Isolation, characterization, inheritance and linkage of ... - Springer Link

Mol Gen Genet (2001) 264: 871±882 DOI 10.1007/s004380000377

O R I GI N A L P A P E R

O. P. Rajora á M. H. Rahman á S. Dayanandan A. Mosseler

Isolation, characterization, inheritance and linkage of microsatellite DNA markers in white spruce (Picea glauca) and their usefulness in other spruce species Received: 15 May 2000 / Accepted: 10 September 2000 / Published online: 29 November 2000 Ó Springer-Verlag 2000

Abstract Microsatellite DNA/simple-sequence-repeat (SSR) loci were identi®ed, isolated and characterized in white spruce (Picea glauca) by screening both a nonenriched partial genomic library and a partial genomic library enriched for (AG/TC)n-containing clones. Inheritance and linkage of polymorphic SSR loci were determined in F1 progeny of four controlled crosses. We also assessed the compatibility and usefulness of the P. glauca microsatellite DNA markers in ®ve other Picea species. Twenty-four microsatellites were identi®ed by sequencing 32 clones selected from screens of 5400 clones from the two libraries. The (AG/TC)n microsatellites were the most abundant in the non-enriched library. Eight microsatellite DNA loci were of the singlecopy type, and six of these were polymorphic. A total of 87 alleles were detected at the six polymorphic SSR loci in 32 P. glauca individuals drawn from several populations. The number of alleles found at these six SSR loci ranged from 2 to 22, with an average of 14.5 alleles per locus, and the observed heterozygosity ranged from 0.48

Communicated by G. Haughn O. P. Rajora (&) Department of Biology, Life Sciences Centre, Dalhousie University, Halifax, NS B3H 4J1, Canada E-mail: [email protected] Tel.: +1-902-4942400 Fax: +1-902-4943736 O. P. Rajora á M. H. Rahman á S. Dayanandan1 Department of Renewable Resources, University of Alberta, Edmonton, AB T6G 2H1, Canada A. Mosseler Canadian Forest Service, Atlantic Forestry Centre, P.O. Box 4000, Fredericton, NB E3B 5P7, Canada Present address: 1 Biology Department, Concordia University, 1455 de Maisonneuve Blvd. West, Montreal, QC H3G 1M8, Canada Contribution RCGB0001 from the Resource and Conservation Genetics and Biotechnology Group, Dalhousie University

to 0.91, with a mean of 0.66 per locus. Parents of the controlled crosses were polymorphic for ®ve of the six polymorphic SSR loci. Microsatellite DNA variants at each of these ®ve SSR loci followed a single-locus, codominant, Mendelian inheritance pattern. Joint twolocus segregation tests indicated complete linkage between PGL13 and PGL14, and no linkage between any of the remaining SSR loci. Each of the 32 P. glauca individuals examined had unique single or two-locus genotypes. With the exception of non-ampli®cation of PGL12 in P. sitchensis, P. mariana, and P. abies and the monomorphic nature of PGL7 in P. mariana, primer pairs for all six polymorphic SSR loci successfully ampli®ed speci®c fragments from genomic DNA and resolved polymorphic microsatellites of comparable sizes in P. engelmanni, P. sitchensis, P. mariana, P. rubens, and P. abies. The closely related species P. mariana and P. rubens, and P. glauca and P. sitchensiss could be distinguished by the PGL12 SSR marker. The microsatellite DNA markers developed and reported here could be used for assisting various genetics, breeding, biotechnology, tree forensics, genome mapping, conservation, restoration, and sustainable forest management programs in spruce species. Key words Spruce (Picea) á Simple sequence repeats á Microsatellite DNA markers á DNA ®ngerprinting á Cross-species compatibility and conservation of microsatellite loci

Introduction Highly informative genetic markers are required to facilitate various programs in genetics, breeding, genome mapping, tree improvement, tree forensics, and conservation, restoration, and sustainable management of forest genetic resources. Microsatellite DNA or ``simplesequence repeats (SSR)'', a relatively new class of DNA markers, are highly informative genetic markers. Microsatellites are abundant polymorphic elements in

872

nuclear genomes, and consist of tandemly repeated, short DNA sequence motifs (Tautz 1989; Weber and May 1989; Wang et al. 1994). These regions are interspersed throughout eukaryotic genomes, and are generally embedded in unique DNA sequences. Polymorphism among individuals arises from changes in the number of repeats. These markers are also known as simple sequence length polymorphism (SSLP) (Tautz 1989), or sequence-tagged microsatellite sites (STMS) (Beckman and Soller 1990). Due to their hypervariability, codominance and high reproducibility, microsatellites are ideal markers for the construction of high-resolution genetic maps and identi®cation of loci controlling traits of interest (Beckman and Soller 1990; Hearne et al. 1992; Devey et al. 1996; Paglia et al. 1998) for population and conservation genetic studies (Rajora et al. 2000), clonal identi®cation (Dayanandan et al. 1998; Sanchez-Escribano et al. 1999), certi®cation of controlled crosses, identi®cation of species and hybrids (Rajora and Rahman, to be submitted), paternity determination (van de Ven and McNicol 1996), marker-assisted early selection (Weising et al. 1998), assessment of the genetic e€ects of forest management practices, and development of strategies for conservation and sustainable management of forest genetic resources (Rajora et al. 2000). One of the main advantages of microsatellite DNA markers is that protocols and results can be readily exchanged among di€erent laboratories and the data obtained can be evaluated in the context of an international database (Sanchez-Escribano et al. 1999). The development of microsatellite DNA markers in coniferous forest trees is quite dicult, time-consuming, and costly. Due to the limited amount of sequence data available in databases, it is often necessary to construct genomic libraries and screen for speci®c SSRs. Diculties have often been encountered in isolating and developing clean and reproducible single-locus microsatellite DNA markers in conifers, probably due to high proportions of highly repeated sequences in the nuclear DNA and the large size of the conifer genomes. Therefore, our knowledge of microsatellites in conifers is quite rudimentary. A limited number of microsatellite DNA markers have been developed for a few pine (Pinus) species (Smith and Devey 1994; Kostia et al. 1995; Echt et al. 1996; Fisher et al. 1998), and two spruce (Picea) species, Sitka spruce (P. sitchensis) (van de Ven and McNicol 1996) and Norway spruce (P. abies) (Pfei€er et al. 1997; Paglia et al. 1998). Microsatellite DNA markers have also been developed for some angiosperm forest trees, such as Populus (Dayanandan et al. 1998; Rahman et al. 2000) and Eucalyptus (Marques et al. 1999) species. Speci®city of SSRs for a given taxon could also limit the cross-species utility of these markers in forest trees. However, it has been documented that SSR markers developed for a given species can be used to detect polymorphism in related species, minimizing laborious cloning and screening steps (Dayanandan et al. 1997, 1998; Echt et al. 1999; Rajora and Rahman, to be submitted).

White spruce [P. glauca (Moench) Voss], a widespread transcontinental tree species of the Canadian boreal forest, is one of the most important trees for the production of wood pulp and lumber, and is a prime reforestation species in Canada. There are many active programs for genetic improvement, breeding, conservation, and sustainable management of this species throughout Canada. In addition, white spruce is a prime candidate for genome mapping research. Highly informative genetic markers are needed to facilitate these programs. However, at present no published information is available on microsatellite DNA markers in white spruce. The objective of our study was to identify, isolate, and characterize microsatellite DNA markers from white spruce by screening partial genomic libraries. We have isolated and characterized several microsatellite DNA markers from white spruce. We determined the informativeness of microsatellite DNA markers for the eight SSR loci showing single-locus segregation patterns by calculating allelic diversity, observed heterozygosity and allele frequencies in 32 white spruce individuals, as well as inheritance patterns and linkage of ®ve polymorphic SSR loci in progeny of four controlled crosses of white spruce. The utility of the characterized SSRs for DNA ®ngerprinting of P. glauca individuals was assessed by constructing single and multilocus genotypes of each individual. We also examined the cross-compatibility and usefulness of the P. glauca SSR markers in ®ve other Picea species.

Materials and methods DNA extraction DNA was extracted from needle tissue of individual spruce trees following a method described in Dayanandan et al. (1997) for both construction of genomic libraries and microsatellite DNA genotyping. Library construction and screening for SSR sequences Non-enriched and microsatellite-enriched partial genomic libraries were constructed using total genomic DNA extracted from needle tissues of white spruce trees. The non-enriched partial genomic library was constructed following the protocol we developed for isolating microsatellites from Populus tremuloides (Dayanandan et al. 1998). The partial genomic library enriched for (AG/TC)ncontaining microsatellites was constructed by following a protocol modi®ed from Kijas et al. (1994). The genomic DNA was digested with Sau3AI, and size fractionated on a 1% agarose gel. The gel segment containing DNA fragments of about 500±1000 bp was excised, and the fragments were extracted and puri®ed using the Qiaquick gel extraction kit (Qiagen, Santa Clarita, Calif.). Puri®ed DNA fragments were ligated to the M13mp19 vector as described in Dayanandan et al. (1998). The ligated DNA mix was enriched for (AG/TC)n microsatellites by following the protocol of Kijas et al. (1994), except that universal M13 forward and M13 reverse primers, biotinylated (AG)15 oligonucleotide primers and Streptavidin Magnesphere paramagnetic particles (Promega, Madison, Wis.) were used. The non-enriched genomic library and the library enriched for the (AG/TC)n microsatellites were screened for clones containing

873 microsatellites, as described by Dayanandan et al. (1998). We used the oligonucleotides (AAAG)7, (AGAT)7, (AT)15, (AG)15, (CA)15, (TC)15 and (TG)15 as probes to screen the non-enriched, and (AG)15 and (TC)15 probes to screen the microsatellite-enriched library. Sequencing of positive clones, design of primers, and resolution of SSR polymorphisms Positive plaques were picked with a sterile pipette tip and placed in 15-ml culture tubes containing 2 ml of LB/tetracycline and 100 ll of an overnight culture of XL1BlueMRF' bacterial cells. Tubes were incubated at 37 °C with shaking for 6 h. A portion of each culture was transferred to 2-ml microfuge tubes and centrifuged at maximum speed (14,000 rpm) for 5 min. Single-stranded DNA was isolated from the supernatant using the Wizard M13 DNA puri®cation system (Promega). Puri®ed DNA from isolated clones with putative microsatellites was sequenced using the ABI Prism 377 automated DNA sequencer (Applied Biosystems, Foster City, Calif.). Oligonucleotide primers complementary to the regions ¯anking the identi®ed repeats were synthesized and used for DNA ampli®cation by PCR. PCR ampli®cation, primer optimization, and resolution of microsatellites The methodology for both the generation of microsatellite DNA markers and the visualization of these markers was developed and optimized. DNA ampli®cation was performed in a total volume of 10 ll, with 20 ng of puri®ed spruce DNA, 0.2 mM dNTPs, 1.5 mM MgCl2, 5 pmol of each primer, 2 lg of BSA, 1 ´ reaction bu€er and 0.4 U of AmpliTaq DNA polymerase (Applied Biosystems), using the GeneAmp 9600 thermal cycler (Applied Biosystems). The optimal temperature cycling parameters for ampli®cation were either the touchdown protocol (60°¯54 °C) for six primers (see below) as described in Rahman et al. (2000) or initial denaturation at 94 °C for 3 min, followed by two cycles of 30 s each at 94 °C, 50 °C, and 72 °C; and 38 cycles of 15 s each at 94 °C, 50 °C and 72 °C, followed by a ®nal extension step at 72 °C for 3 min for two primers (see below). After PCR, 5 ll of the ampli®ed product was electrophoresed on a 1.5% horizontal agarose gel to check for positive ampli®cation and determine the approximate amount of the product. The rest of the ampli®ed product (5 ll) was diluted with 5±20 ll of loading dye (95% formamide, 10 mM NaOH, 0.05% xylene cyanol and 0.05% bromophenol blue) and 3±4 ll of the mix was electrophoresed on a 6% denaturing polyacrylamide gel containing 6 M urea and 1 ´ TBE bu€er on a S2 electrophoretic apparatus (Owl Scienti®c, Woburn, Mass.) at 80 W for 3±4 h. Following electrophoresis, the gel was silver-stained using a silver sequencing kit (Promega), as described in Rajora et al. (2000), and photographed following the manufacturer's instructions. Where a null allele was observed in an individual or an SSR locus could not be ampli®ed in any sample, the PCR and electrophoretic analysis was repeated two or three times, except for certain F1 seedling samples for which very limited amounts of DNA were available. Characterization and informativeness of microsatellites The frequency of the SSRs in the genome was estimated by dividing the estimated total length of the cloned genomic library by the observed number of clones containing a given SSR. The informativeness of each locus was evaluated by calculating the number and frequency of individual alleles, and the levels of observed heterozygosity in 32 white spruce individuals drawn from several populations in northern Alberta. Single- and multi-locus genotypes of each individual were determined to identify unique genotypes in order to assess the utility of the characterized SSRs for genetic ®ngerprinting of individuals.

Inheritance and linkage We used the parents and F1 progeny of four controlled crosses of white spruce (carried out by Weyerhaeuser Canada), to determine the segregation patterns and linkage of microsatellite DNA markers. F1 hybrid seeds and foliage of the ®ve parents of these four controlled crosses were procured. Seeds were germinated in a growth chamber, and DNA was extracted from emerging individual F1 seedlings as described by Rajora (1999). Thirty-four to 46 F1 individuals per cross were analyzed. The inheritance patterns of microsatellite DNA variants were determined from a comparison of the distribution of progeny genotypes with the expected distribution, based on genotypes of the parents, by means of a Chisquare goodness-of-®t test. Two-point linkage was determined between the two loci of each of the eight informative pairs (PGL7PGL12, PGL7-PGL13, PGL7-PGL14, PGL12-PGL13, PGL12PGL14, PGL12-PGL15, PGL13-PGL14, and PGL14-PGL15) by investigating joint segregation and independent assortment of their microsatellite DNA variants. The PGL7-PGL15 locus pair was not informative for the linkage analysis. The null hypothesis ± that the two loci in a pair assort independently ± was tested by G and v2 goodness-of-®t tests. The G test was performed by applying Williams' (Williams 1976) correction to reduce the Type I error and to obtain a better approximation to the v2 distribution (Sokal and Rohlf 1995). Since the expected number of progeny in genotypic classes of all but one controlled cross was less than the minimum of 5 required for a valid v2 test (Sokal and Rohlf 1995), we pooled adjacent two- or four-progeny genotypic classes to obtain a joint class with an expected progeny number >5 to perform G and v2 tests (Sokal and Rohlf 1995). This resulted in a total of four joint progeny classes. Pooling of adjacent classes was not required to test the linkage between PGL7 and PGL13 in the 16 ´ 19 cross. We also performed G tests without pooling any genotypic classes. Cross-species compatibility and usefulness Five to eight individuals each of the ®ve spruce species Engelmann spruce (P. engelmannii), Sitka spruce (P. sitchensis), black spruce (P. mariana), red spruce (P. rubens), and Norway spruce (P. abies) were used to test the cross-compatibility and usefulness of the microsatellite DNA markers developed from P. glauca in these spruce species. The number of alleles observed at each locus in each of these species and the numbers shared with or di€erent from P. glauca were determined.

Results Isolation and development of microsatellite DNA markers The screening of a non-enriched partial genomic library of P. glauca, consisting of 4028 clones, with an average insert size of 600 bp, with nine di-, tri- and tetra-nucleotide repeat probes, identi®ed 71 positive clones. Screening of the (AG/TC)n-enriched partial genomic library consisting of 1376 clones identi®ed 20 positive clones. The positive clones were rescreened and 32 of the most promising SSR-positive clones (25 from the nonenriched and 7 from the enriched library) were sequenced. Sequencing of these 32 clones revealed SSRs with 2±40 repeat units in 23 clones (Table 1). The PGL13 and PGL14 microsatellite DNA sequences were found in the same clone. The (AG)n di-nucleotide SSR showed the highest abundance in the non-enriched library, with a frequency of once every 161 kb, whereas the frequency

874 Table 1 SSR sequences identi®ed in partial genomic libraries of Picea glauca SSR sequence

SSR sequence/ locus designation

Non-enriched library (AG)7AA(AG)18 T10(TG)4T5AAG(GT)4 (TC)3CC(TC)8 T9(TG)5 (TC)8 (AG)4 (AG)38 (CTT)3 (TC)24 .... (TG)5 (AT)18(AG)40 (GAA)2G(GAA)2(TG)4 (AAG)2(AG)3G4(AG)9 (GT)3GG(GT)10 (CTT)2T(CTT)1 .... (GT)3 (T)9G(T)11 (CT)2G(CT)2 .... (AT)4 (AAAC)3 (AG)3 (CA)3 A (CA) (TA)3 ... (TA)3 (AG/TC)n-enriched library (AG)21 (AG)20 (CT)3C2(CT)3T2(CT)4 .... (CT)11 (TC)16 a

PGL1 PGL2 PGL3 PGL4 PGL5 PGL6 PGL7 PGL8 PGL9 PGL10 PGL11 PGL12

PGL13a PGL14a PGL15 PGL16

These sequences were found in the same genomic clone

of (TG)n and (TC)n repeats was 1/279 kb and 1/322 kb, respectively. Oligonucleotide primer pairs were designed for 16 microsatellite DNA sequences (PGL1- PGL16) from the 23 clones (Table 1). A variety of PCR conditions were used in attempts to amplify microsatellite DNA variants from genomic DNA of white spruce, using 16 pairs of SSR primers.

Among the 16 primer pairs tested, only 11 yielded ampli®cation products, although a range of annealing temperatures and other PCR conditions were tried, including redesigned primers. Only eight of these 11 SSR primer pairs resolved microsatellite DNA variants that were consistent and showed single-locus inheritance patterns (Table 2) and thus were considered further. PCR conditions for these eight SSR primer pairs were optimized (Table 2). Allelic variation at three of these SSR loci is shown in Fig. 1. Three SSR primers either did not show single-locus patterns or were inconsistent. The SSR primer pairs for the PGL1 clone resolved microsatellite DNA variants showing consistent but complex, multilocus, patterns. Primer pairs designed for the PGL5 clone resolved two monomorphic SSR loci. Although microsatellite DNA patterns produced by the PGL11 primers were of the single-locus type, high frequencies of null alleles were observed. Despite trying a variety of PCR conditions and redesigning primers, the resolution of PGL1, PGL5 and PGL11 could not be improved. Informativeness Two SSR loci, PGL8 (a short trinucleotide repeat) and PGL9 [a compound of (TC)n and (TG)n repeats], were found to be monomorphic in 32 white spruce samples surveyed (Table 2). PGL9 was also monomorphic when a redesigned reverse primer for only (TC)24 repeats was used. Between 2 and 22 alleles (87 alleles in all) were observed at the six polymorphic SSR loci in the white spruce samples examined (Tables 2 and 3; Fig. 1). Three of the six polymorphic SSR loci were from the nonenriched and three from the enriched library. With the exception of PGL6, these SSR loci were highly infor-

Table 2 Repeat type, ampli®cation parameters, numbers and size range of alleles detected, and observed heterozygosity for eight microsatellite loci in P. glauca Locus

Repeat

PGL6

(AG)4

PGL7 PGL8 PGL9 PGL12 PGL13 PGL14 PGL15 a b

Primer sequence (5¢®3¢)a

TA (°C)b

TACTTCAGGACTTCAGGATTCAGGG (F) TTTGCAAAGGCCTAAAGACCGTTGG (R) TCACTATTTATTTCCCAAATGCTCGTA (F) (AG)38 TCTCCNCAAGAAATCCNCCCTC (R) (CTT)3 CAGCACCCCTTGCAATAGTGG (F) GATCATCATAGCAGATAAAAAGAGC (R) (TC)24 N40 AGCCAATACAATGCCAAGAGATAAC (F) (TG)5 ATAAATAAGGAGGCTAGTGCCTCAC (R1) AGAGACAAGTTTTGGAGCTGCAGT (R2)d (AAG)2 CCATCTCAAAATATTTAATTGTCCAGT (F) TCATATCTGCATGCAAAGTCTGAAC (R) (AG)3G4(AG)9 (AG)21 AAAAATAGTTTATATTTTCTTTATTACTC (F) TATAAATCATTTTTCTTATGTTGTG (R) AAAAATGATTTATATCTTCTTATTGTCT (F) (AG)20 GNGTCATAAACGCCCATCAATAG (R) (CT)4N16 CATACTCTCACACCCACACCCTCTC (F) (CT)11 CAAGAACAGAAGAGAGGTCAAGATTG (R)

F, forward primer; R, reverse primer Optimum annealing temperature

c d

Alleles

Size (bp)

Hoc

60¯54

2

118±116

0.72

60¯54

22

170±104

0.77

60¯54

1

112

0

60¯54

1

158

0

60¯54

11

250±222

0.48

50

19

160±110

0.58

50

18

180±136

0.91

60¯54

15

248±176

0.48

Observed heterozygosity Reverse primer designed to amplify only (TC)24 repeats

875 Table 3 Alleles, allele size (bp) and respective frequencies observed in white spruce (Picea glauca) samples Locus

Allele

Allele size

Frequency

PGL6

A B

118 116

0.359 0.641

PGL7

A B C D E F G H I J K L M N O P Q R S T U V

170 160 158 156 154 152 140 136 134 130 128 126 124 122 120 118 116 114 112 110 106 104

0.016 0.016 0.016 0.016 0.016 0.016 0.016 0.016 0.032 0.065 0.032 0.048 0.065 0.145 0.113 0.016 0.081 0.048 0.065 0.065 0.081 0.016

PGL12

A B C D E F G H I J K 0

250 248 242 240 238 236 234 232 230 228 222 000

0.016 0.031 0.047 0.016 0.141 0.078 0.047 0.094 0.078 0.078 0.218 0.156

PGL13

A B C D E F G H I J K L M N O P Q R S A B C D E F G H I

160 154 152 146 144 142 140 138 136 134 132 126 124 122 120 118 116 114 110 180 172 170 164 162 160 158 156 154

0.016 0.065 0.032 0.016 0.032 0.097 0.016 0.032 0.065 0.113 0.129 0.016 0.081 0.081 0.032 0.081 0.048 0.016 0.032 0.016 0.016 0.016 0.109 0.016 0.031 0.109 0.047 0.124

Fig. 1 Allelic variation at three SSR loci (PGL12, PGL13 and PGL14) in 32 individuals of P. glauca

mative, with the three (AG)n dinucleotide loci (PGL7, PGL13 and PGL14) being the most informative for allelic diversity or heterozygosity (Table 2; Fig. 1). The highest allelic diversity (22 alleles) was observed for PGL7 and the lowest for the diallelic PGL6. The highest heterozygosity (0.91) was observed for the PGL14 locus and the lowest (0.48) for the PGL12 locus (Table 2; Fig. 1). The frequency of a given allele ranged from 0.016 at each of PGL7, PGL12, PGL13, PGL14 and

PGL14

876 Table 3 (Contd.) Locus

PGL15

Allele

Allele size

Frequency

J K L M N O P Q R

152 150 148 146 144 142 140 138 136

0.094 0.016 0.109 0.031 0.047 0.016 0.078 0.047 0.078

A B C D E F G H I J K L M N O

248 244 236 232 230 228 226 224 220 214 210 208 180 178 176

0.048 0.032 0.016 0.016 0.016 0.048 0.032 0.016 0.032 0.016 0.098 0.032 0.098 0.032 0.468

sharing the same genotype at PGL7 had unique genotypes at PGL13 or PGL15. The number of di€erent microsatellite DNA genotypes (the number of individuals with unique genotypes) observed among the 32 white spruce individuals at the remaining ®ve polymorphic loci were as follows: 28 (25) at PGL14, 27 (23) at PGL13, 21 (16) at PGL12, 20 (14) at PGL15, and 2 (0) at PGL6. Inheritance

PGL15 to 0.641 at PGL6 (Table 3). Except for the two alleles at PGL6 and one allele at PGL15, no allele occurred at a frequency higher than 0.22 (Table 3). Each of the 32 white spruce individuals had unique one- or two-locus genotypes. PGL7 was the most informative for DNA ®ngerprinting of the white spruce individuals. There were 29 unique genotypes among the 32 individuals at this locus. Each of 27 individuals had a unique genotype, while three individuals shared one genotype, and two individuals shared another genotype. The trees

The parents of the controlled crosses were polymorphic for microsatellite DNA variants at PGL7, PGL12, PGL13, PGL14, and PGL15 loci (Table 4, Fig. 2). At each of these loci, the progeny of each of the respective controlled crosses segregated into the expected two to four genotypic classes, with no signi®cant deviation from the ratios expected for single-locus codominant Mendelian inheritance of each of these ®ve SSR loci (Table 4; Fig. 2). Since the microsatellite genotypes of the parents of no two crosses were the same at the SSR locus PGL7, PGL12, PGL13 or PGL14, it was not possible to measure the heterogeneity in segregation ratios over di€erent controlled crosses for any of these SSR loci. No polymorphism was observed among the ®ve parents of the four controlled crosses for the SSR locus PGL6. Therefore, the inheritance pattern of the two microsatellite DNA variants at this locus could not be ascertained. Linkage As inferred from the parental microsatellite genotypes, the female and male parents should produce two to four

Table 4 Chi-square goodness-of-®t tests for single-locus segregation of microsatellite DNA genotypes at ®ve SSR loci in F1 progeny of controlled crosses in white spruce (P. glauca) Locus

Cross

Parental genotypes

No. of progeny

Distribution of progeny genotypes observed

Expected ratio

v2

P>v2

PGL7

11 ´ 13 16 ´ 19 19 ´ 15

BB ´ CD AB ´ EE EE ´ AD

38 34 41

15(BC):23(BD) 18(AE):16(BE) 20(AE):21(DE)

1:1 1:1 1:1

1.68 0.12 0.03

0.19 0.73 0.86

PGL12

11 16 16 19

´ 13 ´ 15 X 19 ´ 15

BE ´ BD C0 ´ AE C0 ´ E0 E0 ´ AE

42 43 45 44

12(BB):8(BD):13(BE):9(DE) 10(AC):10(CE):15(A0):8(E0) 16(CE):9(C0):11(E0):9(00) 11(AE):13(A0):20(EE+E0)

1:1:1:1 1:1:1:1 1:1:1:1 1:1:2

1.62 2.49 2.91 0.83

0.65 0.48 0.41 0.66

PGL13

11 ´ 13 16 ´ 19 19 ´ 15

CH ´ BG FF ´ DI DI ´ AE

44 38 39

10(BC):14(CG):10(BH):10(GH) 20(DF):18(FI) 7(AD):13(DE):13(AI):6(EI)

1:1:1:1 1:1 1:1:1:1

1.09 0.11 4.39

0.78 0.74 0.22

11 16 16 19

13 15 19 15

CD ´ BD DF ´ AC DF ´ DE DE ´ AC

42 41 38 46

10(BC):14(CD):9(BD):9(DD) 9(AD):13(CD):7(AF):12(CF) 10(DD):7(DE):11(DF):10(EF) 15(AD):8(CD):8(AE):15(CE)

1:1:1:1 1:1:1:1 1:1:1:1 1:1:1:1

1.62 2.22 0.95 4.26

0.65 0.53 0.81 0.23

16 ´ 15

AD ´ BB

36

19(AB):17(BD)

1:1

0.11

0.74

PGL14

PGL15

´ ´ ´ ´

877

Fig. 2 Inheritance of microsatellite DNA variants at SSR locus PGL12 in F1 progeny of three controlled crosses of white spruce

kinds of gametes with equal frequencies for each of the eight two-locus combinations (Table 5). Assuming no di€erences in viability among gametes and no meiotic drive, the frequency of the 4 to 16 genotypic classes among the F1 progeny should follow the expectations for independent assortment of two loci in a pair if they are not linked (Table 5). G and Chi-square goodness-of®t tests for joint segregation patterns indicated that the observed numbers of genotypes were in agreement with those expected for independent assortment of two loci in seven out of eight two-locus combinations tested (Table 5). The joint segregation patterns for the PGL13 and PGL14 indicated that the observed genotype frequencies were signi®cantly di€erent from those expected for independent assortment of these loci in each of the three controlled crosses examined (Table 5). No recombination was observed between these two SSR loci, as would be expected for two loci derived from the same genomic clone. Similar results were obtained whether G or v2 tests were performed on individual or pooled joint progeny classes. Cross-compatibility and usefulness of white spruce microsatellite DNA markers in ®ve other spruce species Primer pairs for ®ve SSR loci, PGL6, PGL7, PGL13, PGL14 and PGL15, from white spruce yielded PCR ampli®cation products of comparable size with genomic DNA from, and resolved polymorphic microsatellite

DNA variants in, all other ®ve spruce species tested (Table 6, Fig. 3), with the exception that PGL7 was monomorphic in the P. mariana individuals examined. The SSR locus PGL12 was polymorphic in Engelmann spruce and red spruce, but was not detected in black, Norway, or Sitka spruce (Fig. 3, Table 6). Most of the alleles observed at the six SSR loci in Engelmann, Sitka, black, red and Norway spruce samples were similar in size to those observed in the white spruce samples examined. However, these ®ve spruce species had certain alleles at ®ve SSR loci that di€ered in size from those observed in the 32 white spruce samples. These exceptions were the following: PGL7, 200-bp allele in Norway spruce; PGL12, 244-bp allele in red spruce; PGL13, 162bp allele in black spruce, 150-bp allele in Norway spruce, 130-bp allele in Engelmann, Sitka, black and red spruce, 112-bp allele in Sitka and black spruce, and 108- and 106-bp alleles in Sitka and Norway spruce; PGL14, 186bp allele in Engelmann and red spruce, 168- and 130-bp alleles in black spruce, 166-bp allele in Engelmann spruce, 134-bp allele in Sitka spruce, and 128-bp allele in Norway spruce; PGL15, 280- and 238-bp alleles in Sitka and black spruce, 272-, 266-, and 196-bp alleles in Norway spruce, 258- and 212-bp alleles in Sitka spruce, 256-bp allele in black and Norway spruce, 254-bp allele in red and Norway spruce, 234-bp allele in Engelmann and black spruce, 222- and 216-bp alleles in red spruce, 218-bp allele in black spruce and 206- and 200-bp alleles in Engelmann and Norway spruce.

Discussion The results suggest that (AG)n repeats are quite abundant in the genome of white spruce. The estimated frequency of (AG)n falls within the range of dinucleotide repeat frequencies reported for other plant species, including Norway spruce and other forest trees (Condit and Hubbell 1991; Morgante and Olivieri 1993; Smith and Devey 1994; Terauchi and Konuma 1994; Chase et al. 1996; Pfei€er et al. 1997). The higher abundance of (AG/TC)n repeats than (AC/TG)n repeats in white spruce is in agreement with similar observations in Norway spruce (Pfei€er et al. 1997) and trembling aspen (Dayanandan et al. 1998) and with the general ®nding that (AG/TC)n repeats are more abundant than (AC/ TG)n repeats in plants (Wang et al. 1994; Jarret et al. 1997). The opposite trend, with higher abundance of (AC/TG)n than (AG/TC)n repeats, has been reported in primates (Jerka and Pethiyagoda 1995). A comparison of the non-enriched and enriched partial genomic libraries with respect to the recovery of informative microsatellite DNA loci suggests that the enriched library yielded relatively more microsatellites that were informative in white spruce. The results of our study suggest that ®ve (PGL7, PGL12-PGL15) of the eight SSR loci characterized are highly informative in white spruce. Of these, the three (AG)n dinucleotide repeat loci, PGL7, PGL13 and

878 Table 5 Joint two-locus segregation patterns and Chi-square analyses for tests of linkage between microsatellite DNA loci in white spruce (P. glauca) Locus combination

Cross

Parental genotypes (Locus 1/Locus 2)

Number Distribution of progeny genotypes of observed in each expected genotypic progeny class (Locus1/Locus2)a

PGL7/PGL12

11 ´ 13

BB/BE ´ CD/BD

36

16 ´ 19

AB/C0 ´ EE/E0

33

19 ´ 15

EE/E0 ´ AD/AE

40

11 ´ 13

BB/CH ´ CD/BG

37

16 ´ 19 19 ´ 15

AB/FF ´ EE/DI EE/DI ´ AD/AE

29 34

11 ´ 13

BB/CD ´ CD/BD

36

16 ´ 19

AB/DF ´ EE/DE

29

19 ´ 15

EE/DE ´ AD/AC

39

PGL12/PGL13 11 ´ 13

BE/CH ´ BD/BG

41

16 ´ 19

C0/FF ´ E0/DI

37

19 ´ 15

E0/DI ´ AE/AE

37

PGL12/PGL14 11 ´ 13

BE/CD ´ BD/BD

39

16 ´ 15

C0/DF ´ AE/AC

40

16 ´ 19

C0/DF ´ E0/DE

37

19 ´ 15

E0/DE ´ AE/AC

44

PGL12/PGL15 16 ´ 15

C0/AD ´ AE/BB

36

PGL13/PGL14 11 ´ 13

CH/CD ´ BG/BD 36

PGL7/PGL13

PGL7/PGL14

16 ´ 19

FF/DF ´ DI/DE

35

19 ´ 15

DI/DE ´ AE/AC

38

4(BC/BB):1(BC/BD):6(BC/BE):4(BC/DE): 6(BD/BB):5(BD/BD):6(BD/BE):4(BD/DE) 4(AE/CC):5(AE/EE):5(AE/CE):3(AE/00): 2(BE/CC):4(BE/EE):6(BE/CE):4(BE/00) 5(AE/AE):7(AE/A0):8(AE/E0 or EE): 2(DE/AE):5(DE/A0):13(DE/E0 or EE)

Gb

G-Pooledb v2-Pooledb

5.55 2.74

2.44

2.78 1.09

1.06

4.28 2.62

2.60

4.20 2.11

2.03

0.39 0.39 4.56 2.02

1.67 2.00

4.73 3.12

2.89

4.29 3.70

3.69

5.39 0.28

0.28

15.78 2.27

2.22

4.87 3.29

3.54

11.20 0.08

0.08

20.64 2.32

2.33

21.07 2.02

2.00

16.01 3.58

3.76

9.29 0.56

0.55

5.90 1.49

1.56

8(BC/BC):0(BC/CD):0(BC/BD):0(BC/DD): 100.60 50.12 0(CG/BC):12(CG/CD):0(CG/BD):0(CG/DD): 0(BH/BC):0(BH/CD):8(BH/BD):0(BH/DD): 0(GH/BC): 0(GH/CD):0(GH/BD): 8(GH/DD) 9(DF/DD):0(DF/DE):10(DF/DF):0(DF/EF): 49.62 48.55 0(FI/DD):6(FI/DE):0(FI/DF):10(FI/EF) 0(AD/AD):0(AD/AE):6(AD/CD):0(AD/CE): 13(DE/AD):0(DE/AE):0(DE/CD):0(DE/CE): 0(AI/AD):0(AI/AE):0(AI/CD):13(AI/CE): 0(EI/AD):6(EI/AE):0(EI/CD):0(EI/CE)

36.89

2(BC/BC):4(BC/BH):6(BC/CG):3(BC/GH): 6(BD/BC):4(BD/BH):6(BD/CG):6(BD/DD) 7(AE/FD):8(AE/FI):8(BE/FD):6(BE/FI) 3(AE/AD):3(AE/AI):6(AE/DE):4(AE/EI): 4(DE/AD):7(DE/AI):5(DE/DE):2(DE/EI) 3(BC/BC):6(BC/CD):3(BC/BD):2(BC/DD): 6(BD/BC):6(BD/CD):4(BD/BD):6(BD/DD) 2(AE/DD):2(AE/DE):6(AE/DF):5(AE/EF): 5(BE/DD):3(BE/DE):3(BE/DF):3(BE/EF) 7(AE/AD):3(AE/AE):4(AE/CD):5(AE/CE): 6(DE/AD):3(DE/AE):3(DE/CD):8(DE/CE) 1(BB/BC):2(BB/BH):4(BB/CG):5(BB/GH): 1(BD/BC):1(BD/BH):5(BD/CG):0(BD/GH): 5(BE/BC):3(BE/BH):2(BE/CG):3(BE/GH): 2(DE/BC):3(DE/BH):2(DE/CG):2(DE/GH) 7(CE/DF):7(CE/FI):3(C0/DF): 4(C0/FI): 4(E0/DF):5(E0/FI):5(00/DF):2(00/FI) 1(AE/AD):3(AE/AI):2(AE/DE):3(AE/EI): 2(A0/AD):4(A0/AI):3(A0/DE):0(A0/EI): 4(E0 or EE /AD):5(E0 or EE /AI): 8(E0 or EE /DE):2(E0 or EE /EI) 0(BB/BC):4(BB/CD):2(BB/BD):5(BB/DD): 1(BD/BC):5(BD/CD):1(BD/BD):0(BD/DD): 5(BE/BC):3(BE/CD):3(BE/BD):2(BE/DD): 3(DE/BC):1(DE/CD):2(DE/BD):2(DE/DD) 4(AC/AD):3(AC/AF):2(AC/CD):0(AC/CF): 2(CE/AD):0(CE/AF)5(CE/CD):4(CE/CF): 1(A0/AD):3(A0/AF):3(A0/CD):6(A0/CF): 1(E0/AD):1(E0/AF):3(E0/CD):2(E0/CF) 4(CE/DD):5(CE/DE):4(CE/DF):1(CE/EF): 1(C0/DD):0(C0/DE):3(C0/DF):4(C0/EF): 2(E0/DD):1(E0/DE):2(E0/DF):4(E0/EF): 2(00/DD):1(00/DE):2(00/DF):1(00/EF) 2(AE/AD):3(AE/AE):2(AE/CD):4(AE/CE): 3(A0/AD):2(A0/AE):2(A0/CD):5(A0/CE): 10(E0 or EE/AD):2(E0 or EE/AE): 4(E0 or EE /CD):5(E0 or EE /CE) 2(AC/AB):6(AC/BD):5(CE/AB):4(CE/BD): 8(A0/AB):4(A0/BD):4(E0/AB):3(E0/BD)

35.51

879 Table 5 (Contd.) Locus combination

Cross

Parental genotypes (Locus 1/Locus 2)

Number Distribution of progeny genotypes of observed in each expected genotypic progeny class (Locus1/Locus2)a

PGL14/PGL15

16 ´ 15

DF/AD´AC/BB 34

3(AD/AB):4(AD/BD):7(CD/AB):4(CD/ BD):2(AF/AB):3(AF/BD):6(CF/AB):5(CF/ BD)

a

The expected segregation ratio was 1:1:1:1 .... for each genotypic class in respective crosses, except for the locus combinations PGL7/ PGL12 (1:1:2:1:1:2), and PGL12/PGL13 and PGL12/PGL14 (1:1:1:1:1:1:1:1:2:2:2:2) in cross 19´15, because due to the presence of a null allele at PGL12, E0 heterozygotes could not be distinguished from EE homozygotes at this locus and these two genotypic classes were therefore pooled

Gb

G-Pooledb v2-Pooledb

4.56

3.30

3.18

b

The G and v2 values calculated for tests of the linkage between PGL13 and PGL14 were highly signi®cant (P