Micropropagation of Hibiscus moscheutos L. ‘Luna White’: effect of growth regulators and explants on nuclear DNA content and ploidy stability of regenerants

Hibiscus moscheutos L., also known as hardy hibiscus, is valued for its medicinal and ornamental attributes. It is usually propagated via seeds or cuttings. The purpose of this investigation was to develop a dependable micropropagation for H. moscheutos ‘Luna White’. To that end, the effect of four explant types (leaf, root, node, shoot tip) and two growth regulators 6-benzylaminopurine (BA) and meta-Topolin (mT) (6-(3-hydroxybenzylamino) purine) on in vitro growth of H. moscheutos was investigated. Genetic stability of the in vitro grown plants was assessed using flow cytometry, and chromosome count was investigated. No shoots were obtained from leaf or root explants. An efficient protocol for micropropagation of H. moscheutos using two explant types, 2-node and shoot tip explants, and two cytokinins (BA and mT) capable of producing true-to-type regenerants was established. Both BA and mT can be used at 2 μM or 4 μM using either 2-node or shoot tip explants. No significant difference was found between the nuclear DNA contents of seed-derived and in vitro grown plants (P < 0.05). The mean 2C DNA and monoploid 1Cx-values of seed-derived plants were 3.25 ± 0.08 pg and 1.62 ± 0.04 pg, respectively, compared with 3.26 ± 0.06 pg and 1.63 ± 0.02 pg, respectively, for in vitro grown plants. The chromosome number of both seed-derived plants and regenerants was determined to be 2n = 2x = 38. The mature regenerants obtained were fertile and phenotypically similar to seed-derived plants.


Introduction
Over 300 annual and perennial species belonging to the family Malvaceae are included in the genus Hibiscus (Pfeil et al. 2002;Akpan 2007). Hibiscus species are multipurpose plants used as sources of food, medicines, and/or ornamentals (West and Preece 2004). Among these species is H. moscheutos L. also known as mallow, common rose mallow, giant rose mallow, swamp mallow, and hardy hibiscus. Hibiscus moscheutos L. is a perennial shrub native to wetland areas of North America and desired for its showy white to pink flowers (Barrios and Ruter 2019).
Traditional methods have been used to develop new stable cultivars that can be then mass propagated by sowing F 1 seeds or rooting shoot tip cuttings (West and Preece 2004). However, seed-propagated seedlings take a long time to reach maturity (West and Preece 2004). Limitations to this type of propagation include a lack of readily available shoot tips and/ or sufficient space. Micropropagation, an effective tool for research and commercial production, can be used to alleviate these limitations.
Micropropagation is often achieved with the aid of growth regulators such as thidiazuron (TDZ; N-phenyl-1, 2, 3thidiazole-5ylurea), zeatin (6-(4-hydroxy-3-methylbut-2enylamino) purine), 6-benzylaminopurine (BA), and meta-Topolin (mT) or 6-(3-hydroxybenzylamino) purine, an Hamidou F. Sakhanokho and Nurul Islam-Faridi contributed equally to this work. aromatic cytokinin (Strnad et al. 1997). TDZ, a substituted phenyl urea (Mok et al. 1982;Ricci et al. 2001), was first discovered as a cotton defoliant (Arndt et al. 1976), but it has since been widely used as a growth regulator in a large number of species (Jaiswal and Sawhney 2006). TDZ has been used in micropropagation of H. moscheutos, but higher TDZ concentrations resulted in stunted and chlorotic shoots, leading to shoot tip dieback (West and Preece 2004), as well as the development of structural anomalies such as hyperhydricity, stunting, fasciation, and distortion in Rhododendron mucrunulatum Turcz and cotton (Novikova et al. 2020;Personal observations). Naturally occurring zeatin ($86.9 mg L −1 ) is far more expensive than BA ($0.19 mg L −1 ) (Sigma-Aldrich 2020). There is probably no growth regulator without any level of negative effects on tissue cultured plants, but the negative effects depend on the growth regulator concentration, length of time the explants are in culture, and plant species (Trolinder and Goodin 1983;Bairu et al. 2007). BA is the most widely used in micropropagation because it is one of the most effective and affordable cytokinins available (Werbrouck et al. 1996). Successful micropropagation also may depend on the explant source. Improved in vitro shoot multiplication with rooting and reduced or total elimination of hyperhydricity have been reported in micropropagated plants using mT (Werbrouck et al. 1996;Bairu et al. 2007;Novikova et al. 2020). Micropropagation from preformed structures such as shoot tips or axillary buds is generally assumed to be less prone to genetic instability, but somaclonal variation has been reported in tissue culture systems using this method (Ahuja 1998;Rani and Raina 2000;Bairu et al. 2008). In vitro regenerated plants are not always free from genetic changes due to somaclonal variation (Pramanik and Datta 1986;Trolinder and Gooding 1987;Stelly et al. 1988), so assessing genetic fidelity of regenerants is an important factor in the micropropagation of genetically uniform plants. Methods of evaluating genetic stability include flow cytometry analysis and chromosome counts (Stelly et al. 1989;Currais et al. 2013;Sakhanokho et al. 2020).
The objective of this investigation was to optimize the micropropagation protocol for H. moscheutos; therefore, we evaluated various concentrations of the cytokinins BA and mT as well as four explant sources: leaf, root, nodal segments, and shoot tip. Genetic stability of the regenerants was assessed using flow cytometry and chromosome count analysis.

Materials and Methods
Experiment 1 Plant Material F 1 seeds obtained from a controlled self-pollination of a H. moscheutos L. 'Luna White' plant were used. For seed germination, an in vitro germination medium (MS0) was used. The MS0 consisted of Murashige and Skoog (1962) salts supplemented with 30 g L −1 sucrose. For all experiments, the pH of the solution was adjusted to 5.8 before addition of 8 g L −1 agar and autoclaving for 20 min at 121°C and 15 psi. Then, the medium was transferred to 100 × 25 mm Petri dishes. Seeds were scarified for 10 min in 98% sulfuric acid and subsequently thoroughly rinsed at least three times with tap water under a fume hood. Afterwards, the seeds were surface-sterilized by stirring in 70% ethanol for 1 min and 10% (v/v) bleach and 1 drop of Tween-20 (Sigma-Aldrich, St. Louis, MO) for 5 min on a rotary shaker at 130 rpm. The seeds were then rinsed at least 3 times with sterilized deionized water. The surface-sterilized seeds were transferred to 100 × 25 mm Petri dishes containing the seed germination medium and placed in an incubator where the temperature was kept at 25°C for 16 h light, 22°C for 8 h dark and light intensity at 50 μmol m −2 s −1 . Relative humidity was maintained at 55%. About 90% of seeds germinated after 5 d in the germination medium. Six days after seed germination, seedlings were transferred to double Magenta GA-7 (Caisson Labs, Smithfield, UT) vessels containing the same MS0 medium for further growth for at least 28 d.

Effect of Growth Regulators and Explant Types
Four types of explants, root, leaf, nodal segment, and shoot tip, were excised from the seedlings growing in double Magenta jars after 28 d. The nodal explants consisted of 2-node segments. The root and leaf explants were transferred into Petri dishes containing MS0 amended with 6-benzylaminopurine (BA) (0, 2, 4, or 6 μM) or meta-Topolin (mT) (0, 2, 4, or 6 μM). All the growth regulators were purchased from Sigma-Aldrich (St. Louis, MO) except mT which was obtained from PhytoTech Labs (Lenexa, KS). The nodal and shoot tip explants were transferred into double Magenta jars also containing media consisting of MS0 amended with BA (0, 2, 4, or 6 μM) or mT (0, 2, 4, or 6 μM). After 90 d, the effect of growth regulators and explant types on plant growth was assessed by counting the number of roots and shoots, measuring the shoot length (cm), and evaluating vigor based on visual rating. Visual vigor rating was based on a 1 to 3 scale, with 1 being poor and 3 being the most vigorous. Overall plant appearance and health was used for rating; a plant rated 1 had no shoots with very little life; an explant rated 1 was dead or mostly brown with very little sign of life; a plantlet with a 2-vigor rating had some green shoots and leaves but its overall health (e.g., dead leaves) was less than desirable; and a rating of 3 indicated a plant that was green with several shoots and looked healthy. The presence or absence of roots was not considered for the visual rating of plants. Each plant was independently rated by two evaluators.
Experiment 2 In experiment 1, explants subjected to higher concentrations (4 or 6 μM) of either cytokinin failed to form roots after 90 d in culture; therefore, we decided to set up a separate second experiment. In experiment 2, explants were treated with the various concentrations of BA (0, 2, 4, or 6 μM) and mT (0, 2, 4, or 6 μM) for 56 d and transferred afterwards to a medium without any growth regulator and allowed to grow for 49 d. Plants were first grown in double Magenta jars containing the same MS0 cytokinin-amended media as in experiment 1. Double Magenta jars were used to reduce any space or physical restriction and thus allow the plants to grow taller and facilitate multiple shoot production. After 56 d, explants consisting of 2-node segments and shoot tips were excised from these plants and transferred to 25 mm × 150 mm test tubes containing MS0, a medium including no growth regulators. After 49 d of growth in MS0, the number of shoots, number of roots, vigor, and shoot length (cm) of the plantlets were determined.
Experimental design and statistical analysis To determine the effect of the cytokinins BA and mT on in vitro micropropagation of H. moscheutos 'Luna White', 4 concentrations (0, 2, 4, 6 μM) of each cytokinin were used. For both experiment 1 and experiment 2, the experimental design was a completely randomized design (CRD). For the leaf and root explants, the experimental unit was a 100 × 25 mm Petri dish containing 2 to 3 leaf or root explants. Five Petri dishes were used for each explant (leaf or root). For the nodal segment and shoot tip explants, the experimental unit was the double Magenta vessel. Each treatment was replicated 10 times, and the whole experiment repeated once. Duncan's multiple range test was used for mean separation at the 5% level, except for the nuclear DNA content for which Student's t-test was used to determine whether the nuclear DNA content of seedderived plants was different from that of in vitro grown plants at the 5% probability level. SAS software (SAS 9.4, Cary, NC) was used for statistical analysis.
Rooting and acclimatization Fifteen plantlets per treatment were randomly selected and transferred to small (approximately 1 L) pots containing finely ground, aged pine bark in the greenhouse under a mist system (approximately 26°C, 70% humidity). Each pot was covered with a 1-gallon (3.8 L) Ziplock plastic bag, and 4 to 5 small holes were punctured in each plastic bag by the third day for gradual acclimatization of plantlets. The plastic bags were removed after 7 d. All rooted and non-rooted plantlets developed roots and were hardened in pots in the greenhouse.
Additionally, ten in vitro grown plants were randomly chosen from among those hardened in the greenhouse and allowed to flower in the greenhouse. All in vitro grown plants flowered, and five plants were randomly chosen and selffertilized via controlled pollination in the greenhouse. Mature, brown seed pods showing signs of splitting were harvested and their seeds collected. Fifty seeds from each plant were sown for a germination test of harvested seeds.
Nuclear DNA content determination Preliminary results on the genome size of BA-or mT-treated in vitro grown H. moscheutos 'Luna White' plants showed no difference: therefore, 10 plants from among the BA-and mT-treated plants were randomly selected to compare their nuclear DNA content with that of seed-derived plants using flow cytometry analysis following a procedure described elsewhere (Islam-Faridi et al. 2020a;Sakhanokho et al. 2020) with minor modifications. Briefly, two fresh leaves of each of the seed-derived or in vitro H. moscheutos plants were placed in a Petri dish and co-chopped with the internal standard, Sorghum bicolor 'Tx623' (2C 1.67 pg) (Price et al. 2005) to an equal size (approximately 0.5 cm 2 ), and resuspended in 500 μL nuclei extraction buffer. The extraction buffer mixture was pipetted through a filter to remove large debris, then a nuclei staining solution (propidium iodide, RNAse, and 5% polyvinylpyrrolidone-40,000) was added. The mixture was covered to protect against light and incubated in a refrigerator at 4°C for 15 min, and nuclear DNA content was determined using a BD Accuri TM C6 flow cytometer and a BD Accuri C6 software version 1.0.264.21 (BD BioSciences, Ann Arbor, MI). At least 5000 events (nuclei) were gated for each run. Fluorescence ratios were calculated and converted to nuclear DNA content and expressed in picograms (pg) as follows: sample 2C-value (picograms) = reference 2C-value × [(sample 2C mean peak)/(reference 2C mean peak)]; and genome sizes were converted to megabases (Mbp) using the formula 1 pg = 978 Mbp . Sample monoploid 1Cx-value (pg) was calculated by dividing the 2C-value by the ploidy level (x = 2) of H. moscheutos 'Luna White' (Greilhuber et al. 2005).
Chromosome Count Chromosome spreads were prepared for 5 seed-derived and 5 in vitro randomly selected from both shoot tip-and node-derived plants using procedures previously described (Jewell and Islam-Faridi 1994;Sakhanokho et al. 2020) with some minor modifications. Actively growing root tips about 1.0 cm long were harvested and immediately pretreated with 2 mM 8-hydroxyquinoline for 4.0 h in the dark at room temperature (RT, 22-24°C), rinsed with doubledistilled H 2 O and then fixed in 4 EtOH:1 GAA (95% ethanol:glacial acetic acid) and stored at RT overnight before processing for enzyme digestion for chromosome spread. The root tips were processed for enzyme treatment within 7 d after harvest, and fixed root tips were rinsed with deionized water, mildly hydrolyzed (0.2N HCl) at 60°C for 15 min, rinsed with deionized water, then rinsed in cold 0.01M citrate buffer (20 min standing at RT) before enzyme digestion. The enzyme mixture consisted of 2% cellulase RS (w/v), 1% macerozyme R10 (w/v) (Yakult Pharmaceutical Industry Company, Tokyo, Japan), 2% pectolyase Y23 (w/v) (Kyowa Chemical Products, Tokyo, Japan), 30% cellulase [(v/v), C2730, 30% pectinase (v/v), P2611, Sigma-Aldrich], and 40% 0.01 M citrate buffer (pH 4.5). The enzyme digestion time varied from 24 to 35 min based on the thickness of root tips. The chromosome spreads were stained with 1% Azure-B (Sigma, St. Louis, MO) and dried overnight in a 37°C incubator and made permanent with a drop of Permount (Fisher Scientific, Fair Lawn, NJ). Chromosome spreads were viewed under a 63X plan apo-chromatic objective, and digital images were recorded under a green filter using bright field microscopy (AxioImager M2, Carl Zeiss, Göttingen, Germany). The chromosome spread images were processed with Adobe Photoshop (Adobe Systems Inc., NY, NY).

Results and Discussion
Effects of Growth Regulators and Explant Source In experiment 1, the effects of BA (0, 2, 4, 6 μM) and mT (0, 2, 4, 6 μM) and four explant types (leaf, root, nodal section, and shoot tip) on morphogenic response of H. muscheutos 'Luna White' were evaluated based on observation of the dependable variables which included shoot length, number of shoots, number of roots, and vigor based on visual rating (Table 1). Among the explant sources evaluated, leaf and root explants failed to form any shoot at all in any of the BA or mT treatments, including the controls. The majority of leaf explants turned brown and died; however, abundant root formation occurred sporadically in some leaf explants after 28 to 42 d in culture (Fig. 1d), but no shoot formation was observed for the leaf explant. This phenomenon has been reported in chrysanthemum (Trigiano et al. 1994) as well as in other members of the family Malvaceae including cotton (Sakhanokho et al. 2004) and kenaf (Hibiscus cannabinus L.) (Susanto and Mat Hussin 2019). Although not the intent of the current study, this phenomenon of rhizogenesis in leaf tissue culture, if optimized, could be useful for studies focused on better understanding, for example, root formation and development under various environmental conditions. As for the root explants, no changes in appearance, including color change and growth, occurred for the entire duration of the experiment despite multiple transfers to fresh media every 28 d.
For all the growth factors evaluated, the analysis of variance showed that the effects of BA and mT were highly significantly different (P > 0.001), but the effect of explant and explant*growth regulator interactions were not significantly different (P ≤ 0.05) except for plantlet vigor for which both explant and BA had a significant effect (P > 0.05) ( Table 2). For shoot length, the measurements ranged from 0.88 ± 0.44 cm when shoot tips were treated with 6 μM to 12.50 ± 6.08 cm when nodal segments were exposed to MS0 medium (the control medium with no growth regulators) ( Table 1). Regardless of the source of explant or growth regulator, plantlet length was significantly (P > 0.05) reduced with increasing In Vitro Rooting Explants subjected to higher concentrations of both BA and mT (4 or 6 μM) failed to form roots after 90 d in culture. This was probably due to the effect of these growth regulators similar to that of Bairu et al. (2008) reported of rooting. We decided to set up a second and separate experiment, experiment 2, in which explants were treated with the various concentrations of BA (0, 2, 4, or 6 μM) or mT (0, 2, 4, or 6 μM) for 56 d before transferring directly to a rooting medium without any growth regulator and allowed to grow for 49 d. For the in vitro rooting stage, we chose a medium without growth regulators (e.g., auxins) as plants subjected to multiple subcultures and growth regulators for a long period of time tend to undergo somaclonal variation or genetic changes (Nehra et al. 1992;Bairu et al. 2006). BA-and mTtreated explants continued to produce multiple shoots even after transfer to a medium without growth regulators, and their root formation improved dramatically (Tables 3 and 4). For example, 100% and 25% of explants grown in the presence of 4 μM mT and 6 μM mT, respectively, formed roots (Table 4). Similarly, 100 and 75% of explants previously grown in media containing 4 μM BA and 6 μM BA formed roots.
Interestingly, in addition to root and multiple shoot formation, the other growth factors (shoot length and vigor) of both BAand mT-treated explants were similar to those of control plants (Table 3).
Nuclear DNA Content Determination Genetic stability of the microshoots grown in the presence of BA (0, 2, 4, or 6 μM) or mT (0, 2, 4, or 6 μM) was evaluated using flow cytometry. Flow cytometry analysis showed a single peak for both seedderived leaves and leaves from in vitro grown plants, suggesting there was no mixoploids or change in ploidy level in in vitro grown plants. Analysis of the relative DNA contents of H. moscheutos 'Luna White' with the internal standard S. bicolor 'Tx623' resulted in two peaks representing the G1 nuclei of H. moscheutos and S. bicolor 'Tx623', respectively (Fig. 2 a-b). Only the flow cytometry runs with coefficients of variation within the acceptable range, i.e., lower than 5% (Doležel and Bartoš 2005), were included in the nuclear DNA analysis. The mean 2C DNA and monoploid 1Cxvalues of seed-derived plants were 3.25 ± 0.08 pg and 1.62  ± 0.04 pg, respectively, compared with 3.26 ± 0.06 pg and 1.63 ± 0.02 pg, respectively, for in vitro grown plants (Table 5). There was no statistical difference in nuclear DNA content between seed-derived and in vitro grown plants (P ≤ 0.05). To our knowledge, this is the first report of genome size in H. moscheutos.
Chromosome Count Among genetic changes leading to somaclonal variation are various types of chromosomal   (Bednarek and Orłowska 2020). Flow cytometry may be used to detect aneuploidy (addition or deletion of one or more chromosomes) (Pfosser et al. 1995;Roux et al. 2003) but not to count chromosomes, so we used conventional chromosome count to determine the exact number of chromosomes for both seed-derived and in vitro grown plants. The chromosome number 2n = 2x = 38 is generally reported for H. moscheutos (Skovsted 1935;Small 2004;Wise and Menzel 1971;Barrios and Ruter 2019). The first chromosome count in H. moscheutos was conducted by Skovsted (1935) who reported x = 19-20 as basic chromosome numbers for H. moscheutos and H. palustris, which are considered synonyms by some taxonomists. We counted 2n = 2x = 38 for both seed-derived and in vitro grown plants ( Fig. 3a-b). In general, the plant genome contains a very high amount of heterochromatic DNA [mainly AT-rich (Schweizer 1976)] concentrated in the pericentromeric region as revealed by the dark Azure-B stain (braces in Fig. 3a). Late prophase chromosomes of a seed-derived plant with their euchromatin structures indicated by arrows and braces are shown in Fig. 3a, which also shows two satellites indicated by arrowheads. A satellite chromosome is a chromosome with a segment separated from the main body of the chromosome by a secondary constriction, which is known as nucleolus organizing region (NOR). Sometimes, the satellites might be detached during chromosome preparation from the mother chromosomes and counted as individual chromosomes, which would result in additional chromosome counts (Islam-Faridi et al. 2020b). H. moscheutos contained one pair of satellite chromosomes that are clearly separated by lightly stained secondary constrictions (arrows, Fig. 3b). Our chromosome spreads clearly show that the basic chromosome number for H. moscheutos is x = 19. The earlier report (Skovsted 1935) of x = 20 for H. moscheutos was probably due to erroneous counting of the two satellites as complete chromosomes. Metaphase chromosome spread of an in vitro grown H. moscheutos plant with the same number 2n = 2x = 38 is shown in Fig. 3b, confirming the genetic fidelity of the regenerants at the chromosome level.
The results of both flow cytometry analysis and chromosome spread strongly suggested that the regenerants obtained in this study were genetically stable. However, subtle genomic DNA mutations/rearrangements are not readily easy to detect  (Smýkml et al. 2007;Sun et al. 2013). Reduced or loss of fertility attributed to epigenetic factors such as changes in DNA methylation levels has been reported in in vitro regenerated plants (Sun et al. 2013). Therefore, even though the in vitro regenerated plants grown ex vitro in the greenhouse were phenotypically similar to seed-derived plants in terms of both vegetative growth and flowering patterns ( Fig. 1a-b), a controlled self-pollination of in vitro grown plants was conducted. All regenerants produced flowers similar to those of seed-derived plants (Fig. 1b). Also, a seed germination test conducted on seeds derived from controlled self-pollination yielded a 90% germination rate and normal-looking seedlings (Fig. 1c).

Conclusions
In this study, we provided an efficient protocol for micropropagation of H. moscheutos using two explant types, 2-node and shoot tip explants, and two cytokinins (BA and mT) capable of producing true-to-type regenerants. Both BA and mT can be used at 2 μM or 4 μM using either 2-node or shoot tip explants for 56 d. The protocol developed in this investigation can be used for plant transformation, in vitro selection, and industrial mass micropropagation of improved H. moscheutos selections or cultivars. To our knowledge, this study is the first report on genome size in H. moscheutos. Finally, the chromosome number (2n = 2x = 38) of H. moscheutos was unequivocally confirmed.