An improved embryo-rescue protocol for hybrid progeny from seedless Vitis vinifera grapes × wild Chinese Vitis species

A highly efficient technique of embryo rescue is critical when using stenospermocarpic Vitis vinifera cultivars (female parents) to breed novel, disease-resistant, seedless grape cultivars by hybridizing with wild Chinese Vitis species (male parents) having many disease-resistance alleles. The effects of various factors on the improvement of embryo formation, germination, and plantlet development for seven hybrid combinations were studied. The results indicated that Beichun and Shuangyou were the best male parents. The best sampling time for ovule inoculation differed among the female parents. When hybrid ovules were cultured on a double-phase medium with five different solid medium types, percent embryo formation was highest (11.3–28.3%) on a modified MM3 medium. Percentages of embryo germination (15.4–55.4%) and plantlet development (11.15–44.6%) were all highest when embryos were cultured on Woody Plant Medium + 5.7 μM indole-3-acetic acid + 4.4 μM 6-benzylaminopurine + 1.4 μM gibberellic acid + 2% sucrose + 0.05% casein hydrolysate + 0.3% activated charcoal + 0.7% agar. In the absence of other amino acids, the addition of proline significantly increased embryo formation (36.1%), embryo germination (64.6%), and plantlet development (90.5%). A highly efficient protocol has been developed for hybrid embryo rescue from seedless V. vinifera grapes × wild Chinese Vitis species that results in a significant improvement in breeding efficiency for new disease-resistant seedless grapes.


Introduction
Seedless grapes are generally preferred by consumers for both table fruit and raisins. The quality and nutritional composition of seedless grapes and the breeding of new seedless cultivars have been subjects of significant research over many years (Alleweldt and Possingham 1988;Wang et al. 2002;Ebadi et al. 2009). Many breeders have obtained new, seedless grape cultivars from seedless female parents using embryo-rescue techniques (Ramming and Emershad 1982;Spiegel-Roy et al. 1985;Gribaudo et al. 1993).
China is an important center of origin for Vitis species, having many wild resources that are naturally resistant to fungal diseases. Usefully, this resistance is generally heritable (Wang et al. 1998;Wan et al. 2008). Moreover, many of these wild species are relatively easy to hybridize with V. vinifera as they have the same number of chromosomes (2n=38; Tian et al. 2008). Using the wild Chinese Vitis species as the male parents and seedless V. vinifera cultivars as the female parents, it is possible to obtain potentially valuable new hybrid progeny exhibiting both seedlessness and significantly improved disease resistance.
An efficient embryo-rescue technique is, however, critical for breeding seedless grapes when using stenospermocarpic female parents. An embryo-rescue procedure involves three main steps: (1) ovules are cultured in vitro (embryo formation), (2) embryos are removed from the ovules and cultured (embryo germination and plantlet development), and (3) plantlet roots are elongated, acclimated, and transplanted to soil (Pan 2005;Tang 2010; Wang et al. 2010). Of these three steps, the first, in vitro ovule culture, is the most important. An optimal in vitro ovule culture medium is key to the successful rescue of these potentially abortive hybrid embryos.
The growth regulator composition of the culture medium is a critical factor for embryo germination and plantlet development. Gibberellic acid (GA 3 ) and indole-3-acetic acid (IAA) are most often used to improve embryo rescue rates. For example, Liu et al. (2003) found ovule growth and embryo production in vitro were improved using Bouquet and Davis, and Nitsch and Nitsch media. Supplementation with GA 3 increased embryo recovery rates, and the highest rates achieved were 18. 1%, 9.6%, and 12.2% for Sunmuscat, Merbein Seedless, and Marroo Seedless, respectively. Spiegel-Roy et al. (1985) reported that with stenospermic grapes, exogenous GA 3 in combination with IAA in the medium greatly improved ovule embryo culture. Gribaudo et al. (1993) also confirmed the positive effect of plant growth regulators on embryo formation. At the embryo germination stage, Woody Plant Medium (WPM) is usually used as the basal medium because of its beneficial effect on embryo germination and plantlet development (Emershad and Ramming 1994;Tian et al. 2008;Wang et al. 2010).
The aim of this study was to adapt embryo-rescue methods for use with hybrids between seedless cultivars of V. vinifera×wild Chinese Vitis species by defining optimal conditions, including the embryo rescue time, the ovule culture medium, the plant growth regulator content, and other key factors. The efficacy of the newly adapted technique as a tool for breeding new, disease-resistant, seedless grapes was demonstrated by producing a number of hybrid cultivars.

Materials and Methods
Plant materials. All the parent vines used in this study were more than 5 yr old and grown in the vineyards of the Northwest A&F University, in Yangling, Shaanxi, with a 'T' single vertical trellis with 1.5 m spacing within the row and 2.5 m between rows. The seven male parents were Beichun, Yanshan, Xuefeng, Tangwei, Shuangyou, Flame Seedless, and Red Globe (see Table 1 for details). The four female parents were Thompson Seedless, Flame Seedless, Ruby Seedless, and Crimson Seedless (see Table 2 for details). Optimal embryo rescue conditions were determined for the seven hybrid combinations: Thompson Seedless× Shuangyou (C1), Thompson Seedless×Tangwei (C2), Flame Seedless×Yanshan (C3), Flame Seedless×Red Globe (C4), Crimson Seedless×Flame Seedless (C5), Ruby Seedless× Beichun (C6), and Ruby Seedless×Xuefeng (C7).
Pollen assays. Healthy flower clusters were collected from the seven male parents (Shuangyou, Tangwei, Yanshan, Red Globe, Flame Seedless, Beichun, and Xuefeng) during the initial bloom stage and which opened for the first time between 8 and 10 am each morning. Clusters were dried at room temperature (20-25°C) on a smooth paper surface. The dried material was ground and sieved (three times) to remove plant debris using a single layer of gauze, before being collected into a vial which was film sealed and placed in a 500-mL plastic bottle with 200 mg CaCl 2 (desiccant), capped and stored at −80°C pending use. Pollen samples from the seven male parents were cultured at 25°C, 100% relative humidity and dim white light (30 μmol m −2 s −1 ; Fan et al. 2001) on the culture medium [15% (w/v) sucrose, 0.7% (w/v) agar, 0.01% (w/v) H 3 BO 3 , pH 6.0]. The pollen grain culture was examined after 24 h for germination and measurement of pollen tube lengths. A pollen grain was considered to have germinated when the length of the pollen tube exceeded the grain diameter after 24 h of incubation. To quantify the germination rate, about 100 pollen grains per field were screened at ×40 magnification with an optical microscope (Olympus SZ-DT equipped with a calibrated micrometer in a ×10 ocular; Olympus, Japan). The lengths of 100 pollen tubes were also measured (30 tubes randomly selected from each of three replicate culture plates). Percent pollen germination (percentage)=number of germinated grains/number of grains present×100. Germination percentages and pollen tube lengths were analyzed using an IBM statistical software package, SPSS software, version 13.0 (SPSS Inc., Chicago, IL) using oneway ANOVA.
Artificial pollination employed a soft brush loaded with dry pollen and was performed when the stigma mucus of the emasculated female parents was highest (Fig. 1a, b).
Pollination was repeated three times on each of three consecutive days, followed by bagging of the whole inflorescence.
Embryo rescue. Immature fruits were collected for several successive weeks after pollination. These fruits were surfacesterilized using 75% (v/v) ethyl alcohol for 1 min. Alcohol was poured off and replaced with 10% (w/v) sodium hypochlorite for 15 min and then washed three times with sterile water. Fruits were dissected and in vitro ovules were cultured in Erlenmeyer flasks (100 mL) containing a double-phase embryo formation medium (described below). After 8 wk culture, ovules were dissected and all developed embryos were excised from the micropylar end of the ovule. Each embryo was placed on solid WPM in an Erlenmeyer flask (250 mL). After a further 4-6 wk of culture, germinated embryos and developed plantlets were observed and recorded. All plantlets were transferred to optimal media composed of half-strength Murashige and Skoog (½MS;Murashige and Skoog 1962) salts plus 1.7 μM IAA, for root elongation in Erlenmeyer flasks (250 mL). All were grown in a culture room at 25±2°C, 100% relative humidity, and 40 μmol m −1 s −1 of white light (Fig. 1c-e).
Basal-medium effects on embryo formation. Three hundred ovules of each cross were cultured in 100 mL Erlenmeyer flasks (ten ovules per flask, 30 flasks per cross) on doublephase medium for embryo formation. Sampling time for each cross was based on the best sampling time for its female parent. The five solid medium types of the double-phase media examined were: (1) MS (Gray et al. 1990), (2) MM3 (Pan 2005), (3) MM4 (Tian et al. 2008), (4) ER (Emershad and Ramming 1994), and (5) modified MM3 [i.e., containing the macro-and micro-elements of MM3+ferric salt of MS+ organic ingredients of ER, supplemented with 0.05% casein After autoclaving at 121°C for 20 min and cooling to room temperature, the double-phase culture medium was prepared by adding 10 mL of sterilized liquid ER to the solid medium (40 mL). The volume of the liquid medium was just enough to cover the inoculated ovules. Embryo numbers were counted after ovules had been cultured for 8 wk.
Percent embryo formation (percentage)=number of embryos formed/300×100. Analysis of variance (two-way ANOVA) was conducted for each variable, and means were separated by least significant difference (LSD) tests (P≤.0001).

Results and Discussion
Pollen germination and growth. Pollination and fertilization are necessary for fruit set in crosses of seedless V. vinifera× wild Chinese Vitis species. The essential conditions for  fertilization are perfect development of the flowering organs in the V. vinifera female parent and the production of pollen of high germinability and viability by the wild Chinese Vitis species male parent. Except under special conditions, there is a linear relationship between pollen germinability and viability in many fruit species (Dafni and Firmage 2000). Pollen germinability is also related to cultivar, nutritional conditions, and environmental factors (Parton et al. 2002;Watrud et al. 2011). Pollen samples from seven of the cultivars used in our breeding program were analyzed in vitro and showed significant differences among the cultivars in germination and growth (Table 1). A one-way ANOVA for pollen germination shows that the factor 'cultivar' had a significant effect on pollen germination (P≤.05; Fig. 2). The germination percentage of Red Globe pollen was highest (41.3%). No significant differences in pollen germination were observed between Beichun (38.2±1.6%), Shuangyou (36.4±0.9%), Xuefeng (34.3±1.5%), and Red Globe (41.3±2.8%). Yanshan had a pollen germination percentage of only 8.3±1.2%. Similarly, a one-way ANOVA analysis for pollen tube length showed that the seven cultivars had different pollen tube lengths (P≤.05) ranging from 0.22 to 1.15 mm (Fig. 3). The pollen tube lengths of Red Globe were the longest (1.15 mm). Pollen tube lengths of five of the cultivars (0.85 mm for Flame Seedless, 0.64 mm for Tangwei and Xuefeng, 0.22 mm for Yanshan) were significantly different (P≤.05). There were no significant differences (P≤.05) between the pollen tube lengths of Beichun (0.97±0.06 mm), Shuangyou (0.88±0.06 mm), and Red Globe (1.15±0.06 mm). Kelen and Demirtas (2003) also drew similar conclusions, namely that pollen viability, germinability, and tube growth differed significantly among grape cultivars. These authors also studied the fertilization biology of eight grape cultivars including berry diameter and length, berry weight, seed numbers, and seed weight. In the future, we will investigate the fertilization biology of our seven crosses and investigate further the relationship between pollen production and fruit set. Our research examined the pollen viability of seven seedless cultivars. Of the wild Chinese Vitis species or cultivars, Beichun and Shuangyou are considered to be the best male parents for hybridization with seedless V. vinifera grape cultivars.
Sampling time. Sampling time (the embryo rescue time) had a significant effect on embryo rescue efficiency. Embryos are hard to save if sampling time is too early, but they will abort if sampling time is too late (Bin et al. 1991). Therefore, for  Approximately 45 d was the optimal delay with embryo rescue being limited by immaturity before this time, and by natural abortion afterwards. seedless grape breeding using embryo-rescue techniques, it is critical to determine the optimal sampling time for each cultivar under investigation. This should be identified on the basis of the number of embryos that develop from cultured ovules (Pommer et al. 1995;Midani et al. 2002;Liu et al. 2003;Li et al. 2004). Here, we observed that the optimal sampling time depended on the elapsed time (days) between pollination and the maximum percent embryo formation. The optimal sampling time differed among the seedless grapes examined here as female parents. The sampling time was negatively correlated with percent embryo formation of seedless grapes. The sampling time was significantly and negatively correlated with percent embryo formation of Thompson Seedless (r=0.8883, P≤.05; Fig. 4), Flame Seedless (r=0.9050, P≤.05; Fig. 5), Crimson Seedless (r=0.9487, P≤.05; Fig. 6), and Ruby Seedless (r=0.8173, P≤.05; Fig. 7). The highest ovule formation percentage of Thompson Seedless was 37.5%. Approximately 37 d was the optimal delay with embryo rescue being limited by immaturity before this time, and by natural abortion afterwards. In Flame Seedless, Crimson Seedless, and Ruby Seedless the same values were 35.8, 35.0, and 45.7% occurring at 45, 60, and 65 d, respectively.
Basal medium type effects on embryo formation. The components of the basal culture medium are important for in vitro plant culture. Niedz and Evens (2007) found that the growth of non-embryogenic callus of sweet orange could be regulated via the mineral nutrient components of the medium to optimize callus growth. Greenway et al. (2012) suggested that the salt formulation of a basal medium is a vital, but an often-overlooked component in many in vitro applications, as it regulates the growth and morphology of cultured plant tissues by providing essential nutrients. The type of medium used can also markedly influence embryo formation so it is important to optimize embryo development and prevent early embryo abortion (Mathias et al. 1990;Emershad and Ramming 1994;Pan 2005;Tian et al. 2008;Tang 2010). Two-way ANOVA analysis demonstrated that both genotype and medium had significant effects on percent embryo formation (Table 3). There was a highly significant effect (P≤.0001) of the basal medium on percent embryo formation, with highest values observed on the MM3 (modified) medium and reduced values on the others. Therefore, the MM3 (modified) medium was chosen as the solid phase of the double-phase medium in our subsequent embryo formation work. Approximately 60 d was the optimal delay with embryo rescue being limited by immaturity before this time, and by natural abortion afterwards. Approximately 65 d was the optimal delay with embryo rescue being limited by immaturity before this time, and by natural abortion afterwards.
The seven hybrid combinations examined here differed significantly with regard to percent embryo formation on the different basal media types assessed. Highest percent embryo formations in Ruby Seedless×Beichun (28.3%) and Ruby Seedless×Xuefeng (26.0%) were on the MM3 (modified) medium (these were not significantly different). Highest percent embryo formation of Ruby Seedless× Xuefeng (8.7%) and Ruby Seedless×Beichun (8.3%) were on the MS medium (these were not significantly different). Highest percent embryo formation of Ruby Seedless×-Beichun (20.7%) and Ruby Seedless×Xuefeng (19.3%) were on the MM3 medium (these were not significantly different). Highest percent embryo formation of Flame Seedless×Red Globe (11.7%) and Ruby Seedless×Beichun (11.0%) were on the MM4 medium (these were not significantly different). Highest percent embryo formation in Flame Seedless×Red Globe (17.0%) and Ruby Seedless×Beichun (16.0%) were on the ER medium (these were not significantly different). As female parents, Ruby Seedless and Flame Seedless exhibited higher percent embryo formations than the others.
Plant growth regulators effect on embryo germination and plantlet development. Plant growth regulator composition is  (Ledbetter and Shonnard 1990;Gaspar et al. 1996;Tang et al. 2009). Bruce and Moore (1994) suggested that optimal mixtures of plant growth regulators and their concentrations should be determined for each seedless grape genotype. In 2003, Liu et al. established a medium to rescue embryos and recover hybrid plants. This comprised a modified half-strength MS medium containing 4 mg/L 6-BA and 0.5 mg/L indole-3-butyric acid, and it achieved rescue rates of up to 90%. In earlier studies (Gray et al. 1990;Gribaudo et al. 1993;Pinto et al. 1993), GA 3 and IAA were shown to enhance in vitro embryo rescue efficiency. The experiment reported here aimed to evaluate the use of plant growth regulators on embryo rescue in seedless V. vinifera grapes×wild Chinese Vitis species for promoting embryo germination and plantlet development. Two-way ANOVA for percent embryo germination and plant development demonstrated that both genotype and plant growth regulators had highly significant effects (P≤.0001; Tables 4 and 5). Highest percent embryo germination and plant development were observed on the medium: WPM+5.7 μM IAA+4.4 μM 6-BA+1.4 μM GA 3 with germination and development being comparatively reduced on the other media. The seven hybrid combinations examined here were significantly different (P≤.05) with regard to percent embryo germination and plant development on the media containing various plant growth regulators. In Ruby Seedless×Beichun, the highest percent embryo germination (5.4%) was observed on the medium WPM without IAA, BA, or GA 3 , but this rose to 55.4% on the medium WPM supplemented with 5.7 μM IAA+4.4 μM 6-BA+1.4 μM GA 3 . These were significantly different in terms of percent embryo germination. Meanwhile, the highest percent plantlet development was observed in Ruby Seedless×Beichun (44.6%) and in Flame Seedless×Red Globe (37.0; these were not significantly different).
Among the seven hybrid combinations, Ruby Seedless× Beichun showed both the highest embryo germination and plantlet development percentages with additions of IAA, 6-BA, and GA 3 to the medium increasing both embryo germination and plantlet development.
Amino acid effects on embryo rescue in the cross Ruby×Beichun. Extensive studies on in vitro culture have  demonstrated that the inclusion of certain amino acids (e.g., glycine) stimulates plantlet development. In seedless grapes, some amino acids have been shown to affect embryo rescue as plantlet development proceeds from ovule culture to embryo germination. Thus, Emershad and Ramming (1994) reported somatic embryogenesis and plant development from immature zygotic embryos of seedless grapes when the embryos grew on a medium supplemented with cysteine.
In our study, one-way ANOVA analysis for percent embryo formation, embryo germination, and plant development showed that here too, amino acids had significant effects on embryo formation, embryo germination, and plant development in Ruby Seedless×Beichun (65 d; Table 6). Significant differences (P≤.05) also existed among the amino acids examined. In the absence of amino acids, additions of proline, glutamine, or asparagine significantly increased embryo formation, germination, and plantlet development. Highest responses were observed with additions of 4.0 mM proline, which resulted in 36.1% embryo formation, 64.6% germination, and 90.5% plantlet developments. In the absence of added amino acids, plantlet development was reduced to 85.2%. Additions of the other amino acids examined (serine, phenylalanine, and methionine) reduced plantlet development compared with the controls. The lowest percentages of embryo formation (11.1%), germination (25.0%), and plantlet development (40.0%) were obtained in the presence of 4.0 mM methionine.
Roots elongation, acclimation, and transplantation. Roots elongation, acclimation, and transplantation to a soil-like medium are essential before a plantlet could be moved to the vineyard (Chandra et al. 2010;Halloran and Adelberg 2011). The cultivation of strong in vitro plantlets is the primary requirement for survival. Therefore, an optimized root-elongation medium was used (½MS+1.7 μM IAA). All strongly growing plantlets were trained for 7-14 d in a culture room (Chee and Pool 1988;Genoud et al. 1996;Serret et al. 2001). When plantlets were >10 cm tall, they were selected for acclimation and transplantation. Forceps were used to disrupt the gel media around the roots, which aided their removal from the Erlenmeyer flasks. Sterile water was used to wash away any adhering media. Each plantlet was transplanted to an 18×16-cm pot filled with a synthetic soil matrix (vermiculite/peat soil/coconut husk, 1:4:1). The plantlets were then each covered with a large transparent plastic cup whose internal surface had been sprayed with sterile water. Plantlets were held in a hardening room for 60 d for acclimation. They were irrigated as required with nutrient solution (1/10x MS macro-elements and 1x MS microelements) to maintain moisture levels. Acclimated plantlets were transplanted to a greenhouse (25/15°C, day/night) in mid-April each year and planted out in the vineyard the followed spring ( Fig. 1f-h).

Conclusion
A new protocol has been developed for embryo rescue of progeny from seedless V. vinifera grapes×wild Chinese Vitis species that resulted in a marked improvement in the breeding efficiency of new, disease-resistant, and seedless hybrids. With this new protocol, the highest percentage of hybrid plantlets developed was 90.5%. Evaluation of the characteristics of these hybrids is proceeding using marker-assisted selection, to be followed by field resistance analysis and conventional selection. These hybrid progenies of seedless V. vinifera grapes×wild Chinese Vitis species obtained using embryo rescue are foundational for breeding new seedless grape cultivars.