Gossypium hirsutum L. acc. TM-1 seeds were soaked in sterile water to promote germination. After a 16-h incubation at 30°C, the germinating seeds were transferred to vermiculite-filled pots in a greenhouse (16-h light/8-h dark photoperiod at 23°C). When the cotyledon expanded, uniformly growing seedlings were selected, rinsed clean with running water, and transferred to a hydroponic tank containing Hoagland solution, which was refreshed weekly. To induce salinity stress, the cotton seedlings were grown to the three-leaf stage and then treated with Hoagland solution containing 50, 100, 150, or 200 mM NaCl or 200 mM KCl (treatment group). The seedlings in the control group were treated with Hoagland solution lacking NaCl or KCl. Root, stem, and leaf samples were collected at 0, 1, 3, and 6 h after treatments. The samples were rinsed with distilled water 3–5 times and then immediately frozen in liquid nitrogen and stored at −80°C for the subsequent total RNA extraction. Each treatment was completed with three biological replicates.

The wild-type (WT) A. thaliana plants used in this study were Columbia (Col-0). After a 2-day vernalization at 4°C, the A. thaliana seeds were surface sterilized by soaking in 75% ethanol for 3–5 min. After they were washed with sterile distilled water six times, the seeds were placed in plates containing Murashige and Skoog (MS) medium supplemented with 3% sucrose and 0.7% agar (pH adjusted to 5.8). The plates were incubated at 21°C with a 16-h light/8-h dark photoperiod. After 2 weeks, the A. thaliana seedlings were transferred to pots filled with vermiculite and nutritive soil (1:1, v/v) and placed in a greenhouse (16-h light/8-h dark photoperiod at 21°C).

Total RNA was isolated from TM-1 seedlings using the RNAprep Pure Polysaccharide Polyphenol Plant Total RNA Extraction Kit (Tiangen, Beijing, China). The RNA concentration was determined using the NanoDrop 2000 microvolume spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States), whereas RNA integrity was assessed by 1% agarose gel electrophoresis. First-strand cDNA was synthesized from 1 μg RNA using the PrimeScript^TM II first Strand cDNA Synthesis Kit (TaKaRa, Dalian, China). To clone the vacuolar chloride channel gene in upland cotton, the BlastP and tBlastN programs were performed against the G. hirsutum genome database¹ using the A. thaliana CLCg amino acid sequence as the query. The coding sequences of gene IDs GH_A06G0574 and GH_D06G0541 as the best match were retrieved and gene-specific primers were designed. Full-length cDNAs were amplified using the GhCLCg-1F/R primer pair (Supplementary Table 1) and the KOD-Plus-Neo DNA polymerase (TaKaRa, Dalian, China). The PCR products were purified using the FastPure Gel DNA Extraction Mini Kit (Vazemy Biotech Co., Ltd.) and then cloned using the pEASY^®-Blunt Cloning Kit for the subsequent transformation of Trans1-T1 Escherichia coli competent cells (TransGen Biotech). Positive clones cultured at 37°C were analyzed by sequencing to verify they were correctly transformed.

Arabidopsis thaliana CLC amino acid sequences were downloaded from TAIR database². Seven AtCLC sequences along with G. hirsutum GhCLCg-1A and GhCLCg-1D (Supplementary Table 2) were used to construct a phylogenetic tree according to the neighbor-joining method of the MEGA-X software, with pairwise deletion and 1,000 bootstrap replicates. Conserved domains were detected using InterPro³ and gene structures were examined using TBtools (Chen et al., 2020). Amino acid sequences were aligned using the DNAMAN program.

Total RNA was isolated from the roots, stems, and leaves using the RNA Extraction kit (Tiangen). The HiScript^® II Q Select RT SuperMix for qPCR (+gDNA wiper) (Vazemy Biotech Co., Ltd.) was used to synthesize cDNA from approximately 1 μg total RNA. Gene-specific primers were designed according to conversed sequences in GhCLCg-1A and GhCLCg-1D (Supplementary Table 1). The qRT-PCR analysis was performed using the LightCycler 480 system (Roche, Switzerland) and the ChamQ Universal SYBR qPCR Master Mix (Vazemy Biotech Co., Ltd.). The GhHIS3 gene served as an internal control (Ma et al., 2020). Three biological replicates were analyzed for each sample. Relative gene expression levels were calculated according to the 2^–ΔCT method (Livak and Schmittgen, 2001).

The full-length GhCLCg-1A and GhCLCg-1D open reading frames without a stop codon were amplified by PCR using primers with BamHI and SmaI sites (Supplementary Table 1) and then inserted into the BamHI/SmaI-digested pCAMBIA2300-35S-GFP-HA plasmid, which contains the CaMV 35S promoter. The resulting recombinant plasmids were inserted into Trans1-T1 competent cells (TransGen Biotech). The accuracy of the inserted sequences in the transformants was confirmed by sequencing. The recombinant plasmids carrying the GhCLCg-1A-GFP and GhCLCg-1D-GFP constructs were inserted into Agrobacterium tumefaciens strain GV3101 cells via a heat shock protocol. The E. coli and A. tumefaciens strains containing the overexpression plasmids were stored at −80°C.

A transient expression system using A. thaliana mesophyll protoplasts was prepared as previously described (Sheen, 2001; Feng et al., 2021). Protoplasts were co-transfected with 10 μg GhCLCg-1A-GFP/GhCLCg-1D-GFP recombinant plasmids and 10 μg vacuolar marker protein δ-TIP-RFP (Jauh et al., 1998) according to a polyethylene glycol-mediated transfection method (Yoo et al., 2007). After incubating the protoplasts at room temperature for 12–20 h in darkness, the green fluorescent protein (GFP) and red fluorescent protein (RFP) signals were detected using the FV1200 confocal laser scanning microscope (Olympus, Japan).

Based on the GhCLCg-1A and GhCLCg-1D conserved sequences, a 430-bp fragment was amplified by PCR from cotton cDNA using the GhCLCg-1-p-F/R primers (Supplementary Table 1) and inserted between the SacI and EcoRI restriction sites in the VIGS vector pTRV2 of the ClonExpress^® Ultra One Step Cloning Kit (Vazemy Biotech Co., Ltd.). Plasmids from the positive transformants carrying pTRV1 (helper plasmid) and TRV:GhCLCg-1 were inserted into A. tumefaciens strain GV3101 cells via a heat shock protocol. Similarly, we constructed a visual marker (TRV:CLA, where CLA is cloroplastos alterados) to monitor the silencing efficiency. Additionally, TRV:00 (empty vector) was used as the negative control. The VIGS experiments were performed by the tobacco rattle virus (TRV) system as previously described (Pang et al., 2013).

Two fully expanded cotyledons of cotton seedlings were used for the A. tumefaciens infiltration. At 10 days after the transformation, the true leaves of TRV:CLA-silenced cotton plants exhibited signs of albinism. The second true leaves, stems, and roots were harvested from at least three randomly selected TRV:00 and TRV:GhCLCg-1 cotton plants for an RNA isolation. The efficiency of GhCLCg-1 silencing was evaluated by qRT-PCR, with GhHIS3 used as the internal control (Ma et al., 2020). The TRV:00 and TRV:GhCLCg-1 plants were then treated with 200 mM NaCl for 3-day, the plants were photographed and their roots, stems, and leaves were collected for an analysis of their Cl^–, Na⁺, and K⁺ contents. The assays were completed with three biological replicates.

Because GhCLCg-1A and GhCLCg-1D have highly similar sequences, we used GhCLCg-1A to generate transgenic A. thaliana plants overexpressing GhCLCg-1. The recombinant plasmid carrying the GhCLCg-1A-GFP fusion construct was inserted into A. thaliana plants using the floral dip transformation method (Clough and Bent, 1998). Homozygous plants were selected on MS medium containing 50 μg/ml kanamycin. The T₁ generation plants were examined by PCR to confirm the positive lines were transformed correctly. Homozygous T₃ generation progenies were analyzed by reverse transcription PCR using the GhCLCg-1-Q-F/R primers (Supplementary Table 1). Seeds from homozygous T₃ plants were placed on MS agar medium (control group) or MS agar medium supplemented with 100 mM NaCl (salt treatment group). The seedlings were photographed after 3 weeks. Additionally, the taproot length was measured using a vernier caliper. Nine plants per line were examined and their tissues were harvested for an analysis of the Cl^–, Na⁺, and K⁺ contents.

The VIGS plants and transgenic A. thaliana plants overexpressing GhCLCg-1 were collected and incubated at 105°C for 10 min to denature enzymes and then at 75°C to achieve a constant weight. The samples were subsequently ground to a powder. To measure the Cl^– content, a 50-ml Erlenmeyer flask was washed with 5% HNO₃. After drying, 0.1 g powder was added to the flask and resuspended with 15 ml boiled deionized water. The solution was cooled to room temperature and then filtered. The filtrate was collected in a 25-ml volumetric flask to a constant volume. The solution was analyzed using an ion chromatography system (ICS-5000) (Thermo Fisher Scientific, Waltham, MA, United States) to measure the Cl^– content. The experiment was repeated three times.

To measure the Na⁺ and K⁺ contents, 0.1 g powder was treated with nitric acid (HNO₃), hydrogen peroxide (H₂O₂), and hydrofluoric acid (HF) and then microwaved (i.e., digested) before being analyzed using an inductively coupled plasma optical emission spectrometry system (ICAP-7400) (Thermo Fisher Scientific, Waltham, MA, United States). Three biological replicates were analyzed for each sample.

Upland cotton plants were treated with 200 mM NaCl or KCl to assess the effects of chloride. After 10 days, compared with the mock controls, the growth of the plants treated with NaCl or KCl was significantly inhibited. Additionally, the salt-treated plants had shriveled and wilted leaves, some of which had fallen off the plants (Figure 1A).

An analysis of the plant ion contents indicated that the chloride (Cl^–), sodium (Na⁺), and potassium (K⁺) contents were higher in the salt-treated samples than in the mock controls (Figures 1B,C and Supplementary Figure 1). After the NaCl and KCl treatments, cotton plants had extremely high Cl^– contents in the leaves, roots, and stems (Figures 1B,C). As expected, the Na⁺ or K⁺ contents also increased in the leaves, stems, and roots of plants treated with NaCl or KCl (Supplementary Figure 1). These results suggest that Cl^– can accumulate in cotton tissues, thereby affecting plant growth and development.

The vacuolar chloride channel (CLCg) can protect plant cells from ion-induced damages (Nguyen et al., 2015). Using a homology-based cloning strategy, we cloned the full-length coding sequences of two CLCg-encoding genes in G. hirsutum, which are homoeologs from D-subgenome and A-subgenome of the allotetraploid upland cotton and named GhCLCg-1A/D (suffixes D and A represented the D and A subgenomes, respectively). The GhCLCg-1A and GhCLCg-1D sequences were 2,325 bp long and encoded 774 amino acids. To further characterize GhCLCg-1A/D, we used the MEGA-X program to construct a phylogenetic tree that included seven AtCLCs. In the constructed tree, the CLC family was divided into two clades, with GhCLCg belonging to class I and highly homologous to AtCLCg (Figure 2A).

The results of the GhCLCg-1A and GhCLCg-1D gene structural analysis revealed these two genes have a similar intron-exon arrangement (Figure 2B). Specifically, they both comprise seven exons and six introns as well as domains unique to the CLC family (i.e., voltage_CLC and CBS). To further investigate the similarity in the GhCLCg-1A and GhCLCg-1D sequences, their amino acid sequences were aligned to the AtCLCg amino acid sequence (Supplementary Figure 2). The multiple sequence alignment confirmed that GhCLCg-1A and GhCLCg-1D are highly conserved (98.19% identity) and are similar to AtCLCg at the amino acid level (91.13% identity). The multiple sequence alignment also indicated that GhCLCg-1A and GhCLCg-1D have three regions that are conserved in AtCLCg, namely GxGIPE, GKxGPxxH, and PxxGxLF. In the GxGIPE sequence, x was a serine (S), implying GhCLCg-1A/D are specific for Cl^– (Supplementary Figure 2) (Barbier-Brygoo et al., 2000; Dutzler et al., 2002; Zifarelli and Pusch, 2010). These results imply that GhCLCg-1A and GhCLCg-1D are functionally similar to AtCLCg.

The localization of proteins provides clues regarding their functions (Long et al., 2020). The subcellular localization of GhCLCg-1A/D was determined using fusion proteins with a C-terminal GFP. The fusion proteins were transiently expressed in A. thaliana leaf protoplasts that also contained the vacuolar marker protein δ-TIP with a C-terminal RFP (Figure 3). In the control protoplasts with GFP alone and δ-TIP-RFP, the GFP signal was detected throughout, whereas the RFP fluorescence was restricted to the tonoplast. The GhCLCg-1A/D-GFP fusion proteins co-localized with δ-TIP-RFP in the protoplast, with red and green fluorescent signals overlapping, which confirmed that GhCLCg-1A/D are localized to the vacuolar membrane.

To determine whether GhCLCg-1A/D are involved in salt stress responses, we analyzed the expression of the corresponding genes following an exposure to chloride salts. The substantial similarity in the GhCLCg-1A and GhCLCg-1D sequences makes it very difficult to study the expression of these two genes by qRT-PCR. In this study, the universal qRT-PCR primers GhCLCg-1-Q-F/R were designed to analyze GhCLCg-1A/D expression in the roots, stems, and leaves at various time-points (Figure 4). Following the 200 mM NaCl treatment, GhCLCg-1 expression was up-regulated at 3 h in the roots (approximately two times higher than expression level of the mock control). In the stems, GhCLCg-1 expression was about 2–3 times higher in the NaCl-treated plants than in the mock controls at 1 and 6 h. Unexpectedly, the expression level of GhCLCg-1 at 3 h in stems of the 200 mM NaCl treated plants was significantly lower than in stems of the mock plants. The GhCLCg-1 transcript levels in stems treated with different salt concentration at 3 h were analyzed. The results showed that the expression of GhCLCg-1 at different salt concentration is lower than the control (0 mM NaCl), but the expression is increased by high salt concentration, and the expression level under 200 mM NaCl is higher than under 50, 100, and 150 mM NaCl (Supplementary Figure 3). In the leaves, GhCLCg-1 expression was up-regulated by the salt treatment at 1, 3, and 6 h. More specifically, the GhCLCg-1 transcript level was highest at 3 h, but the increase (relative to expression level of the mock control) was greatest at 6 h (five times higher) (Figure 4A). To further demonstrate that GhCLCg-1 is involved in plant responses to chloride-mediated salt stress, cotton plants were treated with 200 mM KCl (Figure 4B). The changes in the leaf GhCLCg-1 expression levels under 200 mM KCl were similar to those resulting from the 200 mM NaCl treatment. Given the up-regulated GhCLCg-1 expression induced by 200 mM NaCl, the leaf GhCLCg-1 expression levels following 6 h treatments with various NaCl concentrations (0, 50, 100, 150, and 200 mM) were analyzed. The relative GhCLCg-1 expression level increased as the salt concentration increased, peaking at 200 mM NaCl (Figure 4C). Thus, GhCLCg-1 expression in leaves was induced and regulated by chloride-mediated salt stress. Moreover, this expression was positively related to the chloride concentration.

To further clarify the GhCLCg-1 function related to chloride tolerance, GhCLCg-1-overexpressing (OE) transgenic A. thaliana plants were generated and examined (Figure 5). The WT, OE1, OE2, and OE3 A. thaliana plants had similar growth phenotypes on MS medium (Mock), with no significant differences in the root length and shoot growth. However, when treated with 100 mM NaCl, the transgenic A. thaliana plants (OE1, OE2, and OE3) had larger leaves and longer roots than the WT plants (Figure 5A). A quantitative analysis of the root length under salt stress conditions confirmed that the transgenic A. thaliana plants had significantly longer roots than the WT plants (Figure 5C). The GhCLCg-1 expression level increased significantly in the transgenic A. thaliana plants, whereas GhCLCg-1 expression was undetectable in the WT controls (Figure 5B). The changes in the Cl^– contents in A. thaliana plants (WT, OE1, OE2, and OE3) were analyzed. Under normal conditions, there were no significant differences in the Cl^– contents between the WT and transgenic plants. However, the Cl^– contents were significantly higher in the transgenic plants than in the WT controls under salt stress conditions (Figure 5D). These results indicate that GhCLCg-1 confers salt tolerance, while also altering the uptake of Cl^– in plants.

High environmental Na⁺ and Cl^– levels alter the ion homeostasis of plant cells, resulting in an ionic imbalance and toxicity. Thus, re-establishing cellular ion homeostasis is critical for metabolic functions and growth (Niu et al., 1995). The changes in Na⁺ and K⁺ contents as well as the Na⁺/K⁺ ratio in WT and transgenic A. thaliana plants (OE1, OE2, and OE3) were analyzed (Figure 6). There were no significant differences in the Na⁺ and K⁺ contents or the Na⁺/K⁺ ratio among the WT, OE1, OE2, and OE3 plants on MS medium (Mock). An examination of the effects of the 100 mM NaCl treatment revealed that compared with the WT controls, the Na⁺ and K⁺ contents were significantly higher in the OE1, OE2, and OE3 plants, whereas the Na⁺/K⁺ ratios were significantly lower in the OE1, OE2, and OE3 plants. These findings suggest that overexpression of GhCLCg-1 can affect the accumulation of Na⁺ and K⁺, then leads to a decrease of the Na⁺/K⁺ ratios under the NaCl treatment, which is beneficial for plant tolerance to salt.

To more thoroughly investigate the roles of GhCLCg-1 related to the salt tolerance of upland cotton plants, we used the TRV-based VIGS system to generate GhCLCg-1 knockdown cotton plants (Figure 7). At 10 days after the A. tumefaciens infiltration, the cotton seedlings transformed with TRV:CLA exhibited an albino phenotype (Figure 7A). The GhCLCg-1 expression levels were determined by qRT-PCR to assess how efficiently the gene was silenced in the cotton plants. The GhCLCg-1 expression levels in the roots, stems, and leaves were obviously lower in the TRV:GhCLCg-1 plants than in the TRV:00 plants (Figure 7B). Accordingly, GhCLCg-1 was efficiently silenced in the TRV:GhCLCg-1 cotton plants.

There were no observable differences in the growth of TRV:00 and TRV:GhCLCg-1 plants under normal conditions (Mock). However, after a 3-day 200 mM NaCl treatment, several symptoms consistent with salt stress were detected on the TRV:GhCLCg-1 plants (Figure 7C). Specifically, the leaves were shrunken and wilted and some of the leaves even fell from the plants. A decrease in the GhCLCg-1 expression level increased the sensitivity of plants to salt stress, suggesting GhCLCg-1 influences cotton salt tolerance. Furthermore, we analyzed the Cl^– contents in the roots, stems, and leaves of the TRV:00 and TRV:GhCLCg-1 cotton plants (Figure 7D). As expected, in the absence of chloride stress, there were no significant differences in the Cl^– contents of TRV:GhCLCg-1 and TRV:00 plants. However, following the 200 mM NaCl treatment, the root, stem, and leaf Cl^– contents were significantly higher in the TRV:GhCLCg-1 plants than in the TRV:00 plants. A comparison of these three tissues in TRV:GhCLCg-1 plants revealed that the Cl^– contents were highest in the leaves, followed by the stems and then the roots. Compared with the TRV:00 plants, the GhCLCg-1-silenced plants accumulated more Cl^–, especially in the leaves, making them more sensitive to salt stress.

The Na⁺ and K⁺ contents as well as the Na⁺/K⁺ ratio in TRV:00 and TRV:GhCLCg-1 plants were measured. There were no significant differences in the Na⁺ and K⁺ contents or the Na⁺/K⁺ ratio between the TRV:00 and TRV:GhCLCg-1 plants in the absence of salt stress (Figure 8). In contrast, after the 200 mM NaCl treatment, the shoot (stems and leaves) Na⁺ content was higher in the TRV:GhCLCg-1 plants than in the TRV:00 plants, and the Na⁺ content was highest in the TRV:GhCLCg-1 leaves (Figure 8A). The NaCl treatment did not significantly affect the K⁺ contents of the TRV:00 and TRV:GhCLCg-1 shoots, but it obviously decreased the K⁺ contents of the TRV:00 and TRV:GhCLCg-1 roots (Figure 8B). Interestingly, the Na⁺/K⁺ ratios sharply and significantly increased in the stems and leaves of TRV:GhCLCg-1 plants in response to the salt treatment, and the Na⁺/K⁺ ratio was highest in the TRV:GhCLCg-1 leaves (Figure 8C). These results indicate that silencing GhCLCg-1 causes cotton plants to absorb Na⁺ and accumulate in shoots, then leads to an increased Na⁺/K⁺ ratio in stems and leaves under NaCl treatment, which increases sensitivity to salinity stress.