Random amplification of polymorphic DNA with conserved sequences reveals genome-specific monomorphic amplicons: Implications in clad identification MD. ASIM AZFER*, ANU BASHAMBOO, NASSER AHMED** and SHER ALI t Molecular Genetics Laboratory, National Institute of Immunology, Aruna Asaf Ali Marg, New Delhi 110 067, India *Department of Zoology, Aligarh Muslim University, Aligarh 202 002, India **College of Veterinary. Science, Assam Agricultural University, Khanapara, Guwahati 781 022, India tCorresponding author (Fax, 91-11-616 2125; Email, [email protected]).

The enzymatic amplification of genomic DNA with an arbitrary primer generates informative band profile useful for genome analysis. We used a set of synthetic oligodeoxyribonucleotide primers OAT15.2 (GACA)3.75, OAT18-2 (GACA)4.5, OAT24.2 (GACA) 6, OAT36 (GACA)9, comprising 4-9 consecutive units of GACA repeat, 033-15 (CACCTCTCCACCTGCC) and 033.6 (CCTCCAGCCCTCCTCCAGCCCT) for RAPD reactions of genomic DNA from different sources. The GACA based oligos of 15 and 18 base residues generated discernible genome specific amplicons whereas primers larger than 18 bases revealed smeary signals. The other oligos O33.15 and 033.6 also generated genome specific amplicons with more bands compared with those obtained from OAT15.2 or OATI8.2. The presence of OATI5.1 (GATA)3.T5 and OAT15.2 (GACA)3.75 sequences in different genomes were ascertained by independent dot-blot hybridization prior to using them for RAPD reactions. The RAPD amplicons generated by evolutionarily conserved primer(s) or sequences shared by many species may be useful for clad identification in controversial systematics, comparative genome analysis, and for establishing the phylogenetic status of an organism.

1. Introduction The sensitivity, speed and versatility of the polymerase chain reaction (PCR) have facilitated the experimental approaches to genome analysis, forensic sciences and evolutionary biology. Unlike conventional PCR, arbitrarily primed polymerase chain reaction (AP-PCR), also referred to as random amplification of polymorphic DNA (RAPD), performed with a single primer, generates discernible amplicons. The RAPD reactions are dependent on the Mg2§ concentration, length and sequence complexity of the oligo primer(s) and their annealing temperatures with target DNA (Kaemmer et al t992). Unlike restriction fragment length polymorphism (RFLP) which requires at least 100-folds intact DNA, the RAPD reaction can be performed with much smaller quantity of target sub-

strate even without prior information on the organization of genome. The primer sequences used in RAPD reaction, not shared by many species, have their applications confined to a single or limited number of genomes. On the other hand, primers based on evolutionarily conserved or "largely shared" sequences are envisaged to be useful for analysing a large number of genomes (Welsh and McClelland 1990; Williams et al 1990). In earlier studies, using a synthetic 01igo probe OAT36 with 9 consecutive units of GACA repeat motif, genome specific monomorphic band pattern was detected (Ali et al 1993). This observation led us to infer that conserved primer sequences are organized and have evolved in a unique species-specific manner and if the organizational profile of these were simultaneously uncovered in different species, it would give rise to genome specific band

Keywords. Genome analysis, conserved sequences; RAPD; monomorphic amplicons; molecular systematics; clad identification

J. Biosci., 24, No. 1, March 1999, pp 35--41. 9 Indian Academy of Sciences

35

Md. Asim Azfer et al

36

pattern. Working on this surmise, we used a set of synthetic oligodeoxyribonucleotides OAT15-2 (GACA)y75, OAT18.2 (GACA)4.5, OAT24.2 (GACA)6, OAT36 (GACA)9, O33.15 (CACCTCTCCACCTGCC) and 033-6 (CCTCCAGCCCTCCTCCAGCCCT) as primers for RAPD reactions with genomic DNA isolated from different sources. The origin of 033.15 and 033-6 primer sequences has been reported earlier (Jeffreys et al 1985). The GATAJGACA repeat motifs were uncovered during the analysis of satellite DNA in the female snakes, Bungarus fasciatus (Singh et al 1981) and Elaphe radiata (Epplen et al 1982). The amount and organization of these sequences vary considerably, and with few exceptions (John et al 1996; John and Ali 1997), they have been reported to be evolutionarily conserved (Epplen et al 1982; Singh and Jones 1986). Notwithstanding earlier reports, we have ascertained the evolutionary conservation of pure GATA/GACA repeats in different genomes by independent dot-blot hybridization. The GACA repeat sequences were positive with a large number of species that were subsequently used for RAPD reactions. The envisaged potential of this approach in the context of comparative genome analysis and clad delineation is discussed. 2.

Materials and methods

2.1

DNA

isolation

DNA was extracted from peripheral blood samples of humans (Homo sapiens), rhesus monkey (Macaca mulatta), langur (Presbytis ~entellus), cattle (Bos indicus), buffalo (Bubalus bubalis), camel (Camelus dromedarius), sheep (Ovis aries), rhinos (Rhinoceros unicornis), goat (Capra hircus), rat (Rattus norvegicus), mouse (Mus musculus), alligator (Gavialis gangeticus), bird (Columba livia), catfish (Heteropneustes fossilis) and rabbit ( Oryctolagus cuniculus) following standard procedures (Ali et al 1986). DNA from house cricket (Acheta domesticus) was isolated from the whole body excluding the alimentary canal. Rice (Oryza sativa) genomic and chloroplast DNA was obtained from leaves following standard protocols (Walbot 1988). Kangaroo (Tammar wallaby) DNA was a kind gift from Dr D W Cooper of Macquarie University, Sydney, Australia. Blood samples from one-horned rhinos R. unicornis from Jaldapara wild life sanctuary, were collected for DNA isolation with prior permission of the Union Government of India and state Government of the West Bengal. South African black rhino D. bicornis DNA was gifted by Dr Colleen O'Ryan, Department of Chemical Pathology, University of Cape Town Medical School, Cape Town, South Africa. Bacteriophage lambda (Cat. No. 301-IS, New England Biolabs, USA), Hepatitis B virus and Escherichia coti DNA samples were also used as control for dot-blot hybridization.

2.2 Oligodeoxyribonucleotide primers The primers (table 1) were purchased from Rama Biotechnologies, Hyderabad. Oligonucleotides were purified by polyacrylamide gel electrophoresis (PAGE) and reverse phase chromatography as described earlier (Ali and Epplen 1991). 2.3 Dot-blot analysis Approximately, 200 ng of heat denatured DNA samples from different sources were blotted onto nylon membrane (Pharmacia LKB, Sweden) and UV cross-linked for immobilization. The labelling of oligonucleotides using [y~2P]d-ATP, dot-blot hybridization and autoradiography were performed following standard protocols (John and Ali 1997). 2.4 Assessment of RAPD amplicons Initially, DNA samples from buffalo, cattle, goat, sheep and human two each comprising both the sexes were used for RAPD reaction. In subsequent analysis, samples from other sources listed earlier were also included. To ascertain the optimal annealing conditions, amplifications were performed at 42~ 45~ 49~ 55~ 58~ 60~ and 62~ Annealing temperatures of 60~ for oligos O33-15 (CACCTCTCCACCTGCC) and 033.6 (CCTCCAGCCCTCCTCCAGCCCT), 58~ for OATI 5-2 (GACA)3.7s and OAT18.2 (GACA)4.5, were found to be appropriate for obtaining distinct bands. RAPD reactions were carried out in a 25 lal volume containing 25 ng of genomic DNA as template, 20 pmol of primer, 2.5 units of AmpliTaq DNA polymerase (Perkin Elmer Cetus, USA), 2.5 mM MgCI2, 2001aM of each dNTP, 50 mM KCI, 20mM Tris-HCI (pH 8-3) and 0.1% Triton X-100. The reaction mixture was layered with an equal volume of mineral oil and heat denatured at 96~ for 2 min. Table 1. Details of the primers used in dot-blot and RAPD reactions with genomic DNA isolated from different sources. Primers" OAT15-1 OATI5.2 OAT18-2 OAT24.2 OAT36 033.15 033.6

Sequences(5'

3") and their lengthb

ATAGATAGATAGATA (15) ACAGACAGACAGACA (15) CAGACAGACAGACAGACA (18) GACAGACAGACAGACAGACAGACA (24) GACAGACAGACAGACAGACAGACAGA CAGACAGACA (36) CACCTCTCCACCTGC C (16) CCTCCAGCCCTC CTCCAG CCCT (22)

Barring primer OATI5:I, rest all were used for RAPD rc~Iction. b Numbers in parentheses indicate length of the oligo primers.

RAPD mediated genome specific amplicons Following this, amplification was carried out for 35 cycles comprising subsequent steps of denaturation at 94~ for 1 min, annealing at the optimal temperatures as mentioned above for l min and extension at 72~ for 1 min in a Thermal Cycler (Perkirr Elmer, Cetus, USA). On completion of the cycles, the amplified products were further incubated at 72~ for 5 min. 2.5

Agarose gel electrophoresis

Approximately 10 ktl of RAPD products were fractionated on 2% agarose gel (Pharmacia) in TAE buffer. The gels were stained with ethidium bromide and bands were photographed under UV. 2.6

Simultaneous comparison of band profiles from different genomes

The size of the fully resolved bands on the gel and overall band similarities amongst different species were ascertained on an optically enhanced densitometer using Diversity One Version I-6 software package (PDI Inc., New York, USA) at the resolution of 169 microns and tolerance value of 0.75%. Using this software, densitometric tracings of the amplicons in each lane were superimposed on top of one another for simultaneous comparison of the band profile. 3.

Results

The primers used for dot-blot and RAPD reactions are given in table 1. Dot-blot hybridization of DNA samples from different sources with two oligo probes OATI5.1 and OAT15.2 revealed that the former does not crosshybridize with buffalo, cattle, goat, sheep, bacteriophage lambda, Hepatitis B virus and E. coli DNA whereas the latter hybridizes with all the DNA samples except those of bacterial and viral origins (table 2). Thus, for all the subsequent RAPD reactions, GACA based primers in addition to O33.15 and 033.6 were used.

3.1

RAPD based amplicons

Of the several primers used for RAPD reactions with different target DNA, OAT15.2, OAT18-2, O33-15 and 033.6 revealed amplicons with varying degrees of clarity. The OAT15.2 primer with genomic DNA from buffalo, cattle, goat, sheep and human showed genome specific monomorphic amplicons with a few polymorphic bands between the different genera (figure l a). In buffalo sample, the band size ranged from 220-2340 base pairs whereas in cattle, the same ranged from 200-2270 base pairs. The RAPD amplicons in goat were clearly discernible but the same in sheep were present within

37

the smear (figure la, lanes 7 and 8) ranging, in both the species, between 350-2060 base pairs. Human samples showed bands in the range of 400-2040 base pairs within the slight smear (figure la, lanes 9 and 10). A few minor bands present in one individual (figure la, lane 9) were found to be absent in other (figure la, lane 10). This may be due to the difference in the sexes because lanes 9 and 10 represent human male and female DNA samples respectively. Despite this variation, the overall band pattern remained genome specific. Similarly, the OAT18-2 primer used for RAPD reaction with the above mentioned five DNA samples also revealed genome specific band pattern. However, number of bands detected was less compared to those obtained with OAT15.2. Amongst the five species, the band size ranged between 500-2080 base pairs (figure l b). In buffalo, two bands of about 870 and 1330 base pairs were most prominent (figure l b, lanes 1 and 2) whereas most of the other bands common to both the samples were of the reduced intensity. In cattle (figure lb, lanes 3 and 4), four strong bands of about 500, 830, 870 and 990 base pairs were shared by both the samples but the band profile compared to other species were distinctly dissimilar. Similar genome specific band patterns in goat and sheep DNA samples were also seen and two bands of about 750 and 990 base pairs were common to both the species (figure l b, lanes 5, 6, 7 and 8). These RAPD reactions were repeated independently with OAT15.2 and OAT18-2 primers at least four times and each OAT18.2 primer revealed consistent band pattern. Thus, OAT18.2 primer was found to be better in giving rise to genome specific band Table 2. Dot-blot hybridization of genomic DNA from different sources. Probes Source of target DNA

OATI5.1

OATI5.2

Human Langur Buffalo Cattle Goat Sheep Kangaroo Rabbit Rat Mouse Catfish Cricket E. coli Bacteriophage Lambda Rice chloroplast DNA Rice genomic DNA Hepatitis B virus

+ + +++ + ++ ++ ++ + + + -

++ ++ ++ ++ + + ++ + + + + + + + -

(-), Absence of hybridization signal. (+), Cross-hybridization of the probe and the intensity of the signal obtained.

38

Md. Asim Azfer et al

Figure 1. RAPD amplicons of genomic DNA from different vertebrate species including human with primer OAT15.2 (a) and OAT18.2 (b) at 58~ annealing temperature resolved on 2% agarose gel. Note the bands within the smear in sheep and human in (a) detected, with OAT15.2 primer and the same without smear in (b) detected with OAT18.2 primer. Molecular marker q~X174 DNA predigested with HaelII enzyme is given in kb,

Figure ~, Genome specific amplicons in 7 different vertebrate species detected by RAPD with primer OAT18-2 at 58~ aalarafliag.temperature, resolved on 2% agarose gel. Note, similarities in the band profile between two closely related species viz., langur and rhesus compared to others. Molecular marker ~X174 DNA digested with HaellI enzyme is given in kb.

pattern. To ensure the uniqueness of the genome specific pattern, DNA samples from several other sources were subjected to independent RAPD analysis using OAT18-2 primer. In the ethidium bromide stained agarose gel, a total of 5 discernible bands in pig, 8 each in bird and langur, 6 each in rhesus monkey and camel, 14 in rhinoceros, 13 in alligator were recorded and no two samples showed identical band pattern (figure 2). Further, simultaneous comparison of band profiles in different lanes was made by densitometric analysis of the ethidium bromide stained gel shown i n figure 2. As noted above all the samples showed distinctly different and individual specific tracing profile (figure 3). The RAPD reactions with OAT24.2 and OAT36 primers conducted under varying conditions with DNA samples from different species listed earlier showed smeary signals devoid of any band (data not shown). The other primer O33.15 showed several genome specific monomorphic bands in the range of 490-2370 base pairs in all the samples (figure 4). Of the five samples, buffalo cattle and human showed distinct bands without smeary signar whereas goat and sheep DNA samples showed bands within the smear. Monomorphic bands in the human samples were prominent. A band of about 540 base pairs was common to all the samples except cattle (figure 4, lanes 3, 4). Similarly, a 490 base pair band was common to human male and both the samples of cattle (figure 4, lanes 3,

39

RAPD mediated genome specific amplicons

4 and 9) but was absent in the human female (figure 4, lane 1 0 ) o r other animal samples. The other primer 033-6 used for RAPD reaction with DNA from 5 different species revealed almost identical bands in two individuals of the same species giving rise to genome specific band pattern. However, bands so detected even under Varying reaction conditions were present within the heavy smear (data not shown). Thus, of all the primers used for RAPD reactions, OAT18.2 was most appropriate for generating distinct genome specific monomorphic amplicons. In yet another attempt, to substantiate the genome specificity of the amplicons detected by OAT i 8-2 primer, we used 6 random DNA samples from one horned rhino R. unicornis and two f r o m South African Black rhino D. bicornis for RAPD reaction. Members of any one species showed similar monomorphic band profile but the two species taken together showed distinctly different band pattern although some of the bands were common to both the species (figure 5). Several bands of 600, 740, 800, 890, 940, 1000 and 1400 base pairs were detected in all the samples of R. unicornis with copy number variation. Similarly, distinct bands of 650, 720,

870, base 650, both

960, 1050, 1250, 1360, 1890, 2040, 2240 and 2430 pairs were seen in D. bicornis genome of which 720, 870 and 1890 base pairs were common to the species. 4.

Discussion

The RAPD approach has been used to discriminate strains of bacteria (Wang et al 1993), mouse (Welsh and McClelland 1991), fishes (Dinesh et al 1993) and for genetic mapping (Rafalski et al 1991). Recently, this approach was adopted for the assessment of sequence polymorphisms in plant genomes (Kaemmer et al 1992), endangered species (Sulaiman and Hasnain 1996; Ali et al 1998) and tumour tissues (Afroze et al 1998). Despite frequent use of evolutionarily conserved repeat sequences in RAPD reactions (Meyer et al 1993), GATA/GACA repeat motifs have been used only in a few recent studies (Hering and Nirenberg 1995; Lorenz et al 1995). Recent reports indicate that GACA repeat motif is not conserved evolutionarily (John et al 1996; John and Ali 1997). However, this seems to be shared

Figure 3. Simultaneous comparison of band profiles detected by RAPD with OATI8.2 primer (shown in figure 2). Note the unique densitometric tracings of the bands in each lane representing an individual animal species.

40

Md. Asim Azfer et al

by a number of distantly related organisms (table 2), proving this to be an attractive candidate to be used as primer for RAPD mediated genome analyses. Oligo primers OAT15.2, OAT18.2 and O33.15 revealed genome specific monomorphic amplicons in a number of closely as well as distantly related species but distinctly discernible bands devoid of smear were generated with GACA based primer OAT18-2. Detection of unique RAPD amplicons in some lower organisms with GATA/GACA primers (Meyer et al 1993; Hering and Nirenberg 1995; Lorenz et al 1995) and genome specific amplicons with GACA primers in higher vertebrates in the present study suggests that these repeat motifs have evolved in genome specific manner. This corroborates our earlier observation (Ali et al 1993). Minisatellite sequences have been implicated with high rate of recombirmtorial activities leading to sequence polymorphism in the genome (Royle et al 1988). We have no direct evidence to substantiate this point, but our earlier observation (Ali et al 1993) and the present study on RAPD with pure GACA based primer on two species of rhinos (see figure 5) suggest that perhaps this repeat motif does not participate in recombinatorial activities in all the vertebrates. In case of R. unicornis, since all the six samples were obtained from a single

Figure 4. Genome specific RAPD amplicons from different species of vertebrates including human detected with primer O33.15 annealed at 60~ resolved on 2% agarose gel. Note the band profile within the smeary background in goat and sheep samples and clear bands in buffalo cattle and human samples. Molecular size marker r X174 DNA digested with HaelIl enzyme is given in kb.

isolated habitat, the high level of band similarity may be attributed to frequent inbreeding of these animals. However, the minor band variation between the two samples of human (figure la, lanes 9 and 10) indicating putative polymorphism may be due to the sex difference because differential accumulation of the GACA repeat sequences in the two sexes have been reported (Epplen 1988). This negligible intra-species variation, however, does not affect the overall genome specific band profiles. The paucity of GATA repeat motif (table 2) in bacteria and viruses has been reported. However, its absence in the members of Bovidae family (cattle, buffalo, sheep, and goat) corroborating our earlier observation (John and Ali 1997) was startling because these sequences were reported to be evolutionarily conserved in all the eukaryotes (Singh et al 1981; Singh and Jones 1986; Epplen 1988). Owing to genome specific amplicons generated with GACA based primer(s) in RAPD reaction and its presence in a number of species, we envisage its potential use in the comparative genome analysis and clad identification. However, it would be unattainable to accurately position a clad based on the amplicons generated by a single repeat motif. Thus, for more accurate clad positioning or phylogenetic analyses, several coding and non-coding loci may simultaneously be analysed. Nonetheless, present study on generating genome specific

Figure 5. Genome specific monomorphic amplicons in two different species of rhinos detected by RAPD with primer OATI 8.2 at 58~ annealing temperature, resolved on 2% agarose gel. Note the closer affinities in the band profile amongst different animals' within a species and distinctly different band pattern between the two species rhinos if taken together, despite some common bands seen in both the species. Molecular marker ~bX174 DNA digested with HaeIII enzyme is given in kb.

R A P D mediated genome specific amplicons

amplicons by identifying primer(s) shared amongst a large number o f species highlights the importance of this approach in molecular systematics o f advanced eukaryotes.

Acknowledgements This study was supported by grant No. SP/SO/D-51/93 from the Department o f Science and Technology, New Delhi to SA and a core grant from the Department of Biotechnology, to the National Institute of Immunology, New Delhi. We thank Dr Colleen O'Ryan, Department o f Chemical Pathology, University of Cape Town Medical School, Cape Town, South Africa for D. bicornis DNA, Dr Sudhanshu Vrati for Hepatitis B viral DNA, and Dr Seyed E Hasnain for his comments on the manuscript and Mr K S Negi for technical assistance.

References Afroze D, Misra M A, Sulaiman I M, Sinha S, Sarkar C, Mahapatra A K and Hasnain S E 1998 Genetic alteration in brain tumors identified by RAPD analysis; Gene 206 45---48 Ali S and Epplen J T 1991 DNA fingerprinting of eukaryotic genomes by synthetic oligonucleotide probes; Indian J. Biochem. Biophys. 28 1-9 Ali S, Gauri and Bala S 1993 Detection of genome specific monomorphic loci in Bos taurus and Bubalus bubalis with oligodeoxyribonucleotide probes; Anim. Genet. 24 199-202 Ali S, Ansari S, Ehtesham N Z, Azfer M A, Homkar U, Gopal R and Hasnain S E 1998 Analysis of the evolutionarily conserved repeat motifs in the genome of the highly endangered central Indian Swamp deer Cervus duvauceli branderi; Gene 223 361-367 Ali S, Miiller C R and Epplen J T 1986 DNA fingerprinting by oligonucleotide probes specific fbr simple repeats; Hum. Genet. 74 239-243 Dinesh K R, Lim R'M, Chua K L, Chan W K and Phang V P 1993 RAPD analysis: An efficient method of DNA fingerprinting in fishes; Zool. Sci. 10 849--854 Epplen J T, McCarrey J R, Sutou S and Ohno S 1982 Base sequence of a cloned snake W chromosome DNA fragment and identification of a male-specific putative mRNA in the mouse; Proc. Natl. Acad. Sci. USA 79 3798-3802 Epplen J T 1988 On simple repetitive GATAJGACA sequences in animal genomes: a critical appraisal; J. Hered. 79 409417 Hering O and Nirenberg H I 1995 Differentiation of Fusarium sambucinum, Fuckel sensulato and related species by RAPDPCR; Mycopathology 129 ! 59-164

Jeffreys A J, Wilson V and Thein S L 1985 Hypervariable minisatellite regions in human DNA; Nature (London) 314 67-73 John M V, Parwez I, Sivaram M V S, Mehta S, Marwah N and Ali S 1996 Analysis of VNTR loci in fish genomes using synthetic oligodeoxyribonucleotide probes; Gene 172 191-197 John M V and All S 1997 Synthetic DNA based genetic markers reveal intra and inter-species DNA sequence variability in the Bubalus bubalis and related genomes; DNA Cell Biol. 16 369-378 Kaemmer D, Afza R, Weising K, Kahl G and Novak F J 1992 Oligonucleotide and amplification fingerprinting of wild species and cultivars of Banana (Musa spp); Bio-Technol. 10 1030--1035 Lorenz M, Partensky F, Bomer T and Hess W R 1995 Sequencing of RAPE) fragments amplified from the genome of the prokaryote Prochlorococcus marinus (Prochlorophyla); Biochem. Mol. Biol. Int. 36 705-713 Meyer W, Mitchell T G, Freedman E Z and Vilgalys R 1993 Hybridization probes for conventional DNA fingerprinting used as single primer in the polymerase chain reaction to distinguish strains of Cryptococcus neoformans; J. Clin. Microbiol. 31 2274-2280 Rafalski J A, Tingey S V and Williams J G K 1991 RAPD markers--a new technology tbr genetic mapping and plant breeding; Agric. Biotech. News Inform. 3 645--648 Royle N J, Clarkson R E, Wong Z and Jeffreys A J 1988 Clustering of hypervariable minisatellite in the proterminal regions of human autosomes; Genomics 3 352-360 Singh L and Jones K W 1986 Bkm sequences are polymorphic in humans and are clustered in pericentric regions of various acrocentric chromosomes including the Y; Hum. Genet. 73 304-308 Singh L, Purdom I F and Jones K W 1981 Conserved sex chromosome associated nucleotide sequences in eukaryotes; Coil Spring Harbor Symp. Quant. Biol. 45 805-813 Sulaiman I M and Hasnain S E 1996 Random amplified polymorphic DNA (RAPD) markers reveal genetic homogeneity in the endangered Himalayan species Meconopsis paniculata and M. simplicifolia; Theor. Appl. Genet. 93 91-96 Waibot V 1988 Preparation of DNA from single rice seedlings; Rice Genet. News 1,5 149-151 Wang G, Whittam T S, Berg C M and Berg D E 1993 RAPD Arbitrary Primer PCR is more sensitive than multiiocus enzyme electrophoresis for distinguishing related bacterial strains; Nucleic Acids Res. 21 5930-5933 Welsh J and McClelland M 1990 Fingerprinting genomes using PCR with arbitrary primers; Nucleic Acids Res. 18 7213-7218 Welsh J and McClelland M 1991 Genomic fingerprinting using arbitrarily primed PCR and a matrix of pairwise combinations of primers; Nucleic Acids Res. 19 5275-5279 Williams J G, Kubelik A R, Livak K J, Rafalski J A and Tingey S V 1990 DNA polymorphisms amplified by arbitrary primers are useful as genetic markers; Nucleic Acids Res. 18 6531---6535

M S received 20 October 1998; accepted 10 D e c e m b e r 1998

Corresponding editor: SEYED E HASNAIN

41