Protoplast isolation and culture from Kalanchoë species: optimization of plant growth regulator concentration for efficient callus production

A high yield of isolated protoplasts and efficient regeneration protocols are prerequisites for successful development of somatic hybrids. In the present study, protoplast isolation and regeneration were evaluated in 12 Kalanchoë accessions belonging to nine species. The highest protoplast yield was obtained from K. blossfeldiana ‘Charming Red Meadow’ with 10.78 ± 0.51 × 105 protoplasts per gram fresh weight. We observed significant differences of protoplast yield while there was no distinct difference in viability among the accessions. Seven accessions reached the microcolony stage and four developed microcalli in medium supplemented with 1.0 mg/l 1-naphthaleneacetic acid (NAA), 0.5 mg/l 6-benzylaminopurine (BAP) and 0.5 mg/l 2,4-dichlorophenoxy acetic acid (2,4-D). Using five selected accessions we optimized the PGR (plant growth regulators) concentrations using combinations of NAA, BAP and 2,4-D. K. blossfeldiana cultivars ‘Charming Red Meadow’ and ‘Paris’ produced significantly different numbers of calli depending on the PGR concentrations. For plant regeneration, the medium was supplemented with 1 mg/l NAA and 2 mg/l BAP or 2 mg/l zeatin. Shoots were regenerated on medium supplemented with NAA and BAP for K. blossfeldiana ‘Charming Red Meadow’ and K. blossfeldiana ‘Paris’. The plants successfully developed roots on the medium supplemented with IAA. The medium containing zeatin induced root formation directly from callus in K. blossfeldiana ‘Charming Red Meadow’. Our findings have the potential to facilitate the use of Kalanchoë species in somatic hybridization breeding programs. The study revealed a strong genotype-dependent efficiency of colony and microcallus formation. Plants were regenerated from two Kalanchoë blossfeldiana cultivars on medium supplemented with NAA and BAP.


Introduction
Kalanchoë is a widespread flowering plant genus belonging to the Crassulaceae family. It comprises around 140 species that are mainly succulent plants native to Madagascar Communicated by Ming-Tsair Chan.
Jin Cui and Kathryn Kuligowska Mackenzie have contributed equally to this work. and East and South Africa (Descoings 2003). Kalanchoë blossfeldiana is the most exploited species of the genus and it gave rise to the majority of commercial cultivars. Kalanchoë cultivars are widely used as potted plants due to plentiful flowering, longevity of individual flowers, postharvest performance, and low-care requirements (Mackenzie et al. 2018). Kalanchoë is the second most economically important potted plant in Europe with a turnover of 69 million euros in 2017 (Royal Flora Holland 2018). The genus is also increasingly utilized as cut flowers (Mackenzie et al. 2018). Breeding of new cultivars started as early as 1930s aiming at the development of compact cultivars with various flower colours. Currently, the cultivar improvement is achieved using cross-breeding and interspecific hybridization (Izumikawa et al. 2008;Kuligowska et al. 2015a), though genetic engineering has also been reported (Lütken et al. 2011). The expansion of the gene pools and generation of new Kalanchoë cultivars using genetic resources of wild species is of great interest (Mackenzie et al. 2018). Successful interspecific hybridization was accomplished within the Kalanchoë genus (Izumikawa et al. 2008;Kuligowska et al. 2015a;Huang and Chu 2017), however, development of hybrids is limited due to hybridization barriers related to sexual reproduction (Kuligowska et al. 2015b). Fusion of protoplasts from different species provides a practical breeding tool that can bypass sexual hybridization-related barriers. Nevertheless, development of an efficient protoplast-to-plant system is often a limiting step in somatic hybridization (Eeckhaut et al. 2013). Protoplast regeneration can be influenced by a number of factors, including donor plant material and genotype, culture system and media composition. In particular, the concentration and ratios of main plant growth regulators, auxins and cytokinins are of key importance (Davey et al. 2005). Thus, to obtain a high regeneration frequency, the protoplast culture conditions need to be optimized.
Previous studies of Kalanchoë in vitro cultures documented shoot regeneration using different explants and hormone combinations. Varga et al. 1988 examined the effect of auxins and cytokinins in K. blossfeldiana (indole-3-acetic acid-IAA, indole-3-butyric acid-IBA, zetin, kinetin and N6-(2-Isopentenyl)adenine-2iP) on shoot regeneration from leaf-derived callus. Khan et al. (2006) evaluated effects of BAP, NAA and PGR-free medium on direct regeneration of Kalanchoë tomentosa from shoot tips. The successful shoot regeneration from leaf and internode explants using thidiazuron (TDZ) was also reported in K. blossfeldiana (Sanikhani et al. 2006;Lütken et al. 2011). Furthermore, one study reported regeneration of plants from protoplasts of a single K. blossfeldiana cultivar (Castelblanque et al. 2010). Electric fusion was reported between isolated vacuoles and protoplasts of K. daigremontiana (Vienken et al. 1981), while neither chemical fusion nor successful somatic hybridization was obtained in Kalanchoë genus.
In the present study we present a modified protocol for isolation of mesophyll protoplasts from nine Kalanchoë species. The results reveal the regeneration ability of several Kalanchoë species and the role of plant growth regulators in protoplast regeneration.

Axenic plant cultures
Stem segments with axillary buds were used as in vitro explant source. The plant stems were collected from 1to 2-month-old plants and surface sterilized according to Lütken et al. (2011). Explants of 3 ± 0.2 cm were cultured on MS medium (Murashige and Skoog 1962) containing 30 g/l sucrose, 0.5 g/l 2-(N-morpholino)ethanesulfonic acid (MES), 0.66 g/l CaCl 2 , 0.08 g/l Ethylenediamino-N,N′bis(2-hydroxy-phenyl)acetic acid iron (EDDHA-Fe), 1 ml/l Atamon™ (sodium benzoate), 100 mg/l timentin, 0.05 mg/l IAA and 3 g/l gelrite. The pH was adjusted to 5.8 prior to autoclaving. All cultures were maintained in a growth chamber at 24 °C with a 16 h photoperiod and a light intensity of 150 μmol s −1 m −2 . Plants were subcultured monthly/every second month depending on the cultivar, and maintained as a source of leaf material for protoplast isolation.

Yield and viability assessment
The purified protoplasts were counted using a Fuchs-Rosenthal haemocytometer chamber. The yield was expressed as number of protoplasts per gram fresh weight (gfw). The protoplast viability was determined by fluorescein diacetate (FDA) staining according to Widholm (1972). For viability assessment, 0.5 μl FDA solution (25 mg FDA dissolved in 5 ml acetone, Sigma-Aldrich, St. Louis, USA) was added to 100 μl protoplast suspension. Protoplasts were observed under a fluorescence microscope (Leica DM2000 LED, Leica, Germany; excitation filter BP 450-490 nm, dichromatic mirror: 510 nm and emission filter: LP 515 nm) equipped with a digital camera (Leica DFC420, Leica, Germany). The protoplast viability was determined as the ratio between the green fluorescent protoplasts and the total number of observed protoplasts. The mean protoplast yield and viability were calculated based on two independent isolations per genotype. At least three microscopic fields per isolation with minimum 50 protoplasts were screened.

Protoplast culture
Protoplasts were suspended in culture medium 1 (M1 -Table 1) and cultured at a concentration of 10 5 protoplasts ml −1 in 1 ml or 2.5 ml depending on the experiment. Cultures were refreshed weekly by adding 0.4 volume of fresh medium with lower mannitol concentration (M2- Table 1). After 4 weeks, the culture medium was replaced by the same volume of medium 3 (M3- Table 1). In M3, the cultures were first cultured for 2 weeks in dark, followed by 2 weeks of 16 h photoperiod. Cultures were inspected regularly under a microscope (Leica DM2000 LED, Leica, Germany) for cell division and microcolony formation. The number of microcalli (1-2 mm diameter) was determined and they were transferred using tweezers to 5 cm Petri dishes containing solid medium 4 (M4- Table 1). The developed shoots were separated from organogenic callus and cultured on solid medium 5 (M5- Table 1). The media composition, culture conditions and culture duration are shown in Table 1. PGR concentrations differed between experiment 1 and 2 and are shown in Table 2 and 3.

Experiment 1: Protoplast yield and viability screening
Nine Kalanchoë species (12 accessions) described above were used. The protoplast isolation, yield and viability assessment were performed as described above using 0.5 g leaf material per isolation. The protoplast cultures were established with 2.5 ml of culture per 5 cm Petri dish. The media were as described in Table 1 and they were supplemented with PGR according to Table 2. The protoplast cultures were observed daily during the first week and once a week during the remaining culture period. Three independent isolations were carried out for each plant accession.
The mean values and standard deviations were calculated. Analysis of variance (ANOVA) and Tukey's honestly significant difference (HSD) tests were performed using SPSS 9.1 (SPSS Institute, Chicago, USA) to assess the effect of the treatments.

Experiment 2: Optimization of PGR concentrations
To determine the effect of plant PGR on protoplast regeneration, protoplast cultures with 12 different PGR concentrations and one control (without PGR) were established using five Kalanchoë accessions: K. blossfeldiana 'Charming Red Meadow', 'Charlie' and 'Paris', K. marnieriana and K. pinnata WT. The protocols for protoplast isolation and yield assessment were as described above. The quantity of leaf material for protoplast isolation was determined based on the result of experiment 1 to yield 39 × 10 5 protoplasts. The protoplast cultures were set up in 35 mm Petri dishes with 1 ml of protoplast culture at a density of 10 5 protoplasts ml −1 . The cultures were performed under the conditions described in Table 1. The media were supplemented with PGR concentrations according to Table 3. Each treatment consisted of 3 technical replicates (Petri dish) and was repeated twice. After 5 weeks in solid medium M4, if no shoot or root formation was observed, the calli were maintained on M4 for an additional 4 weeks. Analysis of variance (ANOVA) and Tukey's honestly significant difference (HSD) tests were performed using SPSS 9.1 (SPSS Institute, Chicago, USA) to assess the effect of the treatments.

Experiment 2: Optimization of PGR concentrations
Based on the results from experiment 1, the plants with high protoplast yield: K. blossfeldiana 'Charming Red Meadow', 'Paris', 'Charlie', K. pinnata WT and K. marnieriana were included in the optimization of culture conditions for regeneration. In experiment 2, the highest yield was observed for K. pinnata WT at 7.57 ± 0.86 x 10 5 protoplasts per gfw. K. marnieriana had the lowest yield at 2.56 ± 0.60 × 10 5 protoplasts per gfw (Fig. 3).
The culture medium in experiment 2 was supplemented with a tenfold lower NAA concentration than in experiment 1, and different concentrations of BAP and 2, 4-D (Table 3). All protoplasts cultured on medium without PGR died within 3 days. In the cultures supplemented with PGR, protoplasts of K. pinnata WT and K. marnieriana rebuilt a cell wall and underwent first divisions, but did not develop further. K. blossfeldiana 'Charlie' protoplasts formed microcolonies, but they stopped their development in the fifth week of culture after transfer to 16 h photoperiod. K. blossfeldiana 'Charming Red Meadow' and 'Paris' protoplasts exhibited stable mitosis and developed microcalli that were transferred to solid medium after 8 weeks (Fig. 1b-f). The K. blossfeldiana 'Charming Red Meadow' calli obtained in both repetitions of experiment 2 followed a similar pattern. The lowest number of calli was obtained in media supplemented with 0.05 mg/l BAP, and the highest number of calli   was obtained in media supplemented with 0.5 mg/l BAP ( Fig. 4a and b). The results of two-way ANOVA (Table 5) showed that the BAP concentration had a significant influence on the calli number. Moreover, the interaction between the effect of BAP and the effect of 2,4-D was statistically significant. The effect of 2,4-D was significant only in repetition 1 ( Table 5). The K. blossfeldiana 'Paris' protoplast derived calli number differed between repetition 1 and 2. In repetition 1, the lowest calli numbers were obtained in media supplemented with 0.05 mg/l BAP. Repetition 2 showed higher calli numbers. The results of two-way ANOVA revealed that the concentration of BAP had significant influence on the calli numbers in both repetitions. In repetition 1, there was a significant interaction between the effect of BAP and 2,4-D on the calli number while 2,4-D had no significant effect. In repetition 2 however, 2,4-D had significant influence on the calli numbers, but no interaction was observed between the effect of BAP and 2,4-D (Table 5).

K . b . ' P a r i s ' K . p i n n a t a r o l K . b . ' C h a r l i e ' K . m a r n i e r i a n a K . g r a c i l i p e s K . s t r e p t a n t h a K . m i n i a t a K . p u m i l a K . r o t u n d i f o l i a K . a r o m a Ɵ c a
For plant regeneration, all calli were transferred to solid medium containing the same PGR as in the initial cultures. The total and average numbers of calli as well as the percentage of green and yellow calli ( Fig. 1e and f) after 8 and 13 weeks of culture are shown in Supplementary Table 1 and 2. After 13 weeks of culture, the calli reached 0.5-0.8 cm in diameter. The biggest (at least 7 mm) calli were cut and all calli were divided between two media supplemented with either NAA and BAP, or NAA and zeatin (Supplementary Table 3 and 4). After 17 weeks of culture of K. blossfeldiana 'Charming Red Meadow', roots were regenerated on the medium supplemented with NAA and zeatin, from initial PGR treatment F (0.1 mg/l NAA, 0.1 mg/l BAP and 0.1 mg/l 2,4-D) in repetition 1 and 2, and from treatment L (0.1 mg/l NAA, 0.5 mg/l BAP and 1.0 mg/l 2,4-D) in repetition 2 (Table 6). These roots failed to regenerate into plants after further transfer. Shoots (Fig. 1g) were obtained from green calli of K. blossfeldiana 'Charming Red Meadow' on the medium containing NAA and BAP from initial PGR treatment F (0.1 mg/l NAA, 0.1 mg/l BAP and 0.1 mg/l 2,4-D) in repetition 1, and treatment L (0.1 mg/l NAA, 0.5 mg/l BAP and 1.0 mg/l 2,4-D) in repetition 2 ( Table 6). All shoots successfully developed into plants (Fig. 1h). The callus regeneration from K. blossfeldiana 'Paris' yielded shoots only in repetition 2 on medium containing NAA and BAP from initial treatment L (0.1 mg/l NAA, 0.5 mg/l BAP and 1.0 mg/l 2,4-D) ( Table 6). The plants successfully developed roots on the medium supplemented with IAA.

Discussion
Constant development of new decorative cultivars is a prerequisite for breeders and producers to be competitive within floriculture. Interspecific hybrids are important breeding material that can provide novel traits for commercial ornamental plants (Kuligowska et al. 2016). They can be obtained through sexual hybridization or somatic hybridization. The second method provides tools to avoid hybridization barriers related to sexual reproduction and can allow development of hybrids from more distantly related species (Eeckhaut et al. 2013;Kuligowska et al. 2016). The ability to isolate high numbers of healthy protoplasts enables a successful protoplast culture and development of an efficient protoplast-to-plant system (Rahmani et al. 2016). However, its establishment is influenced by many factors, among others: genotype, physiological status and growing conditions of the protoplast source tissue, protoplast isolation and medium composition (Davey et al. 2005;Eeckhaut et al. 2013).
In the present study, we successfully isolated protoplasts from 12 Kalanchoë accessions. A previous study of protoplast cultures in Kalanchoë reported a maximal yield of 6.0 × 10 5 protoplasts per gfw from K. blossfeldiana 'Tenorio' (Castelblanque et al. 2010). However, our preliminary studies did not produce satisfactory results following the described isolation protocol (data not shown). Using a newly developed isolation protocol based on Duquenne et al. (2007), four out of 12 Kalanchoë accessions showed a yield higher than 7.0 x 10 5 protoplasts per gfw, with the highest yield of 10.78 ± 0.51 × 10 5 protoplasts per gfw obtained in K. blossfeldiana 'Charming Red Meadow'. Compared to Castelblanque et al. (2010) we were able to increase the duration of enzymatic digestion from 4 h to overnight by increasing the osmotic pressure of enzyme solution (change of mannitol concentration) and modifying enzyme concentrations. Despite a great increase of protoplast yield in some of the tested accessions, several plant accessions yielded significantly less protoplasts compared to the protoplast yield obtained by Castelblanque. Similar to our findings, Maćkowska et al. (2014) found that commercial plants exhibit higher protoplast yields than wild species. A plausible reason for differences in protoplast yields in our study is the leaf morphology of the tested species. High protoplast yields were generally obtained from plants with large, flat and thin leaves (K. blossfeldiana, K. pinnata). The low protoplast yields were observed in plants with small curly leaves or leaves covered with hair or wax (K. aromatica, K. pumila, K. rotundifolia). Moreover, the low protoplast yield was due to incomplete tissue digestion, suggesting that leaf morphology, at least partly disabled leaf tissue penetration can be attributed to seasonal variation (Keskitalo 2001;Maćkowska et al. 2014). The physiological status of the source tissue is important for successful protoplast isolation (Davey et al. 2005;Lord and Gunawardena 2010). In the present study, plants with fast in vitro growth generally yielded more protoplasts. Plants between 4 and 6 weeks were suitable for protoplast isolation. In contrast, a previous study with K. blossfeldiana 'Tenorio' used 8-week-old plants for protoplast isolation (Castelblanque et al. 2010). Development of an efficient protoplast regeneration method relies on multiple parameters related to the culture conditions in several consecutive steps. Optimization of in vitro culture requires empirical testing of different parameters such as media composition, PGR supplementation or cultivation conditions (Davey et al. 2005). In the present study, we used medium according to Kao and Michayluk (1975) for protoplast culture. In a preliminary study using K. blossfeldiana 'Charming Red Meadow' and K. pinnata WT we compared KAO medium and MS-based medium according to Castelblanque et al. (2010). We only observed protoplast divisions using the first medium (data not shown). Auxins and cytokinins are the major growth regulators of sustained protoplast growth (Davey et al. 2005). In experiment 1, we used the PGR supplementation described by Castelblanque et al. (2010). The protoplast proliferation and regeneration showed genotype specific differences, as previously observed in other plant species (Meyer et al. 2009;Maćkowska et al. 2014;Bruznican et al. 2017). Even though all tested plants underwent the first protoplast divisions and we observed microcolony and even microcallus formation in some accessions, we were not able to regenerate  (Castelblanque et al. 2010) and Robinia pseudoacacia (Kanwar et al. 2009). Zeatin is frequently combined with NAA and used in protoplast regeneration protocols for shoot induction e.g. in Solanum (Borgato et al. 2007) and Albizia julibrissin (Rahmani et al. 2016). Moreover, zeatin has a positive effect on shoot regeneration from leaf-derived callus of K. blossfeldiana (Varga et al. 1988). However, in the present study contrary to the expected results of the of shoot development, medium supplemented with zeatin promoted root development in K. blossfeldiana 'Charming Red Meadow'. The promotive effect of zeatin on root development was previously described in in vitro cultures of Vaccinium vitis-idaea and Fragaria ananassa (Debnath 2005(Debnath , 2006. NAA and BAP were able to induce healthy shoots from both K. blossfeldiana cultivars and further plants were regenerated. In the future research further efforts are needed to accomplish the increase in the shoot regeneration efficiency. The optimal medium supplementation with PGRs for shoot regeneration from protoplast-derived callus varies depending on the plant species and even genotypes (Davey et al. 2005;Rahmani et al. 2016). Sanikhani et al. (2006) and Lütken et al. (2011) reported successful shoot regeneration in Kalanchoë blossfeldiana by TDZ. Furthermore, shoot differentiation was induced from protoplast-derived callus in the presence TDZ in Ulmus americana (Jones et al. 2015). Thus, testing a wider range of PGRs including TDZ for shoot induction in Kalanchoë should be considered. Previously reported protoplast-to-plant system for Kalanchoë mesophyll protoplasts (Castelblanque et al. 2010) as well as this study describe plant regenration only via shoot organogenesis. Chao et al. (2004) described somatic embryogenesis induction in K. blossfeldiana on MS medium supplemented with NAA, BAP and activated carbon. Development of efficient regeneration system through indirect somatic embryogenesis could provide an alternative route to obtain high shoot regeneration efficiency in Kalanchoë. Sucrose is a regular choice of carbon source in the most protoplast cultivation media (Davey et al. 2005). However, a change in the carbon source from sucrose to glucose or maltose could promote shoot regeneration for protoplast-derived cells as demonstrated in case of Lilium and rice (Godo et al. 1996;Jain et al. 1995).
In conclusion, our work presents an efficient protocol of protoplast isolation from fast growing Kalanchoë species. The results demonstrate that protoplast yield, but not viability can be influenced by the genotype of donor plant and their growth in vitro. The PGR concentration used during protoplast culture is critical for callus development. However, the efficiency of colony and microcalli formation is strongly genotype-dependent. Despite the high number of obtained calli, the shoot regeneration rate was low. Further research is needed to optimize this final step of a protoplastto-plant regeneration system in Kalanchoë.