Lhca4 Maintaining Photochemical Efficiency and ROS Homeostasis Contributes to Rice Salt Tolerance

Abstract

Background

Rice (Oryza sativa L.), a staple food for more than half of the global population, is moderately salt-sensitive and increasingly threatened by soil salinization. Deciphering the genetic mechanisms of salt tolerance (ST) is pivotal for accelerating genetic improvement of ST by rice molecular breeding and safeguarding global food security.

Results

A quantitative trait locus for rice ST, qRLS8, was mapped using a BC₂F₇ backcross introgression line population derived from the cross between Minghui 63 and 02428. Combined with transcriptome analysis, Lhca4, encoding a subunit of the light-harvesting complex, was identified as the candidate gene of qRLS8. Lhca4 positively regulates rice ST at seedling stage based on the phenotypic verification using knockout and overexpression transgenic lines of Lhca4. Overexpression of Lhca4 increased the activities of superoxide dismutase and peroxidase, and decreased the accumulation of H₂O₂ and O₂⁻, holding a high photochemical efficiency under salt stress conditions. In contrast, knockout of Lhca4 increased accumulation of reactive oxygen species (ROS) in rice, resulting in disruption of the chloroplast lamellae and a decrease in photosynthetic efficiency. Transcriptome analysis revealed that Lhca4 mediated salt stress response pathway involved in metabolic regulation, enzyme activity regulation, and antioxidant regulation.

Conclusions

Lhca4 confers rice ST by preserving chloroplast integrity, maintaining photochemical efficiency, and systemically modulating ROS homeostasis through enhancing antioxidant defense system. These findings provide a valuable gene target for the development of salt-tolerant rice varieties without compromising photosynthetic capacity.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12284-026-00927-3.

Background

Rice (Oryza sativa L.), a staple food for more than half of the global population, is moderately sensitive to salt stress and is becoming increasingly threatened by the escalating soil salinization (Zhu 2002; Ganie et al. 2021). Therefore, improving salinity tolerance (ST) in rice is important for ensuring food security. ST is a complex trait that is controlled by multiple genes and is involved in various molecular and biological processes (Zhu 2002; Yu et al. 2023). Salt stress is particularly detrimental during the seedling stage, often leading to severe growth retardation or plant death, and consequently resulting in substantial yield losses. A comprehensive understanding of the molecular mechanisms underlying rice salt tolerance is essential to accelerate the improvement of ST in rice varieties through molecular breeding approaches (Zhao et al. 2020; Ganie et al. 2021).

Salt stress constrains plant growth and development mainly by imposing osmotic stress and ionic imbalance. In saline soils, the steep Na⁺ gradient drives excessive Na⁺ influx into root cells, competitively displacing K⁺ and depressing cytosolic K⁺ levels. The consequent K⁺ deficiency disrupts stomatal regulation, skews the Na⁺/K⁺ ratio, and ultimately leads to ionic toxicity (Mao et al. 2020). To mitigate cytosolic Na⁺ toxicity, plants have evolved three core adaptive strategies: limiting Na⁺ uptake, enhancing Na⁺ efflux, and sequestering Na⁺ into vacuoles (Liu et al. 2022). A series of ST-related genes mediating the ion regulation and re-establishing ionic homeostasis have been identified in rice, such as OsHKT2;1, SKC1, OsSOS1, and OsNHX1 (Horie et al. 2007; Wei et al. 2020; Ren et al. 2005; Li et al. 2025; Yang and Guo 2018; Zhu 2016; Zhang et al. 2023; Fukuda et al. 2010; Liu et al. 2010; Yu et al. 2023).

Reactive oxygen species (ROS), including ¹O₂, O₂⁻, H₂O₂ and OH⁻, are rapidly overproduced in plant cells under salt stress. While ROS act as essential signaling molecules in normal physiological processes, their excessive accumulation induces oxidative stress, leading to DNA damage, protein degradation and lipid/carbohydrate peroxidation (Yang and Guo 2018). To cope with ROS-induced oxidative injury, plants have developed sophisticated ROS-scavenging systems consisting of enzymatic and non-enzymatic components (Chan et al. 2016).The non-enzymatic scavengers include multiple small-molecule compounds, such as sugars, ascorbate, alkaloids, carotenoids, flavonoids, glutathione (GSH), et al. (Gill and Tuteja 2010; Borghesi et al. 2011; Ben Abdallah et al. 2016), and the enzymatic agents, including catalases (CAT), superoxide dismutase (SOD), ascorbate peroxidase (APX), and glutathione reductase (GR). There is accumulating evidence that enhancing these antioxidant systems is closely associated with improved ST in plants. For example, under salt stress, over-expressing of OsCu/Zn-SOD, OsCPK12 significantly enhances oxidative-stress tolerance in rice (Zhang et al. 2017; Wang et al. 2024). OsCPK12 phosphorylates OsCATA and OsCATC to modulate H₂O₂ homeostasis and thereby enhance rice oxidative-stress tolerance.

Photosynthesis is fundamental to plant growth. Under salt stress, both osmotic stress-induced stomatal closure and excessive Na⁺ accumulation impair the photosynthetic machinery. As a result, the amount of absorbed light exceeds the demand for photosynthesis, which leads to ROS formation in green tissues (Asada 2006). Light-harvesting chlorophyll a/b-binding proteins (LHCs) constitute a pivotal family that plays a central role in photosynthesis by regulating chlorophyll content, photosynthetic efficiency, and ATP supply (Liu et al. 2013). In rice, Lhca1–Lhca4 are intrinsic to the photosystem I (PSI) light-harvesting complex, whereas Lhcb members belong to the photosystem II (PSII) complex (Puthiyaveetil 2011; Perveen et al. 2020). Acting as antennae, these proteins harvest and funnel light energy to the respective reaction centers for efficient photosynthesis. Notably, emerging evidence indicates that LHCs also play broad roles in stress responses (Lan et al. 2022). For example, in Arabidopsis, the ELIP1 and ELIP2 genes (AT4G14690 and AT3G22840) exhibit stress-specific expression patterns under temperature, salt, and drought stresses (Luo et al. 2022). Similarly, in rice, several LHC family members display distinct transcriptional responses to abiotic stresses (Ishida et al. 2011). However, the specific role of LHC genes in rice salt tolerance remains unclear.

In this study, we mapped a major QTL qRLS8 for rice salt tolerance using 226 BC₂F₇ backcross introgression lines (ILs) derived from the cross between indica variety Minghui 63 (MH63) and japonica variety 02428. Based on phenotypic evaluation using transgenic lines, Lhca4 positively regulating rice salt tolerance was verified as the candidate gene of qRLS8. Lhca4 encodes a subunit of PSI light-harvesting protein. Overexpression of Lhca4 increased the activities of superoxide dismutase and peroxidase, and decreased the accumulation of H₂O₂ and O₂⁻, holding a high photochemical efficiency under salt stress conditions. In contrast, knockout of Lhca4 increased accumulation of reactive oxygen species in rice, resulting in disruption of the chloroplast lamellae and a decrease in photochemical efficiency. Transcriptome analysis revealed that Lhca4 mediated salt stress response pathway involved in metabolic regulation, enzyme activity regulation, and antioxidant regulation. Our findings uncover a novel LHC-mediated pathway underlying salt stress response and provide valuable gene sources for the development of salt-tolerant rice varieties through molecular breeding.

Result

Identification of Lhca4 as a Key Putative Gene Underlying the Major ST-QTL

A BC₂F₇ population of 226 ILs was developed using salt-tolerant indica variety MH63 (recurrent parent) and salt-sensitive japonica variety 02428 (donor parent) to evaluate salt tolerance at seedling stage. The shoot length (SL), shoot weight (SW), root length (RL) and root weight (RW) of the 226 ILs were measured under salt stress. All four traits showed a continuous distribution among the ILs, suggesting that these traits should be controlled by multiple QTLs (Additional file 2: Fig. S1A–D). A total of 36 additive QTLs affecting the four traits in 29 regions of the rice genome were identified in the IL population (Additional file 1: Table S1, Additional file 2: Fig. S1E). A major QTL, between Bin3105 and Bin3106 (62.5–65.5 cM or 20939461–23082866 bp) on chromosome 8 was detected with the largest LOD score (15.6) and phenotypic variation explained (PVE) (14.3%) for RL under salt stress (Fig. 1A; Additional file 1: Table S1). The favorable allele with increased RL was from the donor parent 02428. The QTL mapping plots for the other three traits are shown in Additional file 2: Fig. S2.

Fig. 1.

QTL mapping and candidate gene analysis. A The LOD score profiles for root length under salt stress (RLS). B MSU_IDs and annotations of differentially expressed genes located within the qRLS8 interval. C Expression profiling of candidate genes located in the qRLS8 interval between tolerant (‘T’ indicated the DQ46) and sensitive (‘S’ indicated the DQ138) materials under both normal and salt stress conditions. * and ** indicate the significant difference at the P < 0.05 and P < 0.01 level between two different materials, respectively; ns indicates no significant difference (two-tailed Student’s t-test)

To identify the candidate genes of two contrasting ILs, the salt-tolerant IL DQ46 (carrying the MH63 allele at qRLS8) and the salt-sensitive IL DQ138 (carrying the 02428 allele at qRLS8) were used for RNA-seq. By integrating expression profiles with gene annotation, six candidate genes were identified (Fig. 1B). Among them, the expression of LOC_Os08g33820 (Lhca4) was increased by salt stress in both ILs (Fig. 1C). Notably, the expression level of Lhca4 in the salt-tolerant IL (DQ46) was more than twice that in the salt-sensitive IL (DQ138). This expression difference was significantly greater than those observed for the other five candidate genes (Fig. 1C), indicating that Lhca4 is a key putative gene underlying salt tolerance in rice.

Lhca4 is a Positive Regulator of ST in Rice

To verify the function of Lhca4 in the response to salt stress in rice, we examined its expression dynamics at different time points after salt treatment using RT-qPCR. The results showed consistent Lhca4 expression patterns under salt stress in both MH63 and 02428 (Fig. 2E). Specifically, the expression level of Lhca4 was rapidly upregulated to a peak at 6 h after salt stress treatment. At this peak time point, the relative expression level of Lhca4 in MH63 was significantly higher than that in 02428. From 6 to 48 h after salt stress treatment, Lhca4 expression gradually declined in both parents with a slightly higher expression level in MH63 than in 02428, though this difference was not statistically significant (Fig. 2E). We performed next-generation sequencing (NGS) on the parental lines. The genome-wide variations between the two parents are shown in Additional file 2: Fig. S3A, and the variations in the upstream and downstream regions of the Lhca4 gene are illustrated in Additional file 2: Fig. S3B.

Fig. 2.

Phenotypic analysis of Lhca4 transgenic material under normal and salt stress conditions. A, B Both WT and KO materials under normal and salt-treated conditions were photographed on the 8th day of salt treatment, bar = 5 cm. C, D Both WT and OE materials under normal and salt-treated conditions were photographed on the 14th day of salt treatment, bar = 5 cm. E Expression profile of Lhca4 in response to salt stress in rice by RT-qPCR. F–H Chlorophyll content, shoot length, and primary root length of KO and WT under normal conditions and salt treatment. I–K Chlorophyll content, shoot length, and primary root length of OE and WT under normal conditions and salt treatment. *, ** and *** indicate the significant differences at the P < 0.05, P < 0.01 and P < 0.001 level between transgenic material and wild type, respectively; ns indicates no significant difference (two-tailed Student’s t-test)

To clarify the role of Lhca4 in regulating rice ST, we examined phenotypic differences between Lhca4-transgenic lines (KO and OE lines) and wild-type (WT) MH63 under salt stress conditions (Fig. 2, Additional file 2: Fig. S4). Under control conditions, no significant morphological differences were observed between transgenic lines and WT. In contrast, distinct differences among KO lines, OE lines and WT were observed under salt stress conditions (Fig. 2A–D). On the 8th day after salt stress treatment, the KO lines exhibited significantly lower levels of chlorophyll content, plant height, and root length than WT, with average reductions of 44.8%, 21.3%, and 29.6%, respectively (Fig. 2F–H). On the 14th day after salt treatment, the OE lines displayed stronger growth, with chlorophyll content and plant height increasing by 57.6% and 13.8%, respectively, compared to WT (Fig. 2I–K). These phenotypic results confirm that Lhca4 positively regulates ST in rice.

Lhca4 Influence Reactive Oxygen Species (ROS) Accumulation and ROS-Scavenging Machinery of Rice

To explore the physiological mechanism of ST mediated by Lhca4, we analyzed the ion (K⁺ and Na⁺) content under salt stress and growth performance under PEG6000-simulated osmotic stress between the transgenic lines and WT. There were no significant differences in K⁺ and Na⁺ contents in shoots and roots among KO lines, OE lines and WT under either condition (Additional file 2: Fig. S5). Under PEG6000-simulated osmotic stress, WT and transgenic lines exhibited comparable growth performance and survival rates after 7 days of rewatering (Additional file 2: Fig. S6). These results indicated that the enhanced ST conferred by Lhca4 is unlikely to be attributed to the response to ionic or osmotic stress.

Reactive oxygen species (ROS) overaccumulation is a common cellular response to abiotic stress in plants. Consistent with the phenotypic observations, in shoot tissues, KO lines exhibited a 70% increase in MDA (a marker of lipid peroxidation) content compared to WT, and OE lines showed MDA content comparable to that of WT (Fig. 3A). Compared to the WT, ROS content was increased in KO lines but decreased in OE lines (Fig. 3C, E). Specifically, the contents of H₂O₂ and O₂⁻ in shoots were increased by 10% and 29% in KO lines and decreased by 16% and 14% in OE lines compared to WT, respectively (Fig. 3C, E). Furthermore, under salt stress, KO lines exhibited a 12% decrease in SOD activity and OE lines showed 30% and 21% increases in SOD and POD activities compared to WT, respectively (Fig. 3B, D). No significant differences in these physiological traits were observed in roots among genotypes (Additional file 2: Fig. S7). These results demonstrate that Lhca4 regulates the ROS-scavenging system to influence ROS homeostasis under salt stress.

Fig. 3.

Physiological indices of Lhca4 transgenic material under normal and salt stress conditions. A–F MDA content in the shoot, SOD activity in the shoot, H₂O₂ content in the shoot, POD activity in the shoot, O₂⁻ content in the shoot, ¹O₂ content in the shoot. Different letters in the graph indicate significant differences between materials within the same group according to Duncan’s multiple range post-hoc test (P < 0.05). G DAB staining, dyeing by oscillation for 24 h. H NBT staining, dyeing by oscillation for 24 h

DAB and NBT staining of WT, KO-1 and OE-1 lines further confirmed the effect of Lhca4 on ROS accumulation (Fig. 3G, H). The intensity of H₂O₂ (brown) and O₂⁻·(blue) followed the order KO > WT > OE under salt stress, while no staining differences were observed under control conditions (Fig. 3G, H). These histochemical results indicated that Lhca4 overexpression suppresses ROS accumulation by regulating both ROS production and scavenging. These results demonstrated that Lhca4 enhances rice ST mainly by strengthening the ROS-detoxifying machinery to reduce cellular oxidative damage, rather than by modulating ion homeostasis or osmotic adjustment.

Lhca4 Maintains PSII Photochemical Efficiency Under Salt Stress

We conducted GUS activity assays to investigate the spatial expression pattern of Lhca4. The results showed that Lhca4 was mainly expressed in shoots, with no expression detected in roots or seeds (Fig. 4A, B). Leaf blade cross-section staining revealed that Lhca4 is mainly expressed in vascular tissue (Fig. 4C). Subcellular localization analysis showed that the green fluorescence emitted by Lhca4-GFP colocalized with the intrinsic red autofluorescence of chloroplasts, demonstrating that Lhca4 is localized to chloroplasts (Fig. 4E).

Fig. 4.

Expression patterns of Lhca4 and its subcellular localization in rice. A–C Graph showing the results of GUS staining of rice tissues. Scale bars = 3 cm (seedling), 500 μm (roots) or 200 μm (leaf cross sections). D Fv/Fm determination of Lhca4 transgenic material under normal and salt stress conditions. E Subcellular localization of Lhca4 with GFP in rice protoplasts. Scale bars = 5 μm. F Leaves were taken for 6 h of salt stress, the main observation site is the photosynthetic lamellae of chloroplasts, scale bar = 2 μm

Intact chloroplast structure is essential for efficient ROS scavenging within the photosynthetic machinery and for preventing ROS burst into the cytosol. Transmission electron microscopy observations showed that under control conditions, all genotypes exhibited chloroplasts with distinct envelopes and regularly stacked thylakoid membranes (Fig. 4F). Under salt stress, the chloroplasts of WT and OE lines retained intact ultrastructure and clearly stacked photosynthetic lamellae, maintaining photosynthetic competence. In contrast, KO lines displayed severe cellular shrinkage, distorted chloroplast morphology, and obvious disruption to the photosynthetic lamellae (Fig. 4F).

Given that Lhca4 encodes a light-harvesting component of the photosynthetic system, the maximum photochemical efficiency of photosystem II (PSII), represented by the Fv/Fm ratio, was measured under salt stress. Under normal conditions, no significant differences in Fv/Fm were observed between the three genotypes. Salt stress induced a decrease in Fv/Fm in all lines, but this decrease was significantly smaller in the OE lines than in the KO lines (Fig. 4D). These results suggest that overexpression of Lhca4 mitigates the salt-induced loss of PSII maximum efficiency (Fv/Fm) by balancing excitation pressure between PSII and PSI. This protective effect helps to sustain overall photosynthetic performance and growth under salt stress.

Transcriptomic Analysis of Lhca4 Transgenic Rice in Response to Salt Stress

To dissect the early biological processes and transcriptional regulatory networks mediated by Lhca4 in response to salt stress at seedling stage, we performed RNA-seq on OE, KO and WT lines exposed to salt stress for 6 h (a time point corresponding to the peak Lhca4 expression). Differential expression analysis identified 2942 differentially expressed genes (DEGs) between KO and WT (1389 upregulated and 1553 downregulated) and 1378 DEGs between OE and WT (746 upregulated and 632 downregulated) under salt stress, respectively (Additional file 1: Table S2). When salt-treated and control plants were compared within each genotype, OE lines exhibited 2497 salt-responsive DEGs (1543 upregulated and 954 downregulated), KO lines displayed 7751 salt-responsive DEGs (3849 upregulated and 3902 downregulated), and WT showed 5540 salt-responsive DEGs (2873 upregulated and 2667 downregulated). These results indicate that Lhca4 has a significant influence on the gene regulatory network of rice rapid response to salt stress.

To identify genes whose salt-responsive expression is specifically regulated by Lhca4, we performed pairwise comparisons under both control and salt stress conditions, and common DEGs were visualized in Venn diagrams (Fig. 5A, B). A total of 610, 391, 3,064 and 679 specific DEGs were identified based on the comparisons of S_KO_6h vs S_WT_6h, S_OE_6h vs S_WT_6h, S_KO_6h vs C_KO_6h, and S_OE_6h vs C_OE_6h, respectively (“S” denotes salt stress and “C” denotes control). Further screening of genes with opposite expression patterns in KO and OE lines under salt stress revealed that 21 genes were downregulated in KO lines and upregulated in OE lines, and that 16 genes exhibited the opposite pattern of expression (Fig. 5C, D).

Fig. 5.

Venn diagrams and GO enrichment analysis of differentially expressed genes. A Venn diagram of all DEGs between KO and WT. B Venn diagram of all DEGs between OE and WT. C Venn diagrams of the DEGs upregulated in OE and downregulated in KO. D Venn diagrams of the DEGs upregulated in KO and downregulated in OE

Gene ontology (GO) enrichment analysis showed that specific DEGs in S_KO_6h vs S_WT_6h were significantly enriched in terms related to photosynthesis, antioxidant activity and oxidoreductase activity (Additional file 2: Fig. S8A). DEGs specific to S_KO_6h vs C_KO_6h were prominently enriched in peptide metabolism, amide metabolism and zinc-ion binding (Additional file 2: Fig. S8B). For S_OE_6h vs S_WT_6h, the specific DEGs were mainly enriched in the peptide biosynthesis, amide metabolism and ribosome (Additional file 2: Fig. S8C). DEGs specific to S_OE_6h vs C_OE_6h were mainly enriched in oxidoreductase activity (Additional file 2: Fig. S8D).

Further KEGG pathway analysis obtained similar findings (Additional file 2: Fig. S9). Specific DEGs in S_KO_6h vs S_WT_6h were significantly enriched in valine, leucine and isoleucine biosynthesis and photosynthesis pathways (Additional file 2: Fig. S9A). DEGs specific to S_KO_6h vs C_KO_6h were significantly enriched in ribosome biogenesis and aminoacyl-tRNA biosynthesis pathways (Additional file 2: Fig. S9B). For S_OE_6h vs S_WT_6h, the specific DEGs were enriched in ribosome, aminoacyl-tRNA biosynthesis and tyrosine metabolism pathways (Additional file 2: Fig. S9C), while DEGs specific to S_OE_6h vs C_OE_6h were enriched in phosphatidylinositol signaling, pyruvate metabolism, linolenic acid metabolism and tyrosine metabolism pathways (Additional file 2: Fig. S9D). Together, the transcriptomic analysis results suggest that Lhca4 maintains photosynthetic efficiency and enhances rice ST by regulating multiple biological processes, including ROS homeostasis, chloroplast development, redox balance, and signal transduction pathways.

In order to identify the different co-expressed modules under salinity stress in rice, we conducted a weighted gene co-expression network analysis (WGCNA) using 18 transcriptome samples. To construct a scale-free network, a soft-thresholding power of 10 was selected to cluster genes into distinct modules based on their expression profiles (Fig. 6A). To identify the specific gene co-expression networks closely associated with salt responses, a module-trait relationship analysis was performed (Fig. 6B). Among the seven identified modules, the MEblue module exhibited the most striking and significant correlations with several key physiological traits. Specifically, the MEblue module showed a highly significant negative correlation with SOD_R (r =  − 0.88) and H₂O₂_R (r =  − 0.67). Conversely, it displayed strong positive correlations with POD.L (r = 0.7), ¹O₂.R (r = 0.7), MDA_L (r = 0.64), and SOD_L (r = 0.62). These findings strongly suggest that the MEblue module is deeply involved in regulating antioxidant activities and ROS homeostasis under stress conditions. Given its critical role, the MEblue module was selected as the target module for the subsequent mining of key co-expressed regulatory genes. To further elucidate the biological functions of the co-expressed genes within the MEblue module, GO enrichment analysis was performed (Fig. 6C). These co-expressed genes were predominantly enriched in the intracellular signaling cascade, the electron transport chain, and the response to abiotic stimulus, exhibiting highly significant enrichment. Because the electron transport chain is a major source of ROS and the MEblue module is highly correlated with antioxidant traits, we selected representative candidate genes from these pathways. Their expression profiles associated with ROS homeostasis and stress responses were subsequently validated via RT-qPCR. The expression profiles of five genes associated with ROS or stress responses are illustrated in Fig. 6D–H.

Fig. 6.

Identification and validation of key co-expression modules and candidate genes associated with salt stress response. A Construction of the different modules. B Heatmap showing the module-trait relationships. C GO enrichment analysis of the co-expressed genes within the MEblue module. D–H Expression profiles of five representative ROS- or stress-associated genes (OsAIM1, PH8, OsGR2, OsP5CS, and OsGLR1) in WT, KO, and OE lines under 140 mM NaCl treatment. The bars and dots represent the RNA-seq (FPKM) and relative expression levels, respectively. Asterisks indicate statistically significant differences compared with the WT (*P < 0.05, **P < 0.01, ***P < 0.001; Student’s t-test). ‘ns’ indicates no significant difference

Discussion

Rapid industrialization has led to a reduction in arable land and aggravated soil salinization (Naveed et al. 2018), posing a major obstacle to meeting the ever-increasing global grain demand. Therefore, the development of high-yield salt-tolerant crop cultivars has become a critical priority for sustainable rice production. In this study, we identified and functionally characterized Lhca4, encoding a subunit of the PSI light-harvesting complex, as a key gene for ST in rice. Functional analysis suggested that Lhca4 positively regulates ST by maintaining the structural integrity and efficiency of the photosynthetic machinery and enhancing the capacity to mitigate oxidative stress in rice. These findings highlight a core component of the photosynthetic machinery as a new and potent target for developing salt-tolerant rice varieties.

Salt-induced overproduction of ROS is a primary cause of cellular damage, leading to oxidative stress that impairs macromolecular function and membrane integrity (Liang et al. 2018). Our results establish a direct link between Lhca4 and ROS homeostasis. Lhca4 OE lines showed significantly lower H₂O₂ and O₂⁻ levels with higher activities of SOD and POD, while KO lines accumulated excessive ROS and exhibited compromised antioxidant defenses (Fig. 3). This is consistent with recent studies showing that enhanced ROS scavenging improves ST in plants (Zhao et al. 2020), but our work uncovers a novel role for an LHC protein in modulating ROS homeostasis.

Salt stress impairs the photosynthetic function of chloroplasts and causes excess light absorption that exceeds photosynthetic demand, leading to ROS formation in plant green tissues (Asada 2006). Chloroplasts contain thylakoid membranes housing LHCs and the photosynthetic electron transport chain, as well as the stroma where the Calvin cycle fixes CO₂. As a core component of the photosynthetic machinery, Lhca4 was localized to chloroplasts (Fig. 4E), consistent with observations in Arabidopsis and cotton (Zhang et al. 2021). Our GUS assay showed strong Lhca4 promoter activity in aerial parts (especially vascular tissues) but not in roots or seeds (Fig. 4A–C), consistent with abundant leaf-specific LHC transcripts (Lan et al. 2022). LHC proteins are well-known for their roles in photosynthetic electron transport, which contributes to increased photosynthetic rate and grain yield (Umate 2014). In addition, the LHC family modulates physiological and metabolic networks to adapt to biotic stress (e.g., pathogen attack) and abiotic stress (e.g., extreme temperature, salinity, drought) (He et al. 2020). For example, LOC_Os07g08150, an LHC member encoding a light-induced protein, is downregulated under cold stress (Jeong et al. 2018), while Lhcb5 dissociates from PsbS during blast infection to trigger ROS-dependent defense (Liu et al. 2019). Here, our data suggested Lhca4 as a critical protector of chloroplast integrity under salt stress. Lhca4 KO lines exhibited profound thylakoid disorganization, while Lhca4 OE lines maintained elevated PSII maximum photochemical efficiency (Fv/Fm) under salt stress (Fig. 4D). Given Lhca4 plays a key role in stabilizing the PSI supercomplex for efficient light harvesting and electron transfer (Rankelytė et al. 2024), its protective function in photosynthetic structure may directly underpin enhanced photosynthetic efficiency by minimizing photo-oxidative damage. Although Lhca4 is expressed in the shoots, it likely functions by maintaining chloroplast integrity (Fig. 4F). This, in turn, indirectly mitigates the salt-induced growth inhibition in the roots, ultimately resulting in the observed differences in root growth.

Transcriptome analysis further supported this mechanism. DEGs regulated by Lhca4 were enriched in pathways related to photosynthesis, antioxidant defense, and metabolic regulation (Additional file 2: Fig. S8). Among them, oxidoreductase activity-related DEGs likely reinforce ROS scavenging, while amide metabolism (centered on glutamine and asparagine) may support osmotic adjustment and antioxidant enzyme activation under salt stress. KEGG enrichment also highlighted pathways involved in osmolyte synthesis and stress signaling (Additional file 2: Fig. S9), which amplify stress responses and interact with hormone signaling to coordinate tolerance (He et al. 2020).

By using WGCNA, we precisely identified the MEblue module as the core co-expression network highly associated with the salt stress response in rice. The MEblue module showed obvious tissue-specific correlations with antioxidant enzyme activities (Fig. 6B), and the genes in this module were highly enriched in the electron transport chain and intracellular signaling cascade pathways (Fig. 6C). To ensure the reliability of this regulatory network, we selected five representative genes enriched in ROS and stress response pathways for RT-qPCR validation (Fig. 6D–H).

Among them, OsAIM1 plays a crucial role in salicylic acid (SA) biosynthesis. Previous studies have shown that certain WRKY transcriptional repressors inhibit the expression of redox and ROS-scavenging genes, and SA promotes ROS accumulation by inducing the activity of these WRKY repressors (Xu et al. 2023). In our study, the expression level of OsAIM1 was significantly decreased in the OE lines (Fig. 6D), suggesting a relief of the inhibition on ROS-scavenging gene expression. Furthermore, OsP5CS is essential for proline biosynthesis. Its expression is typically higher in salt-tolerant varieties compared to salt-sensitive ones, leading to greater proline accumulation and consequently enhanced salt tolerance (Sripinyowanich et al. 2013). We found that the expression of OsP5CS was higher in the OE lines than in the KO lines (Fig. 6G), implying that Lhca4 might regulate rice salt tolerance by modulating OsP5CS expression. Additionally, GLR1 knockout lines are known to possess a stronger capacity to scavenge ammonia and ROS. Evidence indicates that GLR1 directly binds to the promoters of OsGS1, OsCYP51G3, and OsCATA to repress their expression, which prevents the timely clearance of accumulated ammonia and ROS, ultimately causing plant death (Ren et al. 2023). Here, the expression of GLR1 was lower in the OE lines compared to the KO lines (Fig. 6H), indicating a higher expression level of ROS-scavenging-related genes in the overexpression lines. Therefore, Lhca4 not only physically protects the photosystem but also indirectly orchestrates a broader transcriptomic network to optimize chloroplast function and systemic salt tolerance. Thus, our findings identified Lhca4 as a valuable gene target acting on the photosynthetic machinery to sustain energy production for defense and repair and to regulate ROS homeostasis, offering a promising strategy for developing salt-tolerant rice varieties.

Conclusion

We integrated linkage mapping with transcriptome profiling to identify Lhca4 as a key gene in the salt stress response of rice. Functional validation revealed that Lhca4 acts as a positive regulator for tolerance to salt stress in rice. Physiological analyses indicated that under salt stress conditions, Lhca4 modulates both the accumulation of reactive oxygen species (ROS) and the activity of the ROS-scavenging system. Moreover, overexpression of Lhca4 enhanced the maximum photochemical efficiency of PSII. Transcriptomic profiling further revealed that Lhca4 likely activates oxidoreductase-related genes through specific signaling cascades under salt stress, while simultaneously influencing chloroplast ultrastructure and the rate of photosynthetic electron transport. Taken together, these findings suggest that Lhca4 integrates the regulation of ROS homeostasis with photosynthetic performance to confer salt tolerance in rice.

Materials and methods

Plant Materials for QTLs Identification and Assessment of Salt Tolerance

We constructed a backcross introgression population consisting of 226 lines (BC₂F₇), derived from the recurrent salt-tolerant parent Minghui 63 (MH63) and the donor salt-sensitive parent 02428 (Qiu et al. 2017). Seedling-stage salt-tolerance assays were conducted in the greenhouse of the Institute of Crop Sciences, Chinese Academy of Agricultural Sciences (ICS, CAAS). Seeds were surface-sterilized in 4% (w/v) sodium hypochlorite for 20 min and rinsed five times with sterile distilled water, then soaked at 28°C for 36 h followed by 16 h of germination at 37°C. Germinated seeds were sown in 96-well PCR plates and cultured in distilled water for 7 d, after which the solution was replaced with Yoshida nutrient medium (Yoshida et al. 1976). At the two-leaf-one-heart stage, seedlings were subjected to salt stress by adding NaCl to a final concentration of 100 mM (electrical conductivity = 12 dS ·m^-1). Following 7 days of treatment, root length (RLS) and shoot length (SHLS) were measured. Subsequently, the seedlings were oven-dried at 55°C to constant weight and shoot dry weight (SHWS) and root dry weight (RWS) were determined. Each replicate comprised 3–5 randomly selected plants.

QTL analysis

As we reported previously, the 226 BC₂F₇ lines and two parents were genotyped by resequencing, and were used to construct a high-resolution map of 4568 bins (Qiu et al. 2017). After genotyping the 226 BC₂F₇ lines, redundant bins were removed with the BIN module of IciMapping (Meng et al. 2015), and a genetic linkage map was generated using the MAP function. Additive QTLs were mapped using the interval mapping of the additive (ICIM-ADD) mapping method. The threshold for the logarithm of odds (LOD) scores was set to 2.5.

Parental NGS and SNP Identification

The DNA sample of MH63 and 02428 was used as input material for the DNA library preparations. The sequencing library was generated and index codes were added to each sample. Briefly, genomic DNA sample was fragmented by sonication to a size of 350 bp. Then DNA fragments were end-polished, A-tailed, and ligated with the full-length adapter for Illumina sequencing, followed by further PCR amplification. After PCR products were purified. Subsequently, library quality was assessed and quantified by RT-qPCR. The qualified libraries were pooled and sequenced on Illumina platforms with a PE150 strategy, according to effective library concentration and data amount required, and the average sequencing depth was 10× (Novogene, Beijing, China). The Genome Analysis Toolkit (GATK) was used to detect single-nucleotide polymorphisms (SNPs) in individual samples (McKenna et al. 2010), and the chromosomal density map was plotted by the R package CMplot.

Generation of Transgenic Plants

Transgenic lines targeting Lhca4 were generated with Minghui 63 background. Single-guide RNAs were designed to target exonic sequences located 5′-proximal to, or within, conserved functional domains of Lhca4. The binary vector pYLCRISPR/Cas9Pubi-H carrying the selected gRNA expression cassettes was used for CRISPR/Cas9-mediated mutagenesis. The full-length Lhca4 coding sequence was inserted into pMDC43 under control of the CaMV 35S promoter; the resulting vector confers hygromycin (HYG) resistance. Both constructs, together with Agrobacterium-mediated transformation, were commissioned from Biogle (Wuhan, China). Four homozygous T₄ transgenic lines were employed in this study: two independent knock-out lines (designated KO-1 and KO-2) and two independent over-expression lines (OE-1 and OE-2). The conventional parents Minghui 63 (indica) served as controls (WT). Seedlings of a Minghui 63-derived GUS reporter line were utilized for histochemical GUS assays.

Plant Growth and Treatments for Transgenic Lines

The transgenic lines and MH63 were grown in the greenhouse at ICS, CAAS and cultured as the aforementioned above. At the three-leaf-one-heart stage, salt stress or osmotic stress were imposed by 140 mM NaCl or 20% PEG6000, this solution was renewed every 5 d. The control and treatment groups were arranged in three to five biological replicates.

Expression Analysis and Histochemical Analysis of GUS Activity

The MH63 and 02428 were used as materials for Lhca4 expression profile analysis. Total RNA was extracted from shoot samples (3, 6, 12, 24, 48, and 72 h, after salt treatment), using an RNA simple Total RNA kit (Tiangen, Beijing, China). Reverse transcription into cDNA was performed using the Fast King One-step RT-qPCR kit (Tiangen, Beijing, China). RT-qPCR analysis was conducted as previously described (Liu et al. 2024). Primers used for RT-qPCR are listed in Additional file 1: Table S3. GUS activity was detected using the GUSblue kit (Huayueyang, Beijing, China) and samples were destained with pure ethanol prior to examination.

Subcellular localization

To capture the full-length CDS of Lhca4, cDNA of Minghui 63 was amplified with gene-specific primers listed in Additional file 1: Table S3. In-fusion cloning was designed on the basis of the Hind III and EcoRI sites of the pYBA-1132-GFP vector; Positive colonies were verified by Sanger sequencing. Transient expression assays in rice protoplasts were performed as described previously (Gu et al. 2022). GFP fluorescence was visualised and photographed with a Zeiss LSM 980 laser-scanning confocal microscope.

Physiological Analysis

Phenotypic evaluation: Phenotypic evaluation was carried out when phenotypic differences between genotypes were most pronounced. The investigation for KO and OE lines was conducted on the 8th day and 14th day after salt treatment, respectively. For each replicate, three to five randomly chosen seedlings were photographed and measured for shoot height, root length and chlorophyll content. Chlorophyll content was measured using a SPAD meter (Minolta, Japan).

Ion content determination: On the 6th day of salt stress, shoot and root samples were collected, oven-dried to constant weight, and weighed. A thermostatic shaking water bath was pre-heated to 90°C. Dried samples (50 mL centrifuge tubes) received 10 mL of ice-cold glacial acetic acid for roots or 20 mL for shoots, shaken for 2 h, cooled to room temperature, and supernatants were transferred to 2 mL tubes. K⁺ and Na⁺ contents were determined with an S2 flame atomic-absorption spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) at 766.5 nm (K⁺) and 589.0 nm (Na⁺). Calibration standards (1, 5, 10, 15, 20, 25 ppm) were prepared from 1000 μg mL⁻¹ K⁺/Na⁺ stock. Ion content was calculated as:

X: mass fraction (mg kg⁻¹); c: measured concentration (mg L⁻¹); V: acetic-acid volume (mL); N: dilution factor; m: dry weight (g).

Determination of reactive oxygen species (ROS) accumulation and ROS-scavenging activity: 5th day of salt stress, the shoot and root samples of MH63 and transgenic lines were sampled from both stress and control plants for ROS analyses. Tissues were ground to a fine powder in liquid-nitrogen-chilled mortars. Activities of superoxide dismutase (SOD) and peroxidase (POD), as well as contents of malondialdehyde (MDA), hydrogen peroxide (H₂O₂), superoxide anion (O₂⁻) and singlet oxygen (¹O₂), were quantified using commercial kits (Greens Bio-Science, Suzhou) according to the manufacturer’s instructions.

ROS staining and chloroplast ultrastructure observation: For ROS staining, 5-cm leaf segments at the same position were excised from treated and control plants and immediately immersed in 30 mL of freshly prepared DAB solution (50 mM Tris–HCl, pH 3.8) or 30 mL of NBT solution (50 mM sodium phosphate buffer, pH 7.8) in 50-mL Falcon tubes. After 24 h of gentle shaking in darkness, leaves were cleared in 95% ethanol at 80°C (solution changed every 10 min) until complete bleaching, then photographed. For ultrastructural analysis, 1 mm × 1 mm leaf squares were fixed in 3% (v/v) glutaraldehyde under vacuum until fully submerged. Samples were dehydrated, critical-point dried, sputter-coated, and examined under a transmission electron microscopy to document chloroplast morphology.

The maximum photochemical efficiency of PSII (Fv/Fm) measurement: The maximum photochemical efficiency of PSII (Fv/Fm) was monitored daily from 0 to 6 d after initiating salt stress. For each time-point, leaves were dark-adapted for 30 min and Fv/Fm was recorded with a FluorPen FP100 fluorometer (Photon Systems Instruments, Drasov, Czech Republic). This parameter reflects the potential maximum quantum efficiency of PSII. Three biological replicates were maintained for each treatment.

RNA-seq and Analysis

At 6 h after salt treatment, the leaf samples of DQ46 (salt-tolerant IL), DQ138 (salt-sensitive IL), MH63 (WT), the lines KO-1, and OE-1 were collected from both control and stress plants, respectively, and immediately frozen in liquid nitrogen and stored at − 80°C until RNA extraction. RNA extraction, sequencing, and data analysis were conducted as previously described (Zhang et al. 2019). Differentially expressed genes (DEGs) (P_adj < 0.05 and fold change ≥ 2) were identified using DEGseq2 software (Love et al. 2014). Functional classification and GO enrichment analysis of differentially expressed genes were performed with the GOseq package in R. Venn diagrams were constructed with Venny 2.1 (bioinfogp.cnb.csic.es/tools/venny). WGCNA was conducted based on 47,549 genes using the WGCNA package in R (Langfelder and Horvath 2008). Missing values were inspected, and lowly expressed genes were filtered out. Specifically, only genes with expression levels higher than 1 in at least 25% of the samples were retained. The first soft threshold that made the scale-free topology R² reach 0.85 was selected, which was set to 10 in this study. The module merging threshold was set to 0.4.

Statistical Analysis

Phenotypic differences among traits were tested by one-way ANOVA using the “agricolae” package in R. Differences were considered significant at p ≤ 0.05. Figures were prepared using GraphPad Prism version 9.0.0 for Windows (GraphPad Software, San Diego, California, USA) or Excel 2019.

Abbreviations

QTL

Quantitative trait locus

PVE

Phenotypic variation explained

ROS

Reactive oxygen species

ST

Salt tolerance

GSH

Glutathione

CAT

Catalases

SOD

Superoxide dismutase

APX

Ascorbate peroxidase

GR

Glutathione reductase

LHCs

Light-harvesting chlorophyll a/b-binding proteins

PSI

Photosystem I

PSII

Photosystem II

SL

Shoot length

SW

Shoot weight

RL

Root length

RW

Root weight

WGCNA

Weighted gene co-expression network analysis

Author Contributions

Conceived and designed the experiments: XZ FZ; Performed the experiments: YQ ZX JA KC GZ JJ; Analyzed the data: JA TZ WW JX; Wrote the paper: JA XZ FZ.

Funding

This work was funded by the Nanfan Special Project, CAAS (YBXM2426), the Central Public-interest Scientific Institution Basal Research Fund (No.S2026QZ23), the National Natural Science Foundation of China (31471429), the Central Public-interest Scientific Institution Basal Research Fund (Y2025YC13), and the CAAS Innovative Team Award.

Availability of Data and Materials

RNA-Seq data were deposited into the NCBI database under accession number PRJNA1392147 and are available at the following URL: https://www.ncbi.nlm.nih.gov.

Declarations

Ethics Approval and Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Competing interests

The authors declare no competing interests.