Genomic and Cytogenetic Analysis of Synthetic Polyploids between Diploid and Tetraploid Cotton (Gossypium) Species

Abstract

Cotton (Gossypium spp.) is the most important natural fiber source in the world. The genetic potential of cotton can be successfully and efficiently exploited by identifying and solving the complex fundamental problems of systematics, evolution, and phylogeny, based on interspecific hybridization of cotton. This study describes the results of interspecific hybridization of G. herbaceum L. (A₁-genome) and G. mustelinum Miers ex Watt (AD₄-genome) species, obtaining fertile hybrids through synthetic polyploidization of otherwise sterile triploid forms with colchicine (C₂₂H₂₅NO₆) treatment. The fertile F₁C hybrids were produced from five different cross combinations: (1) G. herbaceum subsp. frutescens × G. mustelinum; (2) G. herbaceum subsp. pseudoarboreum × G. mustelinum; (3) G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum; (4) G. herbaceum subsp. africanum × G. mustelinum; (5) G. herbaceum subsp. euherbaceum (variety A-833) × G. mustelinum. Cytogenetic analysis discovered normal conjugation of bivalent chromosomes in addition to univalent, open, and closed ring-shaped quadrivalent chromosomes at the stage of metaphase I in the F₁C and F₂C hybrids. The setting of hybrid bolls obtained as a result of these crosses ranged from 13.8–92.2%, the fertility of seeds in hybrid bolls from 9.7–16.3%, and the pollen viability rates from 36.6–63.8%. Two transgressive plants with long fiber of 35.1–37.0 mm and one plant with extra-long fiber of 39.1–41.0 mm were identified in the F₂C progeny of G. herbaceum subsp. frutescens × G. mustelinum cross. Phylogenetic analysis with 72 SSR markers that detect genomic changes showed that tetraploid hybrids derived from the G. herbaceum × G. mustelinum were closer to the species G. mustelinum. The G. herbaceum subsp. frutescens was closer to the cultivated form, and its subsp. africanum was closer to the wild form. New knowledge of the interspecific hybridization and synthetic polyploidization was developed for understanding the genetic mechanisms of the evolution of tetraploid cotton during speciation. The synthetic polyploids of cotton obtained in this study would provide beneficial genes for developing new cotton varieties of the G. hirsutum species, with high-quality cotton fiber and strong tolerance to biotic or abiotic stress. In particular, the introduction of these polyploids to conventional and molecular breeding can serve as a bridge of transferring valuable genes related to high-quality fiber and stress tolerance from different cotton species to the new cultivars.

1. Introduction

Cotton belongs to the genus Gossypium which consists of approximately 46 diploid species and seven allotetraploid species [1,2,3]. All diploid (2n = 2x = 26) cotton species belong to eight genome groups (A-G and K), and all tetraploid (2n = 4x = 52) species are classified into one genome group (AD) [4,5]. Gossypium species spread through large geographical and ecological areas and has a wide range of morphological and genetic diversity, mainly preserved in germplasm collections and as genetic materials for cotton breeding programs worldwide. These resources can be successfully used to transfer valuable traits from wild species into elite cultivars [6,7]. Polyploidization is one mechanism that introduces the genetic diversity of diploids to polyploids, which allows easy crossing with tetraploids and obtaining fertile hybrids. Unfortunately, issues arise when diploid wild species are crossed with tetraploids. It is often impossible to obtain fertile hybrids through crossing between tetraploid and wild diploid species [8,9]. Completely sterile triploid (2n = 3x = 39) hybrids are obtained from the natural hybridization of tetraploid elite cultivars with wild diploid species [4]. Nevertheless, the fertility of hybrids can be recovered by polyploidization, resulting in hexaploid (2n = 6x = 78) plants [10].

The process of ancient genome duplication is called paleopolyploidy, which occurred at least several million years ago (MYA) [11]. In paleopolyploidy, the duplication of the genome of a single species is called autopolyploid, while combining the genomes of two species is allopolyploidy. Polyploidy is common in plants, fish, and some amphibians. Polyploids have more than two complete sets of chromosomes, which are inheritable. Diploid gametes can result in increased chromosome sets because of a failure to divide during meiosis. A triploid zygote is formed when these diploid gametes fuse with a monoploid gamete. These triploids can be unstable and sterile. When these diploid gametes fuse with another diploid gamete, a stable tetraploid zygote is formed. Polyploids can be tetraploid (4x), hexaploid (6x), and other aneuploids, higher order polyploids, or multiple chromosome pairs. The chromosome doubling through polyploidization can enhance the genetic diversity of plants.

The first comprehensive study in this field was conducted 250 years ago by Carl Linnaeus [12]. Polyploidization event was first discovered in 1907 and was considered responsible for increasing the number of chromosomes [13]. The development of plants that change as a result of polyploidization increases their adaptability to adverse conditions and environmental factors [14,15]. Currently, several chemical and gaseous agents are used in research to obtain polyploids. The most used agents are colchicine and oryzalin [16]. Colchicine (C₂₂H₂₅NO₆) is obtained from the bulb-like corms of the plant Colchicum autumnale L. [17]. This substance is an alkaloid agent used to induce polyploidy [18]. This type of chemical binds to the tubulin protein and prevents the formation of microtubule spindle fibers (achromatin threads) during the metaphase of mitotic cell division, resulting in the duplication of chromosome bundles without spreading to the poles [19]. Colchicine can effectively stop cell division at the mitotic anaphase. At this stage, the chromosomes have already been duplicated, but mitosis has not yet occurred, and the restriction of cell wall formation at this stage leads to the formation of polyploid cells. They are usually larger than diploid species and often have thickening of the tissues, which leads to the formation of large plant organs [20].

The diversity of Gossypium species is an ideal model for understanding the evolution and domestication of polyploids, as well as the pathways and mechanisms of gene and genome evolution [21]. There are several large Gossypium germplasm collections around the world, and among the largest and richest germplasm collections are those in the USA and Uzbekistan. These germplasm collections consist of cotton accessions with high genetic diversity, and their widespread use in genetics and breeding is of great importance in solving the current problems facing the cotton industry [22,23].

The historical domestication and modern breeding of Upland cotton (Gossypium hirsutum) have significantly improved its yield and fiber quality, but this has led to a dramatically reduced genetic diversity [24,25,26,27]. Upland cotton cultivars have a narrow genetic base of fiber quality and abiotic stress tolerance traits. The gene introgression from species such as G. herbaceum and G. mustelinum allows the increase of genetic diversity by introducing new alleles to improve fiber quality and stress tolerance. G. herbaceum, also known as ‘Levant cotton’, is grown in rainfed areas of Africa, and this crop is tolerant to salinity, drought, and scorching hot weather [28]. G. mustelinum is a wild species native to the northeastern region of Brazil, among existing allotetraploids [29]. The fiber of G. mustelinum is genetically different from that of allotetraploid G. hirsutum. The quantitative trait loci (QTL) for fiber quality and the interesting alleles identified from G. mustelinum can make an important contribution to the improvement of Upland cotton germplasm [25]. Wang et al. [30] dissected the molecular genetic basis of fiber strength and fineness in G. mustelinum and G. hirsutum crosses. They identified 42 QTLs, highlighting the potential of G. mustelinum alleles to improve fiber quality in Upland cotton. The study revealed new allelic variation for cotton breeding. Yang et al. [31] made a chromosome-level genome assembly of G. mustelinum, demonstrating its efficacy in identifying genes for qualitative and quantitative traits. The study laid a foundation for cotton genetics and breeding, emphasizing the rich gene pool of G. mustelinum. Chen et al. [25] used selective genotyping to validate over 75 QTLs for fiber quality traits that were introgressed from G. mustelinum into Upland cotton. The study lays the foundation for fine mapping, marker-assisted selection, and map-based gene cloning.

In the past few decades, the following studies were conducted on obtaining interspecific hybrids exploiting the genetic diversity of Gossypium germplasm: G. hirsutum × G. klotzschianum [32], G. hirsutum × G. trilobum [33], G. arboreum × G. anomalum [34], G. hirsutum × G. anomalum [35], G. herbaceum × G. australe [36], G. hirsutum × G. australe [37], G. hirsutum × G. arboreum [38], G. capitis-viridis × (G. hirsutum × G. australe)² [39], G. hirsutum × G. darwinii [40], G. herbaceum × G. nelsonii [41]. Despite an extensive literature review, no reports were found on obtaining allotetraploids using the cotton species G. herbaceum and G. mustelinum.

In this study, we hypothesized that obtaining fertile hybrids through the polyploidization of triploids derived from the cross between G. herbaceum and G. mustelinum could enable their use in tetraploid cotton breeding. To test this hypothesis, the goals of our research were as follows: (1) to obtain F₁ triploid hybrids of G. herbaceum and G. mustelinum cotton species; (2) to obtain F₁C hexaploids through doubling the genome of triploid hybrids using colchicine treatment; and (3) to select tetraploids among segregating F₂C aneuploid (triploid, tetraploid, and hexaploid) hybrids. The new knowledge and genetic resources would be used in the introduction of valuable traits such as fiber quality and stress tolerance into the elite Upland cotton varieties using marker-assisted selection (MAS) technology.

2. Results

2.1. The Characteristics of Interspecific Cotton Hybrids between G. herbaceum and G. mustelinum

Several crosses were made between G. herbaceum (A₁ genome) and G. mustelinum (AD₄ genome) species to develop interspecific hybrids with valuable traits as initial materials for genetics and breeding studies. As a result of many crossings, the interspecific triploid F₁ hybrids were obtained in five different combinations such as (1) G. herbaceum subsp. frutescens × G. mustelinum, (2) G. herbaceum subsp. pseudoarboreum × G. mustelinum, (3) G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum, (4) G. herbaceum subsp. africanum × G. mustelinum, and (5) G. herbaceum subsp. euherbaceum (variety A-833) × G. mustelinum. The boll-setting rate of the hybrids in the recipient buds after crossing was 13.8–92.2%, and complete seed-setting rate in hybrid bolls was 9.7–16.3%. The highest rate of fertility was observed between variety A-833 (subsp. euherbaceum) and subsp. frutescens (cultivated tropical form), and boll-setting rate was over 50.0%. A relatively low fertility rate was recorded in subsp. pseudoarboreum × G. mustelinum (13.8%). Complete seed-setting rates showed relatively similar among all combinations (9.7–16.3%) (

Table 1. The fertility and hybrid boll-setting rates in interspecific crosses between G. herbaceum and G. mustelinum species.

No.	Hybrid Combination	Number of Crosses	Number of Obtained Hybrid Bolls	Fertilization Rate, %	Boll Setting Rate, %
					X ± Sx	Range	S	V
1.	G. herbaceum subsp. africanum × G. mustelinum	62	15	24.2	10.9 ± 1.0	7.7–18.2	3.06	28.09
2.	G. herbaceum subsp. pseudoarboreum × G. mustelinum	59	8	13.8	10.0 ± 1.9	4.3–23.1	5.9	58.7
3.	G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum	65	15	24.6	9.7 ± 1.1	5.9–15.4	3.39	34.95
4.	G. herbaceum subsp. frutescens × G. mustelinum	63	58	92.2	13.1 ± 3.2	4.0–41.7	10.12	77.25
5.	G. herbaceum subsp. euherbaceum (A-833) × G. mustelinum	62	33	54.5	16.3 ± 1.7	10.0–25.0	5.41	33.18

Note: S stands for the relative error of the results of a series of x measurements at a confidence level of 95%; and V stands for the coefficient of variation, as described by Dospekhov [42].

).

Typically, triploids are infertile, some of them are semi-fertile and can produce both aneuploid and euploid gametes. Taking this into account, the resulting F₁ triploid hybrid seeds were treated with colchicine before sowing. Triploid and polyploid hybrids were obtained as following scheme (Figure 1).

Only, G. herbaceum subsp. africanum × G. mustelinum, subsp. pseudoarboreum f. harga × G. mustelinum, subsp. frutescens × G. mustelinum (Figure 2) combinations were grown after colchicine treatment to produce viable F₁C hybrids.

2.2. Inheritance of Morphological and Economically Valuable Traits in Interspecific Polyploid Hybrids F₁C and F₂C

2.2.1. The Number of Bolls per Plant and the Fertility of Seeds

It is known that parameters such as the number of bolls in a plant, fiber yield, weight of 1000 seeds, weight of cotton in one boll and the number of fully born seeds in one boll play an important role in determining the level of productivity. Since these indicators are inextricably linked with each other and have a correlative effect on the development of the character. In addition, qualitative and quantitative indicators of seeds born in one bag are used as a basis for determining the phylogenetic relationships of species based on classical methods. Therefore, the initial sources used in the rese—arch and the yield indicators of the polyploid hybrids obtained on their basis were analyzed. In terms of the number of bolls per plant among the subspecies of G. herbaceum, the highest rate (36.0 pieces) was recorded in the form of subsp. pseudoarboreum f. harga. In the analyzed bolls, the rate of complete seed (or, fertility of seeds) was 89.2 ± 1.9%. Also, the number of bolls per plant showed positive indicators on subspecies of subsp. africanum—27.0 pieces, and the yield of complete seeds in the boll was 94.5 ± 1.8%, in the variety A-833 (subsp. euherbaceum)—28.0 pieces, and the rate of complete seeds was 86.9 ± 1.9%. In G. mustelinum—a Brazilian endemic species, the average number of bolls per plant was—16.0, and the rate of complete seed was—86.0 ± 2.1% (

Table 2. The number of bolls per parental plant and the fertility of cotton seeds.

No.	Plant Samples	Number of Bolls		Number of Seeds Per Boll			Percentage of Complete Seeds Per Boll, %
		Per Plant	Analyzed	Total	Complete	Empty	$\overline{\mathit{x}}$ ± S$\overline{\mathit{x}}$	Range	S	V %
Parental Lines
1.	G. herbaceum subsp. frutescens	21	10	25.4	23.5	1.9	93 ± 2	84–100	6.17	6.7
2.	G. herbaceum subsp. pseudoarboreum	20	10	19.5	18.0	1.5	92 ± 2	85–100	6.00	6.5
3.	G. herbaceum subsp. pseudoarboreum f. harga	36	10	18.7	16.7	2.0	89 ± 2	80–100	5.90	6.6
4.	G. herbaceum subsp. africanum	27	10	18.5	17.5	1.0	95 ± 2	84–100	5.70	6.0
5.	G. herbaceum subsp. euherbaceum A-833	28	10	17.6	15.3	2.3	87 ± 2	82–100	6.08	7.0
6.	G. mustelinum Miers ex Watt	16	10	27.1	23.2	3.9	86 ± 2	77–96	6.62	8

Note: S stands for the relative error of the results of a series of x measurements at a confidence level of 95%; and V stands for the coefficient of variation, as described by Dospekhov [42].

).

In F₁C subsp. frutescens × G. mustelinum polyploid hybrids, the number of bolls per plant was—21.0, and the “fertility rate of seeds per bolls” was very low (19.4 ± 1.9%) (

Table 3. The number of bolls per hybrid plant of F₁C and F₂C polyploids.

No.	Hybrid Combination	Number of Bolls		Number of Seeds Per Boll			Percentage of Complete Seeds Per Boll, %
		Per Plant	Analyzed	Total	Complete	Empty	$\overline{\mathit{x}}$ ± S$\overline{\mathit{x}}$	Range	S	V %
F1C polyploid hybrids
1.	G. herbaceum subsp. frutescens × G. mustelinum	21	10	14.3	2.8	11.5	19 ± 2	8–27	6.13	31.6
2.	G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum	1	1	9	2	7	-	-	-	-
3.	G. herbaceum subsp. africanum × G. mustelinum	-	-	-	-	-	-	-	-	-
F2C polyploid hybrids
4.	T 51-13 G. herbaceum subsp. frutescens × G. mustelinum	26	10	22.9	7.0	15.9	30 ± 1	25–36	3.75	12.3
5.	T 30-8 G. herbaceum subsp. frutescens × G. mustelinum	24	10	27.2	11.6	15.6	43 ± 1	38–46	2.50	5.9
6.	T 38-3 G. herbaceum subsp. frutescens × G. mustelinum	21	10	26.0	9.5	16.5	36 ± 1	31–43	3.69	10.3
7.	T 30-6 G. herbaceum subsp. frutescens × G. mustelinum	19	10	18.7	9.2	9.5	49 ± 2	41–61	6.60	13.4
8.	T 38-12 G. herbaceum subsp. frutescens × G. mustelinum	18	10	22.8	7.8	15.0	34 ± 1	28–39	3.78	11.1

Note: S stands for the relative error of the results of a series of x measurements at a confidence level of 95%; and V stands for the coefficient of variation, as described by Dospekhov [42].

). An average of 14.3 seeds were set in one boll, of which 2.8 whole seeds and 11.5 empty seeds were set. In F₁C subsp. pseudoarboreum f. harga × G. mustelinum hybrids, the number of bolls per plant was—1.0, the number of complete seeds was 2.0, and the number of empty seeds was 7.0. F₁C subsp. africanum × G. mustelinum hybrid plants were sterile, and no yield components were observed.

In hybrid plants of subsp. frutescens × G. mustelinum F₂C generation, there was a variation of 18.0–26.0 units in terms of “number of bolls per plant”. Relatively low values (between 30.6 ± 1.2% and 49.2 ± 2.1%) were found in the “seeds fertility” indicator, also. According to the results of the analysis, in polyploid hybrids of the F₁C and F₂C generations, there was a slight decrease in the number of pods per plant compared to the parental samples, as well as cases of sterility in the subsp. africanum × G. mustelinum combination. A sharp decrease was noted in terms of the rate of complete seed fertility, which is the main factor of productivity. This indicates that the parental lines are phylogenetically distant and isolated species.

2.2.2. Vegetative Growth Duration

According to the results of the research, wild (subsp. africanum) and ruderal (subsp. pseudoarboreum f. harga) forms of G. herbaceum of which, the vegetative growth duration was 137.4 and 127.2 days, therefore they would require breeding for short day conditions. The vegetative growth duration in the cultivated-tropical subsp. frutescens, ruderal subsp. pseudoarboreum subspecies and cultivated subsp. euherbaceum A-833 variety is 117.2, 119.6, and 117.4 days respectively, in field conditions since they do not require breeding for short days. The wild tetraploid species G. mustelinum also requires breeding for short days since the duration of its vegetative growth duration was 151.8 days.

F₁C hybrid plants obtained by cross-species with representatives of the diploid G. herbaceum species and tetraploid G. mustelinum species was between 128.6–145.2 days. Including F₁C subsp. frutescens × G. mustelinum the vegetative duration of polyploid hybrid plants was 128.6 days, the coefficient of variation was 2.3%. The coefficient of dominance for the sign is equal to hp = 0.34, and the state of partial dominance of the form with a positive indicator (subsp. frutescens) was noted. The vegetative growth duration of F₁C subsp. pseudoarboreum f. harga × G. mustelinum polyploid hybrid combination plants was 139.6 days on average, the variation coefficient was 1.5%, and the dominance coefficient was hp = −0.01. In this combination, the character was inherited with a partial dominance of the negative indicator form (G. mustelinum). The average vegetative growth duration of polyploid hybrid plants was 139.6 days, the variation coefficient was 1.5%, and the dominance coefficient was hp = −0.01. In this combination, the character was inherited in the partial dominance of the G. mustelinum species with a negative indicator. The polyploid hybrid plants of F₁C subsp. africanum × G. mustelinum combination had an average vegetative growth duration of 145.2 days, and the coefficient of variation was 2.2% (

Table 4. Heredity and variability of the vegetative growth duration in F₁C and F₂C—polyploid hybrids and parental lines.

No.	Plant Samples		Vegetative Growth Duration, Day
			Range	$\overline{\mathit{x}}$ ± S$\overline{\mathit{x}}$	S	V %	hp
Parental lines
1.	G. herbaceum subsp. frutescens		116–119	117.2 ± 0.4	1.3	1.1	-
2.	G. herbaceum subsp. pseudoarboreum		118–122	119.6 ± 0.4	1.5	1.3	-
3.	G. herbaceum subsp. pseudoarboreum f. harga		126–129	127.2 ± 0.4	1.3	1.0	-
4.	G. herbaceum subsp. africanum		135–139	137.4 ± 0.4	1.8	1.3	-
5.	G. herbaceum subsp. euherbaceum A-833		116–119	117.4 ± 0.4	1.3	1.1	-
6.	G. mustelinum		150–153	151.8 ± 0.4	1.3	0.9	-
F1C polyploid hybrids
7.	G. herbaceum subsp. frutescens × G. mustelinum		125–133	128.6 ± 0.9	2.8	2.3	0.34
8.	G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum		137–142	139.6 ± 0.7	2.1	1.5	−0.01
9.	G. herbaceum subsp. africanum × G. mustelinum		142–150	145.2 ± 1.3	3.1	2.2	−0.08
F2C polyploid hybrids
10.	G. herbaceum subsp. frutescens × G. mustelinum	81 plants n = 4	115–119	116.6 ± 0.5	1.8	1.6	-
			120–124	121.8 ± 0.7	2.1	1.7	-
			125–129	127.8 ± 0.5	1.6	1.3	-
			130–134	131.4 ± 0.6	2.0	1.5	-
			135–139	136.8 ± 0.7	2.1	1.5	-
			140–144	141.6 ± 0.5	1.5	1.1	-
			145–149	147.2 ± 0.5	1.6	1.1	-
			150–155	152.2 ± 0.8	2.6	1.7	-

Note: S stands for the relative error of the results of a series of x measurements at a confidence level of 95%; and V stands for the coefficient of variation, as described by Dospekhov [42].

). The coefficient of dominance for the studied character is equal to hp = −0.08, and in this combination, the character was inherited in the case of partial dominance of the negative indicator form (G. mustelinum).

From the above-mentioned F₁C polyploid hybrids, only F₂C hybrids of subsp. frutescens × G. mustelinum combination were obtained. Also, in this generation, the degree of variability of the indicators of the period from the germination to the opening of the first bud was studied in 50% of the plants. A total of 134 plants were obtained in the F₂C generation of the subsp. frutescens × G. mustelinum combination, of which 53 strongly required breeding for short days and thus affected yield components under field conditions. In the remaining 81 F₂C hybrids, the growth duration showed a wide range of variability between 116.6–152.2 days, which were divided into eight classes. Among the studied hybrid combinations, the growth duration was 115–119 day and faster recombinant forms around 120–124 were isolated.

2.2.3. Fiber Length and Strength

Cotton has been cultivated for centuries mainly for its fiber and has undergone natural and artificial selection. The wild diploid and tetraploid species are the natural resources with high genetic diversity for improving the fiber quality traits of elite cultivars. Therefore, fiber length is considered one of the major indicators of valuable economic traits in the breeding process.

According to the results of fiber quality tests the fiber length was in the range of 17.2–25.2 mm in G. herbaceum subspecies (

Table 5. Fiber length and strength in F₁C and F₂C—polyploid hybrids and parental lines.

No.	Plant Samples	Fiber Length, mm					Fiber Strength, cN/tex
		Range	$\overline{\mathit{x}}$ ± S$\overline{\mathit{x}}$	S	V %	hp
Parental lines
1.	G. herbaceum subsp. frutescens	16.0–18.0	17.2 ± 0.2	0.63	3.7	-	35.5
2.	G. herbaceum subsp. pseudoarboreum	24.0–27.0	25.1 ± 0.4	1.20	4.8	-
3.	G. herbaceum subsp. pseudoarboreum f. harga	19.0–21.0	20.1 ± 0.2	0.74	3.7	-	21.4
4.	G. herbaceum subsp. africanum	24.0–26.0	24.7 ± 0.3	0.82	3.3	-	32.6
5.	G. herbaceum subsp. euherbaceum A-833	24.0–26.0	25.2 ± 0.3	0.79	3.1	-	27.2
6.	G. mustelinum	25.0–27.0	25.6 ± 0.2	0.70	2.7	-	17.3
F1C polyploid hybrids
7.	G. herbaceum subsp. frutescens × G. mustelinum	25.0–32.0	28.7 ± 0.9	0.91	3.1	1.74	-
F2C polyploid hybrids
8.	G. herbaceum subsp. frutescens × G. mustelinum	23.1–41.0	29.7 ± 1.2	3.3	11.0	57.3	-

Note: S stands for the relative error of the results of a series of x measurements at a confidence level of 95%; and V stands for the coefficient of variation, as described by Dospekhov [42].

). In this case, a relatively low result was observed in subsp. frutescens (17.2 ± 0.2 mm) and a relatively high result was observed in cultivated variety subsp. euherbaceum A-833 (25.2 ± 0.3 mm). As well as 25.1 ± 0.4 mm was in ruderal subsp. pseudoarboreum, 20.1 ± 0.2 mm was in subsp. pseudoarboreum f. harga and 24.7 ± 0.3 mm was in wild-type subsp. africanum. In the case of tetraploid G. mustelinum, the fiber length was 25.6 ± 0.2 mm.

The fiber strength varied in all of the cotton samples including the highest values of 35.5 cN/tex and 32.6 cN/tex, respectively, in subsp. frutescens and subsp. africanum. The fiber strength in A-833 variety of subsp. euherbaceum was 27.2 cN/tex, in subsp. pseudoarbareum f. harga was 24.1 cN/tex, and in G. mustelinum was 17.3 cN/tex (Table 5).

Due to two F₁C allohexaploid combinations were dead after the germination and the other two being male sterile, therefore all analyses were conducted only with G. herbaceum subsp. frutescens × G. mustelinum F₁C allohexaploids since they were fertile. The fiber length was 28.7 ± 0.9 mm, and a positive heterosis (hp = 1.74) was observed.

Variability was observed in the plants of F₂C subsp. frutescens × G. mustelinum hybrid combinations. The analyzed results were divided into eight classes. Accordingly, two transgressive plants with long fiber of 35.1–37.0 mm, and one transgressive plant with extra-long fiber of 39.1–41.0 mm were identified. Since the fiber of the Brazilian endemic species G. mustelinum differs from that of the allotetraploid species G. hirsutum in terms of strength, maturity and length, a genetically rich source for improving fiber qualities of cultivars.

The identification of major QTL loci for fiber quality and application of a set of beneficial alleles from G. mustelinum can contribute significantly to the long-term improvement of cultivated cotton germplasm. To introduce important alleles from G. mustelinum species into the genome of G. hirsutum species for fiber quality improvement, interspecific populations consisting of plants obtained by crossing G. mustelinum species with G. hirsutum species were created [43,44,45]. In the studied families, it was determined that the genes responsible for fiber strength and fineness in alleles of G. mustelinum are dominantly inherited. Based on the previous work with G. barbadense, G. tomentosum, and G. darwinii, which included introgression of G. mustelinum alleles [43,44,45]. These hybrid genotypes identified in our study would serve as a valuable source for the development of long-fiber cultivars and future breeding programs.

2.3. Cytogenetic and Genomic Studies of Interspecific Hybrids

Chromosome conjugation, spore and pollen viability analysis at metaphase I (MI) stage of meiosis is one of the effective methods for determining the ploidy level of intergenomic hybrids (allopolyploids, autopolyploids). Cytogenetic analyzes of hybrids resulting from the crosses between diploid and tetraploid cotton species were conducted in the study, and interesting data were obtained regarding their genomic structure and their partial homology.

In the three hexaploid plants studied, disturbances in the process of the first division of meiosis (MI) were detected. F₁C subsp. frutescens × G. mustelinum 38.22 ± 0.25 bivalents, 0.45 ± 0.29 univalents (the number of univalents was up to six) and 0.36 ± 0.14 quadrivalents. F₁C subsp. pseudoarboreum f. harga × G. mustelinum and F₁C subsp. africanum × G. mustelinum combinations had a higher number of univalents than the above combination (up to 2–10) (

Table 6. Conjugation of cotton chromosomes at metaphase-I stage of meiosis.

No.	Hybrid Combinations	The Number of Studied Maternal Anther Cells	Average Number Per Cell
			Univalents	Bivalents	Quadrivalents
	F1C polyploid hybrids
1.	G. herbaceum subsp. frutescens × G. mustelinum	22	0.45 ± 0.29	38.22 ± 0.25	0.36 ± 0.14
2.	G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum	15	1.73 ± 0.85	37.53 ± 0.44	0.46 ± 0.24
3.	G. herbaceum subsp. africanum × G. mustelinum	16	1.81 ± 0.84	37.56 ± 0.43	0.50 ± 0.24
	F2C polyploid hybrids
4.	T 37-14 G. herbaceum subsp. frutescens × G. mustelinum	30	1.46 ± 0.55	24.73 ± 0.29	0.26 ± 0.09
5.	T 38-3 G. herbaceum subsp. frutescens × G. mustelinum	23	-	26.00 ± 0.00	-
6.	T 37-4 G. herbaceum subsp. frutescens × G. mustelinum	28	0.64 ± 0.25	38.18 ± 0.23	0.25 ± 0.11
7.	T 49-2 G. herbaceum subsp. frutescens × G. mustelinum	22	0.54 ± 0.2 7	38.00 ± 0.2 7	0.27 ± 0.13
8.	T 30-8 G. herbaceum subsp. frutescens × G. mustelinum	18	-	26.00 ± 0.00	-

, Figure 3). The analysis showed that due to structural differences in chromosomes and the impossibility of normal chromosome conjugation in hybrids, there was a desynaptic effect due to early and asynchronous divergence in individual bivalents (Figure 3 a,b).

In general, the number of univalents found in maternal pollen cells was recorded from two to ten. It was known that there is weak desynapsis (meeting of several univalents along with bivalents in the female pollen cell), medium desynapsis (meeting of many univalents along with bivalents in the female pollen cell), and complete desynapsis (meeting of mainly univalents and sometimes several bivalents in the female pollen cell) [46]. Such univalents were formed as a result of early divergence of chromosomes. In the cotton samples of our study, this level—from two to ten univalent encounters—corresponded to the medium desynapsis.

As a result of research, the MI phase of meiosis was observed in some plants of the F₂C subsp. frutescens × G. mustelinum hexaploid hybrid (Figure 4). The cotton plant samples with different levels of ploidy (24.73^II; 38.18^II; 38.00^II) were identified in second generation (F₂C). Disruptions in the mentioned chromosomal conjugation led to partial sterility of anthers in hexaploid hybrids and abnormal spore formation at the sporulation stage. It should also be noted that the MI phase of meiosis was carried out normally (26.00^II) in two hybrid plants (Table 6).

Thus, the cause of partial infertility observed in allopolyploids was likely due to the presence of disturbances in the meiosis phase (disruption of the synchronous distribution of chromosomes in the anaphase or desynapsis phenomenon), instability of the number of chromosomes. One of the unique features of hybrids derived from the crosses between cotton species far from each other was with chromosomal aberrations. In such cases, plants of with chromosomal aberrations are due to the absence of homologous chromosomes of different species or the presence of their only partial homologue. Hybrids with homeologous genomes are viable, but sterile, i.e., sterile, due to genome duplication [47]. According to the conclusions of Sanamyan [48], hybrid combinations obtained from the diploid species (G. thurberi × G. raimondii, G. arboreum × G. thurberi, and G. herbaceum × G. thurberi) are productive in generations, on the contrary, intergenomic hybrids are very a large percentage of plants without flowers to appear.

There were difficulties in spore analysis to determine the required stage due to the very low number of spikelets and pollen grains in polyploid hybrids. As a result, ten additional samples had to be analyzed. According to the analysis of the conducted tetrads, 90.3–96.8% meiotic stage in three hybrid forms (F₁C subsp. frutescens × G. mustelinum, F₁C subsp. pseudoarboreum f. harga × G. mustelinum, F₁C subsp. africanum × G. mustelinum) index was defined. It was noted that certain disturbances of the meiotic index in the form of micronuclear tetrads and polyads were observed in all forms. In F₁C subsp. frutescens × G. mustelinum polyploid hybrid, the meiotic index was 90.36%, the rate of micronuclear tetrads was 3.27% and that of polyads was 6.37% (

Table 7. Cytological observation of F₁C generation polyploid hybrids.

No.	Hybrid Combinations	Total Number of Spores	Meiotic Index, %	Micronuclear Tetrads, %	Polyads, %
	F1C polyploid hybrids
1.	G. herbaceum subsp. frutescens × G. mustelinum	612	90.36 ± 1.19	3.27 ± 0.72	6.37 ± 0.98
2.	G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum	219	96.80 ± 1.18	-	3.19 ± 1.18
3.	G. herbaceum subsp. africanum × G. mustelinum	481	96.05 ± 0.88	0.83 ± 0.41	3.12 ± 0.79
	F2C polyploid hybrids
4.	T 1-8 G. herbaceum subsp. frutescens × G. mustelinum	411	96.15 ± 0.12	0.89 ± 0.41	3.07 ± 0.72
5.	T 38-12 G. herbaceum subsp. frutescens × G. mustelinum	325	97.05 ± 0.64	-	2.62 ± 0.29
6.	T 30-6 G. herbaceum subsp. frutescens × G. mustelinum	354	97.21 ± 0.47	-	2.78 ± 0.71
7.	T 51-13 G. herbaceum subsp. frutescens × G. mustelinum	298	97.45 ± 0.88	0.69 ± 0.23	2.11 ± 0.41
8.	T 30-7 G. herbaceum subsp. frutescens × G. mustelinum	341	98.51 ± 0.48	-	1.48 ± 0.44
9.	T 30-8 G. herbaceum subsp. frutescens × G. mustelinum	405	96.05 ± 0.88	-	3.12 ± 0.79
10.	T 38-3 G. herbaceum subsp. frutescens × G. mustelinum	357	98.54 ± 0.81	-	1.42 ± 0.38

).

Micronuclear tetrads were not found in F₁C subsp. pseudoarboreum f. harga × G. mustelinum polyploid hybrid unlike other plants. F₂C subsp. frutescens × G. mustelinum polyploid hybrids T 38-12, T 30-6, T 30-7, T 30-8, T 38-3 forms meiotic index 96.05–98.54%, polyads 1.42–3.12%, micronuclear tetrads were not observed. In T 1-8 and T 51-13 forms, meiotic index was 96.15–97.45%, polyads 2.11–3.07%, micronuclear tetrads 0.69–0.89%.

In tetrads with identified micronuclei, up to 1–8 micronuclei (Figure 5c), and from polyads up to pentad, hexad, heptad, octad (Figure 5d) aneuploid spores were noted.

It was known that pollen grains formed as result of meiosis are not equal in terms of genetic characters and functional capabilities. When the chromosomal complex of pollen grains was insufficient (i.e., when the frequency of meeting homologous chromosomes was low, the pollen grains were underdeveloped, and the viable pollen decreased significantly). When a complete chromosome complex was obtained, that is, when the homologous row of chromosomes was restored, the pollen grains became pink and fertile again. As a result of cytogenetic analysis, pollen graininess in hybrid plants F₁C subsp. pseudoarboreum f. harga × G. mustelinum 63.89% in the combination of, F₁C subsp. frutescens × G. mustelinum combination had a low yield—32.67%, F₁C subsp. africanum × G. mustelinum combination, pollen sterility was determined (

Table 8. Development of cotton pollens in F₁C polyploid hybrids.

No.	Hybrid Combinations	Total Number of Pollens	Pollen Fertility, %
1.	G. herbaceum subsp. frutescens × G. mustelinum	324	32.7 ± 3.6
2.	G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum	432	63.9 ± 3.6
3.	G. herbaceum subsp. africanum × G. mustelinum	541	0

, Figure 6).

Since most of the hybrid plants showed short day and late tolerance in F₂C generation of the combination subsp. frutescens × G. mustelinum, 24 out of 134 plants were studied for pollen graininess (

Table 9. Analysis of pollen quality in amphidiploid hybrids of F₂C G. herbaceum subsp. frutescens × G. mustelinum.

No.	Sample	Total Number of Pollens	Pollen Fertility, %	No.	Sample	Total Number of Pollens	Pollen Fertility, %
1.	T 1-13	1199	33.9 ± 1.9	13.	T 30-8	2955	55.7 ± 0.8
2.	T 1-4	700	40.4 ± 3.4	14.	T 37-4	482	50.2 ± 5.2
3.	T 1-8	482	51.0 ± 5.2	15.	T 37-8	796	23.1 ± 2.2
4.	T 30-23	538	27.1 ± 3.7	16.	T 38-12	846	60.9 ± 2.8
5.	T 30-6	666	60.2 ± 3.6	17.	T 38-3	1315	67.8 ± 1.7
6.	T 30-7	738	65.9 ± 3.0	18.	T 38-4	367	38.4 ± 6.4
7.	T 37-14	114	7.9 ± 6.4	19.	T 38-7	352	10.2± 2.6
8.	T 30-1	267	22.9 ± 6.6	20.	T 42-2	159	14.5 ± 7.8
9.	T 33-2	690	46.8 ± 3.6	21.	T 49-2	606	1.5 ± 0.2
10.	T 33-6	248	5.6 ± 21.1	22.	T 51-15	426	34.3 ± 5.3
11.	T 36-1	76	15.8 ± 17.5	23.	T 51-19	363	25.1 ± 5.2
12.	T 37-2	228	17.5 ± 6.3	24.	T 51-13	1536	51.8 ± 1.6

). Until late autumn, reproductive organs were not formed in many hybrids, abnormal forms were noted without pollen development in hybrids that started flowering. There were a few cases where dust grains were not formed.

The pollen fertility in the studied hybrids had different indicators. The pollen fertility of 14 hybrids was low (1.49–40.43%). Among them T 33-6, T 30-8, T49-2, T 38-7 plants were very low (1.49–10.23%), T 33-2, T 37-4, plants up to 50% or more, T 38-3, T 30-6, T 30-7, T 38-12, more than 60% (Table 8, Figure 7) indicators of dustiness were found in plants.

2.4. Molecular and Phylogenetic Analysis of Allopolyploid Forms of Cotton

The PCR analysis was conducted using 72 simple sequence repeat (SSR) markers (Table S1) associated with economically important traits to determine the genetic polymorphisms between parental genotypes as well as to determine some genomic changes in allohexaploid hybrids (Figure 8). As a result of PCR analysis, 57 DNA markers were determined as polymorphic markers (Table S2), while 14 DNA markers were monomorphic, and one DNA marker was not amplified.

Based on the results of PCR analysis, the genomic changes were determined in the genomic regions of 18 (31.6%) out of 57 polymorphic DNA markers or in 40 (26.7%) PCR amplicons (alleles) of 150 alleles in allohexaploid hybrids. In particular, 31 (26.0%), 29 (24.3%), and 27 (22.6%) alleles had a change in three F₁C combinations (G. herbaceum subsp. frutescens × G. mustelinum), F₁C (G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum) and F₁C (G. herbaceum subsp. africanum × G. mustelinum) hybrids, of 119, 120 and 116 alleles, respectively. In these 3 allohexaploid hybrids, alleles 23, 21, and 20 were missing although they were present in parental genotypes, respectively. On the other hand, non-parental alleles 8, 8, and 7 appeared in allohexaploid hybrids, respectively.

A phylogenetic tree construction indicated that the cotton accessions were clustered into two main groups (Figure 9). The first group included three Gossypium herbaceum subspecies such as subsp. frutescens, subsp. pseudoarboreum f. harga and subsp. africanum. G. herbaceum subsp. pseudoarboreum f. harga and subsp. africanum are genetically more similar to each other than subsp. frutescens. G. mustelinum and three interspecific allohexaploid hybrids F₁C (G. herbaceum subsp. frutescens × G. mustelinum), F₁C (G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum) and F₁C (G. herbaceum subsp. africanum × G. mustelinum) were presented in the second group of the phylogenetic tree.

3. Discussion

To obtain synthetic polyploids, some chemical mutagens, especially colchicine, are most effective. Colchicine prevents spindle formation during mitosis, which prevents the separation of daughter chromosomes in anaphase and cytokinesis, leading to a doubling of the number of chromosomes in the cell [49]. In this study, to increase the number of chromosomes (autopolyploidy), representatives of the diploid cotton species G. herbaceum were treated with 0.1 and 0.2% solutions of colchicine and synthetic tetraploid plants were developed. In studies conducted with various percentage solutions of colchicine, it was found that a 0.2% solution of colchicine was more effective than a 0.1% solution. As an effective concentration, we recommend the use of a 0.2% (20 h) of a solution of colchicine to conduct this kind of research.

A different genetic variability was observed in F₂C hybrid plants. Two transgressive plants with long fiber of 35.1–37.0 mm, and one transgressive plant with extra-long fiber of 39.1–41.0 mm were identified in the F₂C progeny of G. herbaceum subsp. frutescens × G. mustelinum combination. The observed genetic variability, especially in cotton polyploid genotypes with long and extra-long fiber, has direct implications for cotton breeding. These variations would provide opportunities for selecting and developing cultivars with enhanced fiber qualities, potentially leading to the creation of commercially valuable cotton varieties with improved fiber length. According to the data mentioned above [43,44,45], although the F₂C allopolyploid hybrid combination obtained with G. mustelinum (subsp. frutescens × G. mustelinum) is late maturing, fiber yield is low compared to other forms, it can be concluded that the release of long fiber and extra-long fiber plants is the effect of G. mustelinum genes. Another reason for this outcome can be explained by the mutation of some sections (gene alleles) in the genome of G. herbaceum due to the application of the experimental polyploidy method, and the dis-appearance or appearance of F₁C allopolyploid hybrid genome regions. That is, depending on the distribution of the signs of parental forms in the hybrids of the second generation, it is estimated that the number of polymer genes and their constant state in future generations. These isolated forms will serve as a unique genetic resource for the development of long fiber varieties in future breeding programs. For this reason, we have planned to introduce specific F₂C polyploids into conventional and molecular cotton breeding programs to obtain new varieties with desired traits through hybridization and backcrossing approaches. The parental genotypes possess distinct fiber properties, with G. mustelinum exhibiting soft and long fiber, while G. herbaceum has fiber with hygroscopic properties. The fiber characteristics mentioned above, valuable in the global cotton industry, are expected to be effectively incorporated into the future cultivars.

As a result of molecular and phylogenetic studies, it was shown that tetraploid hybrids derived from the G. herbaceum × G. mustelinum cross, are close to the species G. mustelinum. Also, it was found that compared with the first cluster of the phylogenetic tree, the second cluster appeared to be small. It only had one subcluster containing the forms subsp. frutescens, subsp. pseudoarboreum f. harga, and subsp. africanum belonging to the species G. herbaceum L., which indicates their phylogenetic relatedness. At the same time, it was noted that these forms are significantly distant from the type of G. mustelinum. Another important aspect of the research results was that subsp. frutescens from representatives of G. herbaceum are close to their cultivated forms, and that its subsp. africanum is a wilder form.

4. Materials and Methods

4.1. Plant Materials

In this study, wild, semi-wild and cultivated subspecies of diploid cotton G. herbaceum L. and wild tetraploid cotton G. mustelinum Miers ex Watt as well as their five interspecific hybrids (1) G. herbaceum subsp. frutescens × G. mustelinum; (2) G. herbaceum subsp. pseudoarboreum × G. mustelinum; (3) G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum; (4) G. herbaceum subsp. africanum × G. mustelinum; (5) G. herbaceum subsp. euherbaceum (variety A-833) × G. mustelinum were used. The seeds of parental genotypes were taken from the special cotton germplasm collection at the Institute of Genetics and Plant Experimental Biology (IGPEB) Academy of Sciences of the Republic of Uzbekistan (ASRUz). Experiments were conducted during 2020–2022 in the laboratory of Experimental polyploidy and phylogeny of cotton and the nursery at the IGPEB, located in Qibray district, Tashkent.

4.2. Hybridization and Polyploidization

Diploid (2n = 2x = 26) G. herbaceum L. subspecies were crossed with tetraploid (2n = 4x = 52) G. mustelinum Miers ex Watt and obtained five combinations of triploid (2n = 3x = 39) hybrid genotypes. Experiments were conducted under two conditions: (1) F₁ hybrid seeds of triploids were treated with 0.1% colchicine for 24 h, and (2) 0.2% colchicine was applied for 20 h. Both conditions involved darkness at room temperature (22 °C) during seed germination, reaching approximately 1 cm in shoot length. Consequently, the duration of the colchicine treatment was reduced by four hours with an increased concentration. As a result, F₁C (C for colchicine) synthetic allohexaploid (2n = 6x = 78) hybrid (G. herbaceum subsp. frutescens × G. mustelinum; G. herbaceum subsp. pseudoarboreum × G. mustelinum; G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum; G. herbaceum subsp. africanum × G. mustelinum; G. herbaceum subsp. euherbaceum (variety A-833) × G. mustelinum) genotypes have been obtained. Synthetic allohexaploid genotypes were planted aim to obtain F₂C generations. Unfortunately, seed of two allohexaploid genotypes (G. herbaceum subsp. pseudoarboreum × G. mustelinum and G. herbaceum subsp. euherbaceum (variety A-833) × G. mustelinum) died at the germination stage. Two F₁C allohexaploids (G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum and G. herbaceum subsp. africanum × G. mustelinum) were male-sterile. Only, F₁C G. herbaceum subsp. frutescens × G. mustelinum allohexaploids were fertile and obtained F₂C genotypes.

4.3. Cytological Analysis

Analysis of sporads and pollen viability were carried out according to Pausheva [50]. For the analysis of tetrads, 2–4 mm cotton buds (in the pinhead square and match-head square stages) were collected early morning and were fixed in the ethanol-acetic acid mixture (7:3 v/v). The samples were screened using a trinocular microscope (N-300M(MD101), Ningbo Yongxin Optics Co., Ltd., Ningbo, China) after staining with acetocarmine. For the analysis of tetrads, the meiotic index (Mi) was calculated as the percentage of normal tetrads over total sporads. In order to determine the pollen viability, pollen was collected from newly opened cotton flowers in the first half of the day and was stained with acetocarmine. Next cytological analysis was caried out using a Leica CM E microscope with a Leica EC3 camera (The Leica Microsystems Inc., Wetzlar, Germany).

4.4. Phenotypic Observation

During the 2021–2022 planting seasons, G. herbaceum L. subspecies, G. mustelinum Miers ex Watt and their F_1-2C allohexaploids were evaluated for phenotypic traits in the field condition. To study morpho-biological traits, such as (i) vegetative growth period (VP); (ii) number of bolls per plant (NB); (iii) number of seeds per boll (NSB); (iv) boll weight (BW); (v) 1000-seed weight (SW); (vi) fiber length (FL); fiber strength (FS); and (vii) male fertility (MF) and sterility (MS) of pollens.

4.5. DNA Isolation and SSR Analysis

The genomic DNA from cotton plant leaf tissues was isolated using the cetyltrimethylammonium bromide (CTAB) method [51]. The DNA concentration was calculated by measuring the absorbance of 1 µL of the samples at 260/280 nm using the NanoDrop Eight spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). DNA samples were diluted to a working concentration of 25 ng/µL. A total of 72 SSR (simple sequence repeat) markers were selected from CottonGen the cotton marker database (https://www.cottongen.org/data/markers) (accessed on 20 February 2022) [52]. PCR-based SSR genotyping was conducted as described previously [6,53,54]. The construction and visualization of the phylogenetic tree was performed using NCSS 12.

4.6. Statistical Analysis

Statistical analyses of the obtained experimental results were carried out according to Dospekhov [42]. The mean-square deviation from the mean was determined for the number of measurements n = 10 according to the formula:

$$\sigma =\surd \frac{\sum {\left(X_i-\overline{X}\right)}^2}{n-1}$$

Mean-square deviation from the mean value is calculated by the following formula:

$${\sigma }_{\overline{X}}=\frac{\sigma }{\sqrt{n-1}}$$

Substituting the reliability criteria (Student’s coefficient) into the formula, the boundaries of the confidence interval for the arithmetic mean were obtained:

$$∆=t_c{\sigma }_{\overline{X}}$$

The relative error of the results of a series of x measurements at a confidence level of 95% will be equal to:

$$S=\frac{100{\sigma }_{\overline{X}}{t}_c}{\overline{X}}$$

V—the coefficient of variation was calculated by the below-mentioned formula:

$$V=\frac{100{\sigma }_{\overline{X}}}{\overline{X}}$$

(1)

5. Conclusions

In this study, we made several crosses between cotton species and obtained interspecific triploid F₁ hybrids in five different combinations: (1) G. herbaceum subsp. frutescens × G. mustelinum; (2) G. herbaceum subsp. pseudoarboreum × G. mustelinum; (3) G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum; (4) G. herbaceum subsp. africanum × G. mustelinum; (5) G. herbaceum subsp. euherbaceum (variety A-833) × G. mustelinum. Fertile allohexaploid F₁C hybrids were obtained by polyploidization of triploid forms. Cytogenetic analysis showed the existence of univalent, open, and closed ring-shaped quadrivalent chromosomes at the stage of metaphase I in the F₁C and F₂C hybrids. Also, various meiotic anomalies (tetrad cells with micronuclei and polyads), as well as low rates of pollen (36.6–63.8%) and complete sterility (F₁C subsp. africanum × G. mustelinum) were observed in the analysis of tetrads. This indicates the structural heterozygosity of polyploid hybrids.

A total of 57 out of the 72 SSR markers identified in the study were determined polymorphic between cotton species. Genomic changes were observed in allohexaploid hybrids. Allele changes were observed in F₁C (G. herbaceum subsp. frutescens × G. mustelinum), F₁C (G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum) and F₁C (G. herbaceum subsp. africanum × G. mustelinum) hybrids, 31 (26.0%), 29 (24.3%) and 27 (22.6%) out of 119, 120 and 116 alleles, respectively.

As a result of phylogenetic tree construction, the cotton accessions were clustered into two groups. The first group included three Gossypium herbaceum subspecies which are genetically similar to each other and subsp. frutescens. G. mustelinum and three interspecific allohexaploid hybrids F₁C (G. herbaceum subsp. frutescens × G. mustelinum), F₁C (G. herbaceum subsp. pseudoarboreum f. harga × G. mustelinum) and F₁C (G. herbaceum subsp. africanum × G. mustelinum) were presented in the second group of the phylogenetic tree. Transgressive plants with long fiber were identified for potential improvement of fiber length in cultivated cottons. These synthetic allotetraploid cottons could also serve as valuable sources in the introgression of economically important traits including biotic and abiotic stress tolerance into the elite Upland cotton varieties.