Identification of Bottle Gourd (Lagenaria siceraria) OVATE Family Genes and Functional Characterization of LsOVATE1

The OVATE gene family is a class of conserved transcription factors that play significant roles in plant growth, development, and abiotic stress, and also affect fruit shape in vegetable crops. Bottle gourd (Lagenaria siceraria), commonly known as calabash or gourd, is an annual climber belonging to the Cucurbitaceae family. Studies on bottle gourd OVATE genes are limited. In this study, we performed genome-wide identification of the OVATE gene family in bottle gourd, and identified a total of 20 OVATE family genes. The identified genes were unevenly distributed across 11 bottle gourd chromosomes. We also analyzed the gene homology, amino acid sequence conservation, and three-dimensional protein structure (via prediction) of the 20 OVATE family genes. We used RNA-seq data to perform expression analysis, which found 20 OVATE family genes to be differentially expressed based on spatial and temporal characteristics, suggesting that they have varying functions in the growth and development of bottle gourd. In situ hybridization and subcellular localization analysis showed that the expression characteristics of the LsOVATE1 gene, located on chromosome 7 homologous to OVATE, is a candidate gene for affecting the fruit shape of bottle gourd. In addition, RT-qPCR data from bottle gourd roots, stems, leaves, and flowers showed different spatial expression of the LsOVATE1 gene. The ectopic expression of LsOVATE1 in tomato generated a phenotype with a distinct fruit shape and development. Transgenic-positive plants that overexpressed LsOVATE1 had cone-shaped fruit, calyx hypertrophy, petal degeneration, and petal retention after flowering. Our results indicate that LsOVATE1 could serve important roles in bottle gourd development and fruit shape determination, and provide a basis for future research into the function of LsOVATE1.


Introduction
Bottle gourd [Lagenaria siceraria (Mol.) Standl.] (2n = 2x = 22), a member of the Cucurbitaceae family, also known as calabash or long melon, is a cultivated vegetable, medicinal plant, and grafting rootstock [1]. It originated in Africa and was independently domesticated in Asia [2]. With a cultivation history of 7000 years in China, it is one of the characteristic summer vegetables in southern China. Bottle gourd exhibits high genetic variability, especially in fruit shape, which can be round, oblate, pyriform, Hulu (double gourd), dipper, slender straight, tubby, etc. [3,4]. The diverse morphology of the bottle gourd fruit makes it a good candidate for studying variations in fruit shape. Fruit shape is a major horticultural trait for many fruit and vegetable crops that can influence their yield and quality from a crop-breeding perspective [5,6]. It is usually evaluated in terms of fruit diameter (FD), fruit length (FL), and fruit shape index (FSI, the ratio of FL to FD) [7,8]. Changes in fruit shape are not only the result of natural selection but also of artificial domestication to adapt to diverse environments or consumer preferences [9].

Plant Materials
In this study, plant materials included the 'Hangzhou gourd' (hereafter 'HZ') and YD-4 ( Figure S1). HZ is a local cultivar with slender straight fruit, which originated from Southeast China. YD-4 is a landrace with near-round fruit. For OVATE gene family member expression profiling, roots, stems, leaves, male flowers, female flowers, immature fruits, and ovaries (8 days before anthesis [DBA], 6 DBA, 4 DBA, and 0 DBA) from HZ and YD-4 were used. Ten individuals of each accession were grown in 30 m rows, spaced 0.5 m apart, at the Haining experimental station (30 • N, 120 • E).

Identification and Phylogenetic Analysis of the OVATE Gene Family
For identification of the OVATE gene family, the sequences of tomato OVATE proteins were obtained from the Sol Genomics Network (https://www.sgn.cornell.edu/, accessed on 20 March 2020). These sequences were used as queries to search against bottle gourd proteins using the BLASTP program with a threshold e-value of 1 × e −50 . Data on OVATE family genes in bottle gourd were downloaded from the Cucurbit Genomics Database (http://cucurbitgenomics.org/, accessed on 20 March 2020) and Gourdbase (http://www. gourdbase.cn/, accessed on 20 March 2020). The genome data of Arabidopsis thaliana were downloaded from TAIR (https://www.arabidopsis.org/, accessed on 20 March 2020). Phylogenetic and molecular evolutionary analyses were conducted using MEGA version 7, based on OVATE protein alignment [33,34]. The phylogenetic tree was constructed in Toolbox for Biologists software (TBtools, Ver. 1.098693) using the maximum likelihood (ML) method with 10,000 bootstrap number.

Sequence Alignment, Chromosomal Location, and Three-Dimensional Protein Structure Prediction
OVATE gene family members were mapped on bottle gourd chromosomes according to positional information from Gourdbase. BioEdit software [35] was used to compare the homologous proteins between bottle gourd and tomatoes (using amino acid sequence information), and data of the conserved domain were analyzed. The online SWISS-MODEL software (https://swissmodel.expasy.org/, accessed on 25 March 2020), which can predict the three-dimensional structure of proteins upon input of the amino acid sequence of related proteins, was used to predict the three-dimensional structures of homologs in bottle gourd. These tools were used to compare the differences in genes in the OVATE family.

RNA Extraction and RT-qPCR Analysis
Total RNA was extracted using TRIzol Reagent (Plant RNA Purification Reagent for plant tissue) according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA). Reverse transcription was performed using SynScipt™ III cDNA Synthesis Mix (Tsingke, Beijing, China) with 1 µg total RNA. The cDNA samples were stored at −20 • C before being used. Gene expression was examined via RT-qPCR using the SYBR Green method on a StepOne TM real-time PCR System. PCR amplifications were performed on a StepOne TM real-time thermal cycler in a final volume of 20 µL containing 1.0 µL of cDNA, 0.8 µL of each primer (10 µM), 7.0 µL of sterile water, and 10 µL (2×) of TSINGKE Master qPCR Mix (SYBR Green I) (Tsingke). The amplification conditions were as follows: 1 min of denaturation at 95 • C, followed by 40 cycles at 95 • C for 10 s, 60 • C for 30 s, and 72 • C for 30 s, after which a melt curve was produced at 60 • C. Relative gene expressions were estimated according to the 2 −∆∆Ct method [36,37]. The bottle gourd TuB-α (tubulin alpha chain-like) gene (HG_GLEAN_10019204) was used as a reference gene. Primer sequences of the reference and target genes are listed in Table S1. Three replicates were performed for each reaction. The relationship between gene expression and phenotype was further obtained by analyzing gene expression using RT-qPCR.

Transcriptional Profiling
To intuitively compare the expression differences of 20 OVATE family genes in different fruit types and different periods, an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Rosa, CA, USA) was used to assess the quality of the total RNAs and RNA samples with RNA integrity numbers > 7 were selected for library preparation. Multiplexed libraries for next-generation sequencing were prepared using the TruSeq RNA Sample Prep Kit v2 (Illumina, San Diego, CA, USA) according to the manufacturer's instructions. Paired-end sequencing was performed using an Illumina HiSeq 4000 platform. The raw paired-end reads were trimmed and quality controlled using SeqPrep (https://github.com/jstjohn/ SeqPrep, accessed on 1 July 2020) and Sickle (https://github.com/najoshi/sickle, accessed on 1 July 2020) with the default parameters. TopHat (v2.1.0) was used to align the highquality reads to the reference genome. The count of the mapped reads from each sample was derived and normalized to reads per kilobase of exon per million mapped reads (RPKM) for each predicted transcript using Cufflinks (v2.2.1). [38,39].

In Situ Hybridization
Young ovaries at 8, 4, and 0 DBA from HZ and YD-4 were fixed in 3.7% formol-aceticalcohol and stored at 4 • C until use. In situ hybridization was performed as previously described [40]. The LsOVATE1 probe was designed according to the specific gene fragments. Sense and antisense probes were generated via PCR amplification with specific primers using SP6 and T7 polymerase, respectively. The primers for probe synthesis are listed in Table S1.

Subcellular Localization
The LsOVATE1 probe was inserted into the N-terminus of green fluorescent protein (GFP) in a pCAMBIA1305-GFP vector, generating the fusion construct p35S:LsOVATE1-GFP. The fusion expression and empty vectors were transformed into DH5α (Tsingke) competent cells, and after PCR amplification and verification by sequencing, these plasmids were transformed into competent Agrobacterium tumefaciens GV3101 (Tsingke) cells using the freezing method. The A. tumefaciens cells containing the recombinant plasmid and blank control were injected into the abaxial side of tobacco leaf explants using the vacuum infiltration method. The fusion expression vector was transiently expressed in N. benthamiana leaves. Fluorescent signals were detected using a laser confocal scanning microscope (ZEISS Microsystems LSM 700) [41].

Construction of Overexpression Plasmid and Phenotypic Evaluation of Transgenic Plants
After the study of gene expression characteristics, the LsOVATE1 function was further verified through transgenic analysis. The full-length CDS of LsOVATE1 was PCR-amplified using a cDNA sample prepared from an 'Hangzhou gourd' ovary with gene-specific primers (Table S1). The PCR reaction was carried out using kodakarensis (KOD) and DNA polymerase (ToYoBo, Shanghai, China) in a total volume of 50 µL. The procedure was as follows: an initial denaturing step at 95 • C for 5 min, 34 cycles of 95 • C for 30 s, 58 • C for 30 s, and 68 • C for 90 s, and a final extension step at 68 • C for 7 min. The LsOVATE1 CDS was inserted into the BamHI and SpeI sites of the pCAMBIA1305-GFP vector. The primer sequences used to obtain the target gene and overexpression vector are listed in Table S1. The constructs were then transformed into tomato using A. tumefaciens GV3101. Tomato transformation was performed as previously described [42]. Transgenic tomatoes were verified with PCR using the vector-and gene-specific primers listed in Table S1. Wild tomatoes were used as control plants.

Identification of OVATE Family Genes in Bottle Gourd
To identify OVATE family genes in the bottle gourd genome, BLAST searches were performed using tomato OVATE proteins as query sequences. The sequences of candidate proteins identified from the BLAST searches were compared. Sequences encoding very short polypeptides or those that did not contain the conserved motifs (characteristic of OVATE family proteins) were excluded after phylogenetic and conserved domain analysis. After multiple screening and validation steps of the conserved domains, we finally identified 20 putative OVATE genes (Table 1). To determine the chromosomal distribution of the identified bottle gourd genes, their physical locations on the bottle gourd chromosomes were investigated. As result, the 20 genes were mapped on 11 chromosomes. Chromosomes 1, 2, 6, 7, 9, 10, and 11 contained one gene each; chromosomes 8 contained two genes each; and chromosome 4 and 5 contained three genes. Five OVATE homologs were located on chromosome 3, which is the chromosome with the maximum number of homologous genes.

Phylogenetic Relationship of the OVATE Family Genes in Bottle Gourd
To evaluate the evolutionary relationships among OVATE proteins, a phylogenetic tree was constructed using the protein sequences from the 20 putative bottle gourd OVATE gene families, 19 melon OVATE gene families, 20 Arabidopsis OVATE gene families, and 27 tomato OVATE gene families, including OVATE ( Figure 1). Specific information on the other OVATE gene families is provided in Table S2. The evolutionary tree classified the OVATE genes into four clades with well-supported bootstrap values. The results show that bottle gourd OVATE genes were unevenly distributed in three clades, with the gene related to bottle gourd shape (chromosome 7) having high homology with the tomato OVATE gene. The corresponding GO analysis is shown in Table S3.
with the gene related to bottle gourd shape (chromosome 7) having high homology with the tomato OVATE gene. The corresponding GO analysis is shown in Table S3.

Structural Characterization of Bottle Gourd OVATE Family Genes
The structures of the proteins encoded by the 20 OVATE family genes in bottle gourd were predicted, and the amino acid sequences of the OVATE proteins were aligned with the tomato OVATE protein ( Figure 2). An analysis of the conserved domains of the bottle gourd OVATE family genes revealed that corresponding identical amino acids were found at multiple sites. This suggests that the 20 identified genes have similar structures and functions, and are also similar to the tomato OVATE gene. This result indicates that the OVATE gene family is highly conserved among these species. The three-dimensional protein structure predictions showed that bottle gourd OVATE family proteins were mainly composed of α-helices and random coils ( Figure 3). In terms of spatial configuration, HG_GLEAN_10017349, HG_GLEAN_10016946, HG_GLEAN_10015982, HG_GLEAN_10004872, HG_GLEAN_10007946, and HG_GLEAN_10001965 showed high similarity. Among the 20 genes, only HG_GLEAN_10023346 contains introns, likely because the three-dimensional structure of its encoded protein has the highest complexity.

Structural Characterization of Bottle Gourd OVATE Family Genes
The structures of the proteins encoded by the 20 OVATE family genes in bottle gourd were predicted, and the amino acid sequences of the OVATE proteins were aligned with the tomato OVATE protein (Figure 2). An analysis of the conserved domains of the bottle gourd OVATE family genes revealed that corresponding identical amino acids were found at multiple sites. This suggests that the 20 identified genes have similar structures and functions, and are also similar to the tomato OVATE gene. This result indicates that the OVATE gene family is highly conserved among these species. The threedimensional protein structure predictions showed that bottle gourd OVATE family proteins were mainly composed of α-helices and random coils ( Figure 3). In terms of spatial configuration, HG_GLEAN_10017349, HG_GLEAN_10016946, HG_GLEAN_10015982, HG_GLEAN_10004872, HG_GLEAN_10007946, and HG_GLEAN_10001965 showed high similarity. Among the 20 genes, only HG_GLEAN_10023346 contains introns, likely because the three-dimensional structure of its encoded protein has the highest complexity.

Expression Profiles of the OVATE Gene Family in the Bottle Gourd
RNA-seq data were used to observe the expression patterns at five stages of ovarian development in both HZ and YD-4 bottle gourd (8, 6, 4, 2, and 0 DBA), in order to assess the possible roles of the 20 OVATE family genes in the growth and development of bottle gourd (Table S4). A heat map was constructed according to the Log10 values (tpm + 0.0001) of the genes (Figure 4). The levels of expression varied widely between different bottle gourd OVATE family genes, and between different stages in individual bottle gourd. Some OVATE family genes in bottle gourd demonstrated opposing expression patterns; for example, HG_GLEAN_10020458 is highly expressed in YD-4 at 0 DBA, but HG_GLEAN_10022831 and HG_GLEAN_10010244 are the opposite. HG_GLEAN_10016946 is highly expressed in YD-4 at 2 DBA, but HG_GLEAN_10001965 shows the opposite expression trend. These varying expression levels suggest that these genes may play different significant roles during different developmental stages.

Expression Profiles of the OVATE Gene Family in the Bottle Gourd
RNA-seq data were used to observe the expression patterns at five stages of ovarian development in both HZ and YD-4 bottle gourd (8, 6, 4, 2, and 0 DBA), in order to assess the possible roles of the 20 OVATE family genes in the growth and development of bottle gourd (Table S4). A heat map was constructed according to the Log10 values (tpm + 0.0001) of the genes (Figure 4). The levels of expression varied widely between different bottle gourd OVATE family genes, and between different stages in individual bottle gourd Some OVATE family genes in bottle gourd demonstrated opposing expression patterns; for example, HG_GLEAN_10020458 is highly expressed in YD-4 at 0 DBA, but HG_GLEAN_10022831 and HG_GLEAN_10010244 are the opposite. HG_GLEAN_10016946 is highly expressed in YD-4 at 2 DBA, but HG_GLEAN_10001965 shows the opposite expression trend. These varying expression levels suggest that these genes may play different significant roles during different developmental stages.

Expression Profiles of the OVATE Gene Family in the Bottle Gourd
RNA-seq data were used to observe the expression patterns at five stages of ovarian development in both HZ and YD-4 bottle gourd (8, 6, 4, 2, and 0 DBA), in order to assess the possible roles of the 20 OVATE family genes in the growth and development of bottle gourd (Table S4). A heat map was constructed according to the Log10 values (tpm + 0.0001) of the genes (Figure 4). The levels of expression varied widely between different bottle gourd OVATE family genes, and between different stages in individual bottle gourd. Some OVATE family genes in bottle gourd demonstrated opposing expression patterns; for example, HG_GLEAN_10020458 is highly expressed in YD-4 at 0 DBA, but HG_GLEAN_10022831 and HG_GLEAN_10010244 are the opposite. HG_GLEAN_10016946 is highly expressed in YD-4 at 2 DBA, but HG_GLEAN_10001965 shows the opposite expression trend. These varying expression levels suggest that these genes may play different significant roles during different developmental stages.

Expression Characteristics of LsOVATE1 Gene
To provide new insights into the expression characteristics of the LsOVATE1 gene, we first cloned LsOVATE1 (720 bp) ( Figure S2). Next, we performed RT-qPCR and in situ hybridization to analyze LsOVATE1 expression. Specific RNA probes were used to perform in situ hybridization on HZ and YD-4 samples at 8, 4, and 0 DBA ( Figure 5). The results showed that LsOVATE1 is expressed in tissues around the embryonic seat of young ovaries at the 8, 4, and 0 DBA growth stages. LsOVATE1 expression at different stages of growth was consistent between HZ and YD-4 specimens. When hybridization was performed with the sense probe ( Figure 5C), no signal was detected. In addition, RT-qPCR analysis showed that LsOVATE1 was most highly expressed in leaves, followed by roots and young fruits, while male flowers had the lowest expression levels ( Figure 6). The gene was expressed differently at different growth stages, but no significant expression trend was evident. To analyze the subcellular localization of LsOVATE1, p35S:CmOFP13-GFP constructs were transfected into N. benthamiana epidermal cells for transient expression.
We found that the LsOVATE1-GFP fusion protein was localized both in the cytoplasm and nucleus (Figure 7).

Expression Characteristics of LsOVATE1 Gene
To provide new insights into the expression characteristics of the LsOVATE1 gene, we first cloned LsOVATE1 (720 bp) ( Figure S2). Next, we performed RT-qPCR and in situ hybridization to analyze LsOVATE1 expression. Specific RNA probes were used to perform in situ hybridization on HZ and YD-4 samples at 8, 4, and 0 DBA ( Figure 5). The results showed that LsOVATE1 is expressed in tissues around the embryonic seat of young ovaries at the 8, 4, and 0 DBA growth stages. LsOVATE1 expression at different stages of growth was consistent between HZ and YD-4 specimens. When hybridization was performed with the sense probe ( Figure 5C), no signal was detected. In addition, RT-qPCR analysis showed that LsOVATE1 was most highly expressed in leaves, followed by roots and young fruits, while male flowers had the lowest expression levels ( Figure 6). The gene was expressed differently at different growth stages, but no significant expression trend was evident. To analyze the subcellular localization of LsOVATE1, p35S:CmOFP13-GFP constructs were transfected into N. benthamiana epidermal cells for transient expression. We found that the LsOVATE1-GFP fusion protein was localized both in the cytoplasm and nucleus (Figure 7).

Functional Validation of LsOVATE1
To further validate the function of LsOVATE1, we used the CaMV35S prom overexpress the coding sequence. Since transformation technology is commonly to use in bottle gourd, we could not confirm the function of LsOVATE1 by gen transgenic bottle gourd lines. Instead, we transformed the LsOVATE1 overexpress tor (p35S:LsOVATE1) into wild-type tomato. The transgenic tomato plants were fully detected using PCR, and the transcript levels of representative transgenic lin confirmed with RT-qPCR. Compared with wild-type tomato, the fruit of the tra plants showed significant changes and a cone-shaped morphology. In addition to phenotype, phenotypic changes including calyx hypertrophy, petal degenerati petal retention after anthesis were observed (Figure 8). Our results indica LsOVATE1 not only affected the fruit shape, but also played a role in petal develo The relative expression levels of LsOVATE1 in the four transgenic plants with re the wild-type control were determined. It was found that LsOVATE1 expression i genic plants correlated positively with the phenotype (Figure 9).

Functional Validation of LsOVATE1
To further validate the function of LsOVATE1, we used the CaMV35S promoter to overexpress the coding sequence. Since transformation technology is commonly difficult to use in bottle gourd, we could not confirm the function of LsOVATE1 by generating transgenic bottle gourd lines. Instead, we transformed the LsOVATE1 overexpression vector (p35S:LsOVATE1) into wild-type tomato. The transgenic tomato plants were successfully detected using PCR, and the transcript levels of representative transgenic lines were confirmed with RT-qPCR. Compared with wild-type tomato, the fruit of the transgenic plants showed significant changes and a cone-shaped morphology. In addition to the fruit phenotype, phenotypic changes including calyx hypertrophy, petal degeneration, and petal retention after anthesis were observed (Figure 8). Our results indicate that LsOVATE1 not only affected the fruit shape, but also played a role in petal development. The relative expression levels of LsOVATE1 in the four transgenic plants with respect to the wild-type control were determined. It was found that LsOVATE1 expression in transgenic plants correlated positively with the phenotype (Figure 9).

Functional Validation of LsOVATE1
To further validate the function of LsOVATE1, we used the overexpress the coding sequence. Since transformation technolog to use in bottle gourd, we could not confirm the function of Ls transgenic bottle gourd lines. Instead, we transformed the LsOVAT tor (p35S:LsOVATE1) into wild-type tomato. The transgenic toma fully detected using PCR, and the transcript levels of representativ confirmed with RT-qPCR. Compared with wild-type tomato, the plants showed significant changes and a cone-shaped morphology phenotype, phenotypic changes including calyx hypertrophy, p petal retention after anthesis were observed (Figure 8). Ou LsOVATE1 not only affected the fruit shape, but also played a rol The relative expression levels of LsOVATE1 in the four transgeni the wild-type control were determined. It was found that LsOVAT genic plants correlated positively with the phenotype (Figure 9).

Discussion
The OVATE protein family is unique to plants and includes specific transcription factors that regulate plant growth and development [43]. OVATE was first found to be a major QTL in controlling the development of pear-shaped tomato fruit [44]. OVATE was first cloned in tomato, and the OVATE domain was found to be conserved across tomato, Arabidopsis, and rice [14,22]. It was further confirmed that OVATE overexpression can induce the transition from pear-to round-shaped tomato [45].
To date, many OVATE family genes, in addition to OVATE, have been described. Their functions are related to the regulation of fruit shape; however, different phenotypes arise based on which genes are active. When studying the expression characteristics of tomato OVATE, subcellular localization revealed that the gene was localized in the nucleus [46]. In addition, SlOFP20 was shown to be localized in the nucleus and cytoplasm. Phenotypic observation of SlOFP20-overexpressing transgenic tomato plants showed that the tomato fruit shape reverted from pear-to round-shaped [26]. Many OVATE family genes have also been found in the model crop Arabidopsis. Nine AtOFPs have been found to regulate the development of Arabidopsis meristems and leaves. AtOFP1 transgenic Arabidopsis plants were found to develop slowly, with shorter leaves and anthers, thicker filaments, and fewer seeds. Subcellular localization analysis showed that AtOFP1 was localized in the nucleus [21]. A total of 31 transcription factors of the OVATE family have been detected in rice [24]. The internodes of rice stems are shortened and the leaves become shorter and wider when OsOFP22 is overexpressed. OsOFP1 and OsOFP2 overexpressed transgenic lines were dwarf in size, with thicker leaves and a lower seed setting Values are means ± SD of three replicates.

Discussion
The OVATE protein family is unique to plants and includes specific transcription factors that regulate plant growth and development [43]. OVATE was first found to be a major QTL in controlling the development of pear-shaped tomato fruit [44]. OVATE was first cloned in tomato, and the OVATE domain was found to be conserved across tomato, Arabidopsis, and rice [14,22]. It was further confirmed that OVATE overexpression can induce the transition from pear-to round-shaped tomato [45].
To date, many OVATE family genes, in addition to OVATE, have been described. Their functions are related to the regulation of fruit shape; however, different phenotypes arise based on which genes are active. When studying the expression characteristics of tomato OVATE, subcellular localization revealed that the gene was localized in the nucleus [46]. In addition, SlOFP20 was shown to be localized in the nucleus and cytoplasm. Phenotypic observation of SlOFP20-overexpressing transgenic tomato plants showed that the tomato fruit shape reverted from pear-to round-shaped [26]. Many OVATE family genes have also been found in the model crop Arabidopsis. Nine AtOFPs have been found to regulate the development of Arabidopsis meristems and leaves. AtOFP1 transgenic Arabidopsis plants were found to develop slowly, with shorter leaves and anthers, thicker filaments, and fewer seeds. Subcellular localization analysis showed that AtOFP1 was localized in the nucleus [21]. A total of 31 transcription factors of the OVATE family have been detected in rice [24]. The internodes of rice stems are shortened and the leaves become shorter and wider when OsOFP22 is overexpressed. OsOFP1 and OsOFP2 overexpressed transgenic lines were dwarf in size, with thicker leaves and a lower seed setting rate [23]. In melon crops, CmFSI8/CmOFP13 (a fruit shape-related gene), has been characterized as a homolog to SlOFP20 and AtOFP1 [6]. The overexpression of this gene in Arabidopsis led to significantly smaller and curly kidney-shaped leaves, as well as significantly shorter plants. Similar to SlOFP20, CmFSI8/CmOFP13 is localized both in the cytoplasm and nucleus, but its expression is relatively weaker. The protein encoded by OVATE in grapes is also localized in the nucleus. The differential expression of VvOVATE in different varieties may be an important factor in fruit shape [47]. In round fruit and long fruit pepper, CaOVATE was found to negatively regulate the expression of CaGA20ox1, resulting in alterations in the gibberellin synthesis pathway, which led to changes in fruit shape from round to oval [25].
In this study, 20 OVATE homologs were successfully identified in bottle gourd. These OVATE homologs were highly conserved with tomato OVATE, which is consistent with previous findings regarding OVATE family genes in tomato, rice, and other crops. LsO-VATE1, a homolog with the highest homology to tomato OVATE, was successfully cloned. Subcellular localization and in situ hybridization studies indicated that LsOVATE1 might be involved in the regulation of nuclear gene transcription. In addition, subcellular localization results were also consistent with those of OVATE family genes in other crops [48], and the expressed LsOVATE1 was localized in the cytoplasm and nucleus. The function of LsOVATE1 was verified using a transgenic approach, which also proved that LsOVATE1 could affect fruit shape. LsOVATE1-overexpressing transgenic seedlings showed obvious changes in fruit shape and flowering period. The transgenic plants exhibited phenotypic characteristics such as calyx hypertrophy, petal degeneration, and petal retainment after anthesis. During further ripening of young fruits, the transgenic fruit dehisced in the pericarp and flowered further at dehiscence. No mature seeds were found in ripe fruits upon opening. This transgenic phenotype is similar to that of Arabidopsis thaliana overexpressing AtOFP1 [43], which affects petals and pollen, as well as the seed-setting rate of mature seeds. Therefore, it is speculated that this gene affects the pollen activity and reproductive mode of plants. Recently, VvMADS39 was discovered in grapes [49], and is homologous to the SEP2 gene of Arabidopsis. The heterologous overexpression of VvMADS39 reduced the fruit and seed size, as well as the seed number in tomato. Research on VvMADS39 also found that MADS39 is required for the normal development of the inner three whorls of floral organs as well as the maintenance of floral meristem identity. Further research found that SlMADS39, which is homologous to VvMADS39, is expressed in tomatoes, and used CRISPER technology to knock out SlMADS39. It was observed that the SlMADS39-knockout phenotype was similar to that of LsOVATE1 overexpression, which was characterized in this study. Whether there is an interaction between the LsOVATE1 protein and MAD proteins needs further verification.
Homologs of the tomato OVATE, SUN, FASCIATED, and LC family genes have been suggested as candidate genes regulating fruit shape [9,50,51]. Apart from the abovementioned "classical" genes, recently an ethylene biosynthesis gene (ACS2) [52] and FRUITFULL-like MADS-box gene (FUL1) [53] in cucumber, an AP2/ERF transcription factor gene (AP2a) gene and TONNEAU1 Recruiting Motif protein (TRM5) in tomato [26,54], as well as an LRR-RLK family gene (CAD1) in peach [55], have also been associated with fruit shape. The CsFUL1 gene directly depresses the expression of the auxin transporters CsPIN1 and CsPIN7, resulting in decreases in auxin accumulation in fruits [53]. In the bottle gourd, despite the number of fsQTLs mapped [56], no fruit shape gene has been identified thus far. Our results also suggest certain OVATE family members as candidate fruit shape genes in the bottle gourd.

Conclusions
We identified a total of 20 OVATE family genes in bottle gourd and divided them into four clades to help understand the evolutionary relationships between these genes, which all had at least one conserved domain. Our chromosomal distribution, sequence alignment, three-dimensional protein structure prediction, and exon/intron analyses of the OVATE gene family provide a useful basis for understanding the function of the OVATE gene family. The expression pattern analysis demonstrated that OVATE family genes in bottle gourd were differentially expressed, indicating that they played different roles in the growth and development of bottle gourd. We studied the candidate gene LsOVATE1, and found that it is homologous to OVATE and affects fruit shape. We also analyzed the expression characteristics of LsOVATE1 based on in situ hybridization, expression level, and subcellular localization analyses. In situ hybridization further revealed that LsOVATE1 transcripts were detectable in young ovaries, which were largely restricted to the placental area. Additionally, LsOVATE1 overexpression in tomatoes caused fruit shape changes, as well as secondary flower and non-seed phenotypes, highlighting LsOVATE1's role in the regulation of plant development ( Figure S3). Further studies on fruit shape-regulating genes should provide new gene information that would allow the generation of new bottle gourd varieties with desirable phenotypes. By comparing with other genes that have similar effects on fruit shape, it is speculated that there may be some regulation mechanism, which ultimately led to the effect of LsOVATE1 on the fruit shape of bottle gourd. This will guide further study on the regulation mechanism of fruit shape in bottle gourd.
Supplementary Materials: The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/biom13010085/s1, Figure S1: Fruit shape pictures of bottle gourd HZ and YD-4; Figure S2: Cloning of LsOVATE1 and the pCAMBIA1305.1-GFP vector; Figure S3: Framework figure; Table S1: Primer list summary; Table S2: Composition and chromosomal location of the plants; Table S3: GO analysis of OVATE proteins; Table S4: List of RNA-Seq data.