Identification of suitable reference genes for quantitative reverse transcription PCR in Luffa (Luffa cylindrica)

Abstract

Reverse transcription real-time quantitative PCR is widely used to quantify gene expression. Reference genes are usually used as internal controls to measure the target gene expression level. To date, there is no consensus on the use of systematically validated reference genes in different tissues of Luffa. This study evaluated the expression stability of 11 candidate reference genes in different tissues using five algorithms (BestKeeper, comparative delta-Ct method, GeNorm, NormFinder, and RefFinder). Protein phosphatase 2A was the most stable gene, while alpha Tubulin was the least stable. The relative expression of ethylene-related genes in different tissues was also analyzed to reveal their role in sex determination. This study provides the basis for using suitable reference genes to evaluate targeted gene expression.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12298-022-01182-8.

Introduction

Reverse transcription real-time quantitative PCR (RT-qPCR) is widely used in gene expression quantification because it is fast, sensitive, and accurate (Kong et al. 2016; Umadevi et al. 2019). However, its accuracy is dependent on several factors, including RNA content and quality, reverse transcription efficiency, and amplification efficiency (Su et al. 2020). Reference genes are usually used as internal controls to eliminate or reduce the technical variation between samples and accurately measure the expression of the target gene (Kozera and Rapacz 2013). An ideal reference gene should be stable in different tissues, development stages, and under diverse experimental conditions (Huggett et al. 2005; Sinha et al. 2019). However, many studies have indicated that no single reference gene is universally suitable under various experimental conditions (Kim et al. 2003; Nathalie et al. 2005). Therefore, suitable reference genes should be selected based on specific testing conditions to ensure accuracy.

Several algorithms have been developed to validate the performance of candidate reference genes. For instance, GeNorm software evaluates the performance of the candidate reference genes by calculating their stability value (M) (Vandesompele et al. 2002). A low M value denotes a slight variation and vice-versa. It also computes the variation of reference genes used for normalization (V: n/(n + 1)). The BestKeeper program ranks the candidate reference genes based on the standard deviation (SD) and the coefficient of variation (CV) of mean Cq values (Pfaffl et al. 2004). It is used to select the best candidate reference genes via a pairwise correlation analysis (Pfaffl et al. 2004). The NormFinder software is used to assess the best reference gene based on intra- and inter-group variation (Andersen et al. 2004). The comparative delta-Ct method compares relative expression of reference gene pairs within each sample (Silver et al. 2006). The RefFinder is a web-based tool employed to integrate the results of other algorithms and generate a comprehensive ranking (Xie et al. 2012).

Luffa (Luffa cylindrica) is an economically important vegetable widely cultivated in China, India, and Southeast Asia. Immature Luffa fruits are a good source of carbohydrates, vitamins, and various minerals (Wu et al. 2016). Recent studies have indicated that Luffa possesses antimicrobial, anticancer, and antioxidant properties and are widely used medicine (Al-Snafi 2019). The loofah can be used to clean, filter (Oboh and Aluyor 2009) or combined with graphene nanosheets to make new composite materials for electromagnetic shielding (Liu et al. 2018). Cucurbits are commonly used as model plants for molecular biology and biotechnology studies. Notably, Luffa leaves can be transiently transformed with Agrobacterium tumefaciens and can be used as a plant expression system for physiological and biochemical studies in cucurbits (Błażejewska et al. 2017). For instance, Chincinska et al. (2019) used agroinfiltration to transiently express a recombinant human deoxyribonuclease (rhDN) gene in L. cylindrica. This suggests that Luffa leaves could be a good source of functional rhDNase I.

Several Cucurbitaceae plants, including cucumber (Cucumis sativus L.) (Boualem et al. 2015), melon (C. melo L.) (Boualem et al. 2008), watermelon (Citrullus lanatus) (Zhang et al. 2020b), and Luffa (L. cylindrica), exhibit polymorphisms in their sexual system. Notably, ethylene-related genes play a crucial role in sex determination of Cucurbitaceae plants (Li et al. 2019a; Zhang et al. 2014). For instance, the expressions of 1-aminocyclopropane-1-carboxylate oxidase (ACO) and 1-aminocyclopropane-1-carboxylate synthase (ACS) genes are correlated with the formation of female flowers in melon and cucumber (Boualem et al. 2015; Chen et al. 2016; Saito et al. 2007). Moreover, CmACS-7 and CmWIP1 interact to control the development of male, female, and hermaphrodite flowers in melon (Martin et al. 2009). Ethylene receptor 1A (ETR1A) and ETR2B (García et al. 2019) and ethylene-responsive factors (ERF) (Tao et al. 2018) determine the sex of Cucurbita pepo, cucumber, and melon.

To date, there is no consensus on the use of systematically validated reference genes in different Luffa tissues. In the present study, RT-qPCR was used to validate the expression stability of 11 candidate reference genes in nine tissues via the BestKeeper, comparative delta-Ct method, GeNorm, NormFinder, and RefFinder algorithims. Protein phosphatase 2A (PP2A) was the most appropriate reference gene for normalizing target gene expression based on the RT-qPCR results. This study provides a basis for using suitable reference genes to evaluate target gene expression and for preliminary analysis of the function of ethylene-related genes in the sex determination of Luffa.

Materials and methods

Plant materials and tissue collection

P93075, an advanced Luffa inbred line bred through multiple generations of self-pollination and selection, was grown at 26 ± 2 °C with a 16 h/8 h light/dark cycle in a climate-controlled room. The root (R), stem (S), leaf (L), male petals (Pm), female petals (Pf), stamen (St), stigma (Sg), shoot apices (SA), ovary (O), and fruit (F) were harvested separately 60 days post-planting and immediately frozen in liquid nitrogen. There were three biological replicates, each with 5–10 plants.

cDNA preparation

Total RNA was extracted from all samples using the TransZol Plant kit (TransGen, China). The quality and quantity of RNA were assessed using a NanoDrop 2000c (Thermo Scientific, USA). RNA integrity was confirmed using a 1.5% agarose gel electrophoresis. A TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen, China) was used to remove gDNA and synthesize cDNA, following the manufacturer’s instructions.

Selection of reference genes and primer design

Ten genes were selected as candidate reference genes (ACT7, β-ACTIN, DNAJ, EF1α, EIF4A, GAPDH, PP2A, RPL2, TUA, UBQ) based on a literature review. One previously used reference gene (Wang et al. 2020), Lc18SRNA, was also included. Details regarding these genes are provided in Table 1. ACT7, β-ACTIN, EF1α, GAPDH, PP2A, RPL2, UBQ, and TUA have been used as reference genes for normalization in melons (Kong et al. 2016). Additionally, EIF4A and DNAJ have been used as reference genes for expression studies in poplar (Fang et al. 2019). Arabidopsis homologs for each gene were used as query sequences for blast analysis against the Luffa transcripts (Wu et al. 2020) using the BLASTN tool (E-value = 1e−05). The best-matched sequences were the homologous genes. Primers were designed using Primer Premier 5 for a product size of 80–150 bp. The specificity of primers was confirmed via blast analysis against Luffa transcripts using the BLASTN tool. The PCR amplification specificity was verified through electrophoresis on a 1.5% agarose gel. The PCR amplification efficiency and correlation coefficient (R²) of primer pairs were calculated using a set of 5 × serial dilutions of cDNA.

Table 1.

Information of the candidate reference genes

Gene name	Gene ID	Gene description	Arabidopsis homolog	Primer sequence (5′–3′)	Product size (bp)	PCR efficiency (R2)	Correlation coefficients (E)
ACT7	Lcy04g013250	Actin 7	AT5G09810	F:GTACAACTGGTATCGTGCTG R:AGGTCCAAACGGAGAATT	141	0.99	1.01
β-ACTIN	Lcy12g000340	Beta actin	AT3G18780	F:AGATTTGTATGGTAACATTGTTCTCA R:ACACTGTATTTCCTTTCGGGTG	144	0.99	1.01
DNAJ	Lcy03g007880	DnaJ protein	AT5G22060	F:CATCATCCCAACTGACCGAC R:CGTGCATTTCTTCATCTTCCTC	138	0.99	1.07
EF1α	Lcy01g004250	Elongation factor 1-alpha	AT5G60390	F:GCTCTCCTTGCTTTTACCCTTG R:GACGATTTCATCATACCTTGCCT	105	0.99	0.97
EIF4A	Lcy02g003720	Eukaryotic initiation factor 4A	AT3G19760	F:TGTATTGGAGGCAAAAGTGTGG R:TTAATGGCTCTGGTACGCAGTG	128	0.99	1.05
GAPDH	Lcy09g001440	Glyceraldehyde-3-phosphate dehydrogenase 2	AT1G13440	F:GTCCATTCCATCACCGCAAC R:GCAGGCAGCACTTTACCCAC	137	0.99	1.00
PP2A	Lcy13g002280	Protein phosphatase 2A	AT1G69960	F:CTAGTGGCTCTGAAAGTTCGTTATAG R:CATACACTTGAGTTATCTGTCGGC	85	0.99	1.05
RPL2	Lcy02g007670	Ribosomal protein L2	AT1G04480	F:TACAAAACTTCTACCCCGAGCA R:GGCATTACGACCTTTACCACAA	114	0.99	1.04
TUA	Lcy11g004220	Alpha Tubulin	AT1G04820	F:GAATCAGAAAGCTGGCTGATAAC R:TGACAAACGCTCCAAAAGGA	113	0.99	1.02
UBQ	Lcy11g000140	Ubiquitin	AT2G36170	F:AGGCTTCGTGGTGGTATCATTG R:GCACTTGCGACATATCATCTTGT	84	0.99	1.01
Lc18SRNA		18SRNA		F:GTGTTCTTCGGAATGACTGG R:ATCGTTTACGGCATGGACTA	275	0.99	1.03

RT-qPCR analysis

RT-qPCR was performed using QuantStudio 6 Flex (ABI, USA) in a 10 μL reaction volume containing 0.5 μL cDNA, 5 μL SYBR Premix Ex Taq II (Takara, Japan), 0.2 μL forward and reverse primers (10 μM), and 4.1 μL ddH₂O. The PCR conditions were: 95 °C for 30 s, 40 cycles at 95 °C for 10 s, and 60 °C for 30 s, followed by a melting curve analysis at 65–95 °C. The relative expression profiles were determined using the 2^−ΔΔCq method (Livak and Schmittgen 2001). The gene expression data were log2-transformed before analysis. The gene expression levels for each sample were determined based in triplicates.

Stability analysis

The expression stabilities of the candidate reference genes were evaluated separately using the BestKeeper (Pfaffl et al. 2004), a comparative delta-Ct method (Silver et al. 2006), GeNorm (Vandesompele et al. 2002), and NormFinder (Andersen et al. 2004) programs. The recommended comprehensive ranking based on the four programs was then calculated using the RefFinder (Xie et al. 2012) program. RT-qPCR data were exported into an Excel datasheet to convert the Cq values to the software requirements. The geNorm algorithm was used to calculate the average expression stability value (M), defined as the average pairwise variation in a particular gene with all other potential reference genes (Vandesompele et al. 2002). The NormFinder program was used to calculate a stability value (SV) via an ANOVA-based model to consider intra- and intergroup variation in expression. A lower SV indicates increased stability (Andersen et al. 2004). BestKeeper is an Excel-based software tool that selects best-suited reference genes based on Pearson correlation coefficient (r), SD, and CV of average Cq values. The most stable gene has the lowest CV ± SD value (Pfaffl et al. 2004). The comparative delta-Ct method was used to compare the relative expression of reference gene pairs within each sample to identify stable reference genes (Silver et al. 2006). Finally, RefFinder was used to assign an appropriate weight to an individual gene and calculate the geometric mean of their weights for the final ranking based on the rankings from each program (Xie et al. 2012).

Selection of ethylene-related genes

For each gene, the amino acid sequences of cucumber homologs were used as query sequences to blast against the Luffa protein sequences (Wu et al. 2020) using the BLASTp tool (E-value = 1e−05). The homologs were: ACS (Csa4G049610), ACO (Csa6G511860), WIP (Csa4G290830) and ERF110 (Csa6G017030). The retrieved sequences were manually confirmed by searching against the Pfam database (http://pfam.xfam.org/) and CDD-Search (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi).

Statistical analysis

Data were analyzed using IBM SPSS Statistics 25 (IBM Corp., NY, USA). Differences between means were determined using a one-way Analysis of Variance (ANOVA), followed by Dunnett’s test for multiple comparisons. The significance threshold was set at P ≤ 0.05.

Results

Selection of candidate reference genes and amplification specificity

Eleven candidate reference genes (Lc18SRNA, ACT7, β-ACTIN, DNAJ, EF1α, EIF4A, GAPDH, PP2A, RPL2, TUA, UBQ) were selected in this study. Lc18SRNA had been used as a reference gene in a previous study (Zhu et al. 2016). The designed primer pairs of each reference gene, except Lc18SRNA, are outlined in Table 1. All primers produced single bands and single melting peaks, demonstrating that they all generated specific products of expected sizes (Additional file: Fig. S1 and S2).

A standard curve was generated for each primer using a set of fivefold diluted cDNA templates to evaluate its PCR efficiency (E) and correlation coefficient (R²). The E of all the reference genes ranged between 0.97 and 1.07, while the R² of all genes exceeded 0.99 (Table 1). These results suggested that all primers were efficient and specific in the RT-qPCR system.

Expression profiles of the candidate reference genes

The expression levels of the 11 candidate reference genes in ten tissues were examined using RT-qPCR. The Cq values of candidate reference genes ranged between 14.3 (Lc18SRNA) and 33.2 (TUA) in all samples (Fig. 1). Lc18SRNA had the lowest Cq value and thus had the highest expression level, whereas TUA had the highest Cq value and thus exhibited the lowest expression level. Lc18SRNA had the least expression variation, while TUA had the highest expression variation (Fig. 1).

Fig. 1.

Cq value distribution of the candidate reference genes. The lines across the box represent the medians, the boxes represent the 25/75 percentiles, the whiskers represent the 95% confidence intervals, and the dots represent outliers

Expression stability of candidate reference genes

The expression stabilities of the 11 candidate reference genes in ten tissues were evaluated using BestKeeper (Pfaffl et al. 2004), comparative delta-Ct method (Silver et al. 2006), GeNorm (Vandesompele et al. 2002), NormFinder (Andersen et al. 2004), and RefFinder (Xie et al. 2012) programs.

BestKeeper was used to rank the candidate reference genes based on the SD and the CV of mean Cq values. Lc18SRNA had the lowest CV ± SD (3.25 ± 0.50), indicating that it was the most stable gene. It was followed by DNAJ (3.76 ± 0.82), whereas TUA was the most unstable gene with the highest CV ± SD (17.78 ± 4.34) (Fig. 2A, Table 2).

Fig. 2.

Stability values of the candidate reference genes based on BestKeeper (a), comparative delta-Ct method (b), GeNorm (c, d), NormFinder (e), and RefFinder (f)

Table 2.

Ranking of the stability of the candidate reference genes based on different evaluation programs

Genes	RefFinder		Comparative delta CT		BestKeeper		Normfinder		Genorm
	Stability	Rank	Stability	Rank	Stability	Rank	Stability	Rank	Stability	Rank
PP2A	2.00	1	1.45	2	0.92	4	0.51	2	0.51	1
β-ACTIN	2.14	2	1.43	1	1.16	7	0.60	3	0.51	1
DNAJ	3.98	3	1.56	6	0.82	2	1.00	7	0.60	3
Lc18SRNA	4.23	4	1.68	8	0.50	1	1.06	8	0.84	5
EF1α	4.36	5	1.51	5	1.38	9	0.47	1	1.10	8
EIF4A	4.56	6	1.50	3	1.14	6	0.62	4	0.95	6
UBQ	5.14	7	1.51	4	0.97	5	0.67	5	1.04	7
ACT7	6.70	8	1.65	7	1.20	8	1.21	9	0.73	4
RPL2	7.40	9	1.83	10	0.85	3	1.40	10	1.21	10
GAPDH	8.35	10	1.68	9	1.58	10	0.84	6	1.16	9
TUA	11.00	11	4.92	11	4.34	11	4.85	11	1.88	11

Genes with lower SD of mean Cq values had higher expression stability based on the comparative delta-Ct method. β-ACTIN was the most stable gene in all samples, with the lowest stability value (1.43). It was followed by PP2A (1.45), whereas TUA had the highest stability value (4.92) (Fig. 2B, Table 2).

GeNorm was used to evaluate the stability of the candidate reference genes by calculating their stability value (M) derived from the average pairwise variation of a potential reference gene set with all other genes under investigation. The most stable reference genes had the lowest M values. GeNorm was used to rank β-ACTIN and PP2A as the pair of the best reference genes with an M value of 0.51. TUA was the least stable reference gene, with an M value of 1.88 (Fig. 2C, Table 2). These results were consistent with the comparative delta-Ct method results. GeNorm was also used to determine the suitable number of reference genes required for precise normalization of target gene expression based on pairwise variation (V value) (Fig. 2D, Table 2). Pairwise variation analysis revealed that at least seven reference genes could reliably normalize gene expression data (Fig. 2D, Table 2).

The NormFinder program was also used to evaluate the stability of reference genes based on the expression variations of all genes. The most stable reference gene had the lowest stability value. EF1α was the most stable gene with a stability value of 0.47, followed by PP2A and β-ACTIN with stability values of 0.51 and 0.60, respectively (Fig. 2E, Table 2). TUA was the least stable reference gene with a stability value of 4.85 (Fig. 2E, Table 2).

The RefFinder program was used to construct a comprehensive ranking of the tested candidate reference genes based on the rankings of the other four programs. PP2A was the most stable reference gene, with the lowest stability value of 2.00, followed by β-ACTIN with a stability value of 2.14. TUA was the least stable reference gene with the highest stability value of 11.00 (Fig. 2F, Table 2).

Validation of the identified reference genes

The expression profiles of ACS (Lcy03g011910) were measured and normalized using the two most stable reference genes (PP2A and β-ACTIN) and the least stable reference gene (TUA) based on the RefFinder results. These measurements were made to validate the stability of the selected reference genes. The relative expression profiles of ACS in all samples were similar when normalized with PP2A and β-ACTIN, singly or in combination (Fig. 3). However, the expression patterns markedly differed when TUA was used to normalize the data (Fig. 3). These findings demonstrate that the selection of reference genes considerably impact the standardization results. Inappropriate reference genes may generate misleading results.

Fig. 3.

Analysis of the identified reference genes. R (Root), Stem (S), L (Leaf), Pm (Male petals), Pf (Female petals), St (Stamen), Sg (Stigma), and F (Fruit). Error bars represent standard deviations for three replicates

Expression profiles of putative sex-determining genes

Many studies have shown that ACS, ACO, WIP, and ERF play essential roles in the sex differentiation of Cucurbitaceae plants (Li et al. 2019a). Herein, five ACSs, five ACOs, seven WIPs, and three ERF110s were identified in the Luffa genome (Additional file: Table S1, Fig. S3). The expression levels of these genes in all tissues were analyzed using PP2A as the reference gene (Fig. 4). The expression of Lcy01g007220 was significantly higher in the stigma than in the stamen. In contrast, Lcy03g011910 and Lcy04g015760 expressions were higher in the stamen than in the stigma (Fig. 4A). The expression of Lcy12g013770 was significantly low in all tissues (Fig. 4B). Moreover, the expression of Lcy08g006810 and Lcy08g018730 was significantly higher in the stamen than in the stigma. In contrast, Lcy07g013240 expression was substantially higher in the stigma than in the stamen (Fig. 4B). Lcy05g004290 and Lcy09g014750 were highly expressed in flower and fruit development tissues except for stigma (Fig. 4C). Nonetheless, Lcy06g017840 was specifically expressed in stigma (Fig. 4C). Lcy09g010140 and Lcy07g002220 were highly expressed in flower-related tissues (Fig. 4D).

Fig. 4.

The relative expression levels of ethylene-related genes. a–e ACSs; f–j ACOs; k–q WIPs; r–t ERF110s. R (Root), Stem (S), L (Leaf), Pm (Male petals), Pf (Female petals), St (Stamen), Sg (Stigma), and F (Fruit). Error bars represent standard deviations for three replicates. n.d. not detected

Discussion

A high-quality genome sequence of Luffa has been recently determined (Wu et al. 2020) to promote basic research on Luffa. Błażejewska et al. (2017) reported that L. cylindrica plants could serve as a suitable system for transient expression. The expression platform based on the agroinfiltrated Luffa leaves is a model system for plant research and an effective bio-factory for large-scale production of recombinant proteins (Chincinska et al. 2019). However, there are no simple and effective methods for detecting gene expression in transgenic plants. RT-qPCR is one of the simplest strategies for studying various gene interactions in transgenic plants. However, its accuracy is dependent on stable internal reference genes (Kozera and Rapacz 2013). Only a few reports have assessed the stability of reference genes in Luffa to date. This study systematically evaluated 11 candidate reference genes in ten tissues of L. cylindrica to determine their potential use as reference genes.

Previous studies have shown that combining multiple analysis programs can improve internal reference gene evaluation (Die et al. 2010; Hou et al. 2015). Herein, five algorithms were selected to evaluate the stability of candidate reference genes. However, the ranking results of different algorithms were slightly different. β-ACTIN and PP2A were the most stable genes in all samples based on the comparative delta-Ct method and GeNorm. In contrast, Lc18SRNA and EF1α were the most stable reference genes based on the BestKeeper and NormFinder programs, respectively. PP2A was the most stable reference gene in all tissues based on RefFinder's comprehensive evaluation, followed by β-ACTIN. TUA was the least stable reference gene in all algorithms.

PP2A comprises a family of serine/threonine phosphatases. It plays an essential role in cell cycle regulation, cell morphology and development, and the regulation of numerous signaling pathways (Janssens and Goris 2001; Shi 2009). PP2A has been among the best reference genes for normalization in many plants, such as different varieties of cucumber–pumpkin grafted plants (Li et al. 2019b). Moreover, it has been used as a reference gene to validate gene expression in mulberry, corydalis yanhusuo, sorghum, and Nicotiana benthamiana under abiotic and biotic stresses (Bao et al. 2020; Liu et al. 2012; Palakolanu et al. 2016; Shukla et al. 2019). It has also been used to study the response of Glehnia littoralis to abscisic acid (ABA) and NaCl (Li et al. 2020). However, it is the least stable reference gene in poplar regeneration (Fang et al. 2019) and Eleusine coracana (Jatav et al. 2018). These results indicate that the stability of reference genes varies in different species, tissues, and developmental stages.

ACS1G, gene duplication of ACS1, controls female sex determination in cucumber (Trebitsh et al. 1997; Zhang et al. 2020a). The mutant CmACS7 causes andromonoecy in melon (Boualem et al. 2008). ACS7 orthologs are expressed in pistil-bearing flowers but not in cucurbits male flowers (Boualem et al. 2008; Li et al. 2019a). Combined expression of ACS and ACO genes in specific tissues and time can inhibit stamen development and promote pistil development. Therefore, it is essential in flower sex determination (Li et al. 2019a). Herein, the expression levels of Lcy01g007220 and Lcy07g013240 were significantly higher in the stigma than in the stamen. This differential expression could explain how they determine sex in Luffa. WIP1 mutants cause gynoecy in watermelon (Zhang et al. 2020b), melon (Martin et al. 2009), and cucumber (Hu et al. 2017) when they lose their functions. The expression level of WIP is significantly lower in the stigma than in the stamen (Rodriguez-Granados et al. 2017). In this study, the expression level of Lcy09g014750 was remarkably higher in the stamen than in stigma. Lcy09g014750 is also expressed in the ovary and fruit. This observation was inconsistent with its male flower-specific expression in melon (Martin et al. 2009) or cucumber (Hu et al. 2017) but consistent with its expression in bitter gourd (Urasaki et al. 2017). ERF110 binds to at least two sites in the promoter region of ACS11. Therefore, it plays a crucial role in sex determination in cucumbers and melons (Tao et al. 2018). Herein, the expression of Lcy09g010140 and Lcy07g002220 was higher in flower-related tissues than in other tissues. These results suggest that both genes may be involved in the development of flower-related organs.

Conclusions

This study assessed the suitability of 11 genes as reference genes in Luffa tissues. BestKeeper, comparative delta-Ct method, GeNorm, NormFinder, and RefFinder programs revealed that PP2A was the most stable reference gene for normalization of gene expression across different Luffa tissues. TUA was the least stable gene. The expression patterns of Lcy03g011910 further verified the importance of selecting suitable reference genes for normalization. The expression of ethylene-related genes in different tissues was also analyzed to reveal their role in sex determination. Therefore, this study provides insight into the molecular mechanisms involved in the expression profiles of target genes in Luffa tissues for further studies on the transcriptional regulation of sex determination.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

GZ designed experiments. GZ, MW, JL, HG, JL, XZ, and XL carried out experiments. GZ, HW, SZ and YG processed the experiments data. JL offered the study materials. GZ wrote the manuscript and HW edited the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This work was funded by the National Natural Science Foundation of China (31902011, 31872093), Laboratory of Lingnan Modern Agriculture Project (NZ2021008), the Science and Technology Program of Guangdong Province (2021A1515012500, 2019A050520002, 2018B020202007, 201904020012), Discipline team construction projects of the 14th Five-year Plan (202103TD, 202114TD). Special fund for scientific innovation strategy-construction of high-level Academy of Agriculture Science (R2020PY-JG004, 202005, R2018QD-035, R2018QD-038).

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

Code availability

Not applicable.

Declarations

Conflict of interest

The authors declare no conflict of interest.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Ethics approval

Not applicable.