Induction of 2n pollen with colchicine during microsporogenesis in Phalaenopsis

The induction of 2n pollen is an important technique for breeding polyploid plants. Here, we observed meiosis in the pollen mother cells (PMCs) of six Phalaenopsis cultivars and attempted to induce 2n pollen. The meiotic stage was related to flower bud length. During meiosis, Phalaenopsis cultivars with flower widths of approximately 20–40 mm and 50–60 mm had bud lengths of approximately 3–8 mm and 5–13 mm, respectively. The duration of meiosis ranged from 4.2 to 14 d. This was the first study to characterize meiosis of the PMCs of Phalaenopsis. The natural generation frequency of 2n pollen varied from 0.68% to 1.78%. Meiotic stage and colchicine concentration significantly affected the induction of 2n pollen. The most effective treatment for obtaining 2n pollen was 0.05% colchicine in the leptotene to zygotene stage for 3 d, which achieved a 2n pollen frequency of 10.04%.


Introduction
Members of the genus Phalaenopsis are some of the most popular potted plants in the horticultural market (Minh et al. 2016) because of their colorful and long-lasting flowers (Bolaños-Villegas et al. 2008). Orchid breeders have successfully created thousands of Phalaenopsis cultivars (Hsu et al. 2010). Chromosome doubling plays a critical role in perfecting Phalaenopsis horticultural traits, such as flower color and size, plant height, and stress resistance (Wongprichachan et al. 2013). Polyploid orchids usually have stronger stems, thicker leaves, and larger and brighter flowers and show stronger adaptability to stressful environmental conditions (Baduel et al. 2018, Zeng et al. 2020). Most wild Phalaenopsis species are diploid, and the number of chromosomes is 2n = 2x = 38 (Kao et al. 2001, Shindo andKamemoto 1963). However, many popular commercial cultivars on the market are polyploid plants. Approximately 82% of these cultivars are tetraploid, and the rest are either triploid or aneuploid (Di et al. 2013, He et al. 2013, Zhuang et al. 2007). Many cultivars are produced via interspecific hybridization and intergeneric hybridization (Bolaños-Villegas et al. 2017, Chuang et al. 2013, Yuan et al. 2018. Differences in the frequency of diploidy and the size of chromosomes in Phalaenopsis species make transferring desirable genes of wild species to Communicated by Masahiro Nishihara Received December 2, 2021. Accepted April 24, 2022. First Published Online in J-STAGE on August 26, 2022. *Corresponding author (e-mail: jiaruidong@caas.cn) commercial hybrids a major challenge . Most of the progeny produced by crosses between species differing in chromosome size are sterile (Chuang et al. 2013(Chuang et al. , 2014. Chromosome doubling is an effective method for circumventing hybrid sterility. Polyploid plants are mainly formed through somatic chromosome doubling and 2n gametes (Eng and Ho 2019). Tissue culture technology for Phalaenopsis is highly developed (Cardoso et al. 2020, Chen et al. 2009, 2010, Yeung 2017; consequently, polyploid Phalaenopsis can be easily induced through tissue culture with anti-microtubule reagents such as colchicine (Griesbach 1981, 1985, Putri et al. 2019, Tavallaie and Kolahi 2017, oryzalin (Miguel and Leonhardt 2011). and nitrous oxide gas (N 2 O) (Wongprichachan et al. 2013). Somatic chromosome doubling can be used to obtain many polyploid plants, as well as chimeric and aneuploid plants (Eeckhaut et al. 2018, Van de Peer et al. 2017. Polyploid offspring can also be obtained by crossing untreated plants with plants treated with colchicine (Zeng et al. 2020). Phalaenopsis is known to be able to produce 2n male gametes naturally (Griesbach 1985, Zhu et al. 2014. Triploid and tetraploid Phalaenopsis hybrids can be obtained from diploid × diploid crosses; pentaploid plants can be obtained from diploid × triploid crosses; and pentaploid and hexaploid plants can be obtained from tetraploid × tetraploid crosses (Jian et al. 2009). The frequency of unreduced male gametes was significantly higher in triploid Phalaenopsis than in diploid and tetraploid Phalaenopsis (Zhu et al. 2014). However, the natural frequency of 2n gametes is low. Furthermore, 2n pollen weakly competed with haploid pollen during fertilization, which suggests that the probability of polyploid plants being produced under natural conditions is low , Hwang et al. 2020, Zhou et al. 2020a, Younis et al. 2014. Thus, artificial induction is thus required to increase the production of 2n pollen.
Currently, 2n gametes are mainly induced with chemical agents, such as colchicine, which has been widely used in Populus , Zhou et al. 2020a, 2020b, Lilium (Luo et al. 2016, Sato et al. 2010, Ziziphus jujuba (Liu et al. 2016, Lu et al. 2017, and Eucalyptus . Another method for increasing 2n pollen is a high-temperature treatment, which has been widely used in landscape plants, such as Populus , Tian et al. 2018, persimmon (Mai et al. 2019, Sugiura et al. 2000, Xu et al. 2008, and Eucommia ulmoides , Song et al. 2015. The induction of unreduced gametes is closely related to the meiotic period of the initial treatment, the duration of the treatment, and the temperature or concentration of colchicine (Soltis et al. 2015, Song et al. 2012. Thus, the selection of an appropriate stage of meiosis for the application of the inducing agent requires careful consideration. Few studies have examined the induction of 2n gametes in Phalaenopsis. 2n pollen has been obtained by treating Phalaenopsis Sogo Yukidian 'V3' and Phalaenopsis Tai Lin Red angel 'V31' young flower buds with colchicine (Hsu et al. 2010). In addition, polyploid offspring have been obtained by crossing untreated plants with plants treated with colchicine or N 2 O (Azmi et al. 2016, Wongprichachan et al. 2013. However, few studies have characterized the meiotic stages of pollen mother cells (PMCs), let alone evaluated the most appropriate stages for applying inducing agents for 2n pollen induction.
The aim of this study was to characterize the relationship between the meiotic stages of Phalaenopsis PMCs and flower bud length and identify the most effective approach for inducing the formation of 2n gametes via colchicine treatment. Our findings provide key insights that could be used to aid the polyploidy breeding of Phalaenopsis.

Plant materials
Phalaenopsis equestris, Phalaenopsis equestris var. coerulea, Phalaenopsis Anna-Larati Soekardi, Phalaenopsis Tzu Chiang Sapphire, Phalaenopsis Queen Beer 'Red Sky', and Phalaenopsis Purple Crystal ( Fig. 1) which are clones populations were used for meiotic process analysis. P. equestris was used for the induction of unreduced pollen by colchicine. All six Phalaenopsis plants were cultivated in a greenhouse (23-28°C) at the Institute of Vegetables and Flowers, Chinese Academy of Agricultural Sciences.

Flower bud development and meiotic process
Photographs of three Phalaenopsis flower buds of each cultivar were taken at approximately 10:00 am every day from May to August 2020 to characterize changes in flower bud morphology. To prepare meiotic chromosome samples, 10-20 different sizes of flower buds were collected from 10:00 to 14:00 within a 1-2-day period three times and three times collection was completed in half a month. Morphological characteristics of Phalaenopsis samples were measured from the photographs. After photography, pollen samples taken from the buds were fixed in Carnoy's solution (ethanol:acetic acid, 3:1) for 12 to 24 h at 4°C. The fixed pollen samples were then stored in 70% alcohol at 4°C. Samples were washed 2-3 times with sterile water, placed on a microscope slide, and then pressed with tweezers. A drop of lactophenol-acid fuchsin solution was added after the residues were removed. The samples were then covered with a coverslip and quickly passed through a flame 6-7 times. A pencil with an eraser was then used to lightly tap the samples on the slide to make the chromosomes evenly distributed for subsequent analysis with light microscopy (OLYMPUS DP73).

Frequency of 2n pollen in plants grown in greenhouse and induction of 2n pollen by colchicine
Different from many other plants, Phalaenopsis pollen tetrads remain aggregated to each other during pollen maturation. Three flower buds with mature pollen were randomly selected for each Phalaenopsis cultivar, and the number of dyads, triads, and tetrads in three fields of view was counted. The percentage of 2n pollen was calculated using the following formula: 2n pollen (%) = (2Dy + Tr)/ (2Dy + 3Tr + 4Te) × 100%, where Dy is the number of dyads, Tr is the number of triads, and Te is the number of tetrads.
When the PMCs were in the leptotene to zygotene, pachytene to diakinesis, metaphase I to telophase I, prophase II to telophase II, and tetrad stages, the Phalaenopsis flower buds were covered with cotton for 3 days and the cotton was wrapped in tin foil and had been soaked in colchicine. The colchicine concentrations used in the experiment were 0, 0.01%, 0.05%, and 0.1%. A total of 5 to 6 flower buds were used for each colchicine treatment.

Statistical analysis
The basic data were statistically analyzed by Microsoft Excel 2017 and SPSS 2.0. Data were evaluated by analysis of variance and LSD (least significant difference) multiple comparison Significant differences were detected by Tukey's test (≤0.05). Percentage was analyzed after inverse sine conversion before multiple comparisons.

Relationship between meiosis stage in PMCs and the length of flower buds
Cytological observation of Phalaenopsis buds at different developmental stages revealed a significant correlation between the meiotic process and the length of flower buds ( Fig. 2, Supplemental Table 1). The relationship between the length of Phalaenopsis flower buds and stages of meiosis can be clearly observed in Fig. 2. The six Phalaenopsis cultivars examined could be divided into two types according to the sizes of their flowers: mini Phalaenopsis and middle-size Phalaenopsis. The width of mini Phalaenopsis flowers (P. equestris, P. equestris var. coerulea, Phalaenopsis Anna-Larati soekard, and Phalaenopsis Tzu Chiang Sapphire) was approximately 20-40 mm, and the width of middle-size Phalaenopsis flowers (Phalaenopsis Queen Beer 'Red Sky' and Phalaenopsis Purple Crystal) was approximately 50-60 mm. The relationship between PMC developmental stages and flower bud external morphology in Phalaenopsis cultivars is shown in Fig. 3. The color of the flower buds became increasingly lighter from the leptotene stage to the mature pollen stage.
When the bud lengths of P. equestris, P. equestris var. coerulea, P. Anna-Larati soekardi, Phalaenopsis Tzu Chiang Sapphire, Phalaenopsis Queen Beer 'Red Sky', and Phalaenopsis Purple Crystal were 3.3, 3.0, 4.0, 4.2, 4.4, and 8.2 mm, respectively, the PMCs were in the leptotene stage. At this time, the chromosomes circulated, folded into thin and long lines, and aggregated into clusters (Fig. 4B). The PMCs then entered the zygotene stage, and the chromosomes shrank and thickened. The homologous chromosomes began to congregate and form the synaptonemal complex (Fig. 4C). As the length of the flower buds increased, the chromosomes formed into spirals and became shorter and thicker, and the pachytene stage began (Fig. 4D). At the diplotene stage, the chromosomes were further shortened, chromosome pairing was loosened, and synapsis could no longer be observed (Fig. 4E). When the diakinesis stage began, the spiraling of the chromosomes tightened, and the chromosomes reached their shortest length (Fig. 4F). When the flower bud lengths of P. equestris, P. equestris var. coerulea, Phalaenopsis Anna-Larati Soekardi, Phalaenopsis Tzu Chiang Sapphire, Phalaenopsis Queen Beer 'Red Sky', and Phalaenopsis Purple Crystal were 5.4, 5.6, 5.5, 7.4, 10.8, and 12.0 mm, respectively, the PMCs  were in metaphase I, and the chromosomes were neatly arranged in the center of the equatorial plate (Fig. 4G). In anaphase I, homologous chromosomes began to separate and move to opposite poles via the kinetochore microtubules (Fig. 4H). When homologous chromosomes reached the opposite poles, the PMCs were in telophase I (Fig. 4I), and the chromosomes began to gradually unwind, completing the first round of meiosis. When the flower bud lengths of P. equestris, P. equestris var. coerulea, Phalaenopsis Anna-Larati Soekardi, Phalaenopsis Tzu Chiang Sapphire, Phalaenopsis Queen Beer 'Red Sky', and Phalaenopsis Purple Crystal were 5.6, 5.8, 5.9, 7.6, 11.2, and 12.9 mm, respectively, the chromatin in the nucleus was condensed and spiraled, and PMCs were in prophase II (Fig. 4J). At metaphase II, the two groups of chromosomes were arranged in the center of the equatorial plate, and most of them were parallel (Fig. 4K). When anaphase II began, the sister chromatids separated and moved via the two-stage spindle filaments (Fig. 4L). When meiosis entered telophase II, the sister chromatids gradually unscrewed, marking the end of the second round of meiosis. When the flower bud lengths of P. equestris, P. equestris var. coerulea, Phalaenopsis Anna-Larati Soekardi, Phalaenopsis Tzu Chiang Sapphire, Phalaenopsis Queen Beer 'Red Sky', and Phalaenopsis Purple Crystal were 6.2, 6.5, 6.6, 7.9, 12.8, and 13.6 mm, respectively, the PMCs were in the split stage (Fig. 4N). The flower buds then developed and underwent mitosis (Fig. 4O-4S). The flower buds finally became mature pollen, which were always in the form of tetrads (Fig. 4T).
Although most PMCs were meiotically synchronized, meiotic asynchrony between adjacent PMCs could also observed, such as PMCs in telophase I and telophase II stages (Fig. 5B, 5C, 5E, 5F), prophase II and tetrad stages (Fig. 5A), and mitosis and mature pollen stages (Fig. 5D). In Phalaenopsis Queen Beer 'Red Sky', the synapses were in a disordered state. After meiosis was completed, no tetrads were formed, but several small nuclei were observed (Fig. 5E).
The growth of Phalaenopsis buds from the PMC stages to flowering is shown in Fig. 6. The duration of each meiosis stage could be determined based on the duration of flower bud growth and the length of flower buds at different stages (Table 1). The meiosis of Phalaenopsis PMCs was short in duration, and the duration of prophase I was longer compared with other meiotic stages. After meiosis was complete, the duration of the tetrad, mitosis, and mature pollen stages ranged from 4.2 to 12.7 d.

Frequency of 2n pollen in plants grown in greenhouse and induction of 2n pollen by colchicine
With the exception of Phalaenopsis Queen Beer 'Red Sky' (Fig. 7E), dyads and triads were observed in the five other Phalaenopsis cultivars (Fig. 7A-D, F). The cultivars of phalaenopsis ploidy levels were also determined by observing the chromosome number during meiosis (Supplemental Fig. 1). The results are shown in Table 2. The natural generation frequency of 2n pollen significantly differed among cultivars and ranged from 0.68% to 1.78%. With the exception of Phalaenopsis Anna-Larati Soekardi, the natural generation frequency of 2n pollen did not exceed 1%.
Mature pollen was collected after the flower buds of P. equestris were treated with colchicine at different stages of meiosis. The percentages of dyads and triads and the frequency of unreduced male gametes are shown in Table 3.    Vol. 72 No. 4 According to a univariate generalized linear model analysis, we found that the dominant meiotic stage, colchicine concentration, and the dominant meiotic stage × colchicine concentration interaction had highly significant effects on the frequency of colchicine-induced 2n pollen (Table 4). The percentage of dyads and triads was significantly greater when P. equestris was treated with colchicine than without colchicine treatment (Fig. 8). Pollen was not harvested from all treatments (Table 4), as some flower buds withered or dropped early after treatment either because the concentration of colchicine was too high or the flower buds were too small. LSD multiple-comparison tests indicated that the frequency of unreduced male gametes was significantly higher under treatment with 0.05% colchicine from the leptotene to the zygotene.

Discussion
Unreduced gametes are common in plants and one of the main causes of polyploidy (Baduel et al. 2018, Eng and Ho 2019, Soltis et al. 2015, Song et al. 2012, Van de Peer et al. 2017. Compared with somatic polyploidization, sexual polyploidization greatly increases genetic diversity and heterosis (Brownfield and Kohler 2011, Lai et al. 2015, Ramanna and Jacobsen 2003, Sato et al. 2010, Younis et al. 2014, Zeng et al. 2020. Unreduced gametes have been reported in more than 34 plant families to date (Dewitte et al. 2012, Sora et al. 2016. For example, natural 2n gametes in citrus (Ahmed et al. 2020, Aleza et al. 2012 and Cymbidium (Zeng et al. 2020) have been used to obtain polyploid offspring. In our study, the frequency of 2n pollen in Phalaenopsis cultivars grown in greenhouse were determined. In the triploid Phalaenopsis Queen Beer 'Red Sky', the synapses were in a disordered state during meiosis; consequently, no tetrads were produced, which precluded estimation of the frequency of 2n pollen. The frequency of 2n pollen in most phalaenopsis did not exceed 1%, which is consistent with the results of previous studies (Zhu et al. 2014). The frequency of 2n pollen in Phalaenopsis Anna-Larati soekard was 1.78%, which is greater than the value (0.59%) reported in a previous study (Zhu et al. 2014). According to the Search The International Orchid Register website (https://apps.rhs.org.uk/horticulturaldatabase/orchidregister/ orchidregister.asp), the parents of P. Anna-Larati Soekardi are distantly related, which might explain the observed deviation in the frequency of 2n pollen. A previous study that measured the frequency of natural 2n pollen in nine Cymbidium species revealed that 2n male gamete formation frequencies varied from 0.15% to 4.30%, and the frequency of 2n pollen was higher in interspecific hybrids than in traditional cultivars (Zeng et al. 2020). Therefore, the frequency of 2n pollen in Phalaenopsis cultivars is related to plant ploidy and the genetic distance between parents.
The natural frequencies of Phalaenopsis 2n male gametes are low (Lai et al. 2015, Mercado et al. 1997. Therefore, artificial induction is required to increase the frequency of 2n pollen production to meet the needs of breeding programs. The induction of 2n pollen has been studied in several plants, such as Populus (Xi et al. 2011, Zhou et al. 2020b and Eucommia ulmoide , Song et al. 2015. The selection of an appropriate stage of meiosis for application of the inducing agent requires careful consideration (Xiang 2010). Previous studies have shown that plant meiosis is a finely regulated process in which multiple genes participate in expression at different stages and in different parts of the floral organs (Kaul and Murthy 1985). The process of meiosis of PMCs is correlated with external morphological changes in flowers (Andrada et al. 2019, Shao et al. 2019, and this has been confirmed in several plants, such as persimmon (Mai et al. 2019), Populus canescens (Zhou et al. 2020a), Eucommia ulmoides , and Dendrobium officinale . In our study, the meiosis of microspores and flower bud growth of six Phalaenopsis cultivars differing in flower bud size were tracked and observed. Changes in the flower bud length were correlated with stages of meiosis in Phalaenopsis. Therefore, the meiotic period can be obtained by judging the bud length of Phalaenopsis.
The artificial induction of 2n gametes is an effective approach for chromosome doubling. Colchicine induction is one of the most commonly used methods; it has been used in various plants such as Populus (Zhou et al. 2020b), Ziziphus jujuba (Liu and Liu 2011, Liu et al. 2016, Shao et al. 2019, and Dendrobium officinale . Most studies of the 2n pollen induction have administered the inducing agent via injections. Studies of 2n pollen induction in Cymbidium tortisepalum var. longibracteatum and Dendrobium have shown that injection of colchicine solution into the flower buds at prophase Ⅰ resulted in the highest frequency of 2n pollen (2.65% and 6.22%, respectively) (Bin 2012. We attempted to induce 2n pollen in Phalaenopsis through injection of colchicine; however, the flower buds failed to grow normally, and they all withered in 2-3 d after treament. This might be explained by the fact that the buds of Phalaenopsis are delicate; in addition, cells and tissues might have been injured by the injections. For this reason, we chose absorbent cotton wrapping method for treatment.
Factors such as the concentration of colchicine and treatment time have a substantial effect on the efficacy of 2n pollen induction. No studies to date have examined the optimal timing of colchicine application for 2n pollen induction in Phalaenopsis. In Eucalyptus, the optimal period for colchicine application is diplotene to diakinesis and metaphase I to telophase I ; by contrast, the optimal period in Populus alba was at the end of the leptotene to diakinesis , and the optimal period for Populus canescens was at the pachytene stage (Zhou et al. 2020b). In our study, the frequency of 2n pollen in P. equestris was significantly increased by the application of 0.05% colchicine at the leptotene to zygotene stage, and the frequency of 2n pollen was 10.04%. This is consistent with the results of a previous study of Populus alba . Role of 2n pollen were significant in novel Phalaenopsis breeding (Lee et al. 2020). Due to the tardy germination and poor tube growth, 2n pollen cannot compete well with n pollen (Aleza et al. 2012Song et al. 2012. In addition to improving frequency of 2n pollen production, screening of 2n pollen is important to obtain polyploid plants, In case of sieving of pollens were effective for isolation of 2n pollen (Park et al. 2013, Rathore et al. 1991, Weisenseel and Jaffe 1976, but it is not the case in orchids due to the intrinsic nature of orchid sporogenesis, namely tetrad develops without isolation. Isolation by density gradient by Ficoll could be a solution (Baldi et al. 1988, Midoro-Horiuti et al. 1999. .Several conclusions can be made based on observations of 2n pollen induction in P. equestris and meiosis of the PMCs of six Phalaenopsis cultivars. In mini Phalaenopsis (flower width: 20-40 mm), 2n pollen could be induced by covering the flower buds with 0.05% colchicine when the flower bud length was approximately 3-8 mm. In middlesized Phalaenopsis (flower width: 50-60 mm), 2n pollen could be induced by covering the flower buds with 0.05% colchicine when the flower bud length was approximately 8-10 mm. Our findings provide a convenient way to obtain more 2n pollen and a new insights that could aid the polyploidy breeding of Phalaenopsis. The induction of 2n gametes in sterile interspecific hybrids can be used to breed completely new hybrids. The increasing number of studies examining the use of unreduced 2n gametes in different species and their genetic consequences reflect the potential value of this method for plant breeding.

Author Contribution Statement
RDJ and HG guided the research. TW, HG and RDJ conceived and designed the study. HG had oversight and leadership responsibility for the research activity planning. RDJ had management and coordination responsibility for the research activity planning and execution. RDJ provided the research materials. TW performed the experiments. TW, JHY, and JZ collected, analyzed and deposited the data. TW and RDJ wrote the initial draft. HG critically reviewed, commented, specified and revised initial manuscript. XZ, XNY, SHY and YPK revised the manuscript. All authors have read and approved the manuscript.