Introduction

Polyploid plant production is commonly used in traditional plant breeding, producing polyploid fruits and vegetables used as human and animal food for years. Recent progress reveals the great potential of in vitro culture for inducing polyploids. Because there is evidence that plant tissue culture may cause chromosome alterations and ploidy changes, the most probable origin of ploidy variation during micropropagation is somaclonal variation (Predieri 2001; Weber et al. 2008). In addition, polyploids can be obtained by treating plant material with the chemical agent colchicine, a toxic alkaloid extracted from the seeds and bulbs of Colchicum autumnale L., a member of the Liliaceae family. Colchicine has an affinity for tubulin, a microtubule-subunit protein, and inhibits spindle function, thereby preventing both cell and nuclear division during the replication of chromosomes and division to form sister chromosomes. Colchicine treatment of in vitro-cultured plant tissues has been successfully used to artificially produce polyploids in several plant species, such as tubers (Oumar et al. 2011), grasses (Glowacka et al. 2010), forage (Tulay and Unal 2010), medicinals (Omidbaigi et al. 2010; Xing et al. 2011), ornamentals (Zhang et al. 2008) and fruit crops (Blasco et al. 2015).

Tetraploids often generate variants that may contain favorable horticultural characteristics, such as a large fruit size, sturdiness, high productivity and tolerance to environmental stresses (Estilai and Shannon 1993). Moreover, in medicinal plants, polyploids can show increased content of bioactive compounds. For example, tetraploid lines of Catharanthus roseus increase the production of terpenoid indole alkaloids by modulating the expression of genes involved in their biosynthesis pathway (Xing et al. 2011).

Flow cytometry (FCM) has been described as one of the most reliable techniques for estimating DNA ploidy level and nuclear DNA content in plants. In comparison with other methods, such as Feulgen microdensitometry and chromosome counting, FCM provides unsurpassed ease, speed and accuracy (Doležel and Bartoš 2005). The application of FCM in plants has significantly increased in the last decade and has been successfully applied in the analysis of somaclonal variation in a variety of plant species, including cotton (Jin et al. 2008), grapevine (Acanda et al. 2013; Leal et al. 2006; Prado et al. 2010b) and Pinus pinaster (Marum et al. 2009).

Grapevine (Vitis sp.) is one of the most important fruit crops grown worldwide. In Galicia (North-western Spain), quality wine production has a long tradition, with up to five growing areas with a denomination of origin (DO, Rias Baixas, Ribeiro, Ribeira Sacra, Valdeorras and Monterrei) in which some distinguished cultivars, such as Albariño or Mencía, are grown. Polyploidization in grapevine could allow a greater fruit size and a loose berry cluster, characteristics that are very important for table grapes (Alleweldt and Possingham 1988). In fact, an increase in berry size and in the content of sensory and functional compounds in tetraploid table grape cultivars in comparison with their diploid counterparts has been reported (Shiraishi et al. 2008, 2010, 2012), although without a significant variation in soluble solids and total sugar content (Shiraishi et al. 2012). In addition, tolerance to environmental stress and the increased production of secondary metabolites could also be interesting in wine grapes. Tetraploids have been also used in grapevine breeding to obtain seedless triploid cultivars through embryo rescue (reviewed by Li et al. 2015).

Tetraploid, octoploid and mixoploid grapevine plants have been regenerated from anther-, ovary-, and stamen filament-derived somatic embryos (Acanda et al. 2013; Prado et al. 2010b), with percentages of polyploid occurrence ranging from 2.2 to 7.7 %, depending on the cultivar. Similar percentages of polyploid plants were obtained in the grapevine cv. Sinsaut by treating somatic embryos derived from immature zygotic embryos with colchicine (Yang et al. 2006). However, an undesirable effect of treating developed somatic embryos was a reduction in their germination efficiency. Based on that work, it appears that the use of colchicine did not have a significant effect on the production of polyploid plants versus somaclonal variation. More recently, Sinski et al. (2013) reported the induction of tetraploid plants in two seedless grapevine cultivars using cotyledonary somatic embryos treated with colchicine or oryzalin. These authors obtained higher percentages of tetraploid plants by using colchicine concentrations comparable to those used by Yang et al. (2006), although their polyploid detection system was based only on chromosome counts and stomatal size parameter determinations (Sinski et al. 2013).

In this study, we treated an embryogenic suspension culture of grapevine (cv. Mencía) with high concentrations of colchicine and evaluated the effects on the survival of the embryogenic aggregates and their embryogenic potential. The plantlets derived from colchicine-treated embryogenic cultures were analyzed by flow cytometry to determine possible alterations to the ploidy level. Considering the early embryogenic nature of the aggregates of which the suspension culture is composed (Acanda et al. 2013) and that the most probable origin of somatic embryos of grapevine are single cells (Faure et al. 1996), this method is expected to produce polyploid plants with a reduced or null probability of being mixoploid or chimeric.

Materials and methods

Initiation of embryogenic cultures and embryogenic suspension culture establishment

Adult field-grown plants of Vitis vinifera L. cv. Mencía were selected from the grapevine collection at the “Centro de Formación y Experimentación de Viticultura y Enología de Ribadumia” (Galicia, northwestern Spain). Inflorescences at stage H on the Baggiolini (1952) phenological scale (corresponding to separated clusters) were collected when the flowers were at the late binucleate microspore developmental stage (Prado et al. 2010a). This stage was identified microscopically after anther squashing in the presence of 4′,6-diamidino-2-phenylindole (DAPI) with the aid of an MZ10F fluorescence stereomicroscope (Leica Microsystems GmbH, Wetzlar, Germany). The collected flower clusters were washed twice for 5 min with 200 mL of distilled water containing a drop of detergent, chilled at 4 °C for 4 days and then sterilized as described by Kikkert et al. (2005). Embryogenic cultures were induced from stamen filaments in SEIM4 medium (Somatic Embryogenesis Induction Medium 4; Acanda et al. 2013) containing NN (Nitsch and Nitsch 1969) salts, MS (Murashige and Skoog 1962) vitamins, 1 µM 2,4-dichlorophenoxyacetic acid (2,4-D), 4.5 µM thidiazuron (TDZ), 6 % sucrose and 0.1 % casein hydrolysate. The pH of the medium was adjusted to 5.8 before autoclaving at 98 kPa and 121 °C, and the medium was solidified using 0.3 % gelrite (Duchefa Biochemie, Haarlem, The Netherlands). Twenty-five stamens were placed on 90 mm-diameter polystyrene Petri plates containing 25 mL of medium. The cultures were maintained in the same medium in continuous darkness at 23 ± 1 °C and subcultured onto fresh medium at 30 day intervals.

To initiate embryogenic suspension cultures (ESCs), embryo masses (400 mg, fresh weight) were transferred to 100 mL Erlenmeyer flasks containing 50 mL of liquid SEIM4 medium, as described by Acanda et al. (2013). The suspensions were incubated on an orbital rotary shaker (150 rpm) at 24 °C in continuous darkness and maintained by weekly replacement of 75 % of the medium with fresh SEIM4 medium. After 4 weeks, the suspensions were passed through a 850 µm sieve, and the retained masses were discarded, producing a homogeneous ESC composed of embryogenic cell aggregates (ECAs) smaller than 850 µm in diameter. This homogeneous ESC was subsequently used for the colchicine treatments described below.

Colchicine treatments

Colchicine was prepared by dissolving 500 mg in 2 mL of deionized water containing 1 % (v/v) dimethyl-sulfoxide (DMSO). This solution was filter sterilized (0.22 µm pore diameter) and then aseptically added to 15 mL conical polypropylene tubes containing 10 mL of grapevine ESC to result in final concentrations of 0, 0.1, 0.2 and 0.4 % (w/v) colchicine. The cultures were incubated on a tube rotator mixer (150 rpm) at room temperature for 24 h. Following the colchicine treatment, the ECAs were rinsed several times with sterilized SEIM4 liquid medium without colchicine, plated (1 mL) onto SEIM4 solid medium and incubated as described above to generate somatic embryo clusters. The mean number of somatic embryos produced per ECA (embryogenic potential) was determined after 30 days of incubation. To assess the effect of colchicine treatment on the survival of the grapevine ECAs, aliquots (1 mL) of the colchicine-treated ESCs were plated onto SEIM4 solid medium containing 0.1 % tetrazolium chloride (TZC) and incubated at 24 °C in darkness; the red (viable) ECAs were counted after 48 h of incubation.

Plant regeneration

Clusters of globular embryos regenerated from colchicine-treated ECAs were subcultured on 90 mm-diameter Petri plates containing 25 mL of DM1 differentiation medium (Differentiation Medium 1; Prado et al. 2010a) consisting of NN (Nitsch and Nitsch 1969) salts and MS (Murashige and Skoog 1962) vitamins, 6 % Sucrose, 0.25 % activated charcoal, pH 5.8 and 0.3 % gelrite. The cultures were maintained in continuous darkness for 4 weeks under the same conditions described for the embryogenic culture induction. For germination, opaque-white, well-developed embryos were isolated and cultured for 30 days on 90 mm-diameter Petri plates containing 25 mL of germination medium, which consisted of DM1 medium with 3 % sucrose, 0.3 % gelrite and lacking casein hydrolysate (16 embryos per plate). The cultures were maintained for 2 weeks in darkness at 20 ± 1 °C followed by another 2 weeks at 25 ± 1 °C (20 ± 1 °C night temperature) under a 16 h photoperiod with a photon flux density of 45 µmol m−2 s−1 provided by cool white light fluorescent tubes. For plant conversion, germinated embryos with hypocotyls and cotyledons and an apical root axis were transferred to test tubes (25 mm × 120 mm) containing 12.5 mL of MS medium with half-strength macronutrients, 0.15 % sucrose, pH 5.8 and 0.8 % plant agar (Duchefa). The cultures were maintained at 26 ± 1 °C (20 ± 1 °C night temperature) under a 16 h photoperiod with the same photon flux density of 45 µmol m−2 s−1 as above. The plants were propagated by culturing axillary buds on the same MS medium with half-strength macronutrients.

DAPI staining

Actively growing ECAs were collected by centrifuging aliquots of the ESC at 6000×g for 5 min. The supernatant was discarded, and the pellet was washed with phosphate-buffered saline (PBS; 137 mM NaCl, 3 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4) and centrifuged again. The ECAs were fixed overnight in 4 % (w/v) paraformaldehyde in PBS buffer at 4 °C, dehydrated in an acetone series and embedded in Technovit 8100 resin (Heraeus Kulzer GmbH, Wehrheim, Germany) at 4 °C. Semithin sections (2.5 μm) were obtained and stained with 1 μg/mL DAPI. After rinsing and drying, the preparations were mounted in Eukitt mounting medium (Kindler GmbH, Freiburg, Germany) and observed using an excitation wavelength of 405 nm and detection in the range of 415–490 nm with an E800 microscope (Nikon, Tokyo, Japan) equipped with a Nikon DS-U2 digital camera.

Ploidy determination

To determine plant ploidy levels, well-expanded leaves collected from twenty plants from each of the colchicine treatments were analyzed by flow cytometry. In brief, nuclei were isolated by chopping approximately 100 mg leaves with a sharp razor blade in 500 µL of a buffer containing 0.1 M Tris–HCl, 2 mM MgCl2·6H2O, 1 mM disodium EDTA, 43 mM NaCl, 5 mM Na2S2O5, 1 % PVP-10, and 0.5 % Triton-100, pH 7.5 (Acanda et al. 2013). The homogenate was then filtered through a 50 µm nylon mesh, and 50 µg/mL of propidium iodide and 50 µg/mL of RNase (Sigma, St. Louis, MO) were added to the samples. After a 10 min incubation period, the samples (at least 2000 nuclei each) were analyzed using an FC500 cytometer (Beckman Coulter Inc, Miami, FL).

Statistical analysis

The data were statistically analyzed using a Mann–Whitney U test (P ≥ 0.05) performed with the SPSS (Statistical Package for the Social Sciences) software (v18.0 for Windows, IBM Corp., Somers, NY). All of the experiments were repeated at least twice.

Results and discussion

The fact that somatic embryos can originate from single cells suggests that cell suspension culture is a very convenient micropropagation system to induce polyploids while avoiding the formation of unstable and undesirable mixoploids. Faure et al. (1996) found that grapevine somatic embryos initiate directly from anther connective cells or indirectly from connective-derived callus with starchy parenchymatous cells. In both cases, proembryos originate from single predetermined cells that do not show the cytological characteristics of meristematic cells, although they are highly differentiated cells with a large vacuole and numerous starch grains. We histologically identified proembryos as structured islands of cells with prominent nuclei on the surface of ECAs at an early developmental stage (Fig. 1) as previously described (Acanda et al. 2013). We previously found that the peripheral cells of ECAs are also starchy cells (Acanda et al. 2013). These results suggest that the somatic embryos derived from the ECAs in this study also originated from single cells, as described by Faure et al. (1996) for grapevine somatic embryos derived from anther connective tissue.

Fig. 1
figure 1

Histological structure of embryogenic cell aggregates (ECAs) of grapevine. DAPI staining of a grapevine ECA showing the differentiation of proembryos (arrows) on its surface. Bar = 100 µm

Full size image

The stamen filament-derived embryogenic cultures were composed of somatic embryo masses (Fig. 2a) that were easily disaggregated in liquid medium to establish an ESC (Fig. 2b). The viability of the ECAs following treatment with increasing concentrations (0, 0.1, 0.2 and 0.4 %) of colchicine for 24 h was evaluated using TCZ (Fig. 2c). We found that all colchicine treatments caused a significant decrease in the survival of the ECAs compared with the control (Fig. 3a). Colchicine at 0.4 % seriously affected the survival of ECAs, which was reduced to 28.8 %. These results suggest that colchicine had a cytotoxic effect on the suspension culture. Other studies have indicated that colchicine is toxic to several plant materials from different plant species, especially at high concentrations (Adaniya and Shirai 2001; Glowacka et al. 2009, 2010; Omidbaigi et al. 2010). Yang et al. (2006) found that the survival and germination of grapevine somatic embryos after colchicine treatment were dependent on the colchicine concentration and treatment duration.

Fig. 2
figure 2

Embryogenic cultures of grapevine cv. Mencía. a Proliferation of somatic embryos on SEIM4 induction medium; b embryogenic suspension culture established in SEIM4 liquid medium in a 100 mL Erlenmeyer flask; c embryogenic culture treated with 0.2 % colchicine showing non-surviving (white) and surviving (red) embryogenic cell aggregates on SEIM4 medium containing 0.1 % TZC; d necrotic aggregates and globular embryos derived from embryogenic cell aggregates treated with 0.2 % colchicine after plating on SEIM4 solid medium and culturing for 4 weeks. Bars = 1 cm. (Color figure online)

Full size image
Fig. 3
figure 3

Effect of colchicine on the survival a and embryogenic potential b of grapevine embryogenic cell aggregates. The values are presented as the mean ± standard error. Means with the same letter are not significantly different (Mann–Whitney U test, P ≥ 0.05)

Full size image

After plating the colchicine-treated ESCs onto SEIM4 solid medium, a fraction of the ECAs became necrotic and did not differentiate into embryos (Fig. 2d), showing that the embryogenic potential of the aggregates was more affected by the colchicine treatment than was their survival (Fig. 3b). At 0.1 % colchicine, the embryogenic potential decreased to 38.3 %. In the ESC treated with 0.2 % colchicine, although survival decreased to 52.5 %, the embryogenic potential decreased to 13.8 %; at 0.4 % colchicine, the embryogenic potential was almost completely abolished (Fig. 3). The fact that the embryogenic potential was more affected than survival by the colchicine treatment may be due to the higher exposure received by the embryogenic cells located at the surface of the aggregates, causing them to die.

The somatic embryos derived from the colchicine-treated ESC aggregates plated onto SEIM4 solid medium were able to differentiate after 5 weeks of culture on DM1 differentiation medium. Somatic embryos developed asynchronously, with embryos at different developmental stages being easily distinguished (Fig. 4a). The asynchronous development of grapevine somatic embryos typically occurs when using different somatic embryogenesis protocols (Acanda et al. 2013; Prado et al. 2010a). The developed somatic embryos germinated normally (Fig. 4b) on germination medium and converted to plantlets (Fig. 4c) on a modified MS medium, as we previously reported (Acanda et al. 2013).

Fig. 4
figure 4

Plant regeneration from somatic embryos derived from 0.1 % colchicine-treated grapevine ECAs. a Embryogenic cluster showing embryos at different developmental stages after 5 weeks of culture on DM1 differentiation medium; b somatic embryos germinating after 4 weeks of culture on germination medium; c plantlet development after 4 weeks of culture on modified MS medium. Bars = 1 mm (a); 1 cm (b, c)

Full size image

The effects of colchicine on shoot differentiation and embryo germination have been studied in several plant species. In grapevine, the number of germinated somatic embryos was found to decrease with increasing colchicine concentration (Yang et al. 2006), although it must be taken into consideration that these authors treated somatic embryos that were already developed with colchicine. In other species, colchicine causes a decrease in the survival of colchicine-treated explants, in the number of regenerated shoots and in the percentage of embryo germination (Ishfaq et al. 2012; Mahadevamma et al. 2012; Nilanthi et al. 2009; Pande and Khetmalas 2012).

To investigate the effect of colchicine treatment on the ploidy level of somatic embryo-derived plantlets, 20 plantlets derived from ESCs treated with each of the colchicine concentrations and 20 plantlets derived from non-treated ESCs were analyzed by FCM. Five tetraploid plants (25 % of the analyzed plantlets, Table 1) were identified from the 0.2 % colchicine-treated ESC-derived plantlets. Based on the presence of an intermediate G0/G1 peak, most of the polyploid plantlets obtained from the 0.1 and 0.4 % colchicine treatments probably were near-triploids (Fig. 5). Mixoploid plantlets were not detected in this study.

Table 1 Cytotype of the plantlets regenerated from colchicine-treated and non-treated grapevine ESCs
Full size table
Fig. 5
figure 5

Histograms of the relative fluorescence intensity of propidium iodide (PI) obtained after the analysis of nuclei isolated from Vitis vinifera cv. Mencía leaves. a A diploid somatic embryo-derived plant; b a tetraploid somatic embryo-derived plant; c a near-triploid somatic embryo-derived plant showing an intermediate G0/G1 peak

Full size image

Using colchicine at a range from 0.1 to 0.4 % w/v (equivalent to 2.5–10 mM), we obtained a high percentage of polyploids (25 % tetraploids using 0.2 % colchicine) as well as variants showing an intermediate cytotype (Table 1). Yang et al. (2006) achieved a lower percentage of polyploids (4 %) using 10–20 mg/L colchicine (equivalent to 50–100 µM), which is comparable to the percentage of polyploids achieved due to somaclonal variation in previously published grapevine somatic embryogenesis protocols (Acanda et al. 2013; Prado et al. 2010b). These results suggest that the effective concentration for colchicine-mediated polyploidization in grapevine should be on the order of 2.5 mM, as suggested by other studies of the in vitro application of colchicine in other plant species (Dhooghe et al. 2011). However, recent work (Sinski et al. 2013) in other grapevine cultivars (seedless cvs. Crimson Seedless and BRS Clara) demonstrated that it is possible to obtain high rates of polyploid plants (more than 30 %) using colchicine concentrations and target tissues (mature somatic embryos) similar to those used by Yang et al. (2006). In addition, Sinski et al. (2013) improved the induction of polyploids by using oryzalin instead of colchicine at the same concentrations. Based on this and the present research, it is evident that differences in the sensitivity of the target tissues as well as the grapevine genotype account for the differences observed. Hence, it can be concluded that, as frequently occurs in grapevine for every biotechnological method established (Couselo et al. 2006; Péros et al. 1998), the efficiency of the polyploidization method must be specifically tested for each grapevine cultivar in an empirical fashion.

In comparison to tetraploid plants, chimeric plants appear to be of much less agronomic value (Nilanthi et al. 2009) because the ratio of 4× cells is generally low and unstable, as 2× cells divide more rapidly (Notsuka et al. 2000). Chimeric plants are frequently observed in in vitro polyploidization experiments conducted through organogenesis, with many meristematic cells being involved in shoot formation in several species (Ishigaki et al. 2009; Omidbaigi et al. 2010; Sun et al. 2009; Zhang et al. 2008). In the present study, using a somatic embryogenesis system, the formation of chimeric plantlets was not produced, which may be due to the combination of three factors: the single-cell origin of the somatic embryos (Faure et al. 1996); the early developmental stage of the colchicine-treated embryogenic cultures (embryogenic cell aggregates producing proembryos); and the secondary embryogenesis process occurring on solid medium after plating the colchicine-treated embryogenic suspension.