Random Amplified Polymorphic DNA (RAPD) fingerprinting offers a rapid and efficient method for generating a new series of DNA markers in fishes. Three oligonucleotide primers (two 10-mers and one 9-mer) and their paired combinations were found to generate different but reproducible RAPD fingerprints in the guppy. Of these, a 10-mer primer (designated S3D2) was used to detect DNA polymorphisms in two guppy varieties, Green Snakeskin (GSS) and ¾ Black (¾B). High Genetic Similarity (SI) was found among individuals of the GSS and ¾B varieties indicating low intra-variety genetic variability. The average SI values for the Green Snakeskin and ¾ Black varieties were 0.78 ± 0.104 and 0.81 ± 0.083, respectively. The average SI value between individuals of the GSS and ¾B varieties was 0.66 ± 0.066, indicating higher genetic variability between the two varieties. To study the inheritance of RAPD markers, single-pair crosses were set up between males of the GSS variety and females of the ¾B variety. The S3D2 primer was used to generate RAPD fingerprints of the parents and their F1 offsprings. A total of 14 RAPD markers were scored from these crosses. Of these markers, eight (60.0%) of them were polymorphic. The RAPD markers were shown by the F1 to exhibit dominant Mendelian inheritance and could thus be used for subsequent genetic linkage mapping of the guppy.
INTRODUCTION
The recent method of revealing DNA-based polymorphisms by Random Amplified Polymorphic DNA (RAPD) fingerprinting, which involves PCR amplification of genomic DNA using a single primer of arbitrary nucleotide sequence, was reported in 1990 [20, 24]. Since then the RAPD method has been applied to a wide range of organisms and its advantages over RFLP fingerprinting method have been extensively reviewed [1–3, 10, 12, 14, 17, 20, 22, 24, 26]. However, there are few reports on RAPD fingerprinting of fishes. RAPD fingerprinting has been used by Dinesh et al. [56–7] for detection of DNA polymorphisms in some fish species including color varieties of the guppy (Poecilia reticulata) and tiger barb (Barbus tetrazona). Kubota et al. [15] used this method to detect radiation-induced DNA damages in the medaka, Oryzias latipes. Recently, Johnson et al. [13] identified 721 RAPD polymorphisms between two laboratory strains of the zebrafish, Danio rerio, and of these, 401 of them were used to construct a genetic linkage map [19]. The present study so far is the only one which demonstrates Mendelian inheritance of RAPD markers in fish.
The guppy is the most important aquarium fish species cultured in Singapore with many different color varieties being bred [9]. Preliminary work on some guppy varieties showed that, with suitable arbitrary primers, intra- and inter-variety genetic variability could be detected [6, 7]. In this study, RAPD fingerprinting was used for the analysis of genomic polymorphisms in two varieties of guppies. The inheritance of RAPD markers in the F1 progeny of single-pair crosses between the two guppy varieties was studied as it is vital to demonstrate that RAPD markers follow Mendelian pattern of inheritance before they could be used for constructing a genetic linkage map for the guppy.
MATERIALS AND METHODS
Genetic material
Random samples of two guppy varieties, Green Snakeskin (GSS) and ¾Black (¾B), which are maintained as separate stocks in a farm in Singapore, were obtained. Genomic DNA from individual whole fish was extracted according to the method of Dinesh et al. [7] from 30 individuals of each of the GSS and ¾B varieties. Three single-pair crosses of GSS males with ¾B females were set up in separate glass tanks. F1 broods from each cross were separated from their parents and raised for one month. For each cross, DNA was isolated from both parents as well as from two male and two female one-month old F1 progeny.
RAPD primers and their paired combinations
The three single oligonucleotide primers tested were S11D6 (TTGCGTCCA, 9-mer), S10D5 (AGGTCACTGA, 10-mer) and S3D2 (AATCGGGTCG, 10-mer) with a G+C content of 50–60%. Three paired combinations of these single primers were prepared by mixing equal amounts of any two of them. These RAPD primers were evaluated by using them to generate RAPD profiles of DNA from a single GSS guppy. The single primer, S3D2, was used for generating RAPD profiles in subsequent experiments.
RAPD amplification conditions
The PCR reaction mixture consisted of 1 μg genomic DNA template, 1 μM single primer or paired combination, 1X PCR buffer, 1.5 mM MgCl2, 200 μM of each dNTPs and 2 units of Taq DNA polymerase (Promega) in a total volume of 50 μl. The final reaction mixture was overlaid with 40 μl of mineral oil. Amplifications were performed in a GeneAmp PCR System 9600 (Perkin-Elmer Cetus, USA) with the low stringency temperature profile of 94°C for 3 min, 30°C for 3 min and 72°C for 2 min. A total of 30 cycles were carried out, with a final extension at 72°C for 10 min. Each 10 μl sample of amplified DNA products was mixed with 2 μl of PCR dense dye consisting of 50 mM EDTA, 30% glycerol, 0.25% xylene cyanol and 0.25% bromophenol blue for electrophoresis. The amplified products were separated using Urea-SDS-PAGE gel electrophoresis adapted from Dinesh et al. [6]. The samples were loaded onto 3% stacking and 10% resolving polyacrylamide slab gels of 0.50 or 0.75 mm thickness. Electrophoresis was performed at 100 V for 3.5 hr in the Mini-Protean II (Bio-Rad, USA) electrophoresis set or 12 hr in the Sturdier SE 400 (Hoefer, USA) unit. The gels were stained with silver nitrate and photographed under illuminated fluorescent light, using a Nikon F-501 camera which was fitted with a 55 mm f2.8 Micro-Nikkor lens and Kodak TMAX 100 film.
Analysis of RAPD data
Genetic similarity (SI) values between the RAPD profiles of any two individuals on the same gel were calculated from RAPD markers scored manually by three persons, based on the presence or absence of bands of the same molecular sizes according to the method of Nei and Li [18]. RAPD markers not identified by all three persons were considered as non-scorable.
where NAB is the total number of RAPD bands shared by individuals A and B,NA and NB are the total number of bands scored in individuals A and B, respectively.
SI values range from 0 to 1. When SI=1.0, the two DNA profiles are identical and when SI=0.0, there are no common bands between the two profiles. We used the Bioprofil (Bio-1D), a charged couple device (CCD) video camera imaging system (Vilber Lourmat, France), to counter check and verify the molecular weights of the RAPD fragments. The following notations were used for identification of RAPD bands: C refers to common, nonpolymorphic bands which are shared by all individuals in a variety or family; P denotes polymorphic band which is present only in some individuals of a variety or family; and BP is used for RAPD marker present in most of the ¾B variety but absent in GSS.
RESULTS
DNA profiles generated by RAPD primers and their paired combinations
Consistent and reproducible DNA profiles were generated from the three primers and their pairwise combinations (Fig. 1). RAPD products below 2.3 kb were adequately resolved. Among the single primers, the two 10-mers, S3D2 and S10D5 gave DNA profiles with more bands than the 9-mer, S11D6 (Table 1). The banding patterns generated with combination of two primers were different from those generated by the respective single primers. They were more complex, with more lower molecular weight bands as well as ‘minor’ bands. This result is expected as the average distance between the priming sites in the DNA template is likely to be less when two oligonucleotide primers are used than when only one primer is present [21]. The single 10-mer, S3D2, was used for subsequent experiments as it produced the most scorable bands (14 in number).
Fig. 1
RAPD profiles of a single male GSS guppy generated by three single primers and their paired combinations. The PCR products were resolved in Urea-SDS-PAGE gel and visualized by silver staining. Lanes 1 and 2=S3D2+S10D5 primers; lanes 3 and 4=S3D2+S11D6; lanes 5 and 6=S10D5 + S11D6; lane 7=negative control; lanes 8, 9 and 10=single primers S3D2, S10D5 and S11D6, respectively; lane 11=Lambda BstE II markers (bp).
* represents the S3D2 primer (lane 8) that was used for fingerprinting in subsequent experiments in this study.
Table 1
Single primers and paired combinations of primers used to generate RAPD profiles from DNA of a single guppy (refer to Fig. 1)
Intra- and inter-variety polymorphisms
A DNA polymorphism is detected by band presence versus band absence and may be caused by failure to prime a site in some individuals due to nucleotide sequence differences or by insertions or deletions between priming sites [4]. Non-polymorphic or monomorphic refers to a DNA marker which is present in all individuals in the sample. Figure 2 showed the RAPD fingerprints of eight males of the GSS variety generated by the S3D2 primer. They shared six non-polymorphic bands (C1–C5 and C7) (Table 2). The eight GSS males, except for two individuals (Fig. 2, lanes 7 and 8), could be identified from each other by at least five scorable polymorphic bands (P1–P5). The high degree of similarity between the RAPD profiles of individuals 7 and 8 (SI=0.98) could be indicative of their close relatedness. The SI values obtained by pairwise comparisons of the RAPD profiles of any two GSS individuals in Figure 2 ranged from 0.60 to 0.98, with an average of 0.78 ± 0.104 (Table 3). The high SI values indicated low genetic variability among individuals of the GSS variety.
Fig. 2
DNA profiles of eight male GSS guppies (lanes 2–9) generated by the 10-mer, S3D2 (AAT CGGGTCG). Lane 1=negative control and lane 10=Lambda BstE II markers. At least six scorable non-polymorphic bands i.e. present in all individuals, C1–C5 and C7 (arrowhead) and five obvious polymorphic bands (present only in some individuals), P1–P5 (triangle) were identified.
Table 2
Summary of RAPD bands identified in the DNA profiles of individuals in Figs. 2 to 6
Table 3
Genetic Similarity Index (SI) values calculated for Green Snakeskin (GSS) guppy by pairwise comparison of eight individual fishes using the method of Nei and Li [18], (refer to Fig. 2)
Figure 3 showed that the RAPD profiles of five males of the GSS (these individuals were different from those GSS in Fig. 2) and four males of the ¾B variety. These nine individuals shared six non-polymorphic bands (C1–C5 and C7). These non-polymorphic bands may be useful for identification at the species level. The RAPD band identified as P1 in Figure 2 was present in all individuals in Figure 3. The P2 band was polymorphic in both GSS and ¾B. The P3 band was present in all GSS individuals but was polymorphic in ¾B. The P4 band was polymorphic in GSS but absent in ¾B. The P5 band showing high staining intensity was present in four of the five GSS individuals but was present as a band of lower staining intensity in all four of the ¾B individuals. The BP1 and BP2 polymorphic bands of ¾B individuals were absent in GSS guppies (Fig. 3). However, the BP bands were not specific to the ¾B variety since it was absent in some other ¾B individuals in other gels (Figs. 4 and 6).
Fig. 3
DNA profiles of five GSS males (lanes 2–6) and four ¾B males (lanes 7–10) generated by S3D2 primer. Lane 1=Lambda BstE II markers. At least six distinct common bands, C1–C5, and C7 (arrowhead) were present in both GSS and ¾B individuals. In addition, the P1 band (triangle) is also present in all individuals. P2–P5 (triangle) are the four polymorphic bands in GSS and ¾B. The BP1 and BP2 bands (triangle with enclosed dot) polymorphic in ¾B indivuduals were absent in GSS.
Fig. 4
Cross 1: DNA profiles of the first family generated by the S3D2 primer, consisting of the parents (lane 1=GSS male, and lane 2=¾B female) and their F1 progeny (lanes 3 and 4=males; lanes 5 and 6=females). Lane 7=negative control and lane 8=Lambda BstE II marker. Ten distinct common bands, C1–C7 (arrowhead) and P1, P2 and P5 (triangle), present in both parents were inherited by all four F1 offspring. The maternal band, BP1 (triangle with enclosed dot) was present in all the F1 individuals.
To assess the level of genetic variability within the GSS and ¾B varieties (Fig. 3), pairwise comparisons of the DNA profiles of GSS individuals and those of ¾B individuals were made (Table 4). The SI values for the GSS variety were high, ranging from 0.68 to 0.92 (average SI=0.79 ± 0.083) and those for the ¾B variety were also high ranging from 0.68 to 0.92 (average SI=0.81 ± 0.083). The high average SI value for GSS (0.79 ± 0.083) corresponded very well to the average SI value (0.78 ± 0.104) of individuals of the GSS variety obtained from Figure 2. These high SI values reflected low genetic variabilty within both the GSS and ¾B varieties.
Table 4
Genetic Similarity Index (SI) values calculated within Green Snakeskin (GSS) and ¾Black (¾B) and between GSS and ¾B varieties by pairwise comparisons using the method of Nei and Li [18], (refer to Fig. 3)
To estimate the genetic variability between the two varieties, SI values were computed by pairwise comparisons of the RAPD profile of a GSS individual with that of a ¾B individual (Fig. 3). The inter-variety SI values ranged from 0.55 to 0.85 with an average of 0.66 ± 0.066 (Table 4). Thus, there was greater genetic variability between the two varieties than within each of the varieties.
Inheritance of RAPD markers by the F1 progeny
RAPD fingerprints of three single-pair crosses (Cross 1, 2 and 3), each consisting of a GSS male parent, a ¾B female parent and four F1 offsprings (two males and two females) were obtained (Figs. 4–6) (Table 5).
Table 5
Inheritance of RAPD markers in three familes of single-pair crosses of Green Snakeskin (GSS) male and ¾Black (¾B) female and the four F1 offsprings (refer to Figs. 4–6)
In Cross 1, the GSS male parent, ¾B female parent and the four F1 progeny shared ten non-polymorphic bands (C1–C7, P1, P2 and P5) (Fig. 4). The single maternal band (BP1) of Cross 1 was inherited by all four F1 offspring, indicating homozygosity of this dominant band in the ¾B mother of this cross.
Figure 5 showed the parents and F1 offsprings of Cross 2. The bands common to both parents (C1–C5, C7, P1 and P3) were consistently inherited by all four F1 offspring. The single paternal band (P2) were inherited by two F1 progeny. There were three maternal bands in Cross 2 of which BP1 and BP2 segregated among the F1 progeny, while P5 band was inherited by all F1.
Fig. 5
Cross 2: DNA profiles of the second family generated by S3D2. The labelling of the lanes are as in Fig. 4. Eight distinct non-polymorphic bands, C1–C5 and C7 (arrowhead) and P1 and P3 found in both parents, were inherited by all four F1 progeny. The paternal band, P2 (triangle) was inherited by two F1 individuals. The maternal bands, P5, was present in all the progeny while the BP1 and BP2 maternal bands (triangle with enclosed dot) showed segregation among the progeny.
Fig. 6
Cross 3: DNA profiles of the third family generated by S3D2. The labelling of the lanes are as in Fig. 4. Seven common bands, C1–C5 and C7 (arrowhead) and P3 (triangle) present in both parents were inherited by all F1 progeny. The paternal bands P4 (triangle) was inherited by all the progeny, while P2 segregated among the F1. The P1 and P5 maternal bands segregated among the F1.
In Cross 3, there were seven non-polymorphic bands (C1–C5, C7 and P3) shared by mother, father and F1, demonstrating complete penetrance of these dominant RAPD markers (Fig. 6). Of the two paternal bands, P4 was inherited by all the F1 while the P2 band showed segregation among the F1. The P1 and P5 maternal bands segregated among the offsprings.
Although the sample size for the crosses was small, all the non-polymorphic bands (C1–C5 and C7) present in both parents of the crosses were inherited by all the F1 thus showing full penetrance. The polymorphic paternal and maternal bands showed dominant Mendelian pattern of inheritance in the F1 progeny. The polymorphic paternal or maternal bands homozygous in the individual parents were inherited by all F1 progeny while the bands heterozygous in the parents segregated among the F1. The DNA profiles of the F1 progeny of all three crosses did not show any nonparental bands.
DISCUSSION
The results showed that the RAPD technique gave consistent and reproducible DNA markers in different gels. Combinations of primers could potentially be used to increase the number of DNA-based genetic polymorphisms that can be detected by a set of single arbitrary primers [16, 21]. The differences in RAPD profiles generated by a paired combination from its respective single primers may be due to competition among the generated products during extension [14] or the unfavourable presence of hairpins in the RAPD products generated by the single primers [22]. Hence, the RAPD technique can be tailored to produce banding patterns of varying degrees of complexity. Such ‘fingerprint tailoring’ is useful in gene mapping and genotyping [2, 13].
Both intra- and inter-variety genetic polymorphisms were easily detectable and these RAPD markers could be used in the identification of individual guppies which hitherto was not possible except with the availability of known DNA fingerprinting probes. RAPD markers could also be used as strain or variety markers as shown in a recent study on the zebrafish [13]. RAPD profiles could also be used to assess genetic variability and inbreeding levels in stocks and populations [10, 11]. This is particularly relevant to ornamental fishes like the guppy, where numerous varieties are bred and most of them could be highly inbred being usually derived from a few founding parents as well as having undergone many generations of stringent selective breeding for desirable ornamental traits [9]. In Singapore, the guppy fish farmers have to culture 10–15 varieties to meet market demand for fancy varieties and this practice inevitably leads to restriction of population size for each variety due to various economic constraints. Results from this study confirm these expectations since the average SI values for the GSS and ¾B varieties were high (0.78 ± 0.104 and 0.81 ± 0.083, respectively) reflecting low genetic variabilty within each of these two varieties. The SI values between individuals of the two varieties were generally lower than those of intra-variety values with the average value being 0.66 ± 0.066. This again is expected since the GSS and ¾B varieties have distinctly different color patterns having undergone many generations of artificial selection by the farmer for different morphological features especially that of color and are cultured as separate lines to maintain purity of the stocks.
Since RAPD markers from the analysis of the F1 progeny of the three crosses have been demonstrated to show full penetrance and to follow dominant Mendelian inheritance, numerous RAPD markers can be generated readily and used in pedigree studies [8, 11, 16–17, 22] and for rapid construction of a genetic linkage map [19]. The linkage map can also provide DNA markers for sex determination, and also RAPD bands linked to color pattern, disease resistance, immune response and other important qualitative as well as quantitative traits and these can be used for linkage selection.
