Somatic Embryogenesis and Plant Regeneration via Solid-liquid Alternating Culture in Elite Genotypes of Upland Cotton


 Background: Cotton is one of the most genotype-dependent plants for regeneration, in order to expand cotton regeneration genotypes and establish efficient regeneration system platform, Jiwu 2031 (glandless cotton), ND 58 and CAU 102, were selected for studying the highly efficient somatic embryos formation and regeneration via solid-liquid alternating culture system.Results: In present research, the MSB medium (MS salts adding B5 vitamins) containing 0.571 µM indole-3-acetic acid (IAA), 0.465 µM kinetin (KT) and 0.904 µM 2, 4-dichlorophenoxyacetic acid (2, 4-D) was effective for callus initiation of fourteen Upland cotton with ‘Corker 201’ as the control. ‘Jiwu 2031’, ‘ND 58’ and ‘CAU 102’ could form somatic embryos and regenerate fertile plants in a solid MSB medium containing 10.31 mM NH4NO3, and supplemented with 0.027 mM glycin, 2.460 µM indole-3-acetic acid (IBA), 0.930 µM KT, 3.784 mM asparagine and 6.843 mM glutamine. Under the solid-liquid alternating culture system, the non-embryonic callus was induced to form embryonic callus and the identical status somatic embryos were obtained in 42 days, and the period for plants regeneration was shortened to 90~120 days, with the higher proportion of normal plant regeneration.Conclusions: The solid-liquid alternating culture method could increase the rate of embryogenesis and shorten the period of plants regeneration of Upland cotton. This work provides the evidence that the glandless cotton is beneficial for somatic embryogenesis (SE) and plant regeneration.

Progresses have been achieved in Gossypium regeneration and a few cultivars of Upland cotton were reported to be able to regenerate e ciently in vitro, which were 'Coker 312,' 'Coker201', 'Ekangmian No. 9', 'Simian No. 3', 'YZ-1', etc (Trolinder and Goodin, 1987;Jin et al. 2006;Zhang et al. 2000; ). An elite cotton Jin668, with an extremely high regeneration ability, was developed from its maternal inbred Y668 cultivar using a Successive Regeneration Acclimation (SRA) strategy (Li et al. 2019). Different genotypes or explants showed variable results in the process of cotton SE and plant regeneration (Trolinder and Chen, 1989 Suspension culture was used in tissue culture for producing embryos and regeneration plants from 'Coker 312' (Trolinder and Goodin, 1987) and 'Coker 310' (Finer, 1988), and used as a platform for genetic transformation (Ke et al. 2012). A system with suspension culture for synchronization control of SE in 'Coker 312' was developed (Sun et al. 2005;Yang and Zhang, 2010), but the e ciencies of the system relied on inositol starvation stress, were short of identity and easily changed the metabolic conditions in the process of SE and development (Trolinder and Goodin, 1987;Kumar and Tuli, 2004).
In this study, fourteen Upland cotton varieties with desired agronomic traits were used to compare the regeneration capacity, including the status of callus and embryonic callus induction as well as proliferation rate. With the cultivar 'Coker 201' was used as control, an e cacious protocol for cotton SE and plant regeneration has been developed to regenerate plantlets from three commercial Upland cotton varieties in several months. An improved protocol based on solid-liquid alternating culture was used to improve SE frequency and synchronous development of mass somatic embryos.

Medium for induction of callus
Upland cotton cultivar 'Coker 201' was used to select the e ciency of callus induction medium. The explants began to swell at both ends of hypocotyls after cultured on MSB medium (Table 1) 5~7 days, the hypocotyls elongated and intumesced. A mass of callus produced 30~35 days. Induction capacity varied in medium (CIM01-CIM12) with different combinations and concentration of hormones. The colors of the callus diverged from fawn to green or yellow green, etc., and the textures were compact, loose, friable and hardy. The induction frequency of callus was higher in the medium supplemented with high concentration of 2, 4-D (0.904 µM) and higher ratio of 2, 4-D/KT concentration ( Table 1), indicated that 2, 4-D was an important hormone for the induction of callus.
IBA or IAA was employed to meliorate the callus state reported before (Wang et al. 2006). Callus could be improved to friable and loose texture in light yellow or green yellow color when the IAA (0.571 µM) or IBA (2.460 µM) were added to the medium containing 2, 4-D and KT, but higher frequency of callus induction was obtained using IAA than that of using IBA ( Table 1). The ratio of auxin: cytokinin should not be too high in order to reduce the formation of rhixoid.
Callus with loose and friable texture in yellow and light yellow color were easier to differentiate, this result was coincident with previous report (Han et al. 2009). The CIM06, CIM medium supplied with 2, 4-D (0.904 µM), IAA (0.571 µM) and KT (0.465 µM) (Table 1), generated high frequency callus with desirable color and texture, was con rmed to be the most suitable one for callus induction.

Callus induction among different genotypes of Upland cotton
Explants from fteen Upland cotton cultivars were used to induce callus in the selected medium CIM06, containing MS + 2, 4-D (0.904 µM) + IAA (0.571 µM) + KT (0.465 µM). No obvious difference was observed among these varieties in terms of the time of the callus initiation, color and texture of infant callus in the process of callus induction. The induction rate of loose callus varied from 40.22% to 100.00% (data not shown). The induction rate of 13 cultivars was as higher than 82%, which con rmed further that the medium CIM06 was suitable for callus induction, except induction rate of C1013 The fteen genotypes could be classi ed as four types based on the proliferation rates of callus (Fig. 1). Category I: Callus proliferated very slowly in dark gray color and soft and wet texture, such as Yumian No. 1, C1006 and C1013. Category II: Callus proliferated in general rate, which was in light yellow color, soft and wet texture. More than half of the genotypes, such as CAU102, were classi ed in the group. Category III: Callus of Luwu 401, Jiwu 2031, ND 58 and C1013 proliferated fast, and the callus is light yellow in friable and dry texture. Category IV: Callus of C1028 proliferated in a "crazy" mode with rm and hard texture in green color.

Embryogenic callus induction and somatic embryo formation
Callus of different genotypes were subcultured on medium with MSB1 (1.480 µM IBA, 0.465 µM KT), 6.843 mM glutamine and 3.784 mM asparagine to induce embryogenic callus. After one month subcultured on this medium, various callus differentiations were found among the 15 genotypes. The color, texture, number of callus proliferation and embryogenesis callus varied (Table 2), which can be classi ed into three types ( Fig. 2A-C). Category I: 4 varieties, Coker 201, ND 58, Jiwu 2031 and CAU 102 could differentiate into typical embryogenic callus in light yellow or gray color with friable or alveolate structure; category II: 9 varieties formed callus in brown or green color with the compact surface, but were hard to differentiate into embryogenic callus on this medium; category III: callus from C1028 proliferated in a "crazy" mode in dark green color with loose texture, and could not differentiate into embryogenic callus either. Embryogenic callus from the cultivar which had the ability of redifferentiation could produce somatic embryos after two to three times of subculture on this medium.

Development of somatic embryos and plant regeneration
Embryos formed from embryogenesis were transferred onto medium supplemented with seven combinations and concentrations of hormones plus 10.31 mM NH 4 NO 3 , 6.843 mM glutamine, and 3.784 mM asparagine for embryo development. Somatic embryos successfully developed into plants after subculturing on this medium for four to ve weeks, but the growth and development of the embryos were obviously different (Fig. 3). The germination rate of the embryos with normal root on the medium without hormones was 41%, indicating it was the best among the seven media used in embryo germination. showed reduction in somatic embryo dedifferentiation (28.9%) compared to that of MSB5 (34.5%), which indicated that the relatively lower concentrate of IBA would more proper for embryo development and germination.
Somatic embryos also developed from globular embryo, heart shaped embryo, torpedo shaped embryo, cotyledonary shaped embryo to plantlets (Fig. 2D) as previous report (Sun et al. 2009), and the development process was similar to that of zygotic embryo. After the regenerated plantlets grew four to six leaves and normal roots in the ask, all of the rooted plantlets were successfully transplanted to soil pots in green house. All plants recovered, owered and bore seed after two to three months in green house, the regenerated plantlets had normal fertility ( Fig. 2E-K). And these results also indicated that SE in Upland cotton would be more useful with low frequency of chimeras and higher proportion of regenerates.
The solid-liquid alternating culture increased inducing e ciency of embryogenetic callus and somatic embryos The explants of 'Coker 201', 'Jiwu 2031' , 'CAU 102' and 'ND 58' began to swell at both ends of hypocotyls after cultured on CIM (MS medium supplemented with 0.571 µM IAA, 0.465 µM KT and 0.452 µM 2, 4-D), and a mass of callus produced within 25~30 days (Fig. 4A, the abbreviation for Fig. 4A (a), 4B (a) and 4C (a), similar showed below). After subculture of the initial callus on the same medium for 14 to 28 days, the pre-embryogenetic callus was highly variable in color and texture. Typical colors of nonembryogenetic callus were green, light yellow, white, gray, and brown. The types of callus texture include loose, friable and compact.
For embryogenetic callus initiation and maintenance, pre-embryogenetic callus with loose or friable texture and in gray or light yellow ( Fig. 4(b)) were chosen to suspension in MS liquid medium and shake for ve minutes to disperse the callus. After removing the large particles with 30-mesh stainless steel sieves, the pre-embryogenetic callus was re-suspended in EIML (Fig. 6) and cultured for 14 days until the callus with identical color and texture formed. EIML contained halved NH 4 NO 3 , doubled KNO 3 , and supplied with 2.46 µM IBA+0.698 µM KT, not similar to the hormone-free medium used in establishing and maintaining embryogenic suspension cultures of cotton cultivar 'Coker 312' (Trolinder and Goodin, 1987). The callus in identical/similar light yellow and loose texture were chosen for further suspension culture in 50 mL fresh EIML (Fig. 4(c)). After 14~28 days' inoculation, the embryonic callusor embryogenetic tissues were found in suspension cultures. In this study, the induction frequency of embryogenesis callus by suspension culture was 72.14%, 52.86% and 61.62% in 'Coker 201', 'Jiwu 2031' and 'ND 58', respectively (Table 3), while that in solid culture for Coker 201 was 63.5% (Wu et al. 2004).
The induction time to form embryonic callus was shorted from 60~90 days (Wu et al. 2004) to 35~42 days.
For somatic embryo development/maturation, once the embryogenesis callus with similar/identical status of friable texture and yellow color were formed in suspension culture system, embryogenesis callus were ltered with 50-mesh stainless steel sieves, and subcultured onto EMMS (with halved NH 4 NO 3 , doubled KNO 3 , 2.460 µM IBA, 0.698 µM KT, 6.843 mM glutamine and 3.784 mM asparaginate) ( Fig. 4(d)). Within 28 days, large amounts embryos were obtained (Fig. 4(e)). At the initial stage of embryos, the percentage of global-embryos counts for 67.5% which indicated that the highly synchronous of embryos development. The embryos in identical status (global embryos) were selected to subculture on the solid medium EGM for germination ( Fig. 4(f)). Following placement of the mature embryos on EGM, hypocotyl and root elongation was rst observed in 3~7 days ( Fig. 4(g)). The percentages of embryos germination were 47.71%, 39.83% and 46.67% in 'Coker 201', 'Jiwu 2031' and 'ND 58', respectively (Table 3). Within 14 days, the majority of the embryos had undergone obvious elongation, and then the cotyledons developed, and plantlets regenerated with at least 2 euphylla in an additional month ( Fig. 4(g)).

Discussion
If the agronomic traits of the cultivars used as transgenic receptors are not desirable, and it will take years for breeders to improve the yield and ber qualities of the transformants through backcross. In this study, three elite cotton cultivars (CAU 102, Jiwu 2031 and ND 58) and Coker 201 (as a control) had been identi ed that have high e ciency of regeneration. Jiwu 2031, ND 58 and CAU 102 were commercial varieties with desired agronomic traits. Jiwu 2031 was a glandless variety that could be used for the production of cotton seed oil and protein. The genotype of Jiwu 2031 showed excellent quality and yield with the pre-frost yield up to 85%~90%, and had good characteristics of early maturity without premature senescence in eld comparison experiment. During the later period of adult stage, the Jiwu 2031 displayed strong resistance to stress and blight, and the damage of cotton aphid and leaf spot were light.
The successful regeneration of these elite genotypes provides better choices of cotton cultivars for transformation and great potential of expediting process of transgenic breeding. It seems that only 3 months for the control Coker 201 to regenerate, four to ve months for other 3 cultivars using the protocol developed here took longer time (Fig. 5). But considering the time consumed in many backcrossing generations two or three months is really a short time. The results of this research indicated that 2, 4-D was useful for callus induction, which was consistent with the results of previous reports (Zhang, 1991;Wang et al. 2006).

High concentration of exogenous hormones might result in undesirable effects for callus development,
such as the formation of rhizoid, due to the disequilibrium between endogenous and exogenous hormones. In this study, we used seven hormone combinations to optimize the medium with appropriate asparagine and glutamine amounts for embryo germination (Fig. 3 The rate of callus proliferation, which was considered as a useful index in cotton callogenesis, has signi cant difference among genotypes. That normally proliferated callus could advance to embryogenic callus phase and eventually regenerate plantlets, suggesting that the callus proliferated slowly or massively should not be chosen for cotton regeneration. Obvious genotype-dependent differed in embryogenic callus induction, Jiwu 2031 had the highest differentiation e ciency. The low frequency and time-consuming of embryogenetic callus and somatic embryos induction are the main restrict factor during the process of tissue culture of Upland cotton. The liquid suspension medium can high-e ciency promote callus to produced SE and maturation compared to semi solid medium (Juturu et al. 2015). In this protocol with solid-liquid alternating culture, the time to induce and form embryonic callus was reduced from 60~90 days to around 42 days, and the percentage of embryogenesis callus induction was reached as high as 72.14% in 'Coker201'. Within 14 days, the embryo could be formed from the embryonic callus with identical color and texture, and the percentages of embryos germination on solid medium were over 39.00% in all three genotypes. Besides, the ratio of regeneration seedlings in normal development were as high as 17.53%, 18.04% and 18.29% in 'Coker 201', 'Jiwu2031' and 'ND58', respectively, which reduced the high rate of abnormal seedlings during embryo development. And the time for plants regeneration was only cost 90~120 days (Fig. 6), much shorter than 120~150 days cost in the previous method (Sakhanokho et al. 2001;Wu et al. 2004).

Conclusions
This solid-liquid alternating culture protocol provided an e cient method for inducing of embryonic callus and embryos, not only for 'Coker' lines, but also for other excellent genotypes, such as 'Jiwu 2031' and 'ND 58'. And we also found there was a high synchronism for embryos development, which provided an ideal platform for physiological, biochemical, and molecular biological studies of embryogenesis processes in Upland cotton. This method raised the frequency of normal seedlings and shorted the time of plant regeneration in cotton tissue culture, might have the great potential to be used as a platform for genetic transformation.

Plant materials
Fourteen Upland cotton cultivars with desired agronomic trait (data not shown) were used in this study. A Seed sterilization and production of sterile seedlings Matured seeds were delinted with sulfuric acid (98%, v/v), immersed in 70% alcohol for 30 s and shaken in 0.1% HgCl 2 with for 10 min, then rinsed with sterile distilled water for ve times, and soaked in sterile water at 25~30ºC for 48 hours. Seeds coats were removed before the seeds placing on half strength macronutrient MSB medium comprising water, 0.65% (w/v) agar (Amersico) and 2% (w/v) sucrose in dark at 28±2ºC for 5~7 days to germinate (Sakhanokho et al. 2001).

Initiation and proliferation of callus
Hypocotyls from sterile seedlings of 'Coker 201' were cut into 5-8 mm segments in length, and served as the explants to be placed on MSB medium supplemented with twelve combinations of 2,4dichlorophenoxyacetic acid (2,4-D), IAA (3-indoleacetic acid), indole-3-butyric acid (IBA) and kinetin (KT) (CIM01 -CIM12, Table 1) in order to induce the callus. The medium was selected based on 1) the frequency, and 2) recorded status of callus induced, including the emerging time, texture, color, and proliferation rate of callus (g/d) (Fig. 1). The callus proliferated rates were calculated by net weight of proliferated callus/days of subculture. The soft or friable texture callus in grey and yellow were chosen to subculture on MSB medium contained 10.31 mM NH 4 NO 3 , 1.504 M KNO 3 , and supplemented with 1.48 µM IBA and 0.465 µM KT.

Induction of embryogenic callus and somatic embryos
After subcultured once or twice, the proli c and loose non-embryogenic callus were placed to induce embryogenic callus on above MSB medium but added 6.843 mM glutamine and 3.784 mM asparagine. The embryogenic calli with high proliferation rate were transferred onto same medium to induce somatic embryos (Fig. 2). Somatic embryos were transferred to seven kinds of MSB media contained 10.31 mM NH 4 NO 3 , 1.504 M KNO 3 , and supplemented with 2, 4-D, IAA, IBA and KT at different hormone combinations (see details in Fig. 3), along with 6.843 mM glutamine, 3.784 mM asparagine for embryo growth. The status of embryo development and the ratio of different types of embryo were recorded.

Plant regeneration, plantlets recovery and transplantation
The regenerated normal plantlets were transferred onto the same media for root induction. Apart from the half strength macronutrient MSB medium, all above media were supplied with 3% glucose and 0.27% phytagel, and adjusted the pH to 5.95 before being sterilized by autoclaving at 115ºC for 20 min. All of the culture conditions were maintained at 28±2ºC under a 16 h photoperiod with a light intensity of 1600LX. Plantlets with strong roots were challenged by unsealing the erlenmeyer ask for 2~3 days, and then were transplanted to soil pot and grown in the greenhouse. The fertility of the regeneration plantlets was surveyed when owering and the mature seeds were harvested.

Solid-liquid alternating culture
Hypocotyls from sterile seedlings of 'Coker 201', 'Jiwu2031' and 'ND58' were cut into 5~8 mm segments in length and served as the explants to be placed on callus induction medium (CIM, MS salts supplemented with 0.571 µM IAA, 0.465 µM KT and 0.452 µM 2, 4-D, pH=5.85) (Fig. 6). Explants were transferred to Petri dishes containing CIM under light intensity of 1600LX and a 16 h/8 h light/dark cycle at 28±2°C for the callus induction and proliferation.
After 28 days, pre-embryogenic callus with soft or friable texture were transferred to suspend in MS liquid medium and shake (120 rpm) for 5 minutes. The large particles were removed with 30-mesh stainless steel sieves, and then the pre-embryogenic callus were suspended, and cultured in 250 mL Erlenmeyer asks containing 100 mL somatic embryo initiation liquid medium (EIML, MS salts with halved NH 4  After cultured for 14 days, further subculturing involved the replacement of 50 mL of the cell suspension with fresh EIML medium at 7-d interval, and suspensions were ltered through 50-mesh stainless steel sieves to separate the bigger callus and small cell clumps. The separated callus was re-suspended in fresh EIML medium and cultured for 2-4 intervals (7 days for one interval). Samples from the suspension callus were observed under a microscope to decide subculturing period when the right embryonic callus obtained.
The cultured cells were screened with 50-mesh stainless steel sieves to remove the small cell clumps and then embryonic callus with similar status of texture and color were harvested. The embryonic callus with identical loose and friable status were cultured for embryos initiation on solid somatic embryo development/maturation medium (EMMS, pH=5.85), which consisted of the same composition as EIML (MS salts with halved NH 4 NO 3 , doubled KNO 3 , 2.460 µM IBA, 0.698 µM KT), added 6.843 mM glutamine and 3.784 mM asparaginate. Mature embryos, which were obtained within 28 days on the solid medium, were placed on embryo germination medium (EGM, pH=5.85), which was a modi cation of the medium used by Davidonis and Hamilton (1983), and contained modi ed MS salts (no NH 4 NO 3 and 2× KNO 3 ), 6.843 mM glutamine, 3.784 mM asparaginate, 3% glucose, and 0.27% phytagel.

Statistical analysis
The experimental design was completely randomized with three replications per treatment, each treatment consisting of 24 explants or about 3 g of calluses, or 60 somatic embryos for germination. The induction rate of callus was calculated by weight of callus proliferated each day, and induction rates of the loose callus were calculated by the number of hypocotyls with loose calluses/the total number of hypocotyls per treatments. *induction rate was calculated by weight of callus proliferated each day, and detected signi cance at P = 0.05. **Values were the mean of ten replicates, each replicate contains 8 explants. Table 2 The status of embryogenic callus of different cotton genotypes The callus proliferated rates of 15 different genotypes. The callus was subcultured for four times. Values were the means of three replicates, each from the data of callus in 10 bottles. X axis is the different genotypes, and Y axis is the proliferated rates of callus.     Schematic protocol of solid-liquid alternating culture and plant regeneration in Upland cotton.