“Qingyan No. 1,” a pumpkin rootstock with low temperature tolerance, was used as the experimental material in this study. The pumpkin seeds were soaked with 1‰ KMnO₄ for 15min to conduct surface disinfection. Afterward, the seeds were soaked in warm water at 55°C, cooled naturally, soaked for 12h, and placed in a growth chamber at 30°C for germination. Then, the seeds were sown in 10cm×10cm pots with peat–vermiculite–perlite medium (2:1:1). The growth conditions were as follows: photoperiod, 12/12h; day/night temperature, 28/18°C; light intensity, 16,000 Lx; and air humidity, 70–85%. Pumpkin seedlings at three-leaf stage were exposed to 4°C in a growth chamber (Ningbo Saifu DGX-260, China) for cold stress. The root, stem, and third true leaf of each plant were sampled at 0, 3, 6, 12, and 24h after low-temperature treatment. The samples were frozen at −80°C in liquid nitrogen before qRT-PCR analysis.

The full-length coding sequence (CDS) of CmRCC1 was amplified by PCR using 2× High-Fidelity Master Mix (Tsingke, Inc., Beijing, China), and the fragments were inserted into the Bgl II site of the pCAMBIA1305.4-N-GFP vector by using ClonExpress II One Step Cloning Kits (Vazyme, Piscataway, NJ, United States) to generate 35S::GFP-CmRCC1 fusion protein under the control of the Cauliflower mosaic virus (CaMV) 35S promoter. The construct and negative control (pCAMBIA1305.4-N-GFP) were transformed into Agrobacterium tumefaciens strain GV3101 and infiltrated into tobacco leaves according to previously described method (Sheludko et al., 2007). Leica SP8 confocal microscope was used to detect the GFP fluorescence signal with 4,6-diamidino-2-phenylindole (DAPI) as the nucleus marker.

Total RNA was isolated using TransZol reagent (TransGen Biotech Inc., Beijing, China) in accordance with the manufacturer’s protocol. The extracted total RNA was dissolved in diethylpyrocarbonate-treated water. The cDNA template for gene cloning was synthesized from 2μg of RNA using HiScript II One Step RT-PCR Kit (Vazyme, Piscataway, NJ, United States). While for qRT-PCR, the cDNA was synthesized from 1μg total RNA using HiScript II Q RT SuperMix for qPCR (+g DNA wiper; Vazyme, Piscataway, NJ, United States).

The CDS of CmRCC1 was cloned into the pHellgate8 vector to generate the 35S::CmRCC1 construct by ClonExpress II One Step Cloning Kits. The construct was transformed into A. tumefaciens strain GV3101 and then transferred into tobacco plants using the leaf disc method (Horsch et al., 1985). Transgenic tobacco seeds were screened on MS medium suspended with kanamycin (50mg/L). T₂ homozygous lines were used for further experiments.

The roots of three uniform plants from each replicate were harvested and washed with deionized water. The root morphology was scanned using Imagery Scan Screen (Epson Expression 11000XL, Regent Instruments, Canada). Root image analysis was conducted via the WinRHIZO 2003a software (Regent Instruments, Canada).

The open reading frames (ORFs) of CmRCC1 and CmLAZY1 from “Qianyan No. 1” roots were amplified using sequence-specific primers (Supplementary Table S1) and incorporated into pGBKT7 and pGADT7 vectors (Clontech, United States), respectively, to verify the protein–protein interactions of CmRCC1 with CmLAZY1. According to the manufacturer, the recombinant plasmids, pGADT7-CmLAZY1 and pGBKT7-CmRCC1, pGADT7 and pGBKT7-CmRCC1, pGADT7-T and pGBKT7-lam (negative control), and pGADT7-T and pGBKT7-p53 (positive control), were introduced into the yeast strain, Y2H Gold. The transformants were grown on SD/−Leu/−Trp and SD/−Leu/−Trp/−Ade/-His media to evaluate the interactions.

As described previously, the ORF of CmLAZY1 was cloned into pCAMBIA-nLUC to yield the fusion construct, pCAMBIA-CmLAZY1-nLUC, and the ORF of CmRCC1 was cloned into pCAMBIA-cLUC to generate the fusion construct, pCAMBIA-CmRCC1-cLUC (Chen et al., 2008). Agrobacterium tumefaciens GV3101 was transformed with the empty vector and fusion constructs and incubated at 28°C for 16h. Then, the A. tumefaciens cells were collected and resuspended at OD₆₀₀=0.3. The tobacco leaves were then infiltrated with Agrobacterium strains containing the indicated constructs at a ratio of 1:1. After 3days, the leaves were treated with luciferin, and firefly luciferase (LUC) signal was observed according to Xiong et al. (2019).

Chlorophyll fluorescence was measured by pulse amplitude-modulated fluorometry (MAXI; Heinz Walz, Effeltrich, Germany) as previously described (Cheng et al., 2016). The seedlings were adapted to the dark for at least 30min before the measurements, and the whole area of the third leaf from the bottom was used for the experiment. The intensities of actinic light and saturating light were set to 280 and 4,000μmolm⁻²s⁻¹, respectively. The maximum quantum yield of PS II (Fv/Fm) and the effective quantum yield of PS II (Φ_PSII) were measured and calculated in accordance with the following equations (van Kooten and Snel, 1990): Fv/Fm=(Fm−Fo)/Fm and Φ_PSII=(F’m−Fs)/F’m. The quantum yield of regulated energy dissipation (Φ_NPQ) and the quantum yield of nonregulated energy dissipation (Φ_NO) in PS II were calculated according to the equation (Kramer et al., 2004): Φ_PSII+Φ_NPQ+Φ_NO=1. Electron transfer rate (ETR) was measured using a rapid light-response curve.

Lipid peroxidation was determined by measuring malondialdehyde (MDA) content as described by Hodges et al. (1999). Briefly, leaf samples (0.3g) were ground in 3ml of ice-cold 25mmol/L HEPES buffer (pH 7.8) containing 0.2mmol/L EDTA and 2% (w/v) polyvinylpyrrolidone. The obtained homogenates were centrifuged at 4°C for 20min at 10,000rpm, and the resulting supernatants were used to analyze MDA content. The samples were mixed with 10% trichloroacetic acid containing 0.65% 2-thiobarbituric acid (TBA) and heated at 95°C for 25min. MDA content was corrected for non-MDA compounds by subtracting the absorbance at 532nm of a TBA-less solution that contained the plant extract.

We amplified the PCR products for qRT-PCR analysis in triplicate using 2×TransStart™ TOP Green qPCR SuperMix (TransGen Biotech Inc., Beijing, China) in 10μl qRT-PCR assays. PCR was performed using the QuantStudio 7 Flex Real-time PCR System (Applied Biosystems, Foster City, CA, United States). The cycling conditions consisted of denaturation at 95°C for 30s, followed by 40cycles of denaturation at 95°C for 5s, annealing at 58°C for 15s, and extension at 72°C for 10s. The reference genes, CmCAC and NbACTIN, were used as the internal controls (Obrero et al., 2011; Nie et al., 2020). The gene-specific primers for CmRCC1 and the NbPIN gene family are listed in Supplementary Table S1. Relative gene expression was determined as previously described by Livak and Schmittgen (2001).

The experiment involved a completely randomized block design with four replicates. Statistical analysis was performed using the SAS statistical package. The differences between the treatment means were separated using Tukey’s test at a significance level of p<0.05.

CmRCC1 gene (CmaCh15G006130) was predicted to contain a 3,360bp CDS isolated from 4,143bp cDNA and encode the protein of 1,119 amino acids in the Cucurbit Genomics Database. A Pfam domain search was performed to characterize the pleckstrin homology (PH_12), RCC1 repeats, FYVE zinc finger, BRX N-terminal, and BRX domains of the CmRCC1 protein (Supplementary Figure S1A).¹ Moreover, a database (The Arabidopsis Information Resource) search indicated 24 RCC1 family proteins in A. thaliana, among which 15 protein members have been named and functionally annotated. The phylogenetic tree built from the alignment of CmRCC1 with the previously identified Arabidopsis RCC1s revealed the evolutionary distances between the sequences (Figure 1A). Among these sequences, CmRCC1 showed high similarity to the sequences of AtRLD1 and AtRLD4.

The GFP-CmRCC1 fusion construct and GFP control in the pCAMBIA1305.4-N-GFP vector driven by CaMV35S promoter were transiently expressed in tobacco epidermal cells and visualized under a laser scanning confocal microscope to determine the subcellular localization of CmRCC1. The GFP fluorescence signal of GFP-CmRCC1 fusion protein was detected in the nucleus as confirmed by DAPI staining (Figure 1B).

We detected the changes in CmRCC1 expression in the root, stem, and leaf at different time points after 24h cold treatment to evaluate the response characteristics of CmRCC1 to cold stress in pumpkin. The transcription levels of CmRCC1 in the stem and leaf increased slowly with the extension of cold stress treatment, and they reached 2.13 and 3.15 times of the control (0h) after 24h treatment, respectively. However, the expression level of CmRCC1 in the pumpkin root peaked at 3h, and then reached 4.57 times at 24h of cold treatment (Figure 2). These results indicate that CmRCC1 may be involved in the response of pumpkin root to early cold stress.

CmRCC1 was overexpressed in tobacco under the control of CaMV35S promoter to analyze the role of CmRCC1 in root development. The insertion of the CmRCC1 cassette in 28 independent kanamycin-resistant transformants was confirmed by RT-PCR (Supplementary Figure S2). Three transformed lines (OxCmRCC1-1/−3/−6) which showed that the CmRCC1 gene segregated in the Mendelian segregation ratio of 3:1, were subsequently selected to obtain T₂ homozygous lines (Supplementary Table S2). qRT-PCR analysis of the CmRCC1 transcripts in three independent lines revealed variable levels of transgene expression (Figure 3A). Compared with the wild type, all the overexpressed transgenic lines showed increased gravitropic set-point angle (GSA) in lateral roots (Figure 3B). Moreover, CmRCC1 overexpression increased the root diameter and volume of transgenic tobacco but not root length (Figures 3C–E).

In Arabidopsis, the characterized PIN proteins demonstrate specific expression patterns and are involved in polar auxin transport and root patterning (Paponov et al., 2005). Thus, we further measured the expression levels of four PIN genes in the roots of wild-type and CmRCC1 transgenic plants. As shown in Figure 4, PIN3 expression level remarkably increased in the CmRCC1 overexpression lines than in the wild type. However, the expression of PIN2 and PIN6 showed no substantial differences between the transgenic lines and wild type. By contrast, the expression level of PIN1 differentially decreased in the CmRCC1 overexpression lines compared with the wild type.

LAZY1 functions upstream of lateral auxin translocation in gravity signal transduction in the root and shoot of Arabidopsis and rice (Yoshihara and Iino, 2007; Taniguchi et al., 2017). We co-transformed pGADT7-CmLAZY1 and pGBKT7-CmRCC1 in yeast cells and found that the transformants grew on SD/−Leu/−Trp/−Ade/-His media, which was consistent with the results of the positive control yeast cells (Figure 5A). Furthermore, we performed luciferase complementation imaging assay to verify the interaction of CmRCC1 with CmLAZY1 in vivo. We were able to image LUC signals in tobacco leaves that co-infiltrated with Agrobacterium strains that expressed CmLAZY1-nLUC and CmRCC1-cLUC, but no signal was observed in the negative controls (CmRCC1-cLUC/nLUC and nLUC/cLUC, Figure 5B). Together, the results suggest that CmRCC1 interacts with CmLAZY1 protein.

The seedlings of T₂ transgenic lines and wild type were exposed to chilling stress at 4°C for 12h to examine the possible role of CmRCC1 overexpression in the cold tolerance of tobacco. We observed that the leaves in the wild type completely shrank, and the plants were lodging after chilling stress treatment, whereas the transgenic tobacco plants still stood upright with flat leaves and light wilting (Figure 6A). We then measured the chlorophyll fluorescence of PS II in the third leaves of chilling-stressed and non-stressed plants in the wild-type and transgenic lines. The Fv/Fm and Φ_PSII decreased by 28.6 and 56.7%, respectively, in the wild type after chilling stress in comparison with the control. However, Fv/Fm and Φ_PSII decreased by 11.1–14.7 and 6.7–15.3%, respectively, in the CmRCC1-overexpressed lines in response to chilling stress (Figures 6B,C). A high Φ_NO value indicates that photochemical energy conversion and protective regulatory mechanisms are inefficient. Therefore, it indicates that the plant is already damaged or will be photodamaged upon further irradiation. Here, we found Φ_NO increased by 36.6% after chilling stress in wild-type plants, whereas CmRCC1 overexpression compromised the increase in Φ_NO in chilling-stressed plants (Figure 6D). By contrast, Φ_NPQ showed no substantial differences between chilling-stressed and non-stressed plants in wild-type and CmRCC1 transgenic lines, which indicates that the photoprotection ability was not affected under chilling stress (Figure 6E). We also analyzed the ETR versus incident photosynthetic photon flux density. Light-saturated ETR decreased by 55.0% in chilling-stressed wild-type plants. Again, the decrease in ETR was compromised in CmRCC1-overexpressed lines (Figure 6F). Moreover, increased MDA content (62.5%) was observed after 12h of chilling stress in wild-type plants compared with the control. However, no remarkable differences in MDA content were observed between the control and chilling-stressed transgenic lines (Figure 6G). Thus, we conclude that CmRCC1 overexpression increases the cold tolerance of transgenic tobacco.