1. Introduction
Cucumber (
Cucumis sativus L.) is a diploid species (2n = 14), belonging to the
Cucurbitaceae family of the order Cucurbitales, which encompasses 2295 species and 129 genera [
1]. Cucumber originated in India [
2], particularly in between the Bay of Bengal and the Himalayan Mountain ranges, and it is one of the oldest cultivated crops of the Indian subcontinent. Later, it spread eastward to China [
3,
4] and westward to Asia Minor, northern tropical Africa and Egypt. Columbus introduced it to Haiti, and later to the USA. Some authors have also reported their origins to be in the tropical African plains and Egypt [
3,
5].
Both cultivated and wild cucumbers are extensively used in formulating medicines, for corneal disorders to scorpion bites, and it was believed that the fruits could boost fertility in women [
6]. It is an affordable and excellent source of vitamins K, B
1 and C; pantothenic acid; and minerals such as phosphorus, magnesium, potassium, manganese, copper, molybdenum and biotin. Cucumber fruits are antioxidative, anti-inflammatory and anti-diabetic in nature [
7]. Upon realizing the potential value of cucumbers, they are now cultivated as an economically and nutritionally important vegetable crop in the subtropical and tropical parts of the world.
Despite the availability of enough literature on its economical and nutritious status and the pan-genomic information of eleven (11) wild and cultivated accessions [
8], only minimal genetic and genomic resources are available for cucumber in India. As its primary center of origin, India has the highest diversity ever recorded, but the diversity has gradually depleted due to selection pressure [
6]. Owing to this limited genetic diversity and crossability with only a few related species (such as
C. melo,
C. hystrix and its wild relative
C. sativus. var.
hardwickii), there is no noticeable upsurge in the average yields of cucumbers in the past decades. In order to break the stagnation in yield and enhance the quality, the development of hybrid cucumbers, including gynoecious lines, becomes inevitable.
When monoecious, gynoecious and hermaphrodite phenotypes are present in cucumber [
9,
10,
11,
12,
13], the gynoecious cultivars can potentially bear more fruits much earlier in the production season. Under greenhouse conditions, gynoecy coupled with parthenocarpy has further enhanced productivity. Gynoecious hybrid varieties of cucumber are predominantly used in the production system of many developed and developing countries, and it is estimated that the annual growth rate of the cucumber hybrid seed market will increase by 4.2% from 2019 to 2027. Asia Pacific is expected to be the major player in a seed market containing hybrid seeds and open pollinated/desi (Indian native types)/heirloom seeds and have a whopping summation of about USD 1939.60 million in 2027 [
14]. Thus, hybrid development by using gynoecious lines with regional cucumber cultivars (“desi” types, the popular local/indigenous genotypes of India, meaning “native” types in Hindi) possessing improved qualities would be effective in cucumber breeding programs for exploiting hybrid vigor in yield and quality.
On the other hand, the present day parthenocarpic gynoecious lines are very watery and lack crispiness, resulting in poor quality and market value. However, these lines produce more lateral branches, flowers and fruits and, hence, higher yield. Alternatively, Indian accessions produce crispier, firmer and high-quality fruits, but are poor yielders. Hence, there is a need to combine the qualities of gynoecious cucumbers and desi types by evolving desirable high-yielding hybrids that fetch higher market prices. Nevertheless, understanding the inheritance pattern of gynoecy is imperative to design such hybrid development programs. Research in this direction has provided inconsistent results on the genetic control of female locus (F) in cucumber due to the deployment of diverse sources of gynoecious genes in those studies, as well as environmental effects [
15]; consequently, these drawbacks are bottlenecks in conventional breeding programs. Therefore, the breeding of gynoecious lines using molecular markers has been proposed, since selection is performed with genotypes rather than phenotypes [
16].
Cucumber has a relatively small genome (367 Mbp). Ever since the first draft of its genome was published [
17], several genomic and transcriptomic researches have already been conducted on diversity, biotic and abiotic stress resistances in cucumber [
18,
19,
20,
21,
22,
23,
24,
25,
26]. Further, the updated version of the genome sequence of ‘Chinese Long inbred line 9930′ V3 is also publicly available [
27]. However, the investigation on elucidating the molecular mechanism of the gynoecy habit in cucumber and integrating such information into the regular hybrid development program is yet to be demonstrated.
The study reported here was designed with the rationale of identifying gene-specific SNP markers that are related to the gynoecy habit. Genotyping-by-sequencing (GBS) was performed using the three cucumber accessions, comprising the gynoecious female parent (Pant-PC2, a popular parthenocarpic variety released from Pantnagar University, Uttarakand, India), the male parent (CBE-CS-33, a regional ecotype collected from Sathur, Tamil Nadu, India) and their hybrid (TNAU PCH1), developed in this university. This study has identified unique SNPs/InDels embedded inside the candidate genes concerning the gynoecy habit. Upon validating these markers in other gynoecious lines possessing such genes, these markers can be employed in commercial hybrid development in cucumbers.
4. Discussion
Advanced molecular breeding strategies are promising in improving the production per area, as well as improving quality traits of major vegetable crops in a shorter period. The development of gynoecious hybrids in cucumber has enormous importance for its successful cultivation under both open and polyhouse conditions because of the time and cost incurred in hybrid seed production [
33]. Though the present-day gynoecious lines in India (which dominantly produce female flowers) are high yielders, the fruit quality is very low and unmarketable. On the other hand, the regional Indian types produce fruits with more male flowers and high quality, but they are poor yielders (
Table 1). Thus, introgressing the best of both the lines into gynoecious hybrids would be an effective and affordable strategy to increase yield and quality simultaneously.
The yield of cucumber fruits is chiefly determined by the ratio of female and male flowers. Sex expression in cucurbits is influenced by genetic, environmental and hormonal factors [
34]. Monoecious strains of cucumber bear staminate (male) and pistillate (female) flowers. Gynoecious strains normally produce pistillate flowers only. Ethylene is highly correlated with the femaleness by inducing female flowers and has been regarded as the primary sex determination factor [
35,
36] and the molecular mechanism behind its involvement has been well-documented [
37,
38,
39,
40,
41,
42,
43,
44,
45]. Earlier genetics studies indicated that there are three major sex-determining genes in cucumber and melon:
F,
A and
M. Recently, the
A gene in melon and the
M gene in cucumber have been cloned and both encode
1-aminocyclopropane-1-carboxylic acid synthase (
ACS), which is a key enzyme in ethylene biosynthesis. In cucumber, a series of evidence strongly supports that the
F gene also encodes
ACS [
46].
So far, four “sex genes” have been identified:
F/CsACS1 [
40,
41,
47],
M/CsACS2 [
42,
43,
48],
A/CsACS11 [
49] and
G/CsWIP1 [
49,
50]. Another gene,
CsACO2, has also been shown as critical in sex determination by catalyzing the last step of ethylene biosynthesis [
50]. Apart from these four “sex genes”,
ethylene response 1 (ERT1),
ethylene sensitive 3 (EIN3) and
ethylene responsive factor 110 (ERF110) were also reported to regulate sex expressions in cucumber [
45,
51,
52]. Yin and Quinn put forth the “one-hormone hypothesis” to emphasize the pressing role of ethylene in the sex expression of cucumber [
53]. Besides, gibberellins (GAs) have also shown to have a regulatory role in flower development independent of ethylene [
46,
54].
Thus, it is imperative to identify functional markers derived from the gynoecious habit, as it would fasten the development of hybrids in cucumber. This study has identified several SNP markers linked to the genes involved in sex determination in cucumber using the algorithm PathoLogic [
55]. The genes that are identified in this study have already been functionally annotated and shown to have critical roles in female flower development. In the absence of any experimentally validated markers for the trait of concern, the functionally annotated genes present in publicly available plant databases, such as CuGenDB, serve as valuable genomic resources and a solid base for the initial genic screening of the identified SNPs/InDels across all the chromosomes of the genome.
PathoLogic [
55] automatically develops a Pathway/Genome Database (PGDB) describing the metabolic network of an organism. It is the in-built pathway prediction algorithm of the Pathway Tools software suite. Enzymes catalyze metabolic reactions in all organisms, and each enzyme is linked to a reaction. Based on the organization of the reactions and the dataset provided to the algorithm, PathoLogic constructs pathways. It is the simple assumption that experimentally defined metabolic pathways are conserved between organisms; PathoLogic utilizes MetaCyc [
56], the non-redundant reference pathway database, as a template for building the metabolic pathways for a newly sequenced organism. PathoLogic has succeeded with 91% accuracy and an F-measure of 0.786. In this way, SNPs/InDels that are identified by GBS are found to be annotated for discrete functions by the PathoLogic algorithm of CucurbitCyc of CuGenDB by the technique of chromosome strolling. These SNPs should be confirmed through functional gene characterization in the follow-up studies.
For example,
Transcription factor BIM 1 (
CsaV3_1G003270 G3I-29349) at 2028622 on chromosome 1 (
Supplementary Table S2) encodes a basic helix-loop-helix (
bHLH) family protein, BIM1 (
BES1-INTERACTING MYC-LIKE 1), involved in embryonic patterning, brassinosteroid and auxin signaling in Arabidopsis seedlings. They are found in female floral parts such as petals, carpel, petiole, inflorescence and pollens [
57]. The
bHLH family are one of the largest groups of plants TF [
58]. They are involved in wound and stress responses [
59,
60,
61,
62], hormonal regulation [
63,
64] stigma and anther development and fruit development and differentiation [
65,
66,
67,
68]. The
bHLH genes in
Quercus suber floral libraries have shown to be essential for the development of pollen [
69]. Nine groups of
Q. suber transcripts in female tissues were significantly much expressed, with one transcript (
QsBR ENHANCED EXPRESSION 1) being exclusive to the female samples [
70]. In addition, the
bHLH gene families are also found to be related to ovule development and regulation of female reproductive development [
71]. Therefore, the SNP identified in this study can be used as a potential candidate marker in cucumber hybrid development, at least using the parents investigated in this study.
This study has identified an SNP on
Xyloglucan endotransglucosylate/hydrolase (
CsaV3_1G011020 G3I-30124) at 6,842,309 bp (
Supplementary Table S2), which is associated with primary cell walls as a major tension-bearing structure that limits cell expansion in cauliflower florets [
72]. Likewise, SNPs (on
CsaV3_6G042750 G3I-38711 and
CsaV3_6G042750 G3I-38711;
Supplementary Table S2) were also identified in the vacuolar sorting associated proteins (VPSs). They are a division of Endosomal Sorting Complex Required for Transport (ESCRT) engaged in topologically unique membrane bending and scission reactions away from the cytoplasm and VPS4 were found to have roles in centrosome and spindle maintenance in cell division [
73]. Liu and co-workers [
74] have demonstrated that the absence of VPS38 in plants leads to dampened pollen germination and increased chances of seed abortion.
Similarly,
cyclin-dependent protein kinases (CDKs) are chief components in cell division and expansion in the cell division processes. They are involved in cell cycle arrest, cyclin-dependent protein serine/threonine inhibitor activity, cell division and expansion, and are directly correlated to the growth of the fruit and the fruit quality. This happens during the very early stages of fruit growth [
75].
CDKs act as serine threonine kinases in protein complexes along with cyclins in phosphorylating substrates during mitosis. This study identified SNPs in the serine type peptidase gene
CsaV3_7G029970 G3I-47291 (at 18,964,991 bp position on chromosome 7) and the gene code for pectin esterase (
CsaV3_6G038740 G3I-38310, on chromosome 6 at the 22,176,336th position), which plays major roles in cell division, cell wall formation, cell wall modification and pectin esterase activity. Jiang and his team [
75] correlated the upregulation of
CDKs and
cyclins with fruit length in cucumber, and other studies have also documented their role in fruit development in cucumber [
76] and the early phase of fruit development in avocado [
77].
Interestingly, an SNP on the receptor-like kinase 2 (
CsaV3_1G011660 G3I-30188 at the 7,223,169 bp;
Supplementary Table S2) that possessed the function of pollen recognition and serine/threonine protein activity [
78] was also identified in this study. In the same way, the involvement of 6-phosphofructo-2 kinase/fructose 2, 6-bisphosphate (
CsaV3_6G016080 G3I-37232 on chromosome 6 (involved in phospho–fructose biosynthesis), which was found to have an SNP at 12,041,834 bp in this study) in flower development was also confirmed [
79].
Another functional marker that may have vital importance in hybrid development in cucumber is the SNP at 22,706,544 bp on the zinc knuckle CCHC type family protein (
CsaV3_6G026570 G3I-26767) on chromosome 3, which is involved in DNA binding and zinc ion binding. The zinc-finger family is very diverse in the plant kingdom and consists of a large number of proteins with distinct subfamilies [
80]. Proteins containing zinc finger domains regulate diverse signal transduction pathways [
81] and abiotic stress responses [
79,
82,
83,
84]. Radcova and his team [
85] explored the relationship between the function of the gene coding zinc finger CCHC-type protein and flower morphology and seed size in
Medicago truncatula. A modified transcript of the gene has resulted in overexpression in anthers. Similarly, an SNP at 1,521,082 on
CsaV3_2G003180 G3I-42754, which codes for zinc finger BED-Domain containing protein DAYSLEEPER, is involved in DNA binding and protein dimerization and has been involved in immunity responses in wheat [
86]. These genes are found to be an anciently conserved domain in plants, as they were reported in primitive angiosperms [
87] and have not been engaged in any of the modern home-keeping genes; hence, they are designated as SLEEPER genes.
Fascinatingly, the InDels reported in this study also have importance in gynoecious hybrid development in cucumber. An example is the InDel at 16,265,469 bp on chromosome 3 of male parent code for the gene
MADS Box transcription factor (
CsaV3_3G020270 G3I-26344). The MADS family of transcription factors play significant roles in plant development [
70,
88] and are key players of flower development in several angiosperms [
89]. In another study [
90], it was shown that the MADS-box gene encodes a protein similar (85%) to the Short Vegetative Phase (SVP) protein of
Arabidopsis, which is the popular transcriptional regulator of the flowering time gene.
The significant discovery of this study is the identification of an SNP at 21,385,488 bp on chromosome 6, which was detected 400 bp away from the gene encoding
1-aminocyclopropane-1-carboxylate oxidase 1 (ACS; CsaV3_6G037780 G3I-38214) (EC Number—1.14.17.4). This gene is directly involved in the ethylene biosynthesis pathway. A single nucleotide change in the ACS gene specifically inhibits the male reproductive organs in melon [
91].
Ethylene production is reported to increase drastically during many developmental events such as germination, senescence and abscission, and fruit ripening [
92,
93,
94]. Ethylene is derived from methionine, which is converted to S- adenosylmethionine by S-adenosylmethionine synthetase (AdoMet). AdoMet is then converted to 1-aminocyclopropane-1-carboxylic acid (ACC) and 5¢-deoxy-5¢methylthioadenosine (MTA) by the 1-aminocyclopropane-1-carboxylase synthase (ACS) [
95,
96,
97], which is the rate-limiting step in ethylene biosynthesis.
In some transcriptomic studies, the transcript levels of ethylene receptor genes were quantified to be higher in the reproductive organs in plants such as Arabidopsis, roses, rice and tobacco [
98,
99]. The synthesis of ethylene is related to the sensitivity of ethylene in petals. This is also related to the ACC levels in the flowers. ACC was translocated from the bottom of flowers to the top in the case of roses, orchids and petunia, i.e., from the ovary and receptacle to the pistil, stamen and the lower and upper part of the petals within the flower [
99,
100]. Ethylene determines the sex of every floral meristem by inhibiting the development of stamens and pistil primordia. The development of stamens or carpels are destined depending on the different concentrations of ethylene, thanks to the varying levels of sensitivity to ethylene in the floral tissues [
101,
102,
103,
104]. On the contrary, the knock-out-function studies in the orthologous genes, CsACS11 and CmACS11, of cucumber and melon resulted in the complete blocking of the development of female flowers, resulting in androecium [
93,
105].
The monoecious and dioecious flowers have evolved from bisexual plants either by the arrest of stamens or carpels in a primarily bisexual floral meristem [
93,
95,
97,
106,
107,
108]. The three major loci, namely,
ACS1/ACS1G,
ACS11 and
ACO2, are exclusively involved in the biosynthesis of ethylene at the very early stages of floral meristem development. This regulates the fate of the floral meristem to develop into a female flower [
45,
86,
109]. The homozygous alleles (FF) of the original (
ACS1) and the duplicate (
ACS1G) genes produce increased ethylene contents in the floral meristem. This has led to the conversion of monoecious into gynoecious. The heterozygous alleles (Ff) produce decreased levels of ethylene, so the plant turns subgynoecious. The transcript of
CsACS1/
CsACS1G is accumulated in flowers that developed into females, and is present in higher quantities in the shoot apices of gynoecious plants rather than in monoecious ones [
47,
110,
111]
Sex-determining genes were first discovered in cucumber and melon [
45,
46,
50,
89]. Many other important discoveries in sex determinations were in pumpkin and squash [
100,
108,
110] and watermelon [
112,
113], with only slight variations among the mechanisms in all these species [
112,
113]. Ethylene is the primary regulator of sex determination in cucurbits [
95,
97,
102,
111]. External treatments with ethylene are helpful in determining the role of this hormone in the control of sex expression [
109,
110,
111,
114]. Female flowers are promoted by ethylene by arresting the growth of the stamens or carpels [
114,
115]. Ethylene also increases the ratio of female to male flowers in
Cucumis and
Cucurbita [
115,
116].
Many factors and stimuli affect the increase in the level of ethylene biosynthesis. Light, wounding, pathogen attacks, biotic and abiotic adversities, hypoxia, toxic chemicals and plant hormones such as auxins, brassinosteroids and cytokinin, and even ethylene itself, can auto-stimulate or auto-inhibit their own rate of production [
95,
110,
117]. The increase in the levels of ethylene during such adverse conditions could be the result of a biological trigger/alarm for the plant to initiate the reproductive cycle at the onset of environmental threats. This is why there is a correlation observed between the biosynthesis of ethylene and female flower production, which should lead to fruiting and seed development for the multiplication of its own species.
At least sixteen (16) different
ACS and its homologues have been reported in cucumber on chromosomes 2, 3, 4, 5 and 6 (
Supplementary Table S3). However, after the previous reports accentuating the close association of gynoecy to the set of
ACS genes and their homologs in cucumber, we would like to confirm its role in these investigated accessions. Transcriptomic analysis of the same three samples, namely the female and the male parent with the F
1 hybrid employed in this study, has clearly indicated that
ACS was upregulated in the female parent (PC 2) and hybrid (PCH 1), whereas it was downregulated in the male parent (Sathur) (
Supplementary Table S4). This clearly depicted the role of
ACS in the development of female flowers through ethylene biosynthesis, and if this appropriate allele was introgressed into the male parents, it would certainly induce a gynoecious nature. The markers reported in this study (nucleotide variations based on GBS of two parents and their hybrid) are cross-validated across the published reports. Nevertheless, this mere fact may not strongly support any substantial relationship between the markers and the gynoecious phenotype. Furthermore, validation through quantitative trait loci (QTL) mapping or functional genomic analysis in the breeding materials that are generated using the parental lines reported in this study is obligatory.