Background

Cotton is the leading natural textile fiber crop in the world. Approximately 5 % of the world’s arable land is used for cotton planting, generating about $630.6 billion in 2011 [1]. Cotton belongs to the Gossypium genus of Malvaceae, which contains five tetraploid species (2n = 4× = 52, AADD genome) and approximately 45 diploid species (eight genomes from A to G and K, 2n = 2× = 26) [2]. Upland cotton (G. hirsutum) is the most widely cultivated species and its production accounts for over 95% of the world’s cotton production [3]. During the development of its cultivars, cotton has been subjected to long-term artificial selection, which narrowed its genetic base and gave rise to several difficulties in breeding. Cotton breeders face a scarcity of genetically diverse resources, therefore expanding the genetic base of cotton cultivars is imperative. Wild or untapped species have many excellent characteristics and contain abundant desirable genes, which have yet to be unlocked by pre-breeding. G. anomalum (2n = 2× = 26, BB genome) which is native to Africa, mainly Angola and Namibia [2], has the favorable characteristics of drought tolerance and resistance to diseases (cotton wilt, angular leaf spot) and insect pests (springtails, aphids): more importantly, it also possesses genes with the potential to produce high quality fibers (good fiber strength and fineness) [4] and cytoplasmic male sterility [5–7]. However, it is difficult to transfer these desirable genes into cultivated cotton through conventional breeding methods due to the isolation of wild species from cultivated species, which limits chromosome pairing and genetic recombination.

Monosomic alien addition lines (MAALs) contain only one alien chromosome in addition to the receptor background chromosomes. MAALs can be used as a bridge to transfer desired genes from wild species into G. hirsutum [8]. Over the past two decades, MAALs have been widely available for numerous crops [9], and these can be used for effectively identifying favorable genes in wild species, allowing for more accurate and faster transfer of such genes to create introgression lines, the effect of specific alien chromosomes to be examined, homeologies with chromosomes of cultivated species to be compared [10, 11], and physical maps of specific chromosomes to be constructed [12]. In cotton, however, the high number and smaller size of the chromosomes has resulted in difficulties in discriminating chromosomes from wild species in cultivated cotton background, therefore the development of cotton MAALs has lagged far behind many other crops. No set of cotton MAALs was reported until cotton molecular genetic maps were constructed and a genomic in situ hybridization (GISH) technique for cotton was developed. Previously, only one complete set of G. hirsutum-G. australe MAALs had been developed using simple sequence repeat (SSR) markers and GISH [9, 13, 14]. Two G. hirsutum-G. somalense MAALs and several G. hirsutum-G. sturtianum MAALs have also been obtained [11, 15].

In this study, the G. hirsutum-G. anomalum hexaploid was used as a maternal parent in the continuous backcrossing with upland cotton (recipient parent, G. hirsutum acc. TM-1), and eleven MAALs were isolated using GISH and SSR markers. These MAALs may be useful for mining and transferring favorable genes from G. anomalum into G. hirsutum on a genome-wide scale, mapping genes on chromosomes, analyzing genome structure and evolution, and micro-cloning for chromosome-specific library construction.

Results

Alien chromosomes from G. anomalum in G. hirsutum were examined by the improved GISH

The GISH technique used in our previous research was improved as follows. Genomic DNA extracted from G. anomalum and G. herbaceum was labeled with digoxigenin-11-dUTP and Bio-16-dUTP (Roche Diagnostics, Mannheim, Germany) by nick translation, respectively. The labeled DNA was mixed at a variety of ratios for GISH analysis using chromosomes from the mitotic metaphases as target templates. The results indicate that a ratio of 4:3 is the most suitable for discrimination of chromosomes from G. anomalum and the At-subgenome of G. hirsutum. At this ratio the chromosomes from G. anomalum only hybridized with the G. anomalum probe to produce a red signal, while chromosomes of the At-subgenome of G. hirsutum cross-hybridized with both the G. anomalum and G. herbaceum probes to produce a white signal and chromosomes of the Dt-subgenome of G. hirsutum were stained with 4’,6-diamidino-2-phenylindole (DAPI) (Roche Diagnostics), producing a blue color. Therefore, the GISH technique has been improved and can be further used to differentiate chromosomes from G. anomalum and the At-subgenome of G. hirsutum (Fig. 1).

Fig. 1
figure 1

Genomic in situ hybridization of the putative alien chromosomes of G. anomalum in the G. hirsutum background using two G. herbaceum and G. anomalum probes. Genomic DNA from G. anomalum and G. herbaceum was labeled with digoxigenin-11-dUTP and Bio-16-dUTP by nick translation, respectively. Chromosomes of the At-subgenome of G. hirsutum were cross-hybridized with both the G. anomalum and G. herbaceum probes and produced white signals and chromosomes of the Dt-subgenome of G. hirsutum were stained with 4′,6-diamidino-2-phenylindole (DAPI) and produced blue signals. Chromosomes from G. anomalum were hybridized with G. anomalum probe and produced red signals. a mitotic chromosome spread of the 52 chromosomes of G. hirsutum. b mitotic chromosome spread of the 26 chromosomes of G. anomalum. c–l mitotic chromosome spread showing the 52 G. hirsutum (white and blue) chromosomes and three (c), two (d), and one (e, f, g, h, i, j, k and l) individual chromosomes of G. anomalum (red), respectively. Scale bar = 5μm

Full size image

Progenies of the pentaploid of (G. hirsutum × G. anomalum) × G. hirsutum backcrossed with G. hirsutum were subjected to GISH to determine the number of alien chromosomes transferred from G. anomalum to G. hirsutum using visible fluorescent hybridization signals. Thirty eight individuals of the BC1 population were examined by GISH analysis (Additional file 1: Table S1). The analysis demonstrated that 27 (71.05 %) carried 2 to 6 alien chromosomes, and 6 (15.79 %) carried 7 to 9 alien chromosomes. Only two (5.26 %) individuals carried one chromosome, 6Ba and 13Ba of G. anomalum, resepctively. One (2.63 %) plant had no alien chromosomes and the final two (5.26 %) plants had 13 alien chromosomes from G. anomalum (Fig. 1; Table 1).

Table 1 Incidence of alien chromosomes in the BC1 to BC2 G. hirsutum × G. anomalum generations
Full size table

A total of 290 individuals from the BC2 generation were further analyzed by GISH. The results indicated that 106 (36.55 %) individuals had one alien chromosome of G. anomalum and 121 (41.72 %) had no alien chromosomes in the G. hirsutum background. 50 (17.24 %) and 10 (3.45 %) individuals carried two and three alien chromosomes, respectively, and another 1 (0.34 %) carried four alien chromosomes. The results demonstrated that most of the BC2 individuals carried 0-1 alien chromosomes, and only a small number contained multiple alien chromosomes (Fig. 1; Table 1).

Screening of a set of putative G. anomalum chromosome-specific SSR primer pairs

During the evolution of Gossypium, chromosomal translocations occurred between genomes A1, A2, and B1, while genome D remained relatively stable [16]. Numerous recent reports also show that translocations occurred between chromosomes in the At-subgenome of the tetraploids [17], while no large structural variation was found in the Dt-subgenome. Therefore, we only selected SSR primers from the Dt -subgenome of the tetraploid cotton linkage map to screen putative G. anomalum chromosome-specific SSR primer pairs. Of the 1402 pairs of primers we selected, 1072 amplified distinct fragments in G. hirsutum and G. anomalum, including 272 dominant markers of G. hirsutum, 194 dominant markers of G. anomalum and 452 codominant markers, while 154 pairs produced no amplified polymorphic bands and another 330 pairs produced vague bands, which were excluded from further study. Then, based on the tetraploid cotton linkage map constructed by our institute [17], the above 452 codominant markers were located, and of these, 170 well-amplified and evenly distributed codominant markers within an interval of 10 cM were finally selected for use in genotyping the entire BC1F1 and BC2F1 population. The 170 codominant markers were distributed on the Dt-subgenome chromosomes, ranging from 10 to 18 markers per chromosome, with coverage of 80.9–100.0 % and a density of 6.7–15.0 cM of each chromosome (Table 2; Fig. 2). The G. anomalum-specific SSR markers could be used to track and identify the alien chromosomes from G. anomalum in G. hirsutum.

Table 2 SSR primers used for screening G. anomalum chromosomes in the alien addition lines
Full size table
Fig. 2
figure 2

Genetic linkage map of G. anomalum chromosome-specific SSR markers based on the linkage map of tetraploid cotton reported by Zhao et al. (2012)

Full size image

Identity of alien chromosomes from G. anomalum as discriminated by SSR analysis

One hundred seventy G. hirsutum- and G. anomalum-specific codominant markers distributed on 13 Dt-subgenome chromosomes of the tetraploids were used to identify the alien chromosomes in 108 MAALs and multiple alien addition lines. The results demonstrated that 34 (31.48 %) MAAL individuals were MAAL-10Ba (the largest group), followed by 17 (15.74 %) MAAL-4Ba, 16 (14.81 %) MAAL-6Ba, 11 (10.19 %) MAAL-13Ba, 10 (9.26 %) MAAL-2Ba, 7 (6.48 %) MAAL-12Ba, 3 (2.78 %) MAAL-7Ba, 2 (1.85 %) MAAL-11Ba, 1 (0.93 %) MAAL-3Ba, and 1 (0.93 %) MAAL-9Ba (Figs. 3 and 4; Table 1). Two MAALs were not found, MAAL-1Ba and MAAL-5Ba; therefore Chrs. 1Ba and 5Ba were not isolated and remained as multiple addition lines.

Fig. 3
figure 3

Genomic in situ hybridization of the putative monosomic alien chromosomes of G. anomalum in the G. hirsutum background using G. herbaceum and G. anomalum probes. a mitotic chromosome spread of the 52 chromosomes of G. hirsutum, showing 26 chromosomes each of the At- (white) and Dt- (blue) subgenomes. b-l mitotic chromosome spread showing the 52 G. hirsutum (white and blue) chromosomes and different individual chromosomes from G. anomalum (red), corresponding to 2Ba to 4Ga (b, c and d) and 6Ga to 13Ga (e, f, g, h, i, j, k and l), respectively. Scale bar = 5μm

Full size image
Fig. 4
figure 4

A set of G. anomalum-specific SSR markers were used to identify individual alien chromosomes of G. anomalum in G. hirsutum. a-k the G. anomalum-specific amplicons were obtained using 11 individual chromosome-specific primer pairs for markers; NAU5421, BNL2443, NAU7579, NAU3677, dPL0492, BNL2597, BNL3383, NAU4881, NAU9520, dPL0379, and dPL0864. The chromosomes correspond to D2 to D4 and D6 to D13 in cultivated tetraploid cotton. P1, G. hirsutum; P2, G. anomalum; F1, the hexaploid of G. hirsutum and G. anomalum; 1-11 show that each of these plants possesses a single different individual chromosome from G. anomalum, corresponding to 2Ba to 4Ba, and 6Ba to 13Ba. M, molecular size marker (50 bp ladder). Arrows (red) indicate chromosome-specific markers for G. anomalum

Full size image

During the development of MAALs, Chr. 10Ba appeared most frequently, with an incidence of 16.87 %, followed by 15.89 % for 4Ba, 12.47 % for 6Ba, and 9.29 % for 12Ba. Chrs. 5Ba and 9Ba showed very low incidences of 2.69 % and 2.44 %(Table 1).

Morphological traits of MAALs

Morphological data were gathered during the cotton growing stage. The results shown in Tables 3, 4 and 5 indicate that the eleven MAALs differed from one another and also differed from their parents in terms of their morphological traits, such as plant type, leaf shape, size of flower and boll (Figs. 5 and 6; Tables 3, 4 and 5). Most of these MAALs grew slower than the recipient, TM-1. We found that MAAL-8Ba leaves had a very dark green color. We also observed that MAAL-7Ba, MAAL-12Ba and MAAL-13Ba had relatively bigger leaves, while MAAL-8Ba, MAAL-9Ba and MAAL-10Ba had relatively smaller leaves than the other lines (Fig. 5b). In addition, MAAL-6Ba, MAAL-10Ba, MAAL-11Ba and MAAL-12Ba had relatively larger flowers than the others. Only MAAL-7Ba showed petal spots and MAAL-6Ba had very light brown fibers, indicating that genes for petal spots and light brown fibers are located on chromosomes 7Ba and 6Ba (Figs. 5a and 6d), respectively. MAAL-2Ba and MAAL-12Ba had relatively longer bolls and MAAL-7Ba had the widest boll diameter, while MAAL-8Ba had the shortest bolls and MAAL-10Ba had the smallest boll diameter (Fig. 6c). MAAL-6Ba, MAAL-7Ba and MAAL-9Ba had a relatively larger boll weight, while MAAL-8Ba, MAAL-10Ba and MAAL-11Ba had a relatively smaller boll weight than the others (Table 4). We found that MAAL-7Ba had longer fibers than the others (Fig.6d)

Table 3 Morphological characteristics of the eleven MAALs
Full size table
Table 4 The yield-related traits of the eleven MAALs
Full size table
Table 5 Summary of the unique traits of the monosomic alien addition lines
Full size table
Fig. 5
figure 5

Flower and leaf traits for MAALs of G. anomalum individual chromosomes in G. hirsutum. Flower-related traits were photoed on the flowering day (0 day post anthesis, 0 DPA). a (petal), b (top third leaf) and (c) (bract); P1, G. hirsutum. P2, G. anomalum. F1, the hexaploid of G. hirsutum and G. anomalum. 2–4 and 6–13 are plants that carried a single different individual chromosome from G. anomalum, corresponding to 2Ba, 3Ba, 4Ba, 6Ba, 7Ba, 8Ba, 9Ba, 10Ba, 11Ba, 12Ba and 13Ba. Scale bar = 50 mm

Full size image
Fig. 6
figure 6

Flower, boll and fiber traits of MAALs of G. anomalum individual chromosomes in G. hirsutum. Squares, pistils and bolls were photoed at -1 DPA, 0 DPA and 35 DPA, respectively. a (square), b (pistil), c (boll) and d (fiber); P1, G. hirsutum. P2, G. anomalum. F1, the hexaploid of G. hirsutum and G. anomalum. 2–4 and 6–13 are plants that carried a single individual chromosome from G. anomalum, corresponding to 2Ba, 3Ba, 4Ba, 6Ba, 7Ba, 8Ba, 9Ba, 10Ba, 11Ba, 12Ba and 13Ba. Scale bar = 50 mm

Full size image

Discussion

MAALs are powerful tools in crop breeding since they can be used to produce alien translocation and substitution lines, to study interspecific relationships, and to construct single chromosome libraries. They can also be used in gene mining, gene assignment, gene expression pattern analysis, gene function analysis, physical gene mapping, isolation of chromosome-specific probes, selective isolation and mapping of cDNA/gene of a particular chromosome. Numerous reports have shown that the development of MAALs has been successfully achieved in many crops such as wheat [18–21], rice [22] tomato [23], potato [24], cucumber [25], tobacco [26], oat [12], sugar beet [27, 28], and rapeseed [29, 30]. MAALs have played and are playing important roles in numerous types of plant genomic research. The development of MAALs in Gossypium began as early as the 1980s but greatly lagged behind other crops due to the large number (2n = 52) and small size of chromosomes, which led to difficulty in accurately discriminating each chromosome, therefore, little progress has been made in cotton. So far only one set of MAALs has been completed [9], and this work benefited from advances in the development of GISH and molecular markers in cotton.

However, in this study, due to the very close relationship between chromosomes of the At-subgenome in G. hirsutum and those in G. anomalum often leading to cross-hybridization in GISH, we had to first improve the GISH technique by adjusting the ratio of the two different probes used. We tried five different combinations and found that the ratio of 4:3 was more suitable than any others for the discrimination of chromosomes from G. anomalum and the At-subgenome of G. hirsutum. Therefore, using a combination of the improved GISH methodology, G. anomalum chromosome-specific SSR molecular markers and conventional morphological survey, eleven MAALs were isolated and characterized, and two remain to be isolated from multiple addition states by further backcrossing.

Several previous reports showed that G. anomalum contains the favorable characteristics of drought tolerance and resistance to diseases (cotton Verticillium wilt, angular leaf spot) and insect pests (springtails, aphids); and more importantly, it also possesses genes with the potential to produce high quality fibers (good fiber strength and fineness) [4] and cytoplasmic male sterility [5–7]. Our previous reports also demonstrated that using G. anomalum as a donor parent and G. hirsutum as a recipient parent, a series of introgression lines with longer, stronger and finer fibers has been developed [31]. Shen et al. [32] mapped QTLs on Chr. 7 affecting fiber length in an F2 population derived from G. anomalum introgression line 7235 crossed with TM-1. However, in this study, we investigated some agronomic traits of MAALs and observed that most MAALs had poor performances in fiber quality or fiber yield components, implying that the added alien chromosomes had negative effects on most agronomic traits (Tables 4 and 6; Fig. 6). For example, the bolls of all MAALs were lighter than those of the recipient TM-1; and the fibers of all six MAALs were shorter than TM-1 (the fiber properties of the other five MAALs were not measured due to a lack of fiber samples). The resultant phenomena may be caused by linkage drag, which means that there were very close linkages between favorable and unfavorable genes on the same chromosome, even though the fibers of some MAALs were found to be stronger than those of TM-1. Therefore, to enhance the transfer of desirable genes and eliminate undesirable genes from G. anomalum, it is necessary to break the linkage drags to promote chromosome recombination between G. hirsutum and G. anomalum. The development of chromosome translocation lines or introgression lines may be an alternative choice based on the MAALs. We deeply believe that these MAALs of G. hirsutum-G. anomalum would be a powerful tool for systematically transferring desirable genes chromosome by chromosome from G. anomalum into G. hirsutum, as well as for gene mining, gene assignment, gene function analysis, gene physical mapping, isolation of chromosome-specific probes, selective isolation and mapping of cDNAs for a particular chromosome, and genomic research.

Table 6 Fiber quality traits from some MAALs measured by HVI
Full size table

Conclusions

From this study, we draw two conclusions. (1) The GISH technique used in our previous research has been improved by using a mixture of two probes at a ratio of 4:3 (G. anomalum and G. herbaceum) to avoid cross-hybridization caused by the very close relationship between chromosomes from G. anomalum and the At-subgenome of G. hirsutum, which can be suitable for recognizing alien chromosomes of G. anomalum in G. hirsutum background. (2) Eleven out of 13 potential MAALs were isolated, which would be used, at the chromosome level, for effectively identifying favorable genes in G. anomalum, allowing for more accurate and faster transfer of such genes to create introgression lines, the effect of specific alien chromosomes to be examined, homeologies with chromosomes of cultivated species to be compared, and physical maps of specific chromosomes to be constructed.

Methods

Plant materials

In 2012, the amphiploid (allohexaploid) (2n = 6× = 78, AADDBB genome) (previously obtained in our institue) derived from the doubled triploid hybrid of G. hirsutum (2n = 4× = 52, AADD genome) × G. anomalum (2n = 2× = 26, BB genome, obtained from Cotton Research Institute of Chinese Academy of Agricultural Sciences) was backcrossed as a maternal parent with G. hirsutum acc TM-1, the genetic standard line of upland cotton. In 2013, two pentaploid individuals were obtained at Pailou Experimental Station of Nanjing Agricultural University (PES/NJAU) and used as both paternal and maternal parents in the backcross with TM-1. The BC1 seeds obtained were planted in plastic cups with sterilized soil and incubated in the phytotron at Nanjing Agricultural University in 2014 spring at 25–28 °C and with 80% relative humidity. When they reached the fifth true leaf stage, the seedlings were transplanted into clay pots at PES/NJAU. Lastly, 38 BC1 individuals were identified using SSR markers and GISH and consecutively backcrossed with TM-1. The BC2 seeds obtained were planted in the same way in spring 2015. In the winter, all plants were moved into the greenhouse at PES for preservation.

Scheme for developing the monosomic alien addition lines

The interspecific hexaploid was backcrossed with Gossypium hirsutum acc TM-1 (obtained from the Southern Plains Agricultural Research Center, USDA-ARS) to produce the pentaploid (2n = 5× = 65, AADDB genome), then the pentaploid progenies were further consecutively backcrossed with TM-1 to generate backcross progenies (BC1 and BC2). GISH was used to characterize alien chromosomes in all backcross progenies from the BC1 generation. When more than one alien chromosome was added from G. anomalum, the progenies were further backcrossed with TM-1 to produce monosomic alien addition lines. If only one alien chromosome was added to the background of Upland cotton, the progenies were further examined using chromosome-specific SSR markers of G. anomalum to determine the identity of the added chromosome.

G. anomalum, TM-1, BC1, and BC2 chromosome preparation

Cotton seeds were cultivated in an incubator at 29 °C and their root tips were cut off when they grew to 3 cm long (seedling plant). The tips were immersed in 25 μg/ml cycloheximide at room temperature for 2 h to accumulate metaphase cells and then transferred to Carnoy I fixative containing 95% ethanol and acetic acid (3:1, v/v) for at least 2 h, digested in double enzymolysis liquid (4 % cellulose: 1 % pectinase = 1:2) at 37 °C for 45 min, and squashed in a drop of 45 % acetic acid. Finally, slides containing at least 20 well-spread somatic chromosomes at mitotic metaphase were prepared and stored at -70 °C overnight.

Genomic in situ hybridization (GISH)

Due to the very close relationships that exist between chromosomes of the B genome in G. anomalum and those of the At subgenome in G. hirsutum, two probes were employed here to avoid cross-hybridization between these chromosomes. Genomic DNA extracted from G. anomalum and G. herbaceum (2n = 2× = 26, AA genome) were labeled with digoxigenin-11-dUTP and Bio-16-dUTP (Roche Diagnostics, Mannheim, Germany) by nick translation, respectively. The probe fragment size was between 200-500 bp. Fluorescence in situ hybridization was carried out as described by [33] and [9] with some modifications. The mixing ratio of DNA probes from G. anomalum and G. herbaceum were adjusted to five different ratios, 2:1, 4:3, 1:1, 2:3, and 1:2, to determine the optimal ratio for discrimination of chromosomes from G. anomalum and the At-subgenome of G. hirsutum.

DNA extraction and G. anomalum-specific primer screening

Genomic DNA was extracted from young leaves of the two parents, G. anomalum and G. hirsutum acc. TM-1, the interspecific hexaploid, the pentaploid, and the BC1 and BC2 individuals using the method described by [34] with some modifications. A total of 2,168 pairs of SSR primers were selected from the high density genetic linkage map of Sea island and Upland cotton constructed in our institute [17] and employed to screen G. anomalum-specific primers. PCR reactions were performed and their amplified products were separated by PAGE, as described by [35, 36]. The G. anomalum-specific marker primers obtained were further used to characterize each chromosome from G. anomalum.

MAAL nomenclature

Thirteen G. hirsutum-G. anomalum MAALs were named MAAL-1Ba to MAAL-13Ba, according to the method described by [9], in which B represents the B genome of G. anomalum and ‘a’ refers to the initial letter of anomalum. The chromosome numbers 1 to 13 in the B genome of G. anomalum correspond to the homoeologous chromosomes in the Dt-subgenome of tetraploid cotton.

Investigation of agronomic traits of monosomic alien addition line

At the point of transition from the vegetative to the reproductive stage, the shape and size of fully expanded leaves from the same position in TM-1, G. anomalum, hexaploid and MAAL plants were investigated. Floral morphological traits from these MAALs were investigated in the flowering period. The size of cotton bolls at 35 days post-anthesis was also measured by vernier caliper. Finally, the hundred-seed weight, ginning outturn and single boll weight of the matured bolls were investigated. All the data were analyzed using the SPSS software version 18.0.