Development of the Flow Cytometric Protocol for Ploidy Analysis and Determination of Relative Nuclear DNA Content in Cashew (Anacardium occidentale Linn.)

MATERIALS AND METHODS

Planting materials: Matured nuts harvested from six cashew accessions were obtained from the Cocoa Research Institute of Nigeria, Ibadan, south western Nigeria (Latitude 07°10’N, Longitude 03°52’E) and raised in the greenhouse at Leibniz Institute for the Plant Genetics and Crop Research (IPK), Germany for this study. Although, five individual plants were raised per accession, only three healthy plants were eventually used for the analysis. The seedling plants were grown under controlled greenhouse-tropicalized conditions, with average temperature of 25°C, 16 h photoperiod and alternate day automated indoor-aerial-sprinkling-irrigation for 10-15 min. Sampling of the leaves for the analysis did not commence until 6 weeks after sowing when all the nuts had germinated and the plants are of 4-leaf stage.

Plant materials for the internal reference standard: Six plant species whose genome sizes were known and commonly used as internal reference standards (Dolezel and Bartos, 2005) (Table 1) were collected from Dr. Jaroslav Dolezel of the Laboratory of Molecular Cytogenetics and Cytometry, Institute of Experimental Botany, Sokolovska, Olomouc, Czech Republic, (http://olomouc.ueb.cas.cz/). The seeds were sowed and the seedling plants raised under greenhouse condition at the IPK, Germany. The genome sizes of the selected internal reference plants range from 1.11-26.90 pg/2C for Raphanus sativus cv. Saxa and Vicia faba cv. Inovec, respectively.

Different buffer systems screening on cashew: From the literatures, five buffers compositions commonly used were selected for screening on cashew plant (Table 2). The list include (1) Galbraith’s buffer (Galbraith et al., 1983), (2) LB01 (Dolezel et al., 1989), (3) Otto’s buffer (Otto, 1990, 1992; Dolezel and Gohde, 1995), (4) Tris-MgCl₂ (Pfosser et al., 1995) and (5) DAPI-MaVI commonly used in IPK’s laboratory. The buffers were designated as G, L, O, T and M, respectively.

Table 1:	List of six plant species with their nuclear DNA contents tested as internal reference standards for estimation of relative genome size in cashew, A. occidentale

Table 2:	List of five lysis buffer systems tested on cashew A. occidentale for the estimation of nuclear DNA content
*Source: Dolezel and Bartos (2005). MOPS: 4-morpholinepropane sulfonate, DDT: Dithiothreitol, TRIS: Tris (hydroxymethyl)-aminomethane, EDTA: Ethylenediamine tetra acetic acid, HEPS: 4-(hydroxymethyl) piperazine-1-ethanesulfonic acid

Data on the performance of each buffer on cashew was evaluated according to the quality of DNA nuclear suspension described by Loureiro et al. (2006a). The variables include background factor (%), fluorescence, coefficient of variation (%) and nuclear yield.

Leaf sample size for cashew and internal reference standard: Three categories of leaf sample sizes (40-50, 60-80 and 90-120 mg) from young cashew leaf tissue-(not youngest) were tested to determine suitable quantity of leaf sample size that will release sufficient number of nuclei into the lysis buffer with acceptable quality. Leaf sample of 40-50 mg was used for the internal reference plants as reported in literatures (Dolezel et al., 1998; Loureiro et al., 2006a).

Incubation period for the isolated nuclei: Five incubation periods ranging from 0, 10, 20, 40-60 min after nuclei isolation were tested in this study. The incubation was done on cold ice in the dark box.

Flow cytometric procedure for the isolation of nuclei: Extraction of nuclei was done by either chopping/co-chopping of leaf sample from cashew/or cashew and internal reference standard with a sharp razor for 30 sec in a Petri dish containing about 1.0 mL extraction buffer solution (Galbraith et al., 1983). About 0.8 mL of nuclei suspension recovered was filtered through a 30 μm Partec single tube filter to remove cell fragments and large debris. Isolated nuclei in lysis buffer were stained with 4 μg mL^-1 of 4’, 6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich, Germany) and β-mercaptoethanol (2 μL mL^-1) was added to reduce polyphenolic oxidation (Dolezel and Bartos, 2005). For Otto’s buffer, staining was done in ratio of 1 part of nuclear suspension in Otto I extraction buffer to 4 parts of the Otto II staining buffer. Isolated stained nuclear suspensions were incubated for 0 (i.e., without incubation), 10, 20, 40 and 60 min (for different treatments) on cold ice in the dark before cytometric analysis. The nuclear suspension was analyzed on a Partec CyFlow^® Ploidy Analyser II (Partec, Münster, Germany) equipped with laser excitation at 488 nm as well as UV LED by a mercury arc lamp.

Experimental design and data collection: The experimental layout is a spilt-spilt plot design of three treatment factors (5 buffer systems i.e., Galbraith, LB01, Otto, Tris-MgCl₂ and DAPI-MaVI; 3 leaf sample sizes i.e., 40-50, 60-80 and 90-120 mg and 5 incubation periods i.e., 0, 10, 20, 40 and 60 min). The treatment factors were applied to the six accessions. In addition to three individual plants used for the investigation, the analysis was repeated on three different days (Dolezel et al., 2003, 2007). Data on quality of nuclear suspension in terms of fluorescence intensity (FL), coefficient of variation (CV) of G0/G1 peak (a measure of nuclear integrity and variation in DNA staining), background factor (BF) (a measure of a sample quality) and nuclear yield factor (YF) (a proportion of nuclei in suspension in respect to the sample size) were collected using a Partec Flowmax Software, Windows 2.5 version (Loureiro et al., 2006a). The computation analyses of background factor and nuclear yield followed (Loureiro et al., 2007) i.e.:

where, N is total number of particles and n is total number of intact nuclei and YF (nuclei sec^-1 mg^-1) = (n/t)/w, where n is total number of intact nuclei, t is number of run time in second and w is leaf sample size in weight (mg).

Data and statistical analyses: The data collected on the four nuclear suspension variables, i.e., fluorescence intensity, coefficient of variation (%), background factor (%) and nuclear yield (nuclei sec^-1 mg^-1) were summarized and statistically analyzed using a split-spilt plot analysis of variance method of the Genstat software Discovery Edition version 3.0 (http://www.vsni.co.uk/). The treatment means were ranked using a Duncan multiple range test (DMRT) at LSD p<0.05. The degree of difference in the performance of buffers was plotted into hierarchical dendogram using a group average linkage Euclidean distance method.

RESULTS

The analysis of variance (ANOVA) (Table 3) of the four nuclear suspension quality parameters (fluorescence intensity, background factor, nuclear yield and coefficient of variation of G0/G1 peak), showed significant difference (p<0.001) in the performances of the 5 buffers, 3 sample sizes and 5 incubation periods tested. All combination interactions of these three factors i.e., ‘buffer x sample size’, ‘buffer x incubation period’, sample size x incubation period’ and ‘buffer x sample size x incubation period’ were statistically significant. In the overall, the fluorescence intensity varies between 70 and 136 while the background noise of the nuclear suspension was between 26 and 55%. Moreover, the nuclear yield for the cashew sample was between 0.70 and 1.75 with average of 1.02 (data not shown).

Table 3:	Analysis of variance for the quality of nuclear suspension (background noise, fluorescence intensity, coefficient of variation and nuclear yield) of cashew as affected by buffer composition, leaf tissue size and incubation period of isolated nuclei
**,** and * mean statistically significant at p<0.001, p<0.01 and p<0.05, respectively, ns: Not significant

Fig. 1:

Group average linkage dendogram showing grouping of the five 5 lysis buffers based on the degree of similarity in their capability to extract nuclei for the estimation of relative nuclear DNA content and ploidy analysis in cashew. Data were derived from four parameters for grading the quality of the nuclear suspension (see materials and methods)

The coefficient of variation of the histogram peaks averaged 4.4% with a range between 3.02 and 6.02% (data not shown). The overview of descriptive statistics of the quality of nuclear suspension from cashew suggests presence of high content of secondary metabolites characteristic of woody tree plants. Among the five buffers screened a two-step Otto’s buffer system gave the best results in terms of coefficient of variation of G0/G1 peak, nuclear yield, fluorescence intensity and background noise (Fig. 2). By contrast, nuclei isolated with LB01 buffer consistently recorded the least quality. Based on the quality of nuclear suspension the five buffers grouped into two categories (Fig. 1) i.e., (a) Otto, DAPI-MaVI, Tris-MgCl₂ and Galbraith that are comparatively efficient in extracting nuclei as they fused at about 80% similarity index against less than 50% in LB01. Difference between the quality of nuclear suspensions derived from Galbraith and Tris-MgCl₂ buffer was indiscernible (less than 10% dissimilarity index). Similar performance was recorded between Otto and DAPI-MaVI buffers. This implies that either of each of the pair can be used for nuclei isolation for ploidy analysis. The unique characteristics of each of these lysis buffers have been documented (Dolezel et al., 1992, 1998; Dolezel and Bartos, 2005).

For the sample size, leaf tissue of about 70 mg (i.e., 60-80 mg) gave best quality in terms of sufficient nuclei (yield), better fluorescence signal, least background noise and lowest coefficient of variation of histogram G0/G1 peak for cashew (Fig. 2a-d). The interaction effect between buffer and amount of leaf tissue deed show that the efficiency (buffering capacity) of each buffer differs with increasing leaf sample size but quite obvious was that excess amount of leaf tissues compromise the quality of nuclear suspensions (Fig. 2a-d). For example, significant decrease in background noise was recorded for all the buffers when sample size was increased from 40-50 to 60-80 mg but further increase to 90-120 mg resulted to an increase in background noise significantly (Fig. 2a), indicating the sensitivity of lysis buffers to quantity of leaf sample size. However, the response tends to be buffer-specific for nuclei yield, fluorescence intensity and coefficient of variation of G0/G1 i.e., each buffer responded at different rate to changes in quantity of leaf sample.

Fig. 2(a-d):

Interaction effects of the quantity of leaf tissue sample and buffer composition on the quality of nuclear suspension (a) Fluorescence intensity; (b) Background factor (%); (c) Nuclear yield factor and (d) Coefficient of variation (%)) in cashew. Legend: Buffer: O: -Otto, L: LB01, M: DAPI-MaVI, G: Galbraith and T: Tris-MgCl2 Sample size: -1 = 40-50 mg, 2 = 60-80 mg and 3 = 90-120 mg)

Otto’s buffer was most sensitive (positive or negative) to sample size in terms of nuclear yield (Fig. 2c), LB01 for fluorescence intensity (Fig. 2a) and Galbraith and Tris-MgCl₂’s buffers for coefficient of variation G0/G1 (Fig. 2d).

The quality of nuclear suspension was found to improve with incubation as extracted nuclei incubated for 10 min on cold ice in the dark was better than control (0 min i.e., analysed immediately) across the buffer system. However, interaction effect of buffer and incubation time presented in Fig. 3a-d show that the optimum incubation window for efficient DNA staining of isolated nuclei from cashew would be between 10 and 15 min. The interaction effect further showed that each buffer responded differently with increased incubation period until plateau is reached and subsequent decline in quality (Fig. 3a-d).

Fig. 3(a-d):

Interaction effect of period of incubation (min) and buffer composition on the quality of DNA nuclear suspension; (a) Fluorescence intensity; (b) Background factor (%); (c) Nuclear yield factor and (d) Coefficient of variation (%) of G0/G1 peak) in cashew. Legend: Buffer:- O: Otto, L: LB01, M: DAPI-MaVI, G: Galbraith and T: Tris-MgCl2 tested Incubation period:- 1 = 0 min, 2 = 10 min, 3 = 20 min, 4 = 40 min and 5 = 60 min)

Internal reference standards for flow cytometry in cashew: Without attempting to reinvent the will, the five buffers were not tested on the reference standards as numerous reports on their flow cytometry are available in the literatures (Dolezel and Bartos, 2005).

Fig. 4(a-d):

Histograms showing relative fluorescence of nuclei at G0/G1 and G2 histogram peaks of (a) Glycine max; (b) S. lycopersicum; (c) Raphanus sativus and (d) Vicia faba using Otto’s buffer (Table 2)

However, Otto buffer that shown better performance on cashew was used for the determination of suitable reference standard. Interestingly, the buffer gave higher resolution G0/G1 peak for all the reference standards (Fig. 4). Histogram profiles of isolated nuclei from cashew leaf sample co-chopped with each of the six reference plants show that Raphanus sativus cv. Saxa, Glycine max cv. Polanka and S. lycopersicum cv. Stupicke produced G0/G1 peaks within the same range of fluorescence channels (Gains) with cashew, suggesting a relative small genome size i.e., less than 2.0 pg/2C in this plant species. Among the 3 close reference plants mentioned above, histogram G0/G1 peak of S. lycopersicum cv.

Fig. 5(a-b):

Flow cytometric histograms showing relative fluorescence intensity of nuclear suspensions obtained from (a) Cashew, A. occidentale with Raphanus sativus as an internal reference and (b) Cashew, A. occidentale with S. lycopersicum as an internal reference using Otto’s buffer (Table 2). The coefficient of variations of A. occidentale G0/G1 peak = 3.56% and S. lycopersicum G0/G1 peak = 2.85% in Fig. 5a

Stupicke was about 2 folds to that of cashew and shown to be the most suitable internal reference standard plant for the estimation of relative nuclear DNA content (genome size) in cashew (Fig. 5a). On one hand, the G0/G1 histogram peaks of Raphanus sativus cv. Saxa and cashew nearly overlap (Fig. 5b), an indication that the genome sizes of these two plant species are very close and that R. sativus may not be suitable as an internal reference standard for estimation of nuclear DNA content or genome size in cashew (Dolezel and Bartos, 2005). On the other, the histogram G0/G1 peaks of Glycine max and cashew were about 2.5 folds apart (data not shown) which can also be used as reference standard in the absence of S. lycopersicum.

DISCUSSION

Because of dearth of information on the application of flow cytometry to cashew, investigation into the buffer systems, sample size and internal reference standards as a preliminary measure to develop a flow cytometric protocol suitable for the isolation of nuclear suspension for the cashew plant species was carried out. Preparation of suspension of intact nuclei for estimation of relative or absolute DNA amount has been universally performed following the method of Galbraith et al. (1983). This flow cytometric method involves mechanical homogenization of small quantity of fresh tissue (leaf, pollen, root etc.) in the lysis buffer for the extraction of nuclei for fluorescence analysis on a Flow Cytometer. Critical components in this process are buffer and the nature of sample. Plants are known to contain many chemical compounds that can interfere with the process of nuclei release during extraction. Dolezel and Bartos (2005) remarked that the composition of the buffer is important in order to optimally isolate sufficient nuclei free of cytoplasm and secondary compounds that are capable of confounding DNA staining.

Given the different chemical compositions of the five buffers (Galbraith LB01, Otto, Tris-MgCl₂ and DAPI-MaVI) tested, each performed differently with Otto buffer being the most satisfactory in terms of quality of nuclear suspension obtained. Notwithstanding, it will too hasty to draw conclusion on the non-suitability of other buffers as the results might have been exacerbated by the presence of cytosolic compounds released during nuclei isolation that are capable of increasing background noise and give histogram with high coefficient of variation as recorded in LB01. Studies have shown evidence of high cytosolic compounds in woody plants. Particularly, cashew plant is known to contain a high content o f phenolic compounds and tannins (Lowor and Agyente-Badu, 2009). Such compounds are known to interact with nuclear DNA and/or the fluorochrome, thereby causing stoichiometric errors (Loureiro et al., 2006b, 2007) during flow cytometric analysis. In this investigation, Otto’s buffer with a 4’, 6-diamidino-2-phenylindole (DAPI) fluorochrome has shown to be most appropriate lysis medium for ploidy analysis and estimation of relative nuclear DNA content of cashew, because it gave consistent results with very low coefficient of variation of G0/G1 peak (Fig. 2d, 3d). Loureiro et al. (2006b) opined that a reliable estimate of nuclear DNA content would require a balance between the digitized fluorescence signal and the DNA content and such hypothesis would depend on the stoichiometry of dye binding to DNA, accessibility of DNA to the fluorochrome, fluorescence absorption and linearity of the instrument amplification (Bagwell et al., 1989). Plant inhibitors are other important secondary compounds that hinder DNA staining and decrease dye fluorescence of isolated nuclei (Price et al., 2000).

Another compound that may reduce quality of nuclear suspension is tannin, a common phenolic compound mostly found in various tissues of woody plants. Studies have established correlation between tannins and stoichiometric errors in the estimation of genome size in woody plants (Greilhuber, 1986, 2005; Favre and Brown, 1996 ; Loureiro et al., 2005). Incidentally, high content of tannins has been reported for cashew plant (Michodjehoun-Mestres et al., 2009). Moreover, the choice of young leaf for nuclei isolation in this investigation in addition to incorporation of β-mercaptoethanol into the buffer compositions was to minimize the effect of secondary compounds like tannins. However, Loureiro et al. (2006b) studying the effect of tannic acid on plant nuclei and nuclear DNA content had established that different nuclear isolation buffers gave variable resistance to the negative effect of tannins and concluded that the basis for such phenomenon has not been proven. Based on this buffer resistance-hypothesis, it can be adduced that least quality of nuclear suspension derived from LB01 buffer can be attributed to less resistance to confounding effect of secondary compounds like tannins. Furthermore, the improved quality of nuclear suspension from Otto and DAPI-MaVI buffer may be due to the presence of citric acid in their compositions. Dolezel and Bartos (2005) opined that citric acid has the capacity to enhance better chromatin accessibility and improve staining of the DNA in nuclei isolation medium. Corroborating the result obtained in this investigation, Emshwiller (2002) reported Otto buffer as most suitable for ploidy analysis of Oxalis, a highly acidic plant like cashew. The author however, remarked that the failure of Galbraith and LB01 buffers can probably be attributed to the fact that the buffering capacity of these buffers had been exceeded.

Two-step nature in the application of Otto’s buffer i.e., isolation of in lysis Otto I buffer before staining the isolated nuclei in Otto II may confer advantage on it over other buffers, as concentration of intact nuclei available for DNA staining will be higher in comparison to others. Excellent performance of Otto buffer has been widely reported for many species especially those with lower nuclear DNA content (genome size) (Dolezel et al., 1992, 1998; Dolezel and Bartos, 2005; Loureiro et al., 2006a, b; Loureiro et al., 2007). Although, the coefficient of variation (i.e.,<5%) of G0/G peak recorded for the Otto buffer in this investigation is acceptable for the woody plant species like cashew, further improvement is still possible to achieve better results i.e., lower CV with less background noise. Addition of compound like polvinylpyrrolidone would further reduce negative effect of cytosolic and phenolic compounds (Blondon et al., 1994; Yokoya, 2000; Thiem and Sliwinska, 2003) and improve the quality of the nuclear suspension.

In addition to the composition of buffer, size of leaf tissue was found to also influence the quality of nuclear suspension derived from cashew. Leaf tissue of about 70 mg (60-80 mg) was found ideal for the recovery of sufficient nuclei with minimal background noise. It was obvious that further increase in sample size compromised the quality of the nuclear suspension in all ramifications. Increasing cashew leaf tissue (a highly acidic plant with tannins) will amount to increasing cytosolic compounds in the isolation medium and the effect on the quality of nuclear suspension will be dire when the buffering capacity of the buffer is exceeded. This probably explains the trend observed in this study. Arumuganathan and Earle (1991) recorded similar effect of large sample size on quality of nuclear suspension in pea, papaya and citrus. Although, only leaf tissue was used during this investigation, it is however, not impossible that other tissues especially reproductive parts that may contain less phenolic compounds may be more ideal for flow cytometric analysis in cashew. The incubation window for achieving quality nuclear suspension in cashew revolves around 10 and 15 min as obtained in this study. Delay in the cytometric analysis beyond 20 min is not advisable because of the rapid degradation of intact nuclei in the suspension after the incubation time has been exceeded. Such prolonged incubation could cause a reduction in fluorescence intensity and poor-quality measurements, especially in a buffer system where lysis and staining components are not separated as in Otto buffer (Dolezel and Bartos, 2005).

Vindelov et al. (1983) and Bagwell et al. (1989) hypothesized that an ideal DNA reference standard should have a genome size close to the target species in order to avoid risks termed as nonlinearity and offset errors. In this study, of the six plant species screened as reference standards for the estimation of relative nuclear DNA content in cashew, Raphanus sativus cv. Saxa-, Glycine max cv. Polanka and S. lycopersicum cv. Stupicke, with genome sizes of 1.11, 1.96 and 2.50 pg/2C, respectively, gave G0/G1 histogram peaks in the range of 1 to 2.5 folds to G0/G1 histogram peak of cashew, an indication of relative closeness between genome size of cashew and genome sizes of these three species. Although, Raphanus sativus was the closest to cashew in terms of G0/G1 histogram peak but the two peaks almost overlap (Fig. 5b), as such, it may not be an ideal reference standard for the estimation of nuclear DNA content in this plant species. Hence, for a reliable estimation of relative nuclear DNA content of cashew, S. lycopersicum cv. Stupicke with G0/G1 peak at almost 2 folds to cashew (Fig. 5a) would be the most ideal internal reference standard for this tropical tree species. The genome size of this tomato plant is known to be constant and the plants are genetically stable (Dolezel and Bartos, 2005).

The improved performance recorded for Otto’s buffer with DAPI fluorochrome in this study makes it a suitable flow cytometric buffer system for the ploidy analysis and estimation of relative nuclear DNA content in cashew. In addition, the leaf tissue should not exceed 70 mg and isolated nuclei could be incubated for maximum of 15 min. It is important to mention that the Otto buffer and DAPI fluorochrome are not suitable for the determination of absolute genome size. For absolute genome size determination, Galbraith buffer with Propidium Iodide (PI) fluorochome has been reported as most suitable for many plant species.

ACKNOWLEDGMENTS

The author thanks Dr. J. Dolezel of the Laboratory of Molecular Cytogenetics and Cytometry, Institute of Experimental Botany, Sokolovska, Olomouc, Czech Republic for providing seeds of the reference standard plants and helpful suggestions. Thanks also go to Dr. J. Fuchs of Karyotype Evolution Research Group, IPK, Germany, for useful helps.