Pollen viabilities and gene expression profiles across Musa genomes

Abstract

Banana (Musa spp.) is a major global economic fruit crop. However, cross-pollination from other Musa cultivars grown in nearby plantations results in seeded fruit that exceeds market demand. This study investigated pollen viability and germination and examined the expression profiles of pollen development-related genes across seven Musa genomes (AA, BB, AAA, BBB, AAB, ABB and ABBB). Twenty-three Musa cultivars were assessed for pollen viability using lacto-aceto-orcein and triphenyltetrazolium chloride staining methods. Results revealed that pollen viability obtained from both methods was significantly different among all the studied cultivars. Cultivars carrying BB (diploid) genomes had higher viability percentages than AA (diploid), AAA, BBB, AAB and ABB (triploid) and ABBB (tetraploid) genomes. Germination of the studied cultivars was also investigated on pollen culture medium, with results showing significant differences between the pollen of each cultivar. The best germinating cultivar was TKM (11.0 %), carrying BB genome. Expression profiles of pollen development-related genes by RT–qPCR indicated that both TPD1A and MYB80 genes were highly expressed in triploid Musa genomes but the PTC1 gene showed down-regulated expression, resulting in non-viable pollen. Pollen viability, pollen germination and pollen development-related genes differed across Musa cultivars. This knowledge will be useful for the selection of male parents for Musa cross-breeding programs. Pollen viability should also be considered when planning Musa production to avoid seeded fruit.

Bananas (Musa genus) possess high pollen viability and germination rates, resulting in the production of seeded fruit, which is undesirable for the market. Our study found that bananas with a BB genome revealing exceptionally elevated pollen viability and germination rates compared to other variants. Moreover, we investigated the expression of three pollen development-related genes. Notably, the TPD1A and MYB80 genes showed significantly high expression in bananas with a triploid genome. However, the PTC1 gene exhibited down-regulated expression, suggesting its role in promoting non-viable pollen. These findings hold significant promise for selecting superior male parents in cross-breeding programmes. By considering pollen viability, banana production can be optimized to meet market demands for seedless fruit and avoid the presence of unwanted seeded varieties.

Introduction

Banana (Musa spp.), belonging to the monocotyledon Musaceae family, originated in Southeast Asia (Martanti et al. 2022) and is one of the important economic fruit crops in Thailand (Arwatchananukul et al. 2022). Most Musa cultivars have been improved by mutation, natural selection and artificial breeding (Jeensae et al. 2020) through intraspecific or interspecific hybridization between two wild-diploid Musa hybrids, Musa acuminate (carrying AA genome 2n = 22) and Musa balbisiana (carrying BB genome, 2n = 22). These hybrids have resulted in high genetic diversity among Musa cultivars. All Musa cultivars can be divided into three genomic groups based on polyploid level as diploid (AA, BB and AB), triploid (AAA, BBB, AAB and ABB) and tetraploid (AAAA, ABBB and AABB) (De Langhe et al. 2010). Cultivated edible Musa are seedless vegetative parthenocarpic fruits with pollination but not fertilization, whereby the ovary develops by pollinating non-viable pollen (Backiyarani et al. 2021). In Thailand, Musa ‘Kluai Khai’ (AA), ‘Kluai Hom’ (AAA) and ‘Kluai Namwa’ (ABB) are popularly grown for commercial production with high market demand because seedless fruit has a good taste and sweet smell and is viable for industrial processing. However, Musa plantations often face problems with seed formation caused by cross-pollination from other Musa cultivars grown in nearby areas. The seeded Musa fruit has low market demand and is difficult to process. To solve this limitation, Musa pollen viability and development require study as a useful guideline to prevent seed formation and cross-breeding in mass Musa production to create the desired cultivars.

Seed formation in plants generally occurs through fertilization between female (egg) and male (pollen) parents, with success depending on two main mechanisms involved in pollen development (pollen viability) and pollen tube germination. Musa varieties contain A and B genomes and (diploid, triploid and tetraploid) polyploid levels that impact pollen viability and germination percentages. The diploid Musa genome has the highest pollen viability at 11 and 3 folds higher than triploid and tetraploid genomes, respectively (Fortescue and Turner 2004; Jeensae et al. 2020), while the diploid wild Musa (AA and BB genomes) has higher viability and pollen germination than triploid genomes (Adeleke et al. 2004; Soares et al. 2008; Oselebe et al. 2014). Reduced pollen viability results from abnormal processes during meiotic cell division, such as non-reducing chromosome segregation of trivalent or tetravalent pairings in anaphase I (Šimoníková et al. 2020).

The staining of pollen using triphenyltetrazolium chloride (TTC) and lacto-aceto-orcein (LAO) has been widely employed to assess pollen viability due to its simplicity and effectiveness (Munhoz et al. 2008; Coelho et al. 2012; Damaiyani and Hapsari, 2018). Triphenyltetrazolium chloride starts off colourless but turns red when metabolically active pollen undergoes reduction through dehydrogenase enzymes, indicating viability. In contrast, LAO stains chromosomes, exhibiting a red colour that provides insights into their viability. However, there is limited research comparing these two methods across different Musa genomes.

Few reports have assessed the molecular mechanisms of pollen development and formation in Musa. Hu et al. (2020) found that pollen development in Musa itinerans (Ma) was significantly regulated by the functions of three key genes: tapetum determinant 1 (MaTPD1A), persistent tapetal cell 1 (MaPTC1) and myeloblastosis 80 (MaMYB80). The expression of the MaTPD1A gene plays an important role in pollen development by controlling the expression of MaMYB80 and MaPTC1 genes in male flowers. The up-regulated MaTPD1A gene promotes high expression of the MaMYB80 gene but reduces the expression of the MaPTC1 gene, resulting in pollen sterility in male flowers, producing small seedless fruit. However, the expression profile of the MaTPD1A gene in mediating pollen development has not been studied in Musa cultivars of Thai germplasm.

This research determined pollen viability by TTC and LAO staining and in vitro germination, and expression profiles of pollen development-related genes across Musa genome groups (AA, BB, AAA, BBB, AAB, ABB and ABBB). Increased knowledge of pollen viability can benefit the selection of male parents for Musa cross-breeding programs to produce seedless fruit.

Materials and Methods

Plant materials

Twenty-three Thai Musa cultivars consisting of seven genomes (3 AA, 3 BB, 4 AAA, 1 BBB, 4 AAB, 7 ABB and 1 ABBB) (Table 1) were used in this study. All cultivars were planted and conserved in germplasm at the Phitsanulok Agricultural Extension and Development Center, Phitsanulok Province, Thailand. At 30 days post-flowering, the male inflorescence sample of each cultivar was collected during the anthesis stage in the morning between 0800 and 0900 h, and the sample was stored on ice until used, following the procedure of Soares et al. (2015).

Table 1.

Thai Musa cultivars used in this study.

No.	Botanical name ‘Cultivar’	Code	Genome
1	Musa ‘Kluai Leb Mu Nang’	LMN	AA
2	Musa ‘Kluai Nam Thai’	NT	AA
3	Musa ‘Kluai Khai Kamphaengphet’	KKP	AA
4	Musa ‘Tani Buri Ram167’	T167	BB
5	Musa ‘Tani A15’	TA15	BB
6	Musa ‘Tani Kip Ma’	TKM	BB
7	Musa ‘Kluai Khai Pratabong’	KPB	AAA
8	Musa ‘Kluai Nak’	NAG	AAA
9	Musa ‘Kluai Hom Khieo’	HK	AAA
10	Musa ‘Kluai Hom Thong’	HT	AAA
11	Musa ‘Kluai Lep Chang Kut’	LCK	BBB
12	Musa ‘Kluai Roiwi’	RV	AAB
13	Musa ‘Kluai Nom Sao’	NSA	AAB
14	Musa ‘Kluai Nom Sawan’	NSW	AAB
15	Musa ‘Kluai Niu Jorakhe’	NCD	AAB
16	Musa ‘Thepanom’	TPN	ABB
17	Musa ‘Hak Muk’	HM	ABB
18	Musa ‘Kluai Namwa Nuan Chan’	NNJ	ABB
19	Musa ‘Pakchong 50’	NPC50	ABB
20	Musa ‘Kluai Namwa Mali Ong’	NMO	ABB
21	Musa ‘Kluai Namwa Kap Khao’	NKK	ABB
22	Musa ‘Kluai Namwa Dam’	ND	ABB
23	Musa ‘Theparod’	TPR	ABBB

Assessment of Musa pollen viability by histochemical assay

The histochemical analysis of Musa pollen viability was assessed by staining with two assays. The 1 % LAO assay acted as positively stained chromosomes, modified from Syakhril et al. (2019), while the 1 % 2,3,5-triphenyltetrazolium chloride assay acted as an enzymatic test by detecting the dehydrogenase activity in Musa pollen grains, modified from Damaiyani and Hapsari (2018). Healthy male flowers of Musa (Fig. 1B) were removed from the young inflorescence (Fig. 1A), with the anther compartment collected from the male flower (Fig. 1C) by forceps. For the LAO staining assay, individual whole anthers were transferred into 1.5 mL Eppendorf tubes containing 20 µL of 1 % LAO solution diluted in 45 % acetic acid. The anther wall was cut by a sterilized needle to release the pollen, which was immersed in LAO solution. The mixture was adjusted to a final volume of 1 mL with sterilized distilled water (980 µL) and incubated at room temperature for 5–10 min. The TTC histochemical assay procedure was similar to the LAO assay, except that 1 mL of 1 % TTC solution (in Tris buffer containing HCl 0.15 M, pH 7.8) was used instead of LAO solution, and the mixture was incubated at room temperature in the dark for 2 h.

Figure 1.

Morphology of young Musa inflorescence (30 days post-flowering) as one representative of cultivar NKK (listed in Table 1). Bracts and flowers of closed inflorescence (A), separated male flower (B), components of flowers (sm = stigma, at = anthers, fm = filament, fp = free tepal, or = ovary and cp = compound tepal) (C).

To investigate pollen viability, a stained pollen mixture (100 µL) from individual assays was dropped on a glass slide and covered with a glass slip. Pollen viability was observed under a polarized light microscope (Olympus BX53M, Japan) with ×4 magnification and photographed at 25 fields per sample.

Images were recorded to count the total number of pollen grains and viable pollen grains. Pollen grains stained with either LAO or TTC that expressed as a round shape and were dark/light red were considered as viable, while pollen grains that expressed as an irregular or wrinkled shape that were pale red or transparent were considered non-viable.

Numbers of pollen grains from individual samples were collected from five anthers per Musa cultivar, while individual cultivars were performed for at least six biological replicates.

Assessment of Musa pollen germination by in vitro assay

The assessment of Musa pollen germination was accomplished on pollen culture medium following the modified assay of Brewbaker and Kwack (1963). The medium consisting of 0.03 % calcium nitrate, 0.02 % magnesium sulfate, 0.01 % potassium nitrate, 0.01 % boric acid and 15 % sucrose with pH adjusted to 7.0 was sterilized by autoclaving at 121 °C for 15 min. Whole anthers of individual Musa samples were placed into 1.5 mL Eppendorf tubes containing 1 mL of sterilized pollen culture medium, and the pollen was isolated from its anther by cutting with a needle. The mixture was incubated at room temperature in the dark for 2 h. At the end of the culture period, the pollen was well mixed and 100 µL was dropped onto a glass slide and covered with a glass slip. Germination was observed under a polarized microscope (Olympus BX53M) with ×4 magnification and photographed at 25 fields per sample.

Images were taken to count the total number of pollen grains and number of germinated pollen grains (germinated pollen must have pollen tube length greater than or equal to pollen diameter). Pollen grains from the individual samples were collected from five anthers per Musa cultivar, and individual cultivars were performed for at least six biological replicates.

Determining expression profiles of pollen development-related genes in Musa

Of the 23 studied Musa cultivars, 6 with high (T167 and TA15), medium (KKP and KPB) and low (NMO and NNJ) percentages of pollen viability were selected to determine the gene expression profiles involved in Musa pollen development.

Total RNA extraction

Total RNA was extracted from the anther samples collected from six Musa cultivars, following the modified CTAB method of Khairul‑Anuar et al. (2019). Briefly, Musa anthers (100 mg) per sample were ground to fine powder with liquid nitrogen in a mortar and pestle. The powder was transferred into a 1.5 mL Eppendorf tube containing 800 µL of cetyltrimethylammonium bromide (CTAB) buffer (10 % CTAB, 1 M Tris–HCl pH 8.0, 5 M NaCl, 0.5 M EDTA pH 8.0 and 1 % PVP-6000). The mixture was gently vortexed, incubated at 65 °C for 45 min, allowed to cool at room temperature for 5 min and then centrifuged at 7000×g at 4 °C for 5 min. The supernatant was transferred into a new Eppendorf tube and an equal volume of phenol:chloroform:isoamyl alcohol (P:C:I in the ratio 25:24:1, pH 4.5) was added. The mixture was vortexed for 5 min, then centrifuged at 17 500×g at 4 °C for 15 min and the supernatant (top layer) was transferred to a new Eppendorf tube. An equal volume of C:I at a ratio of 24:1 was added, mixed gently by turning the tube back and forth several times, and then centrifuged at 17 500×g at 4 °C for 15 min. This step was repeated once by adding an equal volume of C:I ratio (24:1). The supernatant (top layer) was then transferred to a new Eppendorf tube and 10 M LiCl at 1/3 volume was added to the supernatant. The mixture was mixed well by inverting the tube and incubated at 4 °C overnight. It was then centrifuged at 17 500×g at 4 °C for 30 min. The suspension was discarded and the total RNA pellet was washed and collected two times with 500 µL of 70 % ethanol by centrifuging at 17 500 × g at 4 °C for 15 min. The total RNA pellet was air-dried for 10 min and dissolved in 30 μL of RNase-free water. Subsequently, total RNA was assessed for quality and integrity by gel electrophoresis, with concentration quantified by absorbance measurements at A260 and A280 nm using a NanoDrop Lite Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Purity, assessed by the ratio of A260 nm and A280 nm, ranged from 1.8 to 2.0. The total RNA was treated with DNase l (Thermo Fisher Scientific, Waltham, MA, USA) to remove genomic DNA contamination according to the manufacturer’s instructions.

First-strand cDNA synthesis

The total RNA template (500 ng) was reversed to the first strand cDNA in a 20 µL reaction mixture using a RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Baltics, UAB, Lithuania) containing 1 µL Oligo (dT)₁₈ primer, 4 µL 5X Reaction buffer, 1 µL RiboLock RNase Inhibitor (20 U µL⁻¹), 2 µL 10 mM dNTP Mix and 1 µL RevertAid M-MuLV RT (200 U µL⁻¹), with the final volume adjusted to 20 µL using nuclease-free water. The reaction mixture was incubated as three steps of 25 °C for 5 min, 42 °C for 60 min and 70 °C for 5 min. The first strand cDNA was immediately subjected to RT–qPCR amplification and stored at −80 °C until required for further use.

RT–qPCR analysis

Four primer pairs designed by the Oligo Analyzer program (version 1.2, USA) (Table 2) were used to evaluate the expression profiles of the three candidate genes associated with pollen development including TPD1A (tapetum determinant 1), PTC1 (persistent tapetal cell 1) and MYB80 (myeloblastosis 80), while the CAC (Clathrin adaptor complexes medium) gene was used as the internal control reference for normalization.

Table 2.

Primer sequences used in this study.

GenBank accession number	Primer name	Primer sequences (5ʹ-3ʹ)	Amplicon size (bp)
HQ853240	CAC-F	CTCCTATGTTGCTCGCTTATG	146
	CAC-R	GGCTACTACTTCGGTTCTTTC
XM_009388561.1	TPD1A-F	GACAGCTCAATCGAGGAAG	113
	TPD1A-R	CCTCTCCGGTCATCTCCATC
XM_009416075.2	PTC1-F	AATCAGGAAGCAGCATCGTG	118
	PTC1-R	TCCTCCTTTCCACCACACA
XM_009403755.1	MYB80-F	AAGCCGTTCTCCCATCTCAT	172
	MYB80-R	TGTAAGCGTTGCCTGTGAC

The RT–qPCR reaction (12.5 µL) was performed in Thermo Scientific Maxima SYBR Green qPCR Master Mix (2X) (no ROX, Thermo Fisher Scientific, Baltics, UAB, Lithuania) (6.5 µL), 10 µM of each forward (0.5 µL) and reverse (0.5 µL) primer, first strand cDNA (1 µL) and nuclease-free water (4 µL), using the Eco48 Real-Time PCR system (Eco 48, PCRmax Limited, UK).

The RT–qPCR amplification was carried out under the following conditions: 1 cycle of 50 °C for 2 min and 1 cycle of 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, 60 °C for 30 s and 72 °C for 30 s. Finally, a melting curve was realized by progressively heating the reaction mixture from 55 to 95 °C using 0.3 °C increments every 0.75 s to check the purity of the RT–qPCR product. All reactions were technically repeated five times with three biological replicates.

The baseline correction was automatically calculated to determine the cycle threshold (Ct) value in each reaction using Eco 48 Study Software installed in the instrument. Data were normalized with CAC as the endogenous control (set to 1). Relative expression of the three target genes was analysed using the comparative Ct method (ΔCt1, ΔCt2 and 2^−ΔΔCt), where ΔCt1 represented different expression between target and reference gene in samples of five Musa cultivars (T167, TA15, KKP, KPB and NMO), ΔCt2 represented different expression between target and reference genes in NNJ samples and ΔΔCt showed different expression levels between ΔCt1 and ΔCt2. Normalized target gene expression level was calculated by the comparative Ct method (2^−ΔΔCt) using Microsoft Excel 2010 program.

Data analysis

Pollen viability percentage was calculated using the formula of Ssebuliba et al. (2008).

$$\begin{array}{l}\text{Pollen viability\hspace{0.17em}}(\%)\\ =\frac{\text{number of stained pollen grains}}{\text{total number of pollen grains }}\times 100\end{array}$$

Pollen germination percentage was calculated by the formula of Nyine and Pillay (2007).

$$\begin{array}{l}\text{Pollen germination (}\%)=\\ \frac{\text{number of germinated pollen grains}}{\text{total number of pollen grains}}\times 100\end{array}$$

The diameters in micrometres of viable and non-viable pollen grains were measured using the ImageJ program (version 1.52, USA).

Pollen tube length in micrometres was measured using the ImageJ program (version 1.52, USA).

All data (pollen viability, pollen germination and gene expression) were statistically assessed using ANOVA. Mean comparisons between multiple treatments (Musa genotypes) were assessed by Duncan’s Multiple Range Test (DMRT) at P-value ≤ 0.1 statistical significance using the Statistical Product and Service Solution version 17.0 software (SPSS Inc.; 2008, Chicago, USA). All values were expressed as mean ± standard error (SE) of three to six biological replicates.

Results

Assessment of Musa pollen viability by histochemical assay

Pollen viability among the 23 Musa cultivars (listed in Table 1) was quantified by staining with 1 % LAO assay and 1 % TTC assay. Both assays gave similar results of pollen viability for all the studied Musa cultivars. The LAO staining assay provided higher sensitivity for pollen viability in most Musa cultivars, except for the HK cultivar compared to the TTC assay (Fig. 2A and B).

Figure 2.

Pollen viability percentage among 23 Musa cultivars stained with either LAO (A) or TTC (B) assay. The dot represents individual data, horizontal line displays mean (n = 6 flowers) and grey box indicates the range, different letters in individual box indicate significant differences analysed by Duncan’s Multiple Range Test at P ≤ 0.01.

For the LAO staining assay, pollen viability percentages of all studied Musa cultivars were statistically significantly different, ranging from 3.8 ± 3.8 to 98.7 ± 0.4 (Fig. 2A). The cultivars were divided into three groups according to modified criteria from Damaiyani and Hapsari (2018) as high (51–100 %), medium (16–50 %) and low (0–15 %) pollen viability. Three cultivars (T167, TA15 and TKM) belonging to the BB genome gave the highest levels of pollen viability as 98.7, 98.6 and 97.6 %, respectively (Figs. 2A and 3A), while four cultivars (RV, TPN, KKP and NSA), belonging to AAB, ABB and AA genomes had moderate percentages of pollen viability at 36.4, 28.6, 20.1 and 19.4 %, respectively (Figs. 2A and 3B). The remaining (16) Musa cultivars had low levels of pollen viability percentage, ranging from 3.8 to 14.9 % (Figs. 2A and 3C). Sixteen Musa cultivars exhibited low pollen viability, while nine (TPR, ND, NKK, NMO, NPC50, NNJ, NCD, NSW and NSA) showed complete male sterility with non-pollen viability detected by TTC assay (Figs. 2B and 3F).

Figure 3.

Viability and germination percentage of Musa pollen. High percentage viability level (T167) (A and D), medium (KKP) (B and E) and low (HM) (C and F), after staining with LAO and TTC, respectively. Viable pollen shows as red with non-viable pollen clear. High percentage of pollen germination (TKM) (G), low (KKP) (H) and non-germination (KPB) (I). Germinated pollen must have pollen tube length greater than or equal to pollen diameter. Bars indicate 500 µm.

When comparing the size of the pollen grains (µm) under a polarized light microscope, viable pollen grain diameter (µm) in all studied Musa cultivars was significantly larger than non-viable pollen. Most Musa cultivars had pollen grains larger than 100 µm, except for three (T167, TA15 and TKM) with pollen grains smaller than 100 µm (Fig. 4A and B).

Figure 4.

Viable (A) and non-viable (B) pollen sizes among the studied 23 Musa cultivars. The dot represents individual data, horizontal line displays mean (n = 10 pollen grains) and grey box indicates the range, different letters in individual box indicate significant differences analysed by Duncan’s Multiple Range Test at P ≤ 0.01.

Pollen size was related to the pollen viability percentage. The Musa cultivars carrying the BB genome (T167, TA15 and TKM) showed highest percentages of pollen viability (98.3 % in LAO and 73.1 % in TTC assay) and number of pollen grains per flower (11 682). By contrast, the Musa cultivar TPR, belonging to ABBB genome had the lowest percentage of pollen viability (8.9 % in LAO and 0 % in TTC assay), with pollen grains per flower (16.1) (Table 3).

Table 3.

Pollen viability and germination among the 23 tested Thai Musa cultivars. Data are mean ± SE of six biological replicates. Different letters within the same column indicate significant differences analysed by Duncan’s Multiple Range Test at P ≤ 0.01. N/A (not applicable) indicates pollen non-germination.

Musa cultivar	Genomic group	Pollen grains per flower	Germination (%)	Pollen tube length (µm)
LMN	AA	35.0 ± 5.9 e	0.0 ± 0.0 b	N/A
NT	AA	312.2 ± 48.1 e	0.9 ± 0.9 b	137.2 ± 137.2 d
KKP	AA	575.6 ± 44.1 e	3.4 ± 1.2 b	496.6 ± 208.0 b
T167	BB	9407.2 ± 269.3 c	2.1 ± 0.6 b	802.4 ± 802.4 a
TA15	BB	14 953.9 ± 665.1 a	2.1 ± 1.0 b	413.8 ± 117.8 bc
TKM	BB	10 685.0 ± 437.1 b	11.0 ± 1.4 a	645.5 ± 27.2 ab
KPB	AAA	217.2 ± 37.5 e	0.0 ± 0.0 b	N/A
HK	AAA	196.1 ± 24.3 e	0.0 ± 0.0 b	N/A
HT	AAA	320.6 ± 68.7 e	0.0 ± 0.0 b	N/A
NAG	AAA	612.2 ± 51.4 e	1.0 ± 0.7 b	148.6 ± 94.8 d
LCK	BBB	28.9 ± 1.5 e	0.0 ± 0.0 b	N/A
RV	AAB	208.3 ± 12.3 e	0.0 ± 0.0 b	N/A
NSA	AAB	35.6 ± 5.1 e	0.0 ± 0.0 b	N/A
NSW	AAB	34.4 ± 7.4 e	0.0 ± 0.0 b	N/A
NCD	AAB	6.1 ± 1.8 e	0.0 ± 0.0 b	N/A
TPN	ABB	2320.6 ± 254.9 d	0.8 ± 0.3 b	208.2 ± 71.7 cd
HM	ABB	191.1 ± 42.2 e	0.0 ± 0.0 b	N/A
NNJ	ABB	66.1 ± 8.0 e	0.0 ± 0.0 b	N/A
NPC50	ABB	63.9 ± 5.1 e	0.0 ± 0.0 b	N/A
NMO	ABB	38.3 ± 6.5 e	0.0 ± 0.0 b	N/A
NKK	ABB	12.2 ± 2.5 e	0.0 ± 0.0 b	N/A
ND	ABB	23.3 ± 3.8 e	0.0 ± 0.0 b	N/A
TPR	ABBB	16.1 ± 5.9 e	4.2 ± 4.2 b	104.0 ± 104.0 d

Assessment of Musa pollen germination by in vitro assay

Musa pollen germination was investigated on pollen culture medium following the modified assay of Brewbaker and Kwack (1963). Results showed that each Musa cultivar had significantly different pollen germination percentages. Musa cultivar TKM gave the highest germination percentage at 11.0 % (Fig. 3G), while seven Musa cultivars (TPR, KKP, TA15, T167, NAG, NT and TPN) showed low pollen germination percentage (Fig. 3H) as 4.2, 3.4, 2.1, 2.1, 1.0, 0.9 and 0.8 %, respectively. Fifteen Musa cultivars including LMN, KPB, HK, HT, LCK, RV, NSA, NSW, NCD, HM, NNJ, NPC50, NMO, NKK and ND did not completely germinate on the culture medium (Fig. 3I).

Among the germinated Musa cultivars, T167 had the longest average pollen tube length at 802.4 µm (Table 3), while TKM, KKP, TA15, TPN, NAG, NT and TPR gave average length of pollen tube as 645.5, 496.6, 413.8, 208.2, 148.6, 137.2 and 104.0 µm, respectively (Table 3).

Assessment of gene expression involved in Musa pollen development

Previous results indicated that pollen viability percentages across the studied Musa cultivars by LAO staining assay could be classified into three groups.

Pollen viabilities were regulated by genes involved in the pollen development process. To test this hypothesis, three levels of pollen viability percentage across six Musa cultivars derived from high (T167 and TA15, carrying BB genome), medium (KKP and KPB, carrying AA and AAA genome, respectively) and low (NMO and NNJ, carrying genome ABB) were selected to analyse the expression profiles of TPD1A, MYB80 and PTC1 genes by RT–qPCR assay. All data obtained from these target genes were normalized by the CAC reference gene. Overall, the TPD1A gene (initial pathway of pollen development in microspore mother cell stage) in two Musa cultivars (NNJ and KPB, carrying ABB and AAA genome, respectively) showed higher expression, with statistically significant differences from the other four Musa cultivars (NMO, ABB genome; T167, TA15, BB genome; and KKP, AA genome) (Fig. 5A).

Figure 5.

Expression of TPD1A (A), MYB80 (B) and PTC1(C) genes in six Musa cultivars. The dot represents individual data, horizontal line displays mean (n = 5 technical iterations) and grey box indicates the range, different letters in individual box indicate significant differences analysed by Duncan’s Multiple Range Test at P ≤ 0.01. Data were normalized with the CAC reference gene as the endogenous control (set to 1).

For genes associated with the middle pathway of pollen development (microspore stage), the MYB80 gene showed the highest up-regulation in Musa cultivar NMO (ABB genome), followed by Musa cultivar NNJ (AA genome), which was significantly different at P-value ≤ 0.01 (Fig. 5B). The MYB80 gene had highly decreased expression with statistically significant differences among the four Musa cultivars (T167, TA15, BB genome; KKP, AA genome; and KPB, AAA genome) (Fig. 5B).

The PTC1 gene expression, required in the late stage of pollen development (mature pollen stage), was highly up-regulated in all cultivars with statistical differences. Among the six Musa cultivars, KPB with AAA genome had the highest PTC1 gene expression followed by Musa cultivar T167, which was significantly different from the TA15 cultivar from the BB genome and KKP from the AA genome (Fig. 5C). Two Musa cultivars (NNJ and NMO, genome ABB) gave the lowest PTC1 gene expression (Fig. 5C).

In summary, the TPD1A and MYB80 gene expressions in Musa triploid genome (ABB and AAA) strongly increased compared to the diploid genome (AA and BB) (Fig. 5A and B). By contrast, the PTC1 gene expression in the Musa triploid genome was lower than the diploid genome (Fig. 5C). The relationship of these genes suggested that high expression of either the TPD1A or MYB80 gene might decrease the expression of PTC1 gene, resulting in male sterility (non-viable pollen development). Conversely, low TPD1A or MYB80 expression might not decrease expression of the PTC1 gene, suggesting pollen grain development as male fertility.

Discussion

Musa inflorescences contain two types of female and male flowers. Of these, the male-fertile Musa with viable pollen enabled cross-pollination with the female ovary, resulting in fully seeded fruit that was surplus to commercial market demands but a valuable source for further Musa cultivar improvement in cross-breeding programs (Ortiz and Swennen 2014). In recent years, pollen development and pollen vitality (Damaiyani and Hapsari 2018), including pollen tube germination (Waniale et al. 2021) in Musa have been reported as highly diverse across different genomes. Most pollen developments were studied in Musa AA and AAA genomes but not in the remaining Musa genomes. This research investigated pollen viability among seven Musa genomes. Pollen viability, obtained from LAO and TTC staining assay across 23 Musa cultivars, showed similar percentages; however, the LAO staining assay gave higher percentage values than the TTC assay. One possible explanation was that the LAO (reddish dye) stained specific DNA molecules that were present throughout whole cells, such as mitochondrial and chloroplast DNAs in cytoplasm and chromosomes in the nucleus (Coelho et al. 2012; Robles and Quesada 2021). Most areas inside the pollen were stained as red all over the cytoplasm and nucleus after LAO was absorbed into the cell, while the TTC assay reflected only survival pollen because TTC reduction activity is based on the dehydrogenase function in only living plant cells. Hydrogen ions released from respiration in survival cells reduced colourless TTC that turned into 2,3,5-triphenyl-tetrazolium-formazan with red appearance (Damaiyani and Hapsari 2018). These results concurred with previous findings that the LAO staining test gave a higher percentage of pollen viability than TTC staining, such as in Nepenthes spp. (Kaewkhumpai 2019) and Crotalaria juncea L. (Coelho et al. 2012). Thus, the TTC staining assay was more effective and suitable for evaluating pollen survival viability of Musa samples. These findings were supported by observations that pollen viability values determined by the TTC staining assay gave consistent results of pollen germination in banana (Soares et al. 2016) and Solanum erianthum (Ghosh et al. 2020). Therefore, in this experiment, we considered the results of TTC as an indication of viability.

As reported previously, pollen viability levels among different Musa genotypes can be divided into three groups of high (51–100 %), medium (16–50 %) and low (0–15 %) (Damaiyani and Hapsari 2018). From TTC, three Musa cultivars (T167, TA15 and TKM) belonging to the BB genome gave the highest levels of pollen viability at 77.0, 65.5 and 76.7 %, respectively, while two cultivars (TPN and KKP) belonging to ABB and AA genomes had moderate percentages of pollen viability at 18.4 and 16.0 %, respectively. The remaining 18 cultivars exhibited low pollen viability. Results indicated that Musa cultivars carrying diploid genomes (AA and BB) had higher pollen viability than Musa cultivars containing triploid genomes. Similarly, the pollen viability of diploid M. acuminata ‘Pisang Rejang’ (AA) was higher than triploid, tetraploid and mixoploids (Martanti et al. 2022). Diploid Musa enabled normal meiotic chromosome behaviour with bivalent chromosome pairs and normal balanced genomic segregation, resulting in fertile pollen (Damaiyani and Hapsari 2018; Ahmad et al. 2021). By contrast, high polyploidy with different genome (A and B) compositions in Musa might produce abnormal pollen due to partial homoeologous chromosome pairing of A and B during the prophase I of meiotic cell division (Jeridi et al. 2011; Ngatat et al. 2022), including non-reducing chromosome segregates of trivalent or tetravalent pairings in anaphase I, leading to unbalanced genome transmission in gametic cells (Šimoníková et al. 2020).

Interestingly, our results showed that LMN, NT and KKP cultivars with AA diploid genomes had different levels of pollen viability. These cultivars were produced by interbreeding of Musa acuminate parents with different diploid genomic groups (such as A1 and A2 chromosomes) (Jeensae et al. 2020). The Musa BB diploid had a higher percentage of viable pollen than the AA diploid, indicating that the genomic composition (A and/or B genome) of Musa impacted production of either partially sterile or fertile pollen. This insight into male sterility located on Musa A and/or B genome requires further investigation.

Pollen size across the 23 Musa cultivars was not related to pollen viability, concurring with previous findings that pollen size (most non-viable pollen) in triploid and tetraploid Musa was larger than diploid Musa (most viable pollen) (Panda et al. 2019; Sukkaewmanee 2019; Martanti et al. 2022). One reason given for this was that pollen size was affected by amounts of nuclear DNA contents (C) with equal size of A and B genomes as various polyploidies across Musa genotypes. Previous publications suggested that nuclear DNA content could be used as an indicator of genomic constitution (De Jesus et al. 2013), supporting that higher DNA composition positively related to longer cell size. The nuclear DNA contents (2C) in diploid (1.16–1.27 pg), triploid (1.61–2.23 pg) and tetraploid (1.94–1.2.37 pg) Musa had large, medium and small genome sizes, respectively (Kamaté et al. 2001). Thus, ploidy levels were probably related to pollen size among Musa genotypes.

Similar results for viable pollen were also observed for higher pollen tube germination under culture medium condition. Among the studied Musa genotypes, both diploid M. balbisiana cultivars TKM and TA15 (BB genome composition) and M. acuminata cultivars KKP and NT (AA genome composition) had high pollen viability and pollen germination percentage, with the longest pollen tube compared to the other genotypes. This finding was supported by observations that Musa AA and BB genomes had high pollen viability and pollen tube germination, with a few differences among them (Damaiyani and Hapsari 2018; Panda et al. 2019). This preliminary result suggested that diploid Musa cultivars (carrying AA or BB genome) might be useful as male parents in Musa breeding programs because both diploids had high percentage of pollen viability and pollen tube germination. These diploid cultivars should also avoid plantations nearby the commercial field of edible Musa production (mostly triploid AAA and AAB genomes) to reduce cross-pollination and set seed in their fruit. The role of pollen development-related genes requires further study to better understand the mechanisms of male fertility and sterility in Musa.

Molecular aspects of pollen development in flowering plants involve several transcription factors that are tightly regulated by dynamic changes in gene expression (Pearce et al. 2015). Up-regulated and down-regulated gene expression profiles are impacted by either fertility or sterility of pollen production (Hu et al. 2020). Previous results showed that pollen viability percentages among six studied Musa cultivars could be classified into three groups of high (Musa T167 and TA15, carrying BB genome), medium (Musa KKP and KPB, carrying AA and AAA genomes, respectively) and low (Musa NMO and NNJ, carrying ABB genome). The expression profiles of pollen development-related genes were further evaluated using the RT–qPCR assay. TPD1A, MYB80 and PTC1 were characterized as pollen-specific genes involved in early (microspore mother cell), middle (microspore) and late (mature pollen) pollen development stage, respectively (Parish and Li 2010; Huang et al. 2011; Wu et al. 2022). Results showed that both the TPD1A and MYB80 genes were highly expressed in the Musa triploid genome (ABB and AAA) compared to the diploid genome (AA and BB). By contrast, the PTC1 gene in the Musa triploid genome was less expressed than in the diploid genome. One possible explanation of these gene associations is that up-regulated TPD1A and MYB80 genes in the Musa triploid genome might reduce PTC1 gene expression, resulting in non-viable pollen or abnormal pollen development.

Previous reports suggested that TPD1A overexpression in M. itinerans (AA genome) promoted MYB80 gene expression but suppressed PTC1 gene expression, resulting in absence of pollen in male flowers with the fruit small and seedless (Hu et al. 2020). Mutant rice with 25 nucleotide insertions within the PTC1 gene promoter region showed significantly decreased expression compared to wild-type plants, causing complete pollen sterility (Peng et al. 2020). This result suggested that non-viable pollen development in Musa might be associated with the up-regulated expression of both TPD1A and MYB80 genes but down-regulated expression of the PTC1 gene.

Conclusions

This study assessed 23 Musa cultivars for pollen viability using LAO and TTC staining methods. The LAO-stained pollen gave higher viability percentages than TTC-stained pollen. Percentages of pollen viability across all Musa cultivars were significantly different. The Musa BB genome gave higher percentage pollen viability than the AA, AAA, BBB, AAB, ABB and ABBB genomes, while the Musa TKM (BB genome) gave the highest germination percentage assessed by a pollen culture method. For the expression of pollen development-related genes, the TPD1A and MYB80 genes were up-regulated but the PTC1 gene was down-regulated in the Musa triploid genome, resulting in non-viable pollen. Knowledge of pollen viability is important when selecting male breeders for Musa cross-breeding programs to prevent seed formation in the fruit.

Phenome, Genome and Environment. Chief Editor: Colleen Doherty

Sources of Funding

This research was financially supported by Thailand Science Research and Innovation (TSRI) (Grant No. FRB650022/0179), Naresuan University (Grant No. R2565B002) and by Graduate Research Scholarships in Agriculture and Agro-Industry from the Agricultural Research Development Agency (ARDA, Public Organization) (Grant No. 2565/4).

Contributions by the Authors

T.B. and P.I. were responsible for conceiving and designing the experiments. Y.M. conducted the experiments. Y.M. and P.I. analysed the data and organized the results in figures and tables. Y.M., T.B. and K.S. collaborated on preparing the initial draft of the manuscript. T.B., K.S., K.R. and P.I. contributed materials, reagents and analysis tools. P.I. was involved in editing the final version of the manuscript. All authors have reviewed and approved the final published version of the manuscript.

Data Availability

The raw data have been provided in the Supporting Information, and additional information is available upon request from the corresponding author.

Conflicts of Interest Statement

The authors declare there are no conflicts of interests.