Map order and linkage distances of molecular markers ... - Springer Link

10 downloads 0 Views 541KB Size Report
Map order and linkage distances of molecular markers close to the supernodulation (nts-1) locus of soybean. Received: 18 January 1996 / Accepted: 9 October ...
 Springer-Verlag 1997

Mol Gen Genet (1997) 254:29 – 36

ORIGINAL PAPER

A. Kolchinsky · D. Landau-Ellis · P. M. Gresshoff

Map order and linkage distances of molecular markers close to the supernodulation (nts-1) locus of soybean

Received: 18 January 1996 / Accepted: 9 October 1996

Abstract The molecular characteristics of markers in the chromosome region surrounding the supernodulation gene (nts-1) of soybean (Glycine max L. Merr.) were investigated in 187 F2 plants from a cross of G. max cv. Bragg (nts) and G. soja PI468.397 (wild-type nodulation). RFLP marker pUTG-132a, linked tightly (0.7±0.5 cM) to nts-1, was converted to a PCR marker. The polymorphism resides within a 1.72 kb PstI fragment and consists of an 832 bp insertion in G. max relative to the wild progenitor G. soja. The insertion is flanked by a 35 bp direct duplication that was found only once in G. soja. Data suggest that the pUTG-132a sequence exists only once in the genome, which is compatible with the recessive nature of nts-1. Accordingly, pUTG-132a is a valuable marker for map-based cloning. Another RFLP marker, pA-381, was mapped 4.8 cM distal to nts-1. Marker order, established by Maximum Likelihood Analysis, placed nts-1 between pUTG-132a and pA-381. To generate additional molecular markers, a segregating F2 population was analysed using bulked segregant analysis (BSA) and single oligonucleotide primer-based PCR (DNA amplification fingerprinting; DAF). PCR marker pcr5-4L was mapped to soybean linkage group H and sequenced. The data revealed (i) recombination events and marker order in the nts-1 region; (ii) the molecular nature and cause of polymorphisms in linked molecular markers; (iii) a low density of polymorphisms around nts-1, and (iv) diploidy of the distal region of linkage group H of soybean. Key words Symbiosis · RFLP mapping · Nitrogen fixation · Bulked segregant analysis · Nodulation

Communicated by A. Kondorosi A. Kolchinsky · D. Landau-Ellis · P. M. Gresshoff (&) Plant Molecular Genetics and Center for Legume Research, The University of Tennessee, 269 Ellington Building, Knoxville, TN 37901–1071 USA

Introduction Nodulation of legumes is a process of plant development initiated by complex plant-bacterium interactions (Hirsch 1992; Verma 1992; Spaink 1992). Mutational analysis has demonstrated that several plant genes control this process (Caetano-Anolle´s and Gresshoff 1991; Gresshoff 1993). Supernodulation (Carroll et al. 1985a,b), resulting from the absence (or decrease) of internal autoregulation of nodulation (Gresshoff 1993) and in abundant nodulation, is controlled by a recessive Mendelian mutation (Delves et al. 1988), with twelve separate alleles known. Grafting of shoots onto roots clearly demonstrated that the supernodulation phenotype is controlled by the shoot, and most probably by the leaf (Delves et al. 1986, 1992). Some alleles conferred increased nitrogen carryover and moderate yield increases (Song et al. 1995), providing the basis for a commercial release of the material in Australia (as ‘‘Nitrobean-60’’). Similar supernodulation and nitrate-tolerant mutants of soybean were isolated by others (Gremaud and Harper 1989; Akao and Kouchi 1992); some of these (e.g., ‘En6500’) are allelic to the mutant locus found in nts382 and nts1007, while others complement and may represent another mutant site (J. Harper, Univ. of Illinois, personal communication). Because of the likelihood of another locus controlling supernodulation, we propose to call the locus defined in line nts382 nts-1. Using clones from the RFLP map generated at Iowa State University (Keim et al. 1990; Shoemaker et al. 1992), Landau-Ellis et al. (1991) first linked the nts382 allele to the pA-132 marker on what is now known as Molecular Linkage Group H (MLG H; NB., previously labeled E). However, the pA-132 clone turned out to be a chimera comprising three unrelated and non-contiguous PstI fragments (Landau-Ellis and Gresshoff 1994). One of the subclones of pA-132, labeled pUTG132a, 1.72 kb in length, was the only one linked to nts-1. It gave a relatively simple RFLP pattern in DraI digests

30

compared to the original pA-132 (Gresshoff 1995, 1996), which suggested a single-copy locus. Linkage was consolidated by the further mapping of the allelic nts1007 mutation (see Delves et al. 1988) to the same marker (Landau-Ellis and Gresshoff 1992). Our goal is the map-based cloning of nts-1 (Wicking and Williamson 1991; Gresshoff, 1995); this strategy is dictated by our inability to detect altered gene products using in vitro translation and total protein approaches (Sayavedra-Soto et al. 1995). Map-based cloning has been successful with plant developmental (Leung et al. 1994; Leyser et al. 1993) and resistance genes (Martin et al. 1993; Mindrinos et al. 1994; Bent et al. 1994). Two prerequisites are the close linkage of the gene of interest to molecular markers, and the availability of YACs (yeast artificial chromosomes; e.g., Grill and Somerville 1991; Edwards et al. 1992; Martin et al. 1992) or BACs (bacterial artificial chromosomes; Shizuya et al. 1992, Woo et al. 1995). With this goal in mind we cloned soybean DNA into BACs (Pillai et al. 1996) and YACs (Funke et al. 1994; Funke 1995), providing a partial library of about 200 kb average insert size with a maximum of 900 kb. Although nts-1 was mapped to Linkage Group H close to pUTG-132a, the analysis used only 20 selected plants; we thus needed to determine from a larger and complete population (i) the most likely marker order and exact marker distances; (ii) whether pUTG-132a detects more than one region (the Southern blot revealed three bands) and (iii) whether additional markers close to nts-1 could be isolated.

Materials and methods Plant lines used in this study were Glycine max (L.) Merr. cv. Bragg and its near-isogenic supernodulating mutant nts382. The ancestral soybean G. soja PI468.397 was chosen because of its fertility in crosses with nts382 and high frequency of molecular polymorphism. Other cultivars or breeding lines were provided by Dr. Fred Allen (University of Tennessee) and Dr. K. Gordon Lark (University of Utah). For linkage analysis, a F2 population was made from G. max nts382 and G. soja PI468.397 (Landau-Ellis et al. 1991). DNA was extracted from F2 plants and parental plants and probed with appropriate probes for RFLP mapping as described elsewhere (Landau-Ellis et al. 1991). Plasmids were maintained and propagated in E. coli DH5a. The pBS(+) vector (Stratagene, La Jolla, Calif) was used throughout the work. Plasmid DNA was isolated by alkaline lysis (Sambrook et al. 1989). Plant DNA was isolated according to Dellaporta et al. (1983). Oligonucleotides were synthesized by NBI (Plymouth, Minn) or Integrated DNA Technologies (Gaithersburg, Md) and used without further purification. Sequencing was done by the dideoxy method based on 32Plabeling, using a sequencing kit from USB (Cleveland, Ohio) as recommended by the supplier for the range up to 300 nucleotides; and by a non-radioactive procedure based on silver-staining of PCR products (Promega, Madison, Wis.) within the range of 150– 550 nucleotides from the 5′-end of primer. Analysis of the sequences was carried out with the aid of the GCG software package (University of Wisconsin, Madison). For direct sequencing of PCR products, the band of interest was amplified in 25–50 ll, then purified by electrophoresis in low-

melting point agarose, and used directly for sequencing using the Sequitherm kit from Epicentre Technologies. Standard PCR conditions were as follows. Samples contained 50 ng DNA 0.1 lM of each primer and 0.7 units Taq polymerase (Perkin-Elmer, Emeryville, Calif) in 1 × PCR-buffer (10 mM TRISHCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.02% Triton X-100, 0.02% Tween-20, 100 lg/ml gelatin). The PCR consisted of DNA melting for 3 min at 95° C and 35 cycles of 1 s at 96° C, 1 s at 60° C, 2 min at 72° C using an Ericomp thermocycler. The samples were analysed on 1.2% agarose gels and stained with ethidium bromide. For amplification of DNA with arbitrary primers, three tubes were run for each primer containing 2, 5, and 10 ng of bulked DNA in 5 ll final volume. Therefore, screening of two sample pools with one primer required six samples (costs were compensated by the small amplification volume). Reaction mix contained 1 × PCR buffer; 200 lM dCTP, dGTP and dTTP; 25 lM dATP, 0.7–1.0 lCi [a-32P] dATP, 3 lM of primer (8–30 nucleotides) and 0.3 units AmpliTaq polymerase (Perkin-Elmer). PCR was carried out as follows: 3 min at 95° C; then 35 cycles of 5 s at 95° C, 30 s at 35° C, 1 min at 72° C; then 5 min at 72° C. After the reaction was over, 4 ll of sequencing dye was added, the reaction was heated for 5– 10 min at 95° C, and 2 ll loaded on a 5% sequencing gel. The gel was exposed overnight.

Results Sequences of the RFLP marker pUTG-132a in G. max and G. soja and the nature of polymorphism The pUTG-132a probe detected several restriction fragments, two of which were polymorphic. The possibility existed that the probe recognized at least two genomic regions, which would complicate its use for map-based cloning approaches. [N.B., for clarity in nomenclature, probe or plasmid pUTG-132a is distinguished from the locus designated as pUTG-132a]. We showed that the two polymorphic bands in genomic Southern blots were caused by a single genetic event, demonstrating that pUTG-132a detects an unique region, in agreement with the recessive nature of the supernodulation trait. Restriction mapping of pUTG132a and Southern hybridization to G. max and G. soja DNA demonstrated that the polymorphism between parental lines resided within the cloned fragment. To define the nature of the polymorphism, the 1.72 kb fragment derived from G. max was sequenced, and proved to be AT-rich (70% AT pairs) with several oligoA and oligo-T stretches. The fragment lacked open reading frames exceeding 80 amino acids in length, and did not hybridize to any discrete RNA bands on Northern blots made from growing shoot RNA (data not shown). Thus, the fragment seems to represent noncoding DNA. Two primers located at the ends of this 1.72 kb fragment were used to amplify its homologous sequence from G. soja genomic DNA. The data are presented schematically in Fig. 1. The sequence for G. max was submitted to the EMBL Data Library (Accession No. Z26335). The 1718 bp fragment from G. max differed from the G. soja fragment by a major insertion of 832 bp. This

31

Fig. 1 The organization of the pUTG-132a fragment from G. max DNA and of the corresponding region in G. soja DNA. Direct repeats flanking the insertion in G. max are shown as open triangles. The locations of PCR primers used for screening (see text) are designated by small arrows. The lengths of DNA segments are indicated as well as some restriction sites (D, DraI; P, PstI). LI = 5 CAC GGT CACTCA TGG GCC AT G 3′; LII = 5′ GCA GCA GTG TTG GGC ATG TCT CT 3′; R = 5′ CTG CAG AAT TGG ATT CCC AAA AGC 3′ 0

insertion is flanked in G. max by a direct repeat of 35/36 bp (with 5 point mutations) that was found in only one copy in G. soja. Outside the insertion, there were 17 minor differences between the two sequences (one per 50 bp), including another short insertion of 13 bp in the G. max fragment. The length of the G. soja fragment corresponding to pUTG-132a was 841 bp. This size difference was easily identified by various combinations of PCR primers. The polymorphism used for the original mapping of pUTG-132a was based on DraI digestion of genomic DNA (Landau-Ellis et al. 1991). In G. max, three fragments were detected: one strong band at 2.9 kb, one intermediate band at 0.5 kb and a weak band at 2.6 kb. Since all fragments from the restriction and sequence map of pUTG-132a were accounted for in Southern blots of DraI-digested soybean DNA, this region was assumed to be present only once. Subsequently, a HindIII-derived polymorphism was used, since it produced a simpler pattern, with a 2.8-kb band in G. max and a 2.0 kb band in G. soja, with no overlap (data not shown; see Gresshoff 1996), allowing for easy identification of heterozygotes. Conversion of the RFLP to a PCR marker pUTG-132a was converted to a sequence-tagged site (STS). PCR primers were synthesized (listed in legend to Fig. 1) to detect polymorphisms of pUTG-132a. Primer (R) found homology in both G. max and G. soja genomic DNA, while primer (LI) was homologous to sequences absent in the G. soja, but present in G. max. The more distant primer (LII) was homologous to both G. max and G. soja DNA. The calculated melting temperatures for all three primers were similar. PCR products were generated by different combinations of primers and DNA (Fig. 2). With primers R and LI, G. max DNA yielded the expected 1.1 kb product

Fig. 2 PCR amplification of allelic variants of the marker pUTG132a in G. max and G. soja DNA. Lane 1, G. max DNA, primers R+LI; 2, G. max DNA, R+LII; 3, G. soja DNA, R+LII; lanes 4–6 mixture of G. max and G. soja. DNA (1:1); 4, R+LII; 5, R+LI+LII; 6, identical to 5, but primer LII added after the first five PCR cycles; lanes 7–10, DNA from F1 hybrid G. max/G. soja.have 7, primers R+LI; 8, primers R+LII; 9, R+LI+LII; 10, same as 9, but primer LII added after the first five cycles. M, marker (1 kb ladder from BRL)

(Fig. 2, lane 1). Amplification of G. max DNA with primers R and LII gave a 1.72 kb fragment (Fig. 2, lane 2). Amplification of G. soja DNA with primers R and LII (Fig. 2, lane 3) yielded a 0.85 kb fragment, but no amplification product with primers R and LI because primer site LI is absent in G. soja DNA When DNA from a G. max /G. soja F1 hybrid was amplified with primers R and LI, only the 1.1 kb band was detected (Fig. 2, lane 7); in contrast, the combination of R and LII gave only the 0.85 kb band, and the 1.72 kb band was barely seen (Fig. 2, lane 8). Using a mixture of R, LI and LII, synthesis of the 0.85 kb band suppressed the amplification of the two other (i.e., 1.1 and 1.72 kb) bands (Fig. 2, lane 9). Only by adding primer LI five cycles prior to the common primer LII was it possible to amplify the 0.85 and 1.1 kb bands in the same tube (Fig. 2, lane 10). The 1.72 kb band was still barely visible. Similar results were obtained with an equal mixture of G. max and G. soja DNA (Fig. 2, lanes 4–6). At higher concentrations of the primers (more than 0.1 lM), or low concentration of the enzyme (less than 0.05 units per ll), the 1.72 kb band was not amplified at all. Estimation of the genetic distance pUTG-132a-nts-1 The linkage of pUTG-132a to nts-1 was established originally by RFLP analysis of only 20 F2 plants showing supernodulation (Landau-Ellis et al. 1991; Landau-Ellis and Gresshoff 1992), as wild-type F2 plants were not scored. Here we present an enlarged data set for linkage analysis using Maximum Likelihood Ana-

32

lysis (Lander et al. 1987), which yields marker order as well as genetic distances. Out of 113 F2 plants [from a cross designated C16 (nts382 × G. soja PI 468.397)], 24 were classified as nts homozygotes by nodulation phenotype (four to five weeks after inoculation with Bradyrhizobium japonicum strain USDA110). All 113 were amplified in two separate reactions with PCR-primer combinations LI/R and LII/R. The allelic assignments were verified by Southern hybridization using pUTG-132a and HindIII digests. All 24 nts plants were homozygous for the G. max allele. All other plants (with wild-type phenotype) were either homozygous or heterozygous for the G. soja banding pattern, except for one plant that was homozygous for the G. max allele. These wild-type plants must carry at least one wild-type allele of the nts-1 gene; however, we have no assignment for this as F3 plants were not maintained. Thus it is possible that further hidden recombination events occurred in this region. The one presumed recombinant plant (1 out of 113 from the C16 population) helped us to establish marker order. Linkage analysis using the MAPMAKER software confirmed marker order and calculated a genetic distance of 0.7±0.5 cM between pUTG-132a and nts-1 (see Fig. 3). Two additional groups of 46 supernodulating plants from two other F2 populations (nts382 × G. soja and nts1007 × G. soja) were screened by PCR, confirming marker order and the proximity of pUTG-132a to nts-1.

Fig. 3 Map of markers in vicinity of the nts-1 gene. Data are derived from 113 F2 plants screened for supernodulation phenotype vs. wildtype nodulation and scored for segregating RFLP markers from the USDA Iowa soybean map and PCR-based markers derived either from bulked segregant analysis (BSA) (pcr5-4L). pPV-7 and pPV-1 are random genomic PstI clones of Phaseolus vulgaris provided by Dr. E. Vallejos (U. of Florida, Gainesville). The pA-36 cluster was used for physical mapping to determine that 1 cM is equivalent to about 500 kb (Funke et al. 1993). The LOD score for the arrangement shown is )84.12, compared to –85.82 for the reversed of map order of nts-1 and pUTG-132a

Map positions and genetic distances of two flanking RFLP markers, pA-36 and pA-381, were screened in the same C16 population of 113 plants; partial sequences for both are known, allowing detection by PCR. Detection of additional markers using Bulked Segregant Analysis Bulked segregant analysis (Michelmore et al. 1991) used two DNA pools, one from 17 supernodulating F2 plants homozygous for the G. max allele at the pUTG-132a locus, the other 17 plants homozygous for the G. soja type pUTG-132a locus. These pools were amplified with 160 primers ranging in size from 8 to 30 nucleotides. All reactions were carried out at high primer concentration (3 lM) and low annealing temperatures (30° C) to favor DAF conditions (Caetano-Anolle´s et al. 1991). Products were analyzed by agarose electrophoresis as described for RAPD analysis (Williams et al. 1990) or by acrylamide electrophoresis with silver staining (Bassam et al. 1991). No reproducible polymorphisms were found in this series (data not shown). To increase the chance of detecting a marker the pools were reconfigured to enlarge the window of detection. First, plants homozygous not only for a given pUTG-132a allele but also for the two flanking markers pA-381 and pA-36 (Fig. 3) were combined. This generated a 20 cM region of presumed homozygosity. This strategy increased the window for the detection of potential polymorphisms from approximately 10 cM (5 cM in each direction from a single homozygous marker) to approximately 25 cM (20 cM from the interval plus 5 cM in one direction because the region appears to be terminal). The reasoning behind the estimate of a 5 cM window comes from work by Prabhu (1995), who demonstrated that DNA mixtures showing an average polymorphic DAF band will no longer be detected on silver-stained DAF-PAGE gels, if the polymorphic template DNA of one genotype is diluted about 20-fold relative to the template DNA of another genotype. This gives the stated window size of 5 cM, meaning that polymorphisms further away than that will be present in both bulks and therefore be detected equally as a monomorphic band. Therefore, if one desires to detect markers further than 5 cM from a specific point, the window of selected homozygosity needs to be extended. While these calculations are based on DAF technology using silver staining and PAGE, similar values were obtained for RAPD markers (Williams et al. 1990). All amplification reactions were done in triplicate, with template DNA concentrations of 2, 5, and 10 ng/ll. To increase sensitivity and yield of bands per primer reaction and to avoid staining of bulky sequencing gels, products were separated in long (60 cm) sequencing gels after labeling with radioactive precursor (see Materials and methods). An example of such an analysis is shown in Fig. 4. Large numbers (about 100) of amplification

33

Fig. 4 Bulked segregant analysis near the nts-1 gene (see text for details). Each set of six lanes on the sequencing gel represents three lanes of one combined sample pool and three lanes of another combined sample at different DNA concentrations. These sets were obtained with primers 5′-CAACCTTGCAACAAAGTCCTTG-3′ (21mer, set 1); 5′-AGCAGATGTACCCATTGCTT-3′ (20mer, set 2); 5′-AATGCAGCTGGC-3′ (12mer, set 3); 5′-GTAACGCC-3′ (8 mer, set 4); 5′-TCTGTGCTGG-3′ (10mer, set 5); 5′-GATCGCAG-3′ (8 mer, set 6, failed to produce any bands); 5′-CGCGGCCA-3′ (8 mer, set 7). The arrow at the top indicates the marker lane, the diagonal arrow shows the polymorphic band converted to a SCAR (see Fig. 5). Sizes of some marker bands in nucleotides are given on the left

products of different sizes were routinely detected. Intensities of bands varied as did those on silver-stained gels (cf. Prabhu and Gresshoff 1994). Two polymorphic bands were detected in pools after using 21 oligonucleotide primers. These were reamplified with the original primer, cloned, sequenced and converted into SCARs (sequence characterized amplified regions; Paran and Michelmore 1992). Despite the precautions, one of them turned out to be unreliable. One repeatable polymorphism (pcr5-4L) was detected and linked 7.6 cM distal to pA-36 (Fig. 3). Use of larger primer populations on the same bulk DNA should produce markers within the homozygous interval. Conserved nature of the pUTG-132a sequence Using PCR primers R and LII, pUTG-132a homologous regions were amplified from other varieties of G. max, namely cvs. Peking (an older soybean line), Enrei

(a Japanese commercial line, which gave rise to En6500; Akao and Kouchi 1992), Minsoy, Noir 1 [both parents of recombinant inbred lines (Lark et al. 1993)], and DPS3589 [an experimental line from Delta Pine Company characterized by cyst nematode (race 3) resistance]. The PCR products were sequenced directly using primers depicted in Fig. 1 (approximately 1 kb from each product) and compared to homologous sequences in G. max cv. Bragg and G soja (data not shown). The average incidience of mutations in Peking as compared to Bragg was approximately one per 100 nucleotides two times lower than between Bragg and G. soja. However, not a single difference was found between Bragg, Minsoy, Noir 1, Enrei and DPS3589, giving a calculated mutation rate among soybean genotypes studied in this area of genome of less than one per 1000 nucleotides. The absence of sequence polymorphism prevented the mapping of pUTG-132a on the RIL map (Lark et al. 1993). However, flanking markers pA-381 and pPV-7 (a random genomic clone derived from Phaseolus vulgaris which mapped about 4 cM from pA-381, obtained from Dr. E. Vallejos, University of Florida, Gainesville, Fla.) were mapped to RIL linkage group U23, suggesting that nts-1 maps to this location (P.M. Gresshoff and A. Filatov, unpublished results). In the C16 mapping population, pPV-7 is 7.3 cM from pA-381, comparable to the RIL distance (P.M. Gresshoff and A. Filatov, unpublished data). Using such mapping in two populations, morphological markers can be integrated into other maps, or new molecular markers can be detected close to mapped genes of interest.

Discussion Genetic analysis of plant genes that control the nodulation process was initiated with the eventual goal of cloning of these genes. As a first step in finding symbiotic genes, the RFLP marker pUTG-132a was tightly linked to two alleles (nts382 and nts1007) of the nts-1 gene conferring the supernodulation phenotype (Landau-Ellis et al. 1991; Landau-Ellis and Gresshoff 1992). An analysis of a complete population was needed as the initial mapping of nts-1 was based on a small selected sub-population. Based on 187 plants from several crosses (mainly from a complete population of 113 plants of the C16 cross), the distance between pUTG-132a and nts-1 may be as small as 0.7 cM (Fig. 3). The error associated with this value, based on only one recombinant, is ± 0.5. The map order places nts-1 between pUTG-132a and pA-381. Physical mapping in the nearby pA-36 RFLP cluster suggested that one centimorgan represents about 500 kb (Funke et al. 1993). This may place pUTG-132a about 350 kb from nts-1. In contrast, pA-381, although flanking nts-1 on the other side, may be nearly 2 Mb away, which may be too large a distance to traverse with contiguous clones. We believe that pUTG-132a provides the best available starting point for the isolation of

34

Fig. 5 Polymorphism pcr5-4L converted to a sequence characterized amplified region (SCAR). DNA samples were amplified with primers 5′-ATGCAGCCATGACCAAGAGT3′- and 5′-GTGTATTCTAAA-GTCACCCAAC-3′, derived from the sequence of the polymorphic band (see Fig. 4). M, marker; lane 1, bulked sample corresponding to the nts-1 allele; 2, bulked wild-type sample; 3, G. max parent; 4, G. soja parent; 5 and 6, individual plants homozygous for G. max-type pUTG-132a; 7 and 8 individual plants homozygous for G. soja-type pUTG-132a. The faint band present in lane 2 is due to the presence of recombinants in the bulked sample.

homologous YAC or BAC clones carrying soybean genomic DNA (Funke et al. 1994). Sequencing of pUTG-132a and analysis of the surrounding region in genomic DNA in both parental lines used for the original mapping revealed the nature of its rearrangement in G. soja as compared with G. max. A deletion of 832 bp was detected in G. soja that removes an internal DraI site. As a result, a 0.5 kb DraI fragment disappeared and one of the other two hybridizing fragments was shortened by 0.4 kb. The patterns of hybridization observed in both parental lines suggested that this region is present only once in the soybean genome. This is consistent with the fact that nts-1 is a recessive mutation, and that pUTG-132a detects only a single band in blots with HindIII- digested DNA. Physical mapping (Funke et al. 1993) around the pA-36 marker showed extensive genomic duplication (reflecting ancestral polyploidy), suggesting perhaps that this distal arm of MLG H, but not the region around pA-36, is present only once in the genome. Indeed, fluorescent in situ hybridization (FISH) of the pUTG-132a probe onto soybean metaphase chromosomes demonstrates a single locus, supporting the genetic and molecular data (P. Keim, Univ. of Arizona, Flagstaff, unpublished data). The original RFLP marker pA-132, and its subfragment pUTG-132a, were obtained from a library of soybean DNA fragments produced with a methylationsensitive restriction enzyme, PstI (Keim et al.1990). Such libraries are presumably enriched in transcribed sequences. However, the pUTG-132a fragment did not contain open reading frames longer than 80 amino acids; a search in the EMBL Data Library did not yield any reasonable homologies exceeding 57%; finally, the frag-

ment has an AT content of 70%. Likewise, no mRNA was detectable by Northern analysis. This suggests that the region is not transcribed. The sequence deleted in G. soja is flanked in G. max by a direct repeat of 35/36 bp, with five single-base changes. This short motif occurs only once in the G. soja clone, corresponding exactly to one of the G. max repeats. Duplication of target sequences is a well-known hallmark of transposable element insertions. However, the sequence in between the two direct repeats had no homologies in the soybean genome (data not shown) and did not resemble the only known soybean transposable element, Tgm (Rhodes and Vodkin 1988). However, it may represent a class of retrotransposons. The direct repeat may have been present in a common ancestor and may have promoted the excision of the 832 bp fragment in G. soja, while its extensive mutation in G. max prevented further rearrangements (for a review of possible mechanisms, see Drake 1991). The incidence of minor changes outside the deleted regions comparing G. soja with G. max (20 changes per 800 bp, or one change per 40 bp), was relatively high; however, the fragment showed striking stability among several cultivars and landraces studied, with fewer than one mutation per 1000 bp. We believe that pUTG-132a belongs to a category of stable markers. Indeed, we were not able to find any changes at the nucleotide level within the marker. No RFLP polymorphism was detected in DNA from 12 ancestral lines and four commercial varieties digested with HindIII; only Peking and G. soja showed polymorphism (data not shown). Moreover, this conserved domain could be significantly larger, as a 14 kb insert in a lambda clone complementary to pUTG-132a did not reveal any additional RFLPs with several restriction enzymes (J. Deckert, unpublished data). Also, the failure to detect polymorphisms in our first bulked segregant search, and the external location of pcr5-4L are indicative of the conserved nature of this region. Finally, our modification of bulked segregant analysis offers significant advantages. It combines high reproducibility, achieved by using three template DNA concentrations, with simple cycling and relatively safe and cost-effective labeling for band detection. Acknowledgments The research was supported by the Ivan Racheff Endowment, the Tennessee Soybean Promotion Board, the United Soybean Board (USB), and The University of Tennessee Agricultural Experimental Station. Janice Crockett is thanked for help with the manuscript and Dr. Fred Allen for soybean lines. Technical assistance of Becky Ervin, Eugenia Almeida and Marcia Young is highly appreciated. Dr. E. Vallejos is thanked for providing clones and Dr. A. Filatov is thanked for Phaseolus mapping data.

References Akao S, Kouchi H. (1992) A supernodulating mutant isolated from soybean cultivar Enrei. Soil Sci Plant Nutrit 38:183–187

35 Bassam BJ, Caetano-Anolle´s G, Gresshoff PM (1991) A fast and sensitive silver-staining for DNA in polyacrylamide gels. Anal Biochem 196:80–83 Bent AF, Kunkel BN, Dahlbeck D, Brown KL, Schmidt R, Giraudat J, Leung J, Staskawicz BJ (1994) RPS2 of Arabidopsis thaliana: a leucine-rich repeat class of plant disease resistance genes. Science 265:1856–1860 Caetano-Anolle´s G, Gresshoff PM (1991) Plant genetic control of nodulation in legumes. Annu Rev Microbiol 45:345–382 Caetano-Anolle´s G, Bassam B, Gresshoff PM (1991) DNA amplification fingerprinting using very short arbitrary oligonucleotide primers. Biotechnology 9:553–557 Carroll BJ, McNeil DL, Gresshoff PM (1985a) Isolation and properties of soybean (Glycine max) mutants that nodulate in the presence of high nitrate concentrations. Proc Natl Acad Sci USA 82:4164–4166 Carroll BJ, McNeil DL, Gresshoff PM (1985b) A supernodulation and nitrate tolerant symbiotic (nts) soybean mutant. Plant Physiol 78:34–40 Dellaporta SL, Wood J, Hicks JB (1983) A plant DNA minipreparation. Version II. Plant Mol Biol Rep 1:19–21 Delves AC, Mathews A, Day DA, Carter AS, Carroll BJ, Gresshoff PM (1986) Regulation of the soybean-Rhizobium symbiosis by shoot and root factors. Plant Physiol 82:588–590 Delves AC, Carroll BJ, Gresshoff PM (1988) Genetic analysis and complementation studies on a number of mutant supernodulating soybeans. J Genet 67:1–8 Delves AC, Higgins A, Gresshoff PM (1992) The shoot apex is not the source of the systemic signal controlling nodule numbers in soybean.Plant Cell Environ 15:249–254 Drake JW (1991) Spontaneous mutations. Annu Rev Genet 25:125–146 Edwards KJ, Thompson H, Edwards D, Saizieu A, Sparks C, Thompson JA, Greenland AJ, Eyers M, Schuch W (1992) Construction and characterization of a yeast artificial chromosome library containing three haploid maize genome equivalents. Plant Mol Biol 19:195–201 Funke RP (1995) High molecular weight DNA approaches to study nodulation genes in soybean. PhD. dissertation. The University of Tennessee, Knoxville Tenn. Funke RP, Kolchinsky AM, Gresshoff PM (1993) Physical mapping of a region in the soybean (Glycine max) genome containing duplicated sequences. Plant Mol Biol 22:437–446 Funke RP, Kolchinsky AM, Gresshoff PM (1994) High EDTA concentrations cause entrapment of small DNA molecules in the compression zone of pulsed field gels, resulting in smaller than expected insert sizes in YACs prepared from size selected DNA. Nucleic Acids Research 22:2708–2709 Gremaud MF, Harper JE (1989) selection and initial characterization of partially nitrate tolerant nodulation mutants of soybean. Plant physiol 89:169–173 Gresshoff PM (1993) Molecular genetics analysis of nodulation genes in soybean. Plant Breed Rev 8:275–318 Gresshoff PM (1995) Moving closer to the positional cloning of legume nodulation genes. In: Tikhonovich I, Provorov N, Romanov V, Newton WE (eds) Nitrogen fixation: horizons and applications. Kluwer Academic Publishers, Dordrecht, Netherlands, pp 416–420 Gresshoff PM (1996) Molecular genetic analysis of soybean nodulation mutants. In: Shoemaker RC, Verma, DPS (eds) Soybean: genetics, molecular biology and biotechnology. CAB International, Oxford, UK, pp 189–217 Grill E, Somerville C (1991) Construction and characterization of a yeast artificial chromosome library of Arabidopsis which is suitable for chromosome walking. Mol Gen Genet 226:484– 490 Hirsch AM (1992) Developmental biology of legume nodulation. New Phytol 122:211–237 Keim P, Diers BW, Olson TC, Shoemaker RC (1990) RFLP mapping in soybean: association between marker loci and variation in quantitative traits. Genetics 126:735–742

Landau-Ellis D, Gresshoff PM (1992) Supernodulating soybean mutant alleles, nts382 and nts1007, show no recombination with the same RFLP marker. Mol Plant-Microbe Interact 5:428–429 Landau-Ellis D, Gresshoff PM (1994) The RFLP molecular marker closely linked to the supernodulation locus of soybean contains three inserts. Mol Plant Microbe Interact 7:432–433 Landau-Ellis D, Angermu¨ller S, Shoemaker R, Gresshoff PM (1991) The genetic locus controlling supernodulation in soybean (Glycine max L.) co-segregates tightly with a cloned molecular marker. Mol Gen Genet 228:221–226 Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE, Newburg L (1987) MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1:174–181 Lark KG, Weisemann JM, Mathews BF, Palmer R, Chase K, Macalma T (1993) A genetic map of soybean (Glycine max L.) using an interspecific cross of two cultivars ‘Minsoy’ and ‘Noir 1’. Theor Appl Genet 86:901–906 Leung J, Bouvier-Durand M, Morris PC, Guerrier D, Chefdor F, Giraudat J (1994) Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science 264:1448–1452 Leyer HMO, Lincoln CA, Timpte C, Lamner T, Estelle M (1993) Arabidopsis auxin resistance gene Axl1 encodes a protein related to ubiquitin-activating enzyme E1. Nature 364:161–163 Martin GB, Ganal MW, Tanksley SD (1992) Construction of a yeast artificial chromosome library of tomato and identification of cloned segments linked to two disease resistance loci. Mol Gen Genet 233:25–32 Martin GB, Brommonschenkel S, Chunwogse J, Frary A, Ganal MN, Spivey R, Wu T, Earle ED, Tanksley SD (1993) Mapbased cloning of a protein kinase gene conferring disease resistance in tomato. Science 262:1432–1436. Michelmore RW, Paran I, Kesseli RV (1991) Identification of markers linked to disease resistant genes by bulked segregant analysis: a rapid method to detect markers in specific genomic regions by using segregating populations. Proc Nat Acad Sci USA 88:9828–9832 Mindrinos M, Katagiri F, Yu G-L, Ausubel FM (1994) The A. thaliana disease resistance gene RPS2 encodes a protein containing a nucleotide-binding site and leucine-rich repeats. Cell 78:1089–1099 Paran I, Michelmore RW (1993) Development of reliable PCR markers linked to downy mildew resistance genes in lettuce. Theor Appl Genet 85:985–993 Pillai S, Funke RP, Gresshoff PM (1996) Yeast and bacterial artificial chromosomes (YAC and BAC) clones of the model legume Lotus japonicus. Symbiosis, in press Prabhu RR (1995) Soybean genome analysis using arbitrary primer technology. PhD Dissertation, The University of Tennessee, Knoxville, Tenn. Prabhu RR, Gresshoff PM (1994) Inheritance of polymorphic markers generated by DNA amplification fingerprinting and their use as genetic markers in soybean. Plant Mol Biol 26:105–116 Rhodes PR, Vodkin LO (1988) Organization of the Tgm family of transposable elements in soybean. Genetics 120:597–604 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual (2nd edn). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York Sayavedra-Soto LA, Angermu¨ller SS, Prabhu RR, Gresshoff PM (1995) Polypeptide patterns in leaves of Glycine max (L.) Merr. cv. ‘Bragg’ and its supernodulating mutant during early nodulation. Physiol Mol Biol Plants 1:27–36 Shizuya H, Birren B, Kim U-J, Mancino V, Slepak T, Tachiiri Y, Simon M (1992) Cloning and stable maintenance of 300-kilobase-pair fragments of human DNA in Escherichia coli using an F-factor-based vector. Proc Natl Acad Sci USA 89:8794–8797 Shoemaker RC, Olson TC (1993) Molecular linkage map of soybean. In: O′ Brien SJ (ed) Genetic maps: locus maps of complex genomes. Cold Spring Laboratary Harbor Press, Cold Spring Harbor, New York pp 6.131–6.138

36 Song L, Carroll BJ, Gresshoff PM, Herridge DF (1995) Field assessment of supernodulating genotypes of soybean for yield, N2 fixation and benefit to subsequent crops. Soil Environ. Biochem 27:563–569 Spaink HP (1992) Rhizobia lipo-oligosaccharides: answers and questions. Plant Mol Biol 20:977–986 Verma DPS (1992) Signals in root nodule organogenesis and endocytosis of Rhizobium. Plant Cell 4:373–382 Wicking C, Williamson B (1991) From linked marker to gene. Trends Genet 7:288–290

Williams JGK, Kubelik AR, Livak KI, Rafalski JA, Tingey SV (1990) DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res 18:6531–6535 Woo SS, Rastogi VK, Zhang H-B, Paterson AH, Schertz KF, Wing RA (1995) Isolation of megabase-size DNA from sorghum and applications for physical mapping and bacterial and yeast artificial chromosome library construction. Plant Mol Biol Rep 13:82–94