Subfunctionalization of CpSVP-Y Paralog Contributes to Male Reproductive Fitness in Papaya

TL;DR

A gene on the Y chromosome in papayas helps male plants grow longer flower stalks, improving their reproductive success.

Contribution

The study identifies subfunctionalization of the CpSVP-Y gene as a mechanism contributing to male reproductive fitness in papayas.

Findings

The male allele CpSVP-Y rescues both pedicel elongation and flowering time in Arabidopsis.

The autosomal allele CpSVP-A only affects pedicel elongation and not flowering time.

CpSVP-Y is retained for both functions, while CpSVP-A likely lost the flowering time function.

Abstract

Papayas possess three sex genotypes: female XX, male XY, and hermaphrodite XYh. Only male plants produce long peduncles with numerous flowers, a trait that improves reproductive success. The gene SHORT VEGETATIVE PHASE (CpSVP) is located on the Y chromosome, absent from the X chromosome, and disrupted in the Yh chromosome, making it a key candidate gene for long peduncles in male plants. An autosomal CpSVP allele (CpSVP-A) was also annotated in a papaya genome. The overexpression of the male allele in Arabidopsis increases pedicel length, which supports its role in pedicel elongation. Unexpectedly, the autosomal allele CpSVP-A produced a similar phenotype as the male one, while the hermaphroditic allele (CpSVP-Yh) did not cause any significant change in pedicel length. Additionally, only the male allele rescued early flowering in an Atsvp mutant, indicating that it regulates both pedicel length and flowering time. In contrast, the autosomal allele only affected pedicel elongation and had no effect on flowering time. Together, these results demonstrate that CpSVP-Y underwent subfunctionalization, retaining both peduncle elongation and flowering time functions, while CpSVP-A likely lost the latter. Thus, CpSVP-Y contributes to male reproductive fitness through enhanced peduncle development without affecting sex determination, potentially regulated by Y chromosome-specific cis-elements. This study provides new insights into the genetic basis of sexually dimorphic traits in papayas.

Figures

Full text

1. Introduction

Papayas (Carica papaya L.) are a tropical fruit consumed worldwide with high nutritional value. Papaya trees are tridioecious (males, hermaphrodites, and females), in which their sex types are controlled by a pair of sex chromosomes that are approximately 7 million years old [1]. There are two slight sequence differences: the Y chromosome determines males, and the Y^h^ chromosome determines hermaphrodites. These two Y chromosomes diverged about 4000 years ago, with only a 0.4% sequence difference between them [2,3], making the identification of sex determination gene(s) difficult due to high sequence similarity. Any combination of these Y chromosomes is inviable at the seed development stage (YY, YY^h^, or Y^h^Y^h^) due to the presence of homozygous lethal gene(s) [4]. The sex-determining region (SDR) of the X and Y/Y^h^ chromosome is 3.4 Mb and 8.1 Mb in length, respectively [1,3]. This relatively small region contains the sex determination genes as well as genes responsible for sex-linked traits such as peduncle length. The long peduncle trait always co-segregates with the male plant; therefore, the gene responsible for peduncle length is a gene that is located on the male Y chromosome but not on the hermaphrodite Y^h^ or female X chromosome [5,6]. Meanwhile, the sequence differences in SHORT VEGETATIVE PHASE (SVP) between male and hermaphrodite papayas have led to the hypothesis that SVP is involved in male–hermaphrodite differentiation [7,8]..

SHORT VEGETATIVE PHASE (SVP), a MADS box transcription factor, has been well characterized in Arabidopsis [9]. It acts in a dosage-dependent manner to repress flowering by forming a complex with FLOWERING LOCUS C (FLC). AtSVP, along with AGAMOUS-LIKE24 (AGL24), determines floral meristem identity in Arabidopsis [10]. During the vegetative phase, AtSVP forms a dimer with FLC, which binds to CArG boxes to repress FLOWERING LOCUS T (FT) in the leaves and SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1 (SOC1) in the shoot apical meristem [11]. How SVP is regulated remains unclear, its function may be modulated through interactions with other MADS-box proteins. MADS box transcription factors activate or repress transcription by binding directly to CArG boxes. While the role of SVP genes in flowering time and floral meristem identity seems to be conserved in some species, it has been shown that SVP genes can have diverse functions. For example, the overexpression of Oryza sativa SVP-like genes in rice implicates their involvement in brassinosteroid signaling and shoot development [12]. In addition, pedicels were elongated when overexpressing Medicago truncatula SVP in Arabidopsis, in addition to delayed flowering and a floral defect phenotype [13].

By comparing the male, hermaphrodite, and female papaya chromosome sequences, a SHORT VEGETATIVE PHASE (SVP) coding sequence was found absent on the X chromosome, an intact SVP sequence was annotated on the Y chromosome (CpSVP-Y), and a truncated SVP sequence without the first exon was located on the Y^h^ chromosome (CpSVP-Y^h^). In addition, there is an autosomal copy of SVP (CpSVP-A) in the papaya genome [1]. Since the elongated peduncle trait is sex-linked to male papaya trees, it is thus hypothesized that the intact CpSVP genes located on the Y chromosome are responsible for this trait, while the truncated CpSVP on the Y^h^ chromosome and intact CpSVP on the autosome are not.

The objective of this study is to functionally characterize CpSVP alleles in Arabidopsis to infer their potential roles in peduncle elongation and flowering time regulation in papayas and to discuss implications for sex chromosome evolution. In this study, the functional characterization of CpSVP-Y, CpSVP-Y^h^, and their paralog CpSVP-A was carried out in Arabidopsis. CpSVP-Y in transgenic Arabidopsis is capable of pedicel elongation with altered floral organs, including enlarged sepals and extra petals but no defects in stamen or carpel structure. CpSVP-Y can also rescue the mutant svp phenotype of short pedicels and early flowering. However, CpSVP-A was found to regulate pedicel elongation but not flowering time in Arabidopsis. These findings suggest that subfunctionalization has occurred between CpSVP-Y and CpSVP-A, with functional divergence likely mediated by promoter regulatory elements. This study provides preliminary insights into the regulatory mechanisms of male organ development in papayas and identifies potential targets for molecular breeding.

2. Materials and Methods

2.1. Plant Material

The experiments were carried out using Arabidopsis thaliana (Col-0). The svp-31 mutant (SALK_026551) was obtained from the Arabidopsis Biological Resource Center. Plants were maintained in controlled-environment growth chambers programmed for 16 h of light at 23 °C and 8 h of darkness at 18 °C, with relative humidity held at 50%.

2.2. Flowering Time Measurements

To determine flowering time, the number of rosette leaves was recorded once the primary inflorescence extended to roughly 2 cm. Pedicel length analyses were performed using nine independent T4 lines per genotype. Sixteen plants were grown for each line, and measurements were taken from the uppermost and lowermost six pedicels on each plant.

2.3. Plasmid Construction

Promoter regions for CpSVP-Y and CpSVP-Y^h^ were identified by aligning their coding sequences to male and hermaphrodite papaya pseudomolecules. Based on these alignments, 2 kb (CpSVP Y) and 1 kb (CpSVP-Y^h^) upstream regions were retrieved. Predicted cis-regulatory motifs were annotated using the PLACE database [14]. Promoter fragments were amplified with specific primers (*CpSVP-Y^h^ *(1kb) and CpSVP-Y (2kb); see Table S1) and inserted into pMDC162 GUS [15]. For overexpression constructs, the coding regions of CpSVP-Y, CpSVP-Y^h^, and CpSVP-A were amplified using primers listed in Table S1 and cloned into pMDC32 [15]. All plasmids were sequence-verified by Sanger sequencing.

2.4. Plant Transformation and Analysis of Transgenic Lines

Constructs were introduced into Agrobacterium tumefaciens GV3101 and subsequently used to transform Arabidopsis via the floral dip method [16]. At least twelve independent T1 transformants were generated for each construct. Seeds were surface-sterilized, plated on half-strength MS medium supplemented with hygromycin and cefotaxime, stratified at 4 °C for two days, and then transferred to growth chambers. Seedlings with four leaves were transplanted to the soil. T2 progeny were screened for single-locus insertions based on a 3:1 segregation ratio on selective medium. Genotyping was performed using gene-specific primers and the Kapa3G Plant PCR Kit (Cat. KK7252).

2.5. Quantitative RT-PCR (qPCR)

Total RNA was isolated with Trizol reagent (Invitrogen, Waltham, MA, USA) according to the supplier’s protocol. RNA concentration was determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and RNA quality was evaluated by electrophoresis on a 1% agarose gel. For cDNA synthesis, 1 µg of total RNA was used as input for the Maxima First-Strand cDNA Synthesis Kit with dsDNase (Cat. K1672, Thermo Fisher Scientific, Waltham, MA, USA). For quantitative PCR, each reaction consisted of 1.6 µL of cDNA template, primers at 0.5 µM, and 10 µL of PowerUp SYBR Green Master Mix (Cat. A25741, Applied Biosystems, Waltham, MA, USA), with nuclease-free water added to volume. Amplification was performed on a CFX96 Touch real-time PCR instrument (Bio-Rad, Hercules, CA, USA). Primer sets used in this study (Table S2) included both previously published sequences and additional primers designed using Primer3 software [17,18]. The cycling conditions are 50 °C for 2 min and 95 °C for 2 min, followed by 35 cycles of denaturation at 95 °C for 15 s and annealing/extension at 60 °C for 1 min, ending with a melt-curve analysis. For expression analysis, a minimum of three independent T3 lines were examined for each gene, with two technical replicates per line. Transcript levels were normalized to ACTIN2 and ELONGATION FACTOR 1-α using the ΔΔCt calculation method [19].

3. Results

3.1. Identification of the CpSVP Gene in Papayas

SVP is annotated in the male and hermaphrodite genomes, as well as a paralogous autosomal copy [1]. CpSVP-Y and CpSVP-A share 94% similarity in their coding sequence and contain the MADS-box and K-box domain. CpSVP-Y and CpSVP-A have an open reading frame of 669 bp, which encodes for 223 amino acids. CpSVP-Y^h^ is 483 bp and encodes a 161-amino-acid truncated protein missing the MADS-box domain at its N-terminal. CpSVP-Y and CpSVP-Y^h^ share 71% similarity in their coding sequence, and both contain the K-box domain.

3.2. Male and Autosomal Copies of CpSVP Regulate Pedicel Elongation

To characterize the function of different copies of CpSVP in regulating pedicel length, overexpression lines were generated in Arabidopsis for each allele containing the open reading frame driven under the constitutive cauliflower mosaic virus (CaMV) 35S promoter. Both 35S:CpSVP-Y and 35S:CpSVP-A transgenic plants had a mean pedicel length of 8.9 mm, an increase of 30.9% when compared to the wild-type pedicels (Figure 1). For 35S:CpSVP-Y^h^, there was no significant difference in pedicel length when compared to wild-type. Only the CpSVP-Y and CpSVP-A alleles resulted in pedicel elongation.

However, the Arabidopsis AtSVP knockout line (svp-31) had shorter pedicels when compared to wild-type plants (p < 0.05, Student’s t-test). The average pedicel length for svp-31 was 5 mm, a 26.5% decrease from wild-type pedicels (Figure 2). The overexpression constructs of the three alleles (35S:CpSVP-Y, 35S:CpSVP-Y^h^, and 35S:CpSVP-A) and two translational constructs (native promoter fused with genes, CpSVP-Y^h^ (1kb):CpSVP-Y^h^ and CpSVP-Y (2kb):CpSVP-Y) were introduced into this svp-31 mutant background to test if any of the alleles could rescue the short pedicel mutant phenotype. 35S:CpSVP-Y svp-31 and 35S:CpSVP-A svp-31 are able to complement the mutant phenotype and have a mean pedicel length of 10.7 mm and 8.7 mm, respectively. 35S:CpSVP-Y^h^ svp-31 and CpSVP-Y^h^(1kb):CpSVP-Y^h^ svp-31 did not have any significant difference in pedicel length when compared to svp-31, indicating the reduced pedicel phenotype in the knockout mutant was not rescued by the CpSVP-Y^h^ gene, no matter which promoter sequence (CaMV35S or native promoter) was used (Figure 2). CpSVP-Y(2kb):CpSVP-Y svp-31 could also rescue the mutant phenotype and had a mean pedicel length of 7.6 mm. These results indicate that CpSVP-Y and CpSVP-A play a role in pedicel elongation but not CpSVP-Y^h^.

The text continues here (Figure 1 and Figure 2).

3.3. CpSVP Regulates Flowering Time

SVP has been demonstrated to be involved in regulating flowering time by binding to FT [11]. SVP acts as a flowering repressor, so knocking out SVP should result in an early flowering phenotype. Indeed, we observed that the knockout (svp-31) line flowers earlier than wild-type (p < 0.01, Student’s t-test; see Table 1). We introduced the overexpression constructs of the three alleles (35S:CpSVP-Y, 35S:CpSVP-Y^h^, and 35S:CpSVP-A) and two translational constructs (native promoter fused with genes—CpSVP-Y^h^_(1kb):CpSVP-Y^h^ and CpSVP-Y (2kb):CpSVP-Y) into this svp-31 mutant background, to determine if any of the alleles could rescue the early flowering phenotype. 35S:CpSVP-A svp-31, 35S:CpSVP-Y^h^ svp-31, and CpSVP-Y^h^(1kb):CpSVP-Y^h^ svp-31 had the same early flowering phenotype as svp-31 (Table S1). Only 35S:CpSVP-Y svp-31 and CpSVP-Y_(2kb):CpSVP-Y svp-31 were able to rescue the early flowering phenotype in the knockout mutant, restoring the flowering time that is comparable to wild-type Col-0 (Table S1). Interestingly, overexpression lines of CpSVP-Y in the Col-0 background flower later than wild-type (p < 0.01, Student’s t-test; see Table S1). Besides flowering time, additional morphological floral defects not found in the svp-31 or Col-0 background were observed. In addition to elongated pedicels, both 35S:CpSVP-Y and 35S:CpSVP-A plants in the svp-31 and Col-0 backgrounds had morphological defects, including extra petals and enlarged sepals, but not in CpSVP-Y_(2kb):CpSVP-Y (Figure 3). These enlarged sepals did not dehisce and remained attached to the siliques throughout senescence (Figure 4).

3.4. Deciphering the Mechanism of Pedicel Elongation

A previous field experiment showed that exogenous application of GA3 to papaya flowers caused peduncles to elongate [7]. AtGA20ox2 plays a key role in GA biosynthesis, encoding the rate-limiting enzyme that catalyzes the second-to-last step of bioactive GAs [20,21]. To understand the role of CpSVP in GA signaling, the transcript levels of AtGA20ox2 in the overexpression and complemented lines were examined. In all three overexpression lines of CpSVP, GA20ox2 was downregulated in the junction tissue, especially for CpSVP-Y. However, GA20ox2 was significantly upregulated in young and old pedicels for all overexpression lines (Figure 5). In the complemented lines, GA20ox2 is significantly downregulated in the junction of 35S:CpSVP-Y svp-31 and CpSVP-Y(2kb):CpSVP-Y svp-31 and expressed at very low levels in young pedicels (Figure 6). However, in complemented lines that express CpSVP-A and CpSVP-Y^h^, the GA20ox2 expression remained high in both junction and young pedicel. All lines showed high GA20ox2 expression in old pedicels, suggesting that there is no expression correlation between CpSVP and GA20ox2 in this specific tissue.

4. Discussion

4.1. Specific Alleles of CpSVP Are Involved in Peduncle Elongation

The long peduncle trait co-segregates only with the male plant; therefore, the gene responsible for peduncle length is likely located on the male Y chromosome but not on the hermaphrodite Y^h^ or female X chromosome [7]. SVP has been identified in the male and hermaphrodite genomes, as well as a paralogous autosomal copy in the autosome [1]. To infer their potential functions, these alleles were functionally characterized in Arabidopsis. 35S:CpSVP-A (autosomal copy) showed elongated pedicels and altered floral morphology compared to wild-type. Furthermore, the elongated pedicels were also observed in transgenic Arabidopsis carrying 35S:CpSVP-Y. 35S:CpSVP-Y (male copy), 35S:CpSVP-A (autosomal copy), and CpSVP-Y(2kb):CpSVP-Y were able to complement the svp mutant’s short pedicel phenotype. These findings indicate that both male and autosomal copies of CpSVP have the potential to control pedicel length in the Arabidopsis system. Unexpectedly, the autosomal copy also exhibited the ability to regulate pedicel length in transgenic Arabidopsis. This copy is present in all three sex types of papayas, but only male plants have the long peduncle phenotype, suggesting that the peduncle elongation function of the autosomal allele is not expressed in papayas. This could be achieved through the control of cis-elements or the suppression of expression at the transcription and/or translational level. Further experiments are required to confirm the role of the autosomal copy in papaya plants.

On the other hand, 35S:CpSVP-Y^h^ (hermaphrodite copy) and CpSVP-Y^h^ (1kb):CpSVP-Y^h^ had no effect on pedicel length in transgenic Arabidopsis. These observations coincide with our hypothesis that CpSVP-Y^h^ is a non-functional copy in hermaphroditic papayas that cannot control peduncle length. Since the gene structure of CpSVP-Y^h^ was disrupted by a transposon insertion at its first exon site, it encodes a truncated protein without the MADS domain at its N-terminal, thus presumably affecting its protein function. The authors of [8] hypothesize that SVP is involved in male–hermaphrodite differentiation in papayas based on the SVP coding sequence difference between male and hermaphroditic papaya plants. Based on observations in all three CpSVP allelic overexpression lines, no defects in stamen or carpel were found, so CpSVP is unlikely to be the gene responsible for sex determination in papayas.

In summary, the functional data from Arabidopsis indicate that the CpSVP-Y and CpSVP-A alleles can affect peduncle elongation in heterologous systems. The ancestral peduncle elongation function is retained in both copies, but its phenotypic expression is restricted to male papayas, most likely mediated by differential expression regulation. This expression regulation represents a key hypothesis for explaining sex-specific traits and should be prioritized in our future research.

4.2. Conserved Role of CpSVP in Flowering Time

SVP has been shown to be involved in controlling flowering time [9,11]; for example, 35S:AtSVP lines are late in flowering and convert inflorescences into shoot-like structures [22]. In this study, this conserved role of SVP was observed, particularly in a specific allele of CpSVP. Overexpression lines of CpSVP-Y in the Col-0 background flowered later than wild-type, and this male copy was able to rescue the early flowering phenotype of the Arabidopsis knockout line svp-31 in the complementation experiment, indicating that CpSVP-Y is involved in flowering time regulation. However, the overexpression of CpSVP-Y^h^ and CpSVP-A in Arabidopsis had no effect on flowering time. In nature, male papaya plants flower earlier than their female and hermaphrodite counterparts. CpSVP-A does not affect flowering time in Arabidopsis, unlike its Arabidopsis homolog AtSVP, suggesting the autosomal copy lost the ability to regulate flowering time, but the male allele retained this conserved role. The shared role of male and autosomal CpSVP in pedicel elongation but difference in flowering time regulation indicates that there is a subfunctionalization of these two CpSVP alleles in papayas.

Arabidopsis thaliana and Carica papaya diverged from a common ancestor 72 million years ago [23], and since this divergence, there have been two whole-genome duplication events in Arabidopsis [24]. Since a heterologous system was used to test the function of CpSVP-A, the possibility that CpSVP-A is involved in flowering time regulation in papayas cannot be ruled out. Ectopic expression of Medicago truncatula SVP in Arabidopsis delayed flowering and had floral defects such as long pedicels, but only floral defects were observed when transformed into Medicago [13]. In addition, the overexpression of CpSVP-Y and CpSVP-A caused floral defects, including enlarged sepals and extra petals, suggesting the genetic redundancy of these genes in flower development. However, CpSVP-Y(2kb):CpSVP-Y did not have altered floral organs, unlike the construct driven under the CaMV 35S promoter, suggesting that the promoter specifies the location where CpSVP-Y is expressed to minimize its deleterious effect on floral organs. In summary, CpSVP-Y retains its dual ability to regulate flower stalk structure and flowering time through subfunctionalization, which provides a direct molecular basis for enhancing male reproductive fitness. The specific realization of the trait of long pedicels in male plants is likely to still require the drive of Y-chromosome-specific cis-elements.

4.3. The Preliminary Research on the Hormone Control of the Pedicel Development

Exogenous applications of gibberellic acid 3 (GA3) on female, hermaphrodite, and male papaya flowers resulted in an increase in peduncle length for all sexes [7]. This suggests a potential interaction between GA and peduncle growth. We identified four GA-related cis-elements in 2kb male and hermaphrodite CpSVP promoter regions using PLACE [14]. GA negatively regulates GA20ox2 in a feedback loop. Thus, we surveyed the transcript levels of GA20ox2, a key enzyme in GA biosynthesis, in our transgenic Arabidopsis to see if the three alleles of CpSVP had any effect on its expression. In overexpression lines, GA20ox2 was upregulated in young and old pedicels and downregulated in the junction, no matter which allelic copy of CpSVP was expressed. There is a stronger reduction in GA20ox2 expression (4-fold) in the young pedicels of plants expressing 35S:CpSVP-Y than in the other two lines (1- to 2-fold). Since the endogenous AtSVP was expressed in these transgenic lines, the effect contributing from papaya SVP might be overturned or minimized. Therefore, we surveyed the expression in the knockout line svp-31. When CpSVP-Y was expressed in svp-31 plants, GA20ox2 was significantly downregulated in the junction and young pedicel. This suggests a correlative link between high CpSVP-Y expression and reduced GA20ox2 levels in these specific tissues, potentially implicating GA metabolism in pedicel development. On the other hand, we observed a positive expression correlation between CpSVP-A/CpSVP-Y^h^ and GA20ox2 in both junction and young pedicel tissue, an opposite trend compared to CpSVP-Y. In the old pedicels, the GA20ox2 expression remained high in all constructs, suggesting no direct correlation between CpSVP and GA20ox2. The correlation between CpSVP-Y and GA20ox2 provides preliminary evidence for a specific association between CpSVP-Y and gibberellin (GA) signaling, while the direct mechanism underlying its interaction with gibberellin signaling remains to be validated through genetic and biochemical approaches in subsequent studies.

Optimal inflorescence architecture plays a pivotal role in development by ultimately determining the number of flowers that can set fruit and seed [25]. Peduncle length contributes to an optimal inflorescence structure, which helps increase the competitiveness of plants [26]. Male papaya plants are the only sex type to bear a long peduncle with a large number of flowers situated on pedicels. Therefore, a long peduncle allows male plants to gain a reproductive advantage by providing more space for flower growth per tree and increasing the capacity of flower bearing. In addition, pollen from male flowers has a higher wind pollination success rate with the aid of a long peduncle [27]. CpSVP-Y plays a major role in peduncle elongation and allows males with their elongated peduncles to remain competitive against hermaphrodites. Simultaneously, the CpSVP-Y copy retained its early flowering property like its Arabidopsis homolog, which allows male papaya plants to flower before female and hermaphroditic plants. This increases the outcrossing pollination rate of the Y chromosome in the papaya population and maintains high genetic diversity, while self-pollinated hermaphroditic plants usually have lower genetic diversity and possibly accumulate deleterious alleles. Although the pedicel structures of Arabidopsis thaliana (an annual herb) and papaya (a woody tree) are not directly homologous, in this study, Arabidopsis thaliana was used as a model system with powerful genetic tools and known gene networks to reveal the regulatory principles that may be suitable for papaya peduncle elongation. The mechanism insights obtained from Arabidopsis thaliana provide key directions and candidate genes for targeted gene function verification and breeding applications in papayas.