Meiotic segregation and recombination of the intergenic spacer of the ...

( Springer-Verlag 1996

Curr Genet (1996) 30: 332—337

OR I G I N A L P AP E R

Marc-Andre{ Selosse · Guy Costa · Ce{ line Di Battista Franyois Le Tacon · Francis Martin

Meiotic segregation and recombination of the intergenic spacer of the ribosomal DNA in the ectomycorrhizal basidiomycete Laccaria bicolor

Received: 9 February/24 April 1996

Abstract The aim of this study was to clarify the inheritance of the nuclear ribosomal DNA (rDNA) in the ectomycorrhizal basidiomycete ¸accaria bicolor S238N in order to resolve inter- and within-strain relationships in forest ecosystems. PCR amplification of the intergenic spacer (IGS) was carried out in the dikaryotic mycelium and its haploid progeny. In the dikaryotic mycelium, multiple amplification products were produced for the 25s/5s (IGS1) and 5s/17s (IGS2) intergenic spacers. The 4.5- and 4.0-kb fragments of IGS2 (haplotypes a and b, respectively) were observed to occur in a 1 : 1 ratio within the haploid progeny as a result of divergent IGS haplotypes in the two separate nuclei. Recombinant monokaryons having both types of IGS2 occurred at a low frequency (6.5%; 60 kb per centimorgan) during meiosis. Haplotypes a and b of IGS1 cross-hybridized forming heteroduplexes during the PCR temperature cycle. The two IGS1 haplotypes differed only by the repeat number of a TA C motif 2 3 and co-segregated with the IGS2 haplotypes. Heteroduplex formation and IGS polymorphism provide information that is helpful in distinguishing between introduced exotic ¸. bicolor S238N and indigenous populations of ¸accaria spp. in forest ecosystems. Key words ¸accaria bicolor · Basidiomycete · Intergenic spacers · Ribosomal DNA · Meiotic recombination · Heteroduplex M.-A. Selosse1 · G. Costa2 · C. Di Battista · F. Le Tacon · F. Martin ( ) Equipe de Microbiologie Forestie`re, Centre de Recherche de Nancy, Institut National de la Recherche Agronomique, F-54280 Champenoux, France 1 On leave from Ecole Nationale du Ge´nie Rural, des Eaux et Foreˆts, France 2 Present address: Laboratoire de Biologie Cellulaire Ve´ge´tale et Valorisation des Espe`ces Ligneuses, Universite´ de Limoges, 123, rue Albert Thomas, F-87060 Limoges, France Communicated by K. Esser

Introduction Because of their mutualistic symbiosis with trees, strains of ectomycorrhizal fungi are currently used for large-scale inoculation of seedlings in nurseries and plantations (Le Tacon et al. 1992). PCR amplification of the rDNA intergenic spacers has been used to trace introduced exotic strains (e.g. ¸accaria bicolor S238N) in local populations of ectomycorrhizal fungi (Henrion et al. 1994 b). However, it is as yet unclear whether the introduced strains interbreed with local fungal populations. Knowledge of potential strain introgression would contribute to more effective management of the ectomycorrhizal inoculation. This requires a sensitive measure of genetic relationships between strains and within their progenies. Nuclear rDNA provides useful inter- and intra-specific polymorphisms with which to type eukaryotic organisms. There are multiple copies of the ribosomal genes, which are arranged as head-to-tail repeats separated by non-coding spacers. They are located at one or more loci and have identical sequences in a given organism due to concerted evolution (Srivasta and Schlessinger 1991). In higher fungi, the 25s/17s intergenic spacer (IGS; see Fig. 1 A) is known to present length and sequence variations within a species (Henrion et al. 1992; Molina et al. 1993; Iraiabal and Labare`re 1994). PCR/RFLP analysis of the IGS in ¸. bicolor revealed variation indicative of sequence polymorphism among isolates in this species (Henrion et al. 1992; Di Battista et al. 1996). In the present study, we report a segregation analysis of 25 s/5s (IGS1) and 5s/17s (IGS2) spacers within progeny of ¸. bicolor S238N. We also describe how amplification patterns revealed the presence of different rDNA haplotypes in the two nuclei, and also resulted in the appearance of informative heteroduplexes between IGS1 haplotypes. .

JCG233

333

Materials and methods Strains, growth and breeding. A detailed description of the origin of the American strain S238N of ¸. bicolor (Maire) P.D. Orton used in this analysis was provided by Di Battista et al. (1996). Vegetative dikaryotic mycelium was grown on Pachlewski medium according to Henrion et al. (1992). Spores were obtained from a sporophore collected under Pseudotsuga menziesii, inoculated with ¸. bicolor strain S238N in a glasshouse (Le Tacon et al. 1992) and germinated according to Fries (1983). Clamped mycelia, probably resulting from dikaryotizing events, were discarded and 45 monokaryons were used in this study (see Table 1). Identification of the sporophore and analysis of the origin of the spores were both ascertained by the presence of specific RAPD markers (Di Battista et al. 1996) and PCR/RFLP of IGS1 (see below). To obtain reconstituted dikaryons, agar dishes containing Pachlewski medium were inoculated with single 4]4-mm plugs of the appropriate monokaryons placed 10 mm apart. After mycelial growth, different plugs from the contact zone were subcultured and screened for the appearance of clamped hyphae. Clamped mycelia were then subcultured three times to check eventual overgrowth by remnant monokaryotic hyphae. The presence of the ¸. bicolor S238N-specific RAPD markers (Di Battista et al. 1996) in the three subcultured myceliums was checked to ensure their origin and ploidy, but only one subculture was kept for this study. Five crosses were made, namely H16]H92, H22]H92, H152]H158, H152]H206 and H234]H306. DNA extraction and PCR amplification. Total DNA was extracted by the hexadecyltrimethylammonium bromide (CTAB) /proteinase K method essentially as described by Henrion et al. (1994 a). Amplification of the 25s/5s spacer (IGS1; see Fig. 1 A) was carried out using the primers CNL12 (5@-CTGAACGCCTCTAAGTCAG) and 5SA (5@-CAGAGTCCTATGGCCGTGGAT) with priming sites at the 3@ end of the 25s gene and the 5@ end of the 5s gene, respectively (Henrion et al. 1992). The sequence containing the 5s rDNA plus the 5s/17s spacer (IGS2, see Fig. 1 A) was amplified with the primers rev5SA (5@-ATCCACGGCCATAGGACTCTG), the reverse complement of primer 5SA, and revNS1 (5@-GAGACAAGCATATGACTAC), the reverse complement of primer NS1 (White et al. 1990), with priming sites at the 5@ end of the 5s gene and the 5@ end of the 17s gene, respectively. Primers were supplied by Bioprobe Systems (Montreuil-sous-Bois, France) and the ¹aq DNA polymerase by Applige`ne (Strasbourg, France). For IGS2 amplification, the dNTP concentration was increased to 400 lM. Reactions were performed in an Eurogentec ThermoJet under the following regime: initial denaturation at 94 °C for 3 min, followed by 35 cycles of denaturation at 94 °C for 1 min, annealing at 50 °C for 30 s, and extension at 72 °C for 5 min, with a final extension at 72 °C for 10 min. Mendelian segregation among the progeny was ensured using a 0.93-kb RAPD marker obtained with primer 174 (5@-AACGGGCAGC; see Di Battista et al. 1996). We included negative controls (no DNA template) for each set of experiments to test for the presence of DNA contamination. Denatured PCR products. Amplified IGS1 of the different ¸. bicolor S238N monokaryons was purified using a Qiaquick Spin PCR Purification kit (Qiagen, Dusseldorf, Germany). Pairs of purified IGS1 in de-ionized water were mixed in equal amounts for testing heteroduplex formation. Denaturation-renaturation of mixed PCR products mimicked a PCR cycle of 94 °C for 1 min, 50 °C for 30 s, and 72 °C for 2 min. Enzymatic digestion and electrophoresis of amplified products. The mung bean nuclease assay was carried out according to Jensen and Straus (1993). Amplified IGS1, digestion and denaturation-renaturation products were separated by electrophoresis using 8% acrylamide gels in a 1]Tris-borate EDTA as previously described (Henrion

et al. 1992). The larger IGS2 was separated on a 1% agarose gel in a 1]Tris-borate EDTA. Sequencing. Sequencing reactions were carried out using an automated ABI 373A sequencer (Applied Biosystems, Foster City, Calif., USA) on amplified and purified IGS1. Sequences were edited and aligned using the application SeqApp (anonymous ftp to iubio.bio.indiana.edu). The sequence of IGS1 (haplotype b) has been deposited in GenBank (accession number L25898).

Results Amplification of the Intergenic spacer of ¸. bicolor S238N The amplification priming sites were selected from highly conserved segments found in the 17s, 5s and 25s regions adjacent to the IGS spacers (White et al. 1990; Henrion et al. 1992; Fig. 1 A). Amplification of the 5s/17s spacer (IGS2) of the strain ¸. bicolor S238N produced two fragments of 4.0 and 4.5 kb (Fig. 1 B), whereas amplification of the 25s/5s spacer (IGS1) generated three fragments of 0.8, 2.2 and 2.4 kb (Fig. 1 C). The occurrence of these multiple bands may arise from heterogeneity in the IGS of the dikaryotic strain S238N (i.e. several loci, divergent sequences in the two nuclei or within the same rDNA repeat). Slower-moving fragments of PCR products may also correspond to concatemers or heteroduplexes formed by cross-hybridization between slightly different amplification products (Jensen and Straus 1993). To further study the origin and inheritance of the amplified IGS fragments, 45 monokaryons germinated from S238N spores were analyzed.

Inheritance of IGS2 Amplification of IGS2 of the 45 monokaryons gave rise to three kinds of PCR patterns. Twenty two monokaryons showed a pattern, referred to as the a type, with only the 4.5-kb fragment (Fig. 1 B, lane 2). Twenty monokaryons, referred to as the b type, exhibited a single 4.0-kb fragment (Fig. 1 B, lane 3). The three remaining monokaryons produced the a#b parental pattern of the S238N strain (Fig. 1 B, lane 4). This suggests a meiotic segregation of two IGS2 haplotypes, a and b, leading to two kinds of monokaryons. Good segregation among ¸. bicolor S238N progeny was further ascertained using a RAPD marker segregating independently 20 : 22 in the a- and b-type monokaryons, and 21 : 24 in the whole progeny (Table 1). The occurrence of a few monokaryons (e.g. H123) exhibiting the parental a#b pattern could be explained by crossing-over within the rDNA locus, creating chimeral rDNA with both a and b repeat units. The a and b IGS2 types can therefore be considered as two haplotypes of rDNA encoded in the two nuclei of S238N.

JCG233

334

Inheritance of IGS1 Unexpectedly, all monokaryons except the three putative recombined ones showed only the 0.8-kb parental IGS1 fragment (Fig. 1 C, lane 2 and 3). The three other monokaryons showed the parental pattern with the 0.8-, 2.2- and 2.4-kb fragments (Fig. 1 C, lane 4). The latter monokaryons therefore seemed to have the same rDNA repeats as the parental strain. The lack of the 2.2- and 2.4-kb fragments in the 42 monokaryons exhibiting Mendelian segregation of IGS2 suggested that these fragments did not arise from one of the IGS1 haplotypes. Five reconstituted dikaryons from crosses between monokaryons showing only the 0.8-kb IGS1 fragment were also analyzed (Fig. 2 A, B). The two crosses involving monokaryons of the same rDNA type (a or b) gave rise to a single-banded pattern (Fig. 2 B, dikaryon H152]H158). On the other hand, the three crosses involving a- and b-type monokaryons gave rise to the S238N-like a#b pattern (Fig. 2 B, dikaryon H152]H206), due to the appearance of the 2.2- and 2.4-kb fragments. These lower-mobility fragments were thus only amplified when both a and b rDNA were present in the genome, suggesting that they resulted from an interaction between the two haplotypes during PCR. Amplification of the IGS1 produces heteroduplexes

Fig. 1 A–C Diagram of a rDNA repeat and segregation of the amplified pattern of the intergenic spacers in ¸. bicolor S238N progeny. A A repeated unit of the rDNA in Basidiomycetes, with the coding sequences (rectangles) and the two intergenic spacers (IGS1 and IGS2, not to scale) B amplified IGS2 separated on 1% agarose. C amplified IGS1 separated on 8% acrylamide: lane 1 dikaryotic strain S238N; lane 2 monokaryon HC2 (a type); lane 3 monokaryon HC4 (b type); lane 4 monokaryon HC18 (recombinant with a parental pattern)

Table 1 Independent segregation of rDNA types (a and b) and of a 0.93-kb heterozygous RAPD marker obtained with primer 174 (Di Battista et al. 1996) among 45 monokaryons derived from ¸. bicolor S238N

Amplification of the IGS1 was carried out on mixes of DNA extracts of different monokaryons. The 2.2- and 2.4-kb fragments were always observed when the DNAs of a and b monokaryons were mixed, but not in DNA mixes of monokaryons having the same rDNA type (data not shown). However, the mixing at 25 °C of separately amplified a and b IGS1 types never led to the appearance of the 2.2- and 2.4-kb fragments (Fig. 2 C). Therefore, the occurrence of these additional fragments must have resulted from an interaction between

IGS haplotype

a

b

a#b (recombined)

Presence of the 0.93-kb RAPD fragment

HC2, HC3, HC8, HC9, HC14, HC35, HC44, HC45, H16, H296 (Ten monokaryons)

HC4, HC12, HC16, HC21, HC23, HC34, HC43, H221, H278, H306 (Ten monokaryons)

H123

HC5, HC6, HC10, HC15, HC20, HC22, HC27, HC35, HC37, H22, H28, H152, H158 (Twelve monokaryons)

HC13, HC19, HC28, HC36, HC48, HC50, H92, H119, H206, H234

HC18, HC49

(Ten monokaryons)

(Two monokaryons)

Absence of the 0.93-kb RAPD fragment

(One monokaryon)

JCG233

335 Fig. 2 A–D Heteroduplex formation during PCR amplification of the IGS1 of ¸. bicolor. A Amplified IGS1 of the a or b monokaryons (control). B amplified IGS1 of reconstituted dikaryons. C amplified and purified IGS1 of monokaryons mixed at room temperature. D amplified and purified IGS1 of monokaryons mixed and submitted to denaturation-renaturation after mixing. Products were analyzed on 8% acrylamide gels

the IGS1 types of the two monokaryons during the PCR reaction. To demonstrate further the influence of PCR thermal denaturation, amplified IGS1 types of the 42 single-banded a and b monokaryons were purified, mixed in every permutation and subjected to denaturation-renaturation in the absence of ¹aq polymerase. The 2.2- and 2.4-kb fragments were only formed when amplified IGS1 from a- and b-type monokaryons were mixed (e.g. H152 and H206 in Fig. 2 D). Because the polymerase is not present, the lower-mobility products most likely resulted from cross-hybridization between a and b IGS1 strands and the failure of some singlestranded regions to re-hybridize during the renaturation. The 2.2- and 2.4-kb fragments were removed when the amplification products were treated with mung bean single-stranded endonuclease (data not shown) confirming the presence of single-stranded DNA in these lower-mobility amplification products. The denaturation-renaturation assay enabled us to distinguish between a and b IGS1 types, although they differ neither in their apparent size (Fig. 2 A) nor in RFLP patterns (data not shown). The IGS1 types co-segregated with IGS2 types, further substantiating that a and b are the rDNA haplotypes encoded in the two nuclei of S238N. Sequences of a and b IGS1 Four a and four b IGS1 types were sequenced to clarify the difference between a and b IGS1. All IGS1 regions amplified from monokaryons of the same haplotype were identical in sequence. Most of the sequence of the a and b types was alignable, except for an additional TA C repeat in the a IGS1 (Fig. 3). Amplifications of 2 3 IGS1 sub-regions surrounding this additional TA C 2 3 repeat were performed using various primers internal to IGS1. Amplifications of ¸. bicolor S238N and its recombinant monokaryons always generated two lowermobility fragments when compared to the amplification of other monokaryons. In contrast, amplifications

Fig. 3 The sequence of the 25s/5s intergenic spacer (IGS1) region is shown for a- and b-type monokaryons of ¸. bicolor S238N. The portions of the amplified sequence corresponding to the 25s gene and the 5s gene are not shown. The sequences were aligned to show the insertion/deletion of the TA C repeat at position 308 (boxed). 2 3 Additional TA C repeats are underlined. The GenBank accession 2 3 number is L25898

excluding the additional TA C repeat generated a single 2 3 fragment in every case (data not shown). The TA C 2 3 repeat was thus responsible for heteroduplex formation, and the 2.2- and 2.4-kb fragments are most likely the two heteroduplexes expected from the cross-hybridization of the four single-strand IGS1 fragments obtained after denaturation of the a and b IGS1 types (Fig. 4).

JCG233

336

Fig. 4 Schematic representation of the heteroduplexes resulting from the cross-hybridization of the a and b amplified IGS1 spacer region. The TA C repeats (arrowheads) give rise to a single-strand 2 3 loop in the heteroduplex structure resulting from cross-hybridization of the a and b IGS1 strands

Discussion The use of the IGS region in analyzing the population genetics of ectomycorrhizal fungi will depend on a thorough understanding of the inheritance of this region and its mechanisms of evolution. Earlier, we found inter- and intra-specific length variation in the IGS of ¸accaria spp. (Henrion et al. 1992, 1994 b) and the present study revealed that length variation may also occur in IGS within haploid progeny (Fig. 1). In higher fungi, such as the diploid Saccharomyces (Petes and Botstein 1977) or the dikaryotic Pleurotus (Iraiabal and Labare`re 1994), two forms of rDNA occurred in a 1 : 1 ratio within the progeny due to heterozygosity. In ¸. bicolor S238N, two types of IGS were also observed to occur in a 1 : 1 ratio as a result of divergent haplotypes in the two separate nuclei. In strain S238N, amplification products of the IGS sequences upstream of (IGS1) and downstream from (IGS2) the 5s gene are heterogeneous, showing three and two fragments, respectively. A Mendelian inheritance of the IGS2 amplification pattern (22 : 20) was observed suggesting that all the rDNA arrays are located at a single locus, and that the two IGS2 fragments amplified from S238N are allelic. However, the possibility of two tightly linked rDNA tandems cannot be ruled out. Other basidiomycetes studied to-date also present a unique rDNA locus (Cassidy et al. 1984; Iraiabal and Labare`re 1994). The segregation analysis

of the IGS1 amplification pattern showed that the 2.2and 2.4-kb fragments of lower electrophoretic mobility were heteroduplexes formed by cross-hybridization between the two IGS1 haplotypes (Fig. 4). Length variation in IGS has been reported in many fungi (Anderson et al. 1990; Morton et al. 1995) where the differences are attributed to deletions or insertions in the arrays of subrepeats within the IGS. The two a and b rDNA haplotypes of ¸. bicolor S238N differ by the size of their IGS2 (4.5 and 4 kb, respectively) and by the number of TA C motifs in the repeat-rich region of 2 3 IGS1. Slippage during replication, or unequal sisterchromatid exchange, could lead to deletion or insertion of the TA C repeat, generating a new type of IGS1 (see 2 3 Tautz et al. 1986 for discussion). Concerted evolution of the rDNA may then spread the novel repeat within the rDNA haplotype, thus converting an a-type IGS1 to a b-type IGS1 (or vice-versa). In dikaryons, separation of the haploid genomes would then stabilize the heterozygosity after such a conversion. During the PCR temperature cycles, the crosshybridization of a and b IGS1 strands containing the non-complementary TA C region gave rise to hetero2 3 duplexes of lower electrophoretic mobility. Because multiple bands were observed in the IGS1 amplification products of several ectomycorrhizal basidiomycetes (Martin et al. 1993), heteroduplex formation (and probably rDNA heterozygosity) may be common to many dikaryotic fungi. Heteroduplex formation was also observed in the amplification of bacterial rDNA (Jensen and Strauss 1993). Meiotic recombination within rDNA is thought to be suppressed in many eukaryotes including higher fungi (e.g. Saccharomyces cerevisiae; Petes and Botstein 1977; Coprinus cinereus, Cassidy et al. 1984), and was not observed during segregation of a heterozygous Pleurotus cornucopiae (Iraiabal and Labare`re 1994). In the segregation analysis of ¸. bicolor, 6.5% of recombined monokaryons were, however, detected (Table 1). Assuming about 40 copies per haploid genome of S238N and a repeat length of 10 kb (Henrion 1993) gives a value of 60 kb per centimorgan in the rDNA of ¸. bicolor. Subculturing of these recombined monokaryons did not reveal any mitotic segregation (Selosse, unpublished results). Two kinds of units can thus coexist at least transiently within a repeat, as described in P. cornucopiae (Iraiabal and Labare`re 1994), and recombination frequency is not underestimated. In conclusion, length and sequence variations in the IGS of the rDNA offer considerable potential to investigate interstrain and progeny relationships within ¸. bicolor. Acknowledgements This work was supported by a grant (59226) of the Bureau des Ressources Ge´ne´tiques and a EU contract (DG VI, PL 931742). We thank F. Lapeyrie and W. Yan for providing the monokaryons used in the present study and B. Henrion for advice in the IGS amplification. Finally, we acknowledge the valuable discussion provided by D. Tagu.

JCG233

337

References Anderson JB, Bailey SS, Pukkila PJ (1990) Variation in ribosomal DNA among biological species of Armillaria, a genus of rootinfecting fungi. Evolution 43:1652—1662 Cassidy JR, Moore D, Lu BC, Pukkila PJ (1984) Unusual organization and lack of recombination in the ribosomal RNA genes of Coprinus cinereus. Curr Genet 8:607—613 Di Battista C, Selosse M-A, Bouchard D, Stenstro¨m E, Le Tacon F (1996) Variations in symbiotic efficiency, phenotypic characters and ploidy level among different isolates of the ectomycorrhizal basidiomycete ¸accaria bicolor strain S238. Mycol Res (in press) Fries N (1983) Spore germination, homing reaction, and intersterility groups in ¸accaria laccata (Agaricales). Mycologia 75:221—227 Henrion B (1993) Caracte´risation et identification de champignons ectomycorhiziens par amplification enzymatique (PCR) de l’ADN ribosomal. PhD thesis, University of Nancy Henrion B, Le Tacon F, Martin F (1992) Rapid identification of genetic variation of ectomycorrhizal fungi by amplification of ribosomal RNA genes. New Phytol 122:289—298 Henrion B, Chevalier G, Martin F (1994 a) Typing truffles species by PCR amplification of the ribosomal DNA spacers. Mycol Res 98, 37—43 Henrion B, Di Battista C, Bouchard D, Vairelles D Thompson BD, Le Tacon F, Martin F (1994 b) Monitoring the persistence of ¸accaria bicolor as an ectomycorrhizal symbiont of nursery grown Douglas fir by PCR of the rDNA intergenic spacer. Mol Ecol 3:571—580 Iraiabal B, Labarre`re J (1994) Restriction-site and length polymorphism of the rDNA unit in the cultivated basidiomycete Pleurotus cornucopiae. Theor Appl Genet 88:824—830

Jensen MA, Straus N (1993) Effect of PCR conditions on the formation of heteroduplex and single-stranded DNA products in the amplification of bacterial ribosomal DNA spacer regions. PCR Methods Appl 3:186—194 Le Tacon F, Alvarez IF, Bouchard D, Henrion B, Jackson RM, Luff S, Pera J, Stenstro¨m E, Villeneuve N, Walker C (1992) Variations in field response of forest trees to nursery ectomycorrhizal inoculation in Europe. In: Read DJ, Lewis DH, Fitter AH, Alexander IJ (eds) Mycorrhizas in ecosystems. CAB International, Wallingford, pp 119—134 Martin F, Vairelles D, Henrion B (1993) Automated ribosomal DNA fingerprinting by capillary electrophoresis of PCR products. Anal Biochem 214:182—189 Molina FI, Jong S-C, Huffman JL (1993) PCR amplification of the 3@ external transcribed and intergenic spacers of the ribosomal DNA repeat unit in three species of Saccharomyces. FEMS Microbiol Lett 108:259—264 Morton A, Tabrett AM, Carder JH, Barbara DJ (1995) Sub-repeat sequences in the ribosomal RNA intergenic regions of »erticillium alboatrum and ». dahliae. Mycol Res 99: 257—266 Petes T, Botstein D (1977) Simple Mendelian inheritance of the reiterated ribosomal DNA of yeast. Proc Natl Acad Sci USA 74:5091—5095 Srivastava AK, Schlessinger D (1991) Structure and organization of ribosomal DNA. Biochimie 73:631—638 Tautz D, Trick M, Dover GA (1986) Cryptic simplicity in DNA is a major source of genetic variation. Nature 322:652—658 White TJ, Bruns T, Lee S, Taylor J (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In Innis MA, Gelfand DH, Sninsky JJ, White TJ (eds) PCR protocols. A guide to methods and applications. Academic Press, San Diego, pp 315—322

.

JCG233