell wall. All four germin-like genes were downregulated in the cold-tolerant sub4 mutant seedlings, warranting further exploration of their roles in the cold response. Fig. 6. [105]Fig. 6 [106]Open in a new tab Functional classification of cell wall-related genes exhibiting differential expression between the wild type ZH11 and the sub4 mutant seedlings under cold stress. Subcellular localization of SUB4 Our results suggest that the mutation of SUB4 likely modulates cold tolerance by influencing cell wall biological processes. To test this hypothesis, we conducted a subcellular localization analysis to explore the connection between SUB4 and the cell wall. SUB4–GFP constructs were generated and transiently expressed in tobacco mesophyll cells. Subcellular localization revealed that the SUB4 protein was located along the cell periphery (Fig. [107]7A–C). Meanwhile, GFP fluorescence was observed throughout the cell (Fig. [108]7D–F). Propidium iodide (PI) is usually used to detect cell viability. PI cannot permeate the cell membrane of living cells, but when cells die, PI can embed in double stranded nucleic acid to produce red fluorescence. Thus, PI is widely used to visualize cell walls. We then conducted PI staining at the same time as subcellular localization, and our result showed the overlap between the SUB4 localization and PI signal (Fig. [109]S1). Overall, these results provide additional evidence for the connection between SUB4 and the cell wall. Fig. 7. [110]Fig. 7 [111]Open in a new tab Subcellular localization of SUB4. (A–C) Transient expression of SUB4–GFP driven by the 35S promoter in tobacco mesophyll cells. (D–F) GFP signals in tobacco cells. Scale bars = 20 μm. Expression patterns of the OsSub gene family Rice contains a large subtilisin gene family consisting of 62 members^[112]42,[113]43, suggestive of possible functional redundancy. To investigate any potential interactions between SUB4 and other members of OsSub, we examined the expression patterns of all OsSub genes detected in this study. We performed functional annotation of all expressed genes using the Pfam and Swiss-Prot databases, categorizing genes annotated as “Subtilase family” in Pfam or “Subtilisin-like protease” in Swiss-Prot as members of the OsSub family. In total, we identified 42 OsSub family members expressed in at least one treatment (Fig. [114]8A). Five members exhibited differential expression between the sub4 and ZH11 following cold treatment. Among these, OsSub13 and OsSub20 were downregulated, while OsSub8, OsSub9, and OsSub62 were upregulated. Interestingly, under cold stress, the expressions of OsSub8, OsSub9, and OsSub62 significantly increased in sub4 mutant, while OsSub13 and OsSub20 exhibited no significant changes in their expression (Fig. [115]8B). The functional complementarity of OsSub8/9/62 deserves further investigation. This result suggested five OsSubs may have similar function with SUB4 in the adaption of cold stress, providing clues for future SUB function research. Besides, it is worth noting that these differential expressions all occurred at the 6 h time point, indicating that SUB4 functions in the early response to low temperatures. Fig. 8. [116]Fig. 8 [117]Open in a new tab Expression patterns of OsSub gene family members in response to cold stress. (A) Expression heatmap illustrating the expression levels of OsSub genes in ZH11 and the sub4 mutant under cold stress. Genes marked with asterisks represent differentially expressed genes (DEGs). The inner circle represents a neighbor-joining (NJ) phylogenetic tree of 62 SUBs in rice. The branches of different subgroups are indicated by various colors. (B) FPKM values of the differentially expressed OsSub8/9/13/20/62 at different time points under cold treatment. *p < 0.05. Discussion The sub4 mutant seedlings exhibited enhanced cold tolerance Cold stress during the seedling stage is a common and inevitable phenomenon in rice production, especially in regions such as southern China where early-season rice is cultivated, as well as in high-latitude regions. Abrupt drops in temperature over a short period of time can lead to severe damage to rice plants, resulting in seedling decay, stunted growth, and ultimately decreased yields. Identifying cold-tolerant germplasm during the seedling stage and uncovering the mechanisms of cold tolerance are crucial objectives in rice research. In this study, we identified a loss-of-function mutant that exhibited enhanced cold tolerance during the seedling stage. Following 72 h of exposure to 4 °C, this mutant exhibited a higher survival rate compared to the WT. Subsequently, we sought to elucidate the mutation-mediated mechanisms of cold tolerance from both the physiological and transcriptional perspectives. In plants, cold stress induces a series of physiological and metabolic changes, including alterations in photosynthesis, PM fluidity and integrity, ROS production, and accumulation of osmoprotectants^[118]2. As a result, several physiological indicators can be used to evaluate the cold tolerance of rice seedlings, including electrolyte leakage, chlorophyll content, ROS content, lipid peroxidation, enzymatic antioxidant activity, and carbohydrate content, among others^[119]44. Exposure to cold temperatures reduces chlorophyll synthesis, and transgenic rice with enhanced cold tolerance exhibits a significant increase in chlorophyll content^[120]45. The PM is the primary site of cold signal interception, and is susceptible to phase transitions under cold conditions, resulting in electrolyte leakage. Therefore, a lower REC is often an indicator of plant cold tolerance^[121]46,[122]47. Under adverse conditions such as cold stress, cells experience imbalanced oxygen metabolism, leading to excessive ROS production. The increased accumulation of ROS results in membrane lipid peroxidation, resulting in PM damage. ROS accumulation also promotes the degradation of polyunsaturated fatty acids into MDA, which further damages tissues and cells^[123]48,[124]49. To protect themselves against oxidative stress, rice plants employ antioxidant enzymes to catalyze the removal of ROS. Among these antioxidant enzymes, SOD, CAT, and POD are most effective^[125]50. Higher antioxidant enzyme activities help alleviate ROS-induced cellular damage, thereby enhancing cold tolerance in rice. Here, we observed that several physiological indicators consistently indicated that the sub4 mutants exhibit stronger cold tolerance than WT seedlings. Together with the dramatic increase in the expression of cold-response maker genes, particularly CBF-related genes, these results indicate that the sub4 mutant seedlings are tolerant of cold stress. Transcriptomic analysis highlighted alterations in metabolism and signal transduction Numerous RNA-Seq studies have been conducted to explore cold-induced changes in gene expression of rice seedlings. In addition, comparative transcriptomic analyses have been conducted between cold-tolerant and cold-sensitive rice genotypes. Although the selected genotypes, cold conditions, and treatment durations may vary across these studies, transcriptomic commonalities have nevertheless been observed. Overall, cold-responsive gene expression in rice seedlings appears to be mainly related to metabolite transport and metabolism, signal transduction, and transcriptional regulation^[126]6,[127]51–[128]56. The results of our GO and KEGG enrichment analyses were consistent with these previous studies. In rice, cold induce the accumulation of soluble sugars, primarily trehalose, which serves as both stress-protective metabolite and storage carbohydrate. Notably, trehalose acts as an osmoregulatory substance capable of protecting cells from cold injury^[129]57. Trehalose-6-phosphate synthase (TPS) is a key enzyme related to trehalose biosynthesis. Overexpression of OsTPS1 has been demonstrated to significantly increase trehalose levels in rice, thereby improving cold tolerance^[130]58,[131]59. Consistent with previous transcriptomic studies^[132]60, we observed that LOC_Os08g31980, which encodes the enzyme TPS7, was induced by cold stress in the sub4 mutant (Table [133]S1). This suggests that cold-tolerant plants synthesize higher levels of trehalose to protect cellular components from damage. Furthermore, genes encoding the hydrolytic enzymes beta-1,3-glucanase (LOC_Os01g71340, LOC_Os01g71670, LOC_Os07g35510, and LOC_Os09g32550) and α-amylase (LOC_Os06g49970 and LOC_Os08g36910) were observed to be differentially expressed under cold stress (Table [134]S1). These observations suggest that the enhanced cold tolerance of the sub4 mutant is, at least in part, achieved through alterations in carbohydrate metabolism. Cold stress-related signal transduction primarily involves Ca^2+ signaling, ROS homeostasis, plant hormones, and MAPK pathways^[135]61. In our study, we observed significant enrichment of 17 DEGs related to the KEGG terms “plant hormone signal transduction” and “MAPK signaling pathway” (Fig. [136]5B, Table [137]S2). Several transcriptomic studies have highlighted the pivotal roles of MAPK members and plant hormones in the response to cold stress^[138]62–[139]64. Therefore, alterations in the expression of signal transduction genes are likely a significant contributing factor to the enhanced cold tolerance observed in the sub4 mutant. Additionally, it is worth noting that several genes encoding lipoxygenases (OsLOX1, OsLOX2, OsLOX11, and OsLOX12), which are assigned to the “linoleic acid metabolism pathway” (Table [140]S2), exhibited differential expression in the sub4 mutant. A recent transcriptomic study revealed that linoleic acid metabolism may play a key role in enhancing cold stress adaptation in rice^[141]65. Although LOX genes are well-known for their specific role in rice seed storability, their relationship to cold tolerance deserves further study. Mutation of SUB4 likely modulates cold tolerance by influencing cell wall dynamics Unlike the effects of metabolism and signal transduction, the role of cell wall in plant cold response is commonly underestimated. The plant cell wall determines cell shape and acts as the first line of defense against biotic and abiotic stress^[142]66–[143]68. Its structure is subject to constant modification in response to growth signals and various external stimuli. In general, these modifications typically result from the transcriptional regulation of cell wall-related genes. Several studies have highlighted the extensive role of cell wall-related genes in the plant response to heat, drought, salt, and cold^[144]69–[145]72. In particular, transcriptomic studies have revealed that genes involved in cell wall composition or modification are highly induced under cold stress^[146]73–[147]75. The structure and composition of the cell wall are primarily determined by cellulose and hemicellulose, pectin, and cell wall proteins such as extensins, expansins, xyloglucan endo-transglycosylase/hydrolase (XTH), pectin methylesterases (PMEs), and pectin acetyl-esterase (PAE)^[148]76. Cellulose and pectin biosynthesis may play a crucial role in cold tolerance, as two related genes (LOC_Os08g23780 and LOC_Os09g30130) were upregulated in the sub4 mutant (Fig. [149]6), indicating that maintaining cell wall integrity could be an effective strategy for enhancing cold tolerance in rice. This is supported by a similar transcriptomic analysis which revealed that four cellulose synthase genes and five cellulose synthase-like genes were more highly expressed in cold-tolerant rice genotype^[150]77. In addition, typical cell wall-loosening enzymes, including the expansin, PME, and XTH families, were found to be differentially expressed in rice seedlings following cold exposure, with certain genes exhibiting either up- or down regulation^[151]77–[152]79. In our study, all six related genes were downregulated in the sub4 mutant (Fig. [153]6). Cell wall-loosening enzymes play a role in cell growth following stress relaxation^[154]80. The gene repression observed in this study may suggest that cell wall loosening was inhibited, while the structure of the cell wall was strengthened, in cold-stressed mutant rice seedlings. Such strengthening is further indicated by the induction of cellulose and pectin biosynthesis genes, which may contribute to the enhanced cold tolerance observed in the sub4 mutant. In addition to genes related to cell wall organization, six members of the rice chitinase gene family exhibited differential expression in the sub4 mutant (Fig. [155]6). Among them, five genes were upregulated, suggesting a potential role for chitinase genes in rice cold tolerance. Plant chitinases, categorized as glycosyl hydrolases (GHs), play a crucial role in plant defense against pathogens by catalyzing the hydrolysis of chitin, a major component of fungal cell walls^[156]81. Plant chitinases may also play an important role during normal plant growth or during the abiotic stress response^[157]82,[158]83, although the underlying molecular mechanism remains elusive. As GHs, chitinases are likely involved in cell wall remodeling via the metabolism of cell wall polysaccharides^[159]84,[160]85. The rice genome contains a total of 49 putative chitinase genes, with the majority predicted to be involved in the secretory pathway^[161]86. The rice chitinase-like protein OsCTL1 has been shown to mediate cellulose biosynthesis and cell wall remodeling^[162]82. Interestingly, a recent transcriptional analysis of a cold-tolerant rice seedling revealed the upregulation of five chitinase genes during the recovery period following cold treatment^[163]87, with four of these genes being consistent with those identified in our study. Therefore, we speculate that the differential expression of chitinase genes in the sub4 mutant may contribute to cell wall remodeling, thus maintaining cell wall functionality under cold conditions. Peroxidases and germin-like proteins (GLPs) are reported to be stress-responsive cell wall proteins in maize and rice^[164]88. The role of oxidoreductases in ROS scavenging to maintain cell redox homeostasis under stress conditions was discussed above. Here, we observed that genes encoding six class III peroxidases and ascorbate oxidase 2 were differentially expressed in the sub4 mutant, although the majority of these genes were downregulated (Fig. [165]6). In addition, four GLPs (OsGLP1-1, OsGLP8-7, OsGLP8-10, and OsGLP8-11) were differentially expressed in the sub4 mutants (Fig. [166]6). GLPs are widely involved in the abiotic stress responses of different plant species^[167]89. For example, overexpression of rice GLP1 in tobacco results in the hyper-accumulation of H[2]O[2] and reinforcement of the cell wall through the cross-linking of cell wall components, leading to increased biotic and abiotic stress tolerance^[168]90. Consistent with these results, the OsGLP8 gene is thought to be expressed in the cell wall and possess SOD activity^[169]91. Therefore, we speculate that the cold tolerance observed in the sub4 mutant is partly achieved through alterations in GLP expression. Conclusion and perspective Plant SUBs are protein-hydrolyzing enzymes which play important roles in response to biotic and abiotic stress. The rice SUB gene family contains 62 members but none of them has been functionally characterized. Whether SUB genes participate in the rice cold stress response remains unclear. In this study, we observed that a loss-of-function sub4 mutant exhibited enhanced cold tolerance at the seedling stage. In addition to genes related to metabolism and signal transduction, our comparative transcriptome analysis revealed that the majority of cold-responsive DEGs were associated with cell wall functions. These include genes encoding expansin, PME, XTH, chitinase and peroxidases. Overall, our results suggest that SUB4 negatively regulates the cold response in rice seedlings, possibly by modifying the properties of the cell wall. As crucial enzymes engaged in proteolysis and protein modification, subtilisin-like proteases are typically predicted to be secretory pathway proteins^[170]42. Most SUBs are targeted to the cell wall, where they play a role in controlling growth and development by regulating the properties of the cell wall and the activities of extracellular signaling molecules^[171]31. In maize, 15 out of 18 SUB genes are located in the cell wall^[172]40. Accumulating evidence highlights the contribution of SUBs to the regulation of cell wall structure. In A. thaliana, several SUBs were found to affect cell wall stability by participating in PME processing^[173]92,[174]93. Therefore, SUBs are likely involved in a wide variety of cell wall-related physiological and biochemical processes. In this study, the cold-responsive DEGs identified in the sub4 mutant were mainly enriched in GO terms related to the cell wall. This suggests that the mutation of SUB4 likely enhances seedling cold tolerance primarily by influencing cell wall biological processes (Fig. [175]9). Based on our findings, we suggest the following research should be conducted: 1. Evaluate physiological alterations in cell wall structure and composition in the cold-stressed sub4 mutants. 2. Identify the substrates of the SUB4 enzyme and determine whether they are responsible for the post-translational modification of cold-responsive proteins through limited proteolysis at specific sites. By filling these research gaps, we will gain a deeper understanding of the mechanisms underlying the enhanced cold tolerance of the sub4 mutant. Fig. 9. [176]Fig. 9 [177]Open in a new tab Schematic model of the regulatory network associated with the cold response in the sub4 mutant rice. Cold stress primarily induced transcriptomic changes in genes related to cell wall modification, as well as alterations in metabolism. These changes collectively contribute to the enhanced cold tolerance observed in the sub4 mutant. Materials and methods Plant materials and cold stress treatment The sub4 mutant was derived from the Zhonghua11 (ZH11) rice cultivar (Oryza sativa L. Japonica) through CRISPR-Cas9 targeted gene editing. Seeds of ZH11 were obtained from Tianjin Crop Research Institute, Tianjin Academy of Agricultural Sciences. Seeds of the sub4 mutant were purchased from Biorun BioSciences Co., LTD (Wuhan, China). The OsU3-driven gene editing vector pHK1-Cas9-U3 was constructed by designing the target sequence TCTGCGAAGCTGGGATGAGTTGG in exon 3 of ZH11, with SUB4 (Os01g0769200) sequence as reference. Homozygous T2 generation plants were used for subsequent experiments. For the cold stress treatment, mature and healthy seeds of the wild-type ZH11 and the sub4 mutant were selected, with three replicates for each, consisting of 50 seeds per replicate. The seeds were soaked in a 1% NaClO (Sodium hypochlorite) solution for 15–30 min, rinsed three times with water, and then allowed to germinate in the dark at 28 °C for 2–3 days. Seedlings with similar growth status were selected and transferred to a rice substrate, where they were cultured under a 16-h light/8-h dark photoperiod at 28 °C for 14 days. After 14 days, the seedlings were transferred to a growth chamber with a 4 °C temperature, subjected to treatment for 24–72 h. They were then removed and transferred back to a 28 °C growth chamber for recovery for 3–7 days, after which survival rates were recorded. Determination of the relative electrolyte conductivity Five seedlings of both ZH11 and sub4, after recovery from cold treatment, were placed in separate 50 mL centrifuge tubes. To each tube, 30 μL of distilled water was added, and the tubes were subjected to vacuum for 30 min. They were then shaken at room temperature for 1 h, and the initial electrical conductivity value (S1) was measured. Subsequently, the tubes were placed in a boiling water bath for 15 min, removed, and allowed to shake for 2 h. The electrical conductivity value (S2) was then measured. The electrical conductivity value of distilled water was denoted as S0. REC = (S1 − S0)/(S2 − S0). Determination of chlorophyll content 0.05 g (fresh weight) of leaves from each treatment were cut into small pieces and placed in 15 mL centrifuge tubes. To each tube, 5 mL of 95% ethanol was added, and the tubes were kept in the dark for 48 h until the leaves turned white (shaking 2–3 times daily). Then, 1 mL of the extract was taken from each tube, and the absorbance was measured at 663 nm and 649 nm. Chlorophyll content was calculated using the formula: Chlorophyll content = (18.08OD649 + 6.63OD663) × V/(W × 1000). Determination of MDA content For both ZH11 and sub4, seedlings before and after the 4 °C treatment were weighed and ground into powder in liquid nitrogen. The determination of MDA content was performed using a spectrophotometer following the procedures outlined in the Malondialdehyde Content Assay Kit (Solarbio, Beijing, China). MDA can condense with thiobarbituric acid to form a brownish red product with a maximum absorption wavelength of 532 nm. Determination of antioxidant enzyme activity and H[2]O[2] content For both ZH11 and sub4, seedlings before and after the 4 °C treatment were weighed and ground into powder in liquid nitrogen. The activity of superoxide dismutase (SOD), catalase (CAT), peroxidase (POD) and H[2]O[2] content were identified according to standard methods (Solarbio, Beijing, China) respectively. SOD activity was determined by reduced nitrogen blue tetrazole detection at 560 nm. CAT activity was measured by the change in absorbance of H[2]O[2] at 240 nm, and POD activity was measured by monitor the absorbance change at 470 nm. H[2]O[2] reacts with titanium sulfate to form a titanium peroxide complex, which exhibits characteristic absorption at 415 nm. The H[2]O[2] content was measured by the change in absorbance of titanium peroxide at 415 nm. DAB staining Before and after the 4 °C treatment, seedlings of ZH11 and sub4 mutant lines were used for DAB staining. The staining was performed using a DAB Staining Kit (Solarbio, Beijing, China). Seedlings were rinsed with deionized water and blotted dry. They were then incubated in the prepared DAB solution for 2–6 h in the dark at room temperature until brown coloration indicative of ROS accumulation was observed. After staining, seedlings were destained in a 70–80 °C water bath for 20–40 min to remove chlorophyll. Stained samples were then photographed and stored in DAB preservative solution at room temperature for further analysis. RNA extraction and qRT-PCR Total RNA was extracted from various tissues, including the roots and shoots of ZH11 at the seedling stage, as well as the stems, leaves, and panicles at the heading stage, using the RNAprep Pure Plant Kit (TIANGEN, Beijing, China). The extracted total RNA was then reverse-transcribed into cDNA using the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix (Transgen Biotech, Beijing, China) following the manufacturer’s instructions. OsActin served as an internal control for gene expression in qRT-PCR experiments. The relative gene expression levels were determined using the 2^−ΔΔCT method^[178]94, and the expression pattern of the SUB4 gene was analyzed. Primers used in this study were shown in Table [179]S3. Subcellular localization Transient expression of SUB4 was performed in tobacco leaves. The coding sequence of SUB4 was amplified and inserted into the pCambia1300-GFP vector, then it was transformed into Agrobacterium strain EHA105, and the resulted Agrobacterium cells were grown, harvested and resuspended for 12 h at room temperature before being infiltrated into fully expanded leaves of N. benthamiana plants. After 3 days, the GFP fluorescence was visualized under a confocal microscope (Zeiss, LSM880, Germany). For PI staining, leaves of transient expression tobacco were immersed in PI solutions for 1h, followed by red fluorescence visualization under confocal microscope. Transcriptome sequencing Two-week-old seedlings of both ZH11 and sub4 were sampled after being subjected to 4 °C treatment for 0 h, 6 h, and 24 h, resulting in a total of 18 samples for transcriptome sequencing. Total RNA was extracted from each sample and underwent quality assessment before library construction. After quality control, the qualified libraries were combined based on the pre-designed target data volume and subsequently sequenced on an Illumina sequencing platform by Biomarker Technologies in Beijing, China. The transcriptome sequencing process yielded a total of 122.64 Gb of Clean Data, with each sample generating at least 5.87 Gb of clean data. These clean reads from each sample were then mapped to the rice reference genome (MSU7.0) using HISAT2 software. Differentially expressed gene analysis FPKM (Fragments Per Kilobase of transcript per Million fragments mapped) was applied to measure the expression level of a gene or transcript by StringTie using maximum flow algorithm. Differential expression analysis is processed by edgeR. Criteria for differentially expressed genes were set as Fold Change (FC) ≥ 2 and FDR < 0.05. Gene Ontology (GO) enrichment analysis of the DEGs was implemented by the GOseq R packages based Wallenius noncentral hypergeometric distribution. GO terms with corrected p < 0.05 were considered significantly enriched by differentially expressed genes. The enrichment of differentially expressed genes in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways^[180]95 was analyzed using KOBAS software. Supplementary Information [181]Supplementary Legends.^ (13.1KB, docx) [182]Supplementary Figure S1.^ (808.1KB, tif) [183]Supplementary Table S1.^ (13.5KB, xlsx) [184]Supplementary Table S2.^ (17.2KB, xlsx) [185]Supplementary Table S3.^ (9.4KB, xlsx) Author contributions Conceptualization, Jingyan Liu and Xuejun Liu; Data curation, Jingyan Liu, Fei He and Zhicai Chen; Formal analysis, Jingyan Liu and Fei He; Funding acquisition, Jingyan Liu and Fei He; Investigation, Jingyan Liu, Fei He, Zhicai Chen, Meng Liu, Ying Wang, Yumeng Cai and Jin Du; Supervision, Xuejun Liu; Validation, Jingyan Liu, Fei He, Zhicai Chen, Yingni Xiao, Ying Wang, Yumeng Cai and Jin Du; Writing—original draft, Jingyan Liu and Fei He; Writing—review & editing, Jingyan Liu, Yingni Xiao, Weiwei Jin and Xuejun Liu. All authors have read and agreed to the published version of the manuscript. Funding This work was supported by Open Fund from the Tianjin Key Laboratory of Intelligent Breeding of Major Crops (KLIBMC2304 and KLIBMC2309) and Innovative Research and Experimental Projects for Young Researchers of Tianjin Academy of Agricultural Science (2019001 and 2021007). Data availability Data sets including RNA-seq and detailed information can be viewed and downloaded from the website ([186]http://www.ncbi.nlm.nih.gov/sra/) with accession PRJNA1036792. Our experimental research and field studies on rice plants, including the collection of plant material, complied with relevant institutional, national, and international guidelines and legislation. To collect homozygous T2 seeds, the wild-type and the mutation rice plants were grown in the field of a transgenic test area which belongs to Tianjin Academy of Agricultural Sciences, and the permission was obtained. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Jingyan Liu and Fei He contributed equally to this work. Contributor Information Jingyan Liu, Email: liujingyan826@qq.com. Xuejun Liu, Email: goodrice@263.net. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-024-84491-0. References