Morphology and viability of castor bean genotypes pollen grains

The objective of this work was to characterize the morphology and viability of the pollen of 15 genotypes of castor bean (Ricinus communis L.) and to generate information that can assist in the selection of highly promising male parents for future use in genetic improvement programs aimed at producing seeds for oil extraction. Acetolysis and scanning electron microscopy was used to characterize the morphology of the pollen. The viability of the pollen grains was estimated by in vitro germination and colorimetric analysis (acetocarmine 2% and 2, 3, 5-triphenyltetrazolium chloride 1%). For the in vitro germination, pollen grains were grown in 10 types of solidified culture medium consisting of different concentrations of sucrose, boric acid, calcium nitrate, magnesium sulfate and potassium nitrate. The pollen grains had the following characteristics: medium size, isopolar and subspheroidal shape, radial symmetry, circular ambit, 3-colporate, elongated endoapertures, tectate exine and granulated sexine. The acetocarmine dye overestimated pollen viability. The media M5 and M8 were the most efficient at promoting the germination of pollen grains. The studied genotypes had high levels of viability and can therefore be used as male parents in genetic improvement programs.


Introduction
With growing demand for renewable and cleaner energy sources to produce oil for biodiesel, the castor bean (Ricinus communis L.) is being considered as a good option for farmers.
Plant-breeding techniques can be used to obtain new, more productive varieties that often differ in seasonality; these plants are of great value in selective breeding programs, as they can enable the storage, transportation and maintenance of pollen grains with high viability (Vargas, Souza, Silva, & Bobrowski, 2009).
The in vitro germination of pollen grains is the most commonly used viability assay in genetic improvement programs, as it simulates the stylestigma interaction, induces the growth of pollen tubes (Soares, Jesus, Souza, Santos-Serejo, & Oliveira, 2013), promotes fertilization and allows crossings between high quality genotypes that flower at different times.This technique has been widely studied in several species (Cuchiara et al., 2012;Machado et al., 2014), including oleaginous species such as the castor bean, cotton, soy, canola (rapeseed), oil palm, sunflower, babassu palm and peanuts.Each species requires a specific protocol and culture medium to obtain good germination.The basic medium used for in vitro assays consists of sucrose, boric acid and a variety of other substances (Zambon, Silva, Pio, Figueiredo, & Silva, 2014).
Colorimetric assays are extensively used to monitor the viability of pollen grains, as these assays are simple and faster than direct methods.However, they may overestimate viability, as non-viable grains can be stained due to the presence of enzymes, starch or other substances (Galletta, 1983).Acetocarmine, aniline blue, blue cotton, potassium iodide and 2, 3, 5-triphenyltetrazolium chloride are the most commonly used dyes for these assays; they differentially stain pollen grains, thus providing quick and cost-effective results.
Given this context, the objective of this study was to characterize the morphology and to investigate the viability of pollen from fifteen castor bean genotypes and to generate information that can assist in the selection of highly promising male parents for future use in genetic improvement programs aimed at producing seeds for oil extraction.
For the morphological characterization, the pollen grains were fixed in a modified Karnovsky (1965) solution [glutaraldehyde (2%), paraformaldehyde (2%), calcium chloride (0.001 M), sodium cacodylate buffer (0.05 M)] at pH 7.2 for 48 hours, dehydrated in an ascending ethanol series and dried in HMDS (hexamethyldisilazane).The samples were mounted in metal supports and coated with gold.The images were obtained with a variable pressure scanning electron microscope (LEO 435 VP,Carl Zeiss,Jena,Germany).
Pollen grains were subjected to weak lactic acetolysis to measure the pollen grains and exine (Raynal & Raynal, 1979).Twenty-five randomly selected pollen grains were used to measure the polar diameter, equatorial diameter and exine.The images were obtained with a photomicroscope (BX51, Olympus, Tokyo, Japan) coupled to a Sony camera using the software Image Pro-plus, v. 3.0.The terminology used to describe the pollen follows that of Punt, Hoen, Blackmore, Nilsson and Le Thomas (2007) and Hesse et al. (2009).
For the in vitro germination assays, pollen grains not subjected to any aseptic processes were inoculated in 35 mL of the different culture media: combination of two concentrations of sucrose (150 and 200 g L -1 ), three concentrations of boric acid (0.1, 0.2 and 0.3 g L -1 ), three concentrations of calcium nitrate (0.3, 0.4 and 0.5 g L -1 ), three concentrations of magnesium sulfate (0.214, 0.314 and 0.414 g L -1 ), three concentrations of potassium nitrate (0.1, 0.2 and 0.3 g L -1 ) and a control without the use of the substances.All media were solidified with 0.8% agar, and the pH was adjusted to 7.0.
For each Petri dish, a sample consisting of pollen grains collected at anthesis from three flowers per raceme from thirty plants of each genotype was used.After the inoculation of the pollen grains, the Petri dishes were kept under controlled temperature conditions (27±1 o C) in the dark for 24 hours.The germinated pollen grains were counted and the pollen tube length was measured using a binocular stereomicroscope.
To calculate the in vitro germination percentage, 100 randomly selected pollen grains from the Petri dishes were counted; the lengths of five randomly selected pollen tubes from each Petri dish were measured, for a total of 40 tubes per genotype.The pollen grain was considered germinated when its pollen tube diameter was equal to or larger than the pollen itself (Chagas et al., 2010).
The experimental design was completely randomized, with a 15 x 10 factorial arrangement (genotypes x culture media) with eight replicates (i.e., eight Petri dishes).The percentage data were arc-sine transformed (√x 100 -1 ) and subjected to analysis of variance using the Scott-Knott test (p ≤ 0.01).The analyses were performed using the SAS software (Statistical Analysis System [SAS], 2010).
The colorimetric analyses of pollen grains were performed using acetocarmine (2%) and 2,3,5-triphenyltetrazolium chloride (TTC) (1%) staining.One pollen sample collected from three flowers of each genotype was distributed over a glass slide; a drop of the specific dye was then placed on the slide, and a coverslip was added.In the case of the TTC dye, to allow the enzymatic reaction to occur, the amount of viable and nonviable pollen grains per genotype was determined two hours after the slides were prepared.For the acetocarmine dye, the analysis was performed shortly after staining.
To obtain a random sample of stained pollen grains, the slide-scanning method was used with an optical microscope; 100 pollen grains/slide/ genotype were counted with three replicates each, for a total of 300 pollen grains.
The experimental design was completely randomized in a 15x2 factorial scheme (genotypes x dyes) with three replicates each.Data were subjected to analysis of variance and means were compared by the Scott-Knott test (p ≤ 0.01) using the SAS software (SAS, 2010).
The pollen grain germination rates were highest in the M5 and M8 media, which differ only by the presence of sucrose.The M5 medium, which contains 150 g L -1 of sucrose, formed seven distinct groups.Five genotypes achieved germination rates above 70% (MPA36, MPA39, MPA41, MPA42 and MPA43) and the MPA42 genotype achieved a rate of 90.37% (Figure 2a).The M8 culture medium, which contains 200 g L -1 of sucrose, was defined by four groups and seven genotypes with in vitro germination rates above 70% (MPA26, MPA34, MPA35, MPA36, MPA37, MPA38 and MPA40); the MPA34 and MPA36 genotypes achieved particularly high germination rates of 92.37 and 91.12%, respectively.However, the germination rates of the MPA40 genotype did not differ from those obtained with the M10 culture medium.
The lowest rates of in vitro pollen grain germination were obtained with the M1 culture.Two groups formed with this medium: a group composed of the genotypes MPA11, MPA31, MPA36, MPA39 and MPA42, with germination rates ranging from 3.12 to 5.37%, and a group composed of the remaining genotypes that did not germinate in vitro (Table 2, Figure 2b).This result may be explained by the composition of the culture medium (water and agar only) and suggests that castor bean pollen requires other chemical elements to produce the pollen tube.
Different culture media have been used to germinate the pollen grains of a large number of species in vitro (Chagas et al., 2010;Cuchiara et al., 2012;Soares et al., 2008;2013).Several authors have used carbohydrates and germinationstimulating substances (nutrients and hormones) in culture medium.Several organic and inorganic substances, such as sucrose, boric acid, calcium nitrate, potassium nitrate and magnesium sulfate, affect the in vitro germination of pollen grains (Galletta, 1983).
Sucrose, as a carbohydrate source, is added to culture media to either meet the metabolic needs of the explants by participating in the generation of energy and/or serve as a source of carbon skeletons for biosynthetic processes involved in cellular differentiation (Chagas et al., 2010;Figueiredo, Pio, Silva, & Silva, 2013).Greater pollen tube growth was observed in culture medium M5 (Figure 2c) for all of the genotypes except genotype MPA26, which showed a longer pollen tube (0.236 mm) in medium M7, and genotype MPA18 in media M6 (0.182 mm), M9 (0.174 mm) and M10 (0.185 mm), which belong to the same group (Table 2).
The smallest pollen tubes were observed in the M1 medium, and most of the genotypes produced no germinated pollen in this medium.Germination occurred for genotypes MPA34, MPA41 and MPA43 only in medium M5, indicating that the absence of sucrose completely inhibited germination.According to Scorza and Sherman (1995), good pollen must have 50 to 80% germinated pollen grains with well-developed pollen tubes.Most of the genotypes analyzed in this study could be used as male parents for genetic improvement or conservation programs in germplasm banks; the only exception was genotype MPA17, which achieved low pollen grain germination rates (from 0 to 51% in vitro) (Table 2).This pattern can be explained by several factors, including genetic origin, environmental conditions and an inappropriate germination medium.
In vitro germination provides a controlled experimental system; however, this system does not completely reproduce in vivo pollen tube growth, as interactions may occur between ingredients of the culture medium and different plant materials.Nevertheless, according to Soares et al. (2008), in vitro germination produces results that are relatively close to those that would be expected in vivo.For this reason, it is important to develop a welladjusted methodology for each studied species.
The analysis of viability with acetocarmine identified three groups (Table 3).The first group consisted of eleven genotypes (MPA11, MPA17, MPA31, MPA34, MPA36, MPA37, MPA38, MPA39, MPA41, MPA42 and MPA43) with viabilities greater than 96%.The second group consisted of genotypes MPA26, MPA35 and MPA40 with viabilities of 91.67, 95 and 93.67%, respectively.The third and last group consisted of only one genotype (MPA18) with a viability of 64.66%.It is important to note that the acetocarmine dye is a marker of the integrity of chromatin and stains pollen grains deep red; in the presence of this dye, non-viable grains remain transparent (Figure 2d).The results found in this study corroborate those obtained by Vargas, Souza, Silva and Bobrowski (2009) who analyzed pollen viability in four cultivars of castor bean using acetocarmine and observed viabilities of more than 86.46%.The pollen viability of genotype MPA17 (98%) was overestimated by the acetocarmine assay, as the TTC assay yielded a viability (60%) closer to that observed in the in vitro germination (51.50%) in culture medium M10.Viability results for genotype MPA18 obtained with the dye assays and the in vitro culture system with media M8 (51.87%) and M9 (53%) were similar.
The TTC assay identified five groups of genotypes, with pollen germination rates varying from 58.33 to 99% (Table 3).Pollen grains stained light red were considered viable, whereas non-viable pollen showed grayish tones (Figure 2e).
The viability of a given genotype differed with different dyes, and the acetocarmine appeared to overestimate the viability (over 90% for most genotypes).Similar results were observed by Munhoz, Luz, Meissner Filho, Barth and Reinert (2008), who found that acetocarmine staining overestimated the viability of papaya pollen relative to TTC.
In this study, it was evident that the viability assay using TTC was more reliable, as the results obtained with this dye were similar to those obtained in the in vitro germination system.This dye is a marker for the activity of dehydrogenase enzymes involved in the respiratory activity of living tissues.The enzymatic activity present in the pollen grain is associated with the germination capacity of the pollen.Several authors have argued that the TTC test is a reliable estimate of pollen viability and provides results close to those obtained in in vitro germination tests (Huang, Zhu, Mu, & Lin, 2004;Munhoz et al., 2008).
The in vitro germination rates and the results obtained in the colorimetric analysis are directly related (Scorza & Sherman, 1995).However, it has been suggested that the dye method overestimates and the in vitro germination method underestimates the percentage of germinated pollen (Galletta, 1983).Einhardt, Correa and Raseira (2006) compared the methods used to test the viability of peach pollen and concluded that the use of the in vitro germination method provides satisfactory results relative to the in vivo germination method and that propionic carmine staining overestimates the percentage of viable pollen grains.
Exudates were observed in the exine (external surface) of the pollen grain of the castor bean; these exudates were most likely a lipophilic substance known as pollenkitt (Figure 2f), which is very common in members of the Euphorbiaceae family (Vargas et al., 2009).Pollenkitt protects and minimizes the dehydration of pollen grains and consequently limits the loss of viability in this species; additionally, this compound is involved in promoting the adhesion of grains to the stigma, inducing the volatilization of compounds and attracting pollinators due to its color (Souza, Pereira, Viana, Silva, & Sudre , 2004).

Conclusion
Similar pollen morphology was observed in the fifteen studied genotypes.
The M5 and M8 media were the most efficient media used in the in vitro germination system.
Acetocarmine and 2, 3, 5-triphenyltetrazolium chloride (TTC) staining showed that the castor bean pollen grains of the studied genotypes had high levels of and great variation in pollen viability at anthesis.
All of the studied genotypes have high levels of pollen viability and can therefore be used as male parents in genetic improvement programs.

Figure 1 .
Figure 1.Morphology of pollen grains of the castor bean (Ricinus communis L.).a) Polar view of the acetolyzed pollen grain of genotype MPA38 by light microscopy (LM); the three colpi are indicated (arrows).b) Equatorial view of the acetolyzed pollen grain of genotype MPA40 by light microscopy (LM).c-g) Polar and equatorial views of pollen grains of genotypes MPA11 (c) MPA17 (d) MPA18 (e) MPA26 (f) and MPA31 (g) by scanning electron microscopy (SEM); the colpi are indicated (arrows).h) Detailed view of the colpus of genotype MPA31 by SEM (arrow).i) Detailed view of the exine ornamentation of genotype MPA18 by SEM.Bars: a-g = 10 μm, h-i = 3 μm.

Figure 2 .
Figure 2. In vitro germination (a-c) and colorimetric analysis (d-f) of pollen grains of the castor bean (Ricinus communis L.).a) Genotype MPA42 in M5 culture medium showing high percentages of germination.b) Genotype MPA17 in M1 culture medium, with the absence of germination.c) Genotype MPA17 in M5 culture medium showing longer pollen tube lengths.d) Viable and non-viable (arrow) pollen grains stained with acetocarmine.e) Viable and non-viable (arrow) pollen grains following staining with TTC.f) Presence of pollenkitt (arrow) in pollen grain stained with acetocarmine.Bars: a-c = 0.5 mm; d-e = 200 μm; f = 20 μm.

Table 1 .
Morphology and morphometry of pollen grains of the castor bean (Ricinus communis L.) in equatorial view.

Table 2 .
Percentage of in vitro germination and pollen tube length (mm) of genotypes of the castor bean (Ricinus communis L.) in different culture media.

Table 3 .
Viability of pollen grains of the castor bean (Ricinus communis L.) genotypes by colorimetric analysis.
Means followed by the same column lowercase letter and row uppercase letter belong to the same group by the Scott-Knott test (p ≤ 0.01).