Printer Friendly

Tetrasomic linkage mapping of RFLP, PCR, and isozyme loci in Lotus corniculatus L. (Cell Biology & Molecular Genetics).

BROADLEAF BIRDSFOOT TREFOIL is a tetraploid (2n = 4x = 24) forage legume grown throughout temperate regions of the world. Having a broad natural distribution throughout Europe, Asia, and North Africa, the center (or centers) of origin for L. corniculatus is unclear (Grant and Small, 1996). Particularly adapted to marginal soils, it is grown for pasture grazing, hay, and silage. In North America, L. corniculatus is grown on [approximately equal to] 1.2 million ha (Beuselinck and Grant, 1995). Few genetic studies have been conducted in L. corniculatus, comprised mostly on the inheritance of a limited number of morphological traits that primarily suggested tetrasomic, but occasionally disomic, inheritance patterns (Grant and Small, 1996). Tetrasomic inheritance was reported for cyanogenesis (Dawson, 1941), brown keel tip color (Hart and Wilsie, 1959; Buzzell and Wilsie, 1963; Ramnani and Jones, 1984), and pubescence, chlorophyll deficiency, flower color, and corolla striping (Bubar and Miri, 1965). However, brown keel tip color (in opposition to the above findings) and seed color mottling (Donovan, 1957), tannin production (Dalrymple et al., 1984), and Pgi2 isozyme inheritance (Raelson et al., 1989) was best explained by disomic inheritance. Recent evidence from RFLP markers indicates that only tetrasomic, and not disomic, inheritance could explain phenotypic and genotypic [F.sub.2] progeny classes and double reduction events in L. corniculatus (Fjellstrom et al., 2001).

Advances made by utilizing genetically mapped markers in plant breeding are becoming well documented as they are used for mapping genes controlling valuable characteristics, studying inter- and intraspecific genetic relationships, and monitoring gene introgression. Unfortunately, L. corniculatus cannot benefit from such advances since this species has no genetic map. Mapping genetic markers in a polysomic species like L. corniculatus can be challenging because genetic mapping algorithms and software analysis programs are almost exclusively based on disomic inheritance. Various strategies have been used to construct linkage maps in tetrasomic plants, primarily in alfalfa (Medicago sativa L.) and potato (Solanum tuberosum L.). Since diploid forms of L. corniculatus are either rare or nonexistent (Grant and Small, 1996), it does not appear possible to perform mapping in diploid derivatives of tetrasomic species, as has been done in alfalfa (Brummer et al., 1993; Kiss et al., 1993; Echt et al., 1994; Diwan et al., 2000; Kalo et al., 2000) and potato (Bonierbale et al., 1988; Gebhardt et al., 1989; Freyre et al., 1994; Van Eck et al., 1994).

Another strategy for marker mapping in tetrasomic species is to analyze the inheritance of marker alleles present in single copies within parental lines and scoring them as present or absent in their progeny. Segregation ratios and recombination fractions for such single copy markers in coupling phase from polysomic species are equivalent to those from disomic species. Therefore, mapping programs used in studying disomic inheritance can be used to construct linkage maps of individual chromosomes in polyploid species using single-dose markers. Genetic maps of polyploids have been developed in this manner using RFLP alleles, termed single dose restriction fragments (Wu et al., 1992), for crops such as sugarcane (Saccharum officinarum L.) (Da Silva et al., 1993) and alfalfa (Brouwer and Osborn, 1999). Single-dose polymerase chain reaction (PCR) marker alleles have also been used for mapping polysomic genomes using arbitrarily primed PCR polymorphism or RAPDs in crops like sugarcane (Al-Janabi et al., 1993) and tetraploid alfalfa (Yu and Pauls, 1993), amplified length polymorphic DNA (AFLP) markers in tetraploid potato (Meyer et al., 1998), and simple sequence repeat (SSR) markers in tetraploid alfalfa (Diwan et al., 2000). Markers in repulsion phase linkage have not generally been used in mapping efforts for polysomic species and shown to have low accuracy in estimating recombination fractions (Hackett et al., 1998; Ripol et al., 1999), although formulas for estimating recombination among markers in repulsion phase linkage for polysomic species are available (Yu and Pauls, 1993; Hackett et al., 1998; Meyer et al., 1998; Ripol et al., 1999; Qu and Hancock, 2001).

Although less common, recombination fractions involving duplex or triplex markers, having two or three doses, respectively, can be determined for species with polysomic inheritance. Methods for estimating recombination in polysomic species have focused primarily on dominant duplex markers (Haldane, 1930; De Winton and Haldane, 1931; Mather, 1936; Yu and Pauls, 1993; Hackett et al., 1998; Meyer et al., 1998), although formulas for analyzing recombination involving dominant triplex markers have also been developed (Ripol et al., 1999). More recently, statistical methodology for linkage analysis of codominant and dominant markers in tetrasomic species has been developed (Luo et al., 2001; Wu et al., 2001). Unfortunately, because these authors presented generalized equations for estimating recombination among tetrasomic markers, no integer formulas (such as those for dominantly scored markers provided by Yu and Pauls, 1993; Hackett et al., 1998; and Meyer et al., 1998) are presently available in the literature to directly calculate codominantly duplex marker recombination.

The main objective of this research was to develop the first genetic marker map in L. corniculatus to help establish a foundation for marker assisted selection or trait dissection. Another goal of this research was to characterize the genome of L. corniculatus by depicting genome size and identifying regions of duplication or heterogeneity. Albeit not an original goal, this research also developed and makes available formulas for calculating maximum likelihood estimates of recombination involving codominantly scored duplex markers in tetrasomic species.

MATERIALS AND METHODS

Plant Materials and DNA Isolation

Inheritance of molecular markers was based on the analysis of segregation from 82 [F.sub.2] progeny derived from a cross between two diverse L. corniculatus accessions. The male parent was a vegetatively-propagated rhizomatous Moroccan accession, 'G31276' (Beuselinck et al., 1996) and the female parent was from an autogamous accession, 'AG-S4' (Steiner, 1993). [F.sub.2] progeny were obtained from an autogamous [F.sub.1] plant. Parental, [F.sub.1], and [F.sub.2] progeny DNA was isolated from lyophilized leaf and stem tissue according to Fjellstrom et al. (2001).

RFLP Marker Detection

Genomic clones used for RFLP inheritance analysis were produced from 0.5- to 1.5-kb length fragments of PstI digested G31276 DNA ligated into pSPORT1 (Life Technologies, Gaithersburg, MD) (1). Lotus corniculatus inserts were PCR amplified from plasmid miniprep DNA (Sambrook et al., 1989) using commercially available pUC forward and reverse sequencing primers (Operon Technologies, Alameda, CA). To identify low copy number clones, amplified inserts were first electrophoresed on 1.2% (w/v) agarose gels with 1 x Tris-acetate buffer, Southern blotted onto nylon membranes (Hybond N; Amersham Pharmacia Biotech, Piscataway, NJ), and hybridized with [alpha][P.sup.32]-dCTP radio-labeled total genomic L. corniculatus DNA (Landry and Michelmore, 1985). Hybridized blots were washed for 15 min twice at room temperature and twice at 60[degrees]C with 2 x standard saline citrate (0.3 M NaCl, 0.03 M sodium citrate, pH 7.0) and then exposed to autoradiographic film (X-Omat AR; Eastman Kodak, Rochester, NY) for 24 h at -80[degrees]C. Developed film was analyzed for radiographic signal intensity of each insert.

Inserts showing low radioactive signal strength were presumed to be low copy number and used for screening parental DNA for RFLP polymorphism using the methods of Fjellstrom et al. (2001) on EcoRI, EcoRV, HindIII, and XbaI (New England Biolabs, Beverly, MA) digested DNA. Progeny RFLPs were detected using the methods of Fjellstrom et al. (2001) on EcoRI and EcoRV digested DNA. Phenotypes based on RFLPs were determined from visual inspection of autoradiographs, while RFLP genotypes were scored based on marker allele signal intensity interpreted by visual inspection and using image analysis software (NIH Image 1.61; U.S. National Institutes of Health, http://rsb.info.nih.gov/nih-image/).

PCR Marker and Isozyme Detection

DNA was amplified for RAPD analysis using single 10-base oligomer primers (Operon Technologies) using the method of Steiner et al. (1995). RAPD amplification products were separated by electrophoresis in 1.75% (w/v) NuSieve 3:1 agarose (BioWhittaker Molecular Applications, Rockland, ME) 1 x tris-borate (TBE) gels or in 4% (w/v) 29:1 polyacrylamide:bis 1 x TBE nondenaturing gels, run in 1 x TBE buffer.

DNA was amplified for ISSR analysis (Zietkiewicz et al., 1994) using three single 19-base oligomers listed as follows: CTG (5'-CTC TCT CTC TCT CTC TCT G-3'), GAA (5'-GAG AGA GAG AGA GAG AGA A-3'), and GAT (5'-GAG AGA GAG AGA GAG AGA T-3'). Amplification of ISSRs was performed under the following reaction conditions: 2.0 ng [micro][L.sup.-1] template DNA, 1.0 [micro]M primer, 200 [micro]M of each dNTP, 1.6 mM Mg[Cl.sub.2], 0.04 units [micro][L.sup.-1] Taq polymerase (Life Technologies), 20 mM Tris-HCl (pH 8.4), and 50 mM KCl. The thermal profile for ISSR PCR was: 94[degrees]C initial denaturation for 5.0 min; then 35 cycles of 94[degrees]C for 45 s, 55[degrees]C for 60 s, and 72[degrees]C for 2.0 min; followed by a final extension at 72[degrees]C for 7.0 min. ISSR amplification products were separated by electrophoresis in nondenaturing 5% (w/v) 29:1 polyacrylamide:bis 1 x TBE gels containing 3 M urea, run in 1 x TBE buffer.

Glutathione reductase (GR) STS gene amplification utilized primers based on the GR sequence of Pisum sativum L. (Creissen et al., 1992). Amplification was performed using the method of Steiner et al. (1995) using a forward primer of (5'-GAC AGT GAA CCT AAG CAT GAC TGG-3') and reverse primer of (5'-GGT GAG CGT CCA GTG GCA AAC AT-3'). The GR-STS amplification products were both undigested as well as digested for 3 h with 0.2 [micro]L (2 units) of DpnII or HaeIII (New England Biolabs), 3.8 [micro]L reaction mixture, 5.4 [micro]L water, and 0.6 [micro]L 10 x restriction endonuclease buffer, run on 2.5% (w/v) agarose 1 x TBE gels.

Images of electrophoresed DNA amplification products were obtained by flatbed scanning of instant film (Type 667 Polaroid, Sigma, Ronkonkoma, NY) exposures of UV-fluoresced ethidium bromide stained agarose gels or by digital capture of GelStar (BioWhittaker Molecular Applications, Rockland, ME) stained polyacrylamide gels.

Starch gel electrophoresis assays of isozyme loci from fresh leaf material were performed using the protocols of Raelson et al. (1989) using their buffer system D for analysis of malic enzyme (Me), phosphogluconate dehydrogenase (6pgd), and phosphoglucomutase (Pgm), and Buffer System I for glucose-6-phosphate isomerase (Pgi) and triose-phosphate isomerase (Tpi).

Linkage Analysis

All molecular markers were analyzed for [chi square] goodness of fit to 3:1 (presence:absence of marker) ratios as single-dose dominant markers (Wu et al., 1992). In addition, suitable RFLP alleles were codominantly scored for genotype based on signal intensity and analyzed for [chi square] goodness of fit to 1:2:1 segregation ratios having two or more doses:one dose:zero doses, respectively. Single loci with two or more codominantly scored RFLP alleles that displayed significant difference from 1:2:1 segregation, but not significant from 3:1 segregation were subsequently treated as dominantly scored loci. Single-dose markers were entered as being dominant or codominant and analyzed for linkage using Map Manager XP (v. 0.7, http:// mapmgr.roswellpark.org) and MapMaker (v. 3.0b; Lander et al., 1987) software. All markers were analyzed as maternal type, with codominant alleles labeled as homozygous maternal if two doses were present, as heterozygous if one dose was present, and as homozygous paternal if no copies were present.

Duplex markers were identified as those fitting a codominantly scored 27:8:1 segregation ratio. Linkage distances involving codominantly scored duplex markers were calculated using maximum likelihood estimates of recombination, as developed by Mather (1935a). Equations were first derived for calculating the expected proportion of progeny (m) displaying a particular combination of marker genotypes (for codominant markers) or phenotypes (for dominant markers) at two linked markers as a function of the recombination fraction (p) between markers in an [F.sub.2] segregating population (see Tables A1-A3 in Appendix), similar to the equations derived by De Winton and Haldane (1931) for dominant characters. Given the observed number of offspring (a) for each marker combination (t), where n = [a.sub.1] + [a.sub.2] + [a.sub.3] + ..... + [a.sub.t], the likelihood function of obtaining p can be expressed as:

[1] [THETA] = [n!/([a.sub.1]![a.sub.2]![a.sub.3]! ... [a.sub.t]!)] [([m.sub.1]).sup.a1][([m.sub.2]).sup.a2][([m.sub.3]).sup.a3] .... [([m.sub.t]).sup.at].

This likelihood expression can be transformed into logarithmic form to give:

[2] L = ln[n!/([a.sub.1]![a.sub.2]![a.sub.3]! ... [a.sub.t]!)] + [a.sub.1]ln([m.sub.1]) + [a.sub.2]ln([m.sub.2]) + .... + [a.sub.t]ln([m.sub.t]).

The maximum value for [THETA] can then be obtained by differentiation of the logarithm of probabilities (L) and finding the value of p that equates the result to zero using the equation:

[3] dL/dp = [a.sub.1] [dln[m.sub.1]/dp] + [a.sub.2] [dln[m.sub.2]/dp] + ....... + [a.sub.t] [dln[m.sub.t]/dp] = 0.

LOD scores for the estimated recombination fraction (p) can be calculated by the formula:

[4] LOD = [log.sub.10]([THETA]|p = p) - [log.sub.10]([THETA]|p = 0.5).

The log-likelihood equations used in this report for estimating recombination between codominantly scored duplex markers and dominantly or codominantly scored simplex markers or between two codominantly scored duplex markers, all in coupling phase in [F.sub.2] populations, are provided in Appendix Tables A1 to A3, respectively. Estimates of recombination using these equations were calculated by substituting integer count data of the number of progeny with a given genotype value ([a.sub.t]) into a Microsoft Excel (v. 8.0, Microsoft, Redmond, WA) spreadsheet, multiplying each cell by its appropriate partial derivative (d ln m/dp) term as a function of the estimate of recombination (p), and using iterative trials for p (requiring [approximately equal to] 10 s per trial) to equate the summation of all products to zero. An example of such a spreadsheet used to estimate the recombination fraction between a codominant duplex allele and a dominant simplex allele in coupling in an [F.sub.2] population is available at the internet site: http://usda-ars-beaumont. tamu.edu/biotech.html. Recombination estimates were made to the fourth decimal place to report recombination fractions to the level of 0.1 cM. Maximum likelihood equations for estimating recombination between codominantly or dominantly scored duplex markers in mixed or repulsion phase in [F.sub.1] or [F.sub.2] populations, as well as recombination in [F.sub.2] populations between all four homologs in tetrasomic inheritance are of considerable length and have not been presented, but are available at the above cited internet site. These equations are based on chromosomal-type tetrasomic inheritance of markers, as this mode of inheritance predominates over chromatid-type tetrasomic inheritance in L. corniculatus (Fjellstrom et al., 2001).

Linkage groups were formed using a minimum LOD score criterion of 4.0, and reassessed using a LOD score criteria of 3.0 and 2.5. The four homologous linkage groups of each chromosome were identified by shared RFLP loci and the presence of repulsion linkage calculated from tetrasomic inheritance expectations. Repulsion linkage analysis was performed after assigning interpreted RFLP genotypes to the [F.sub.2] progeny to determine if the presence of one parental allele was linked to the absence of any other alternative parental allele. For this analysis, conversely-scored parental RFLP alleles were scored as a parental homozygote if no copies were present, a heterozygote if one copy was present, and nonparental homozygote if two or more copies were present. Repulsion linkage distances were then measured as the Kosambi centimorgan map distance between conversely-scored parental RFLP alleles and markers that were normally (nonconversely) scored. Homologous linkage groups were combined into a composite linkage group by averaging the distances between markers shared among homologous chromosomes. Maps were generated using the Kosambi distance function.

RESULTS

Parental Polymorphism

The PstI genomic library provided numerous low copy number RFLP probes. In surveying 86 cloned L. corniculatus genomic fragments for copy number when hybridized to radiolabeled total genomic DNA, 82 fragments displayed low autoradiographic signal intensity. Southern blot analysis of parental and [F.sub.1] DNA screened with radiolabeled L. corniculatus genomic fragments revealed four additional medium-to-high copy number sequences, indicating that [approximately equal to] 90.7% of the library was comprised of low copy number sequences. Of the 50 RFLP probes tested for parental polymorphism that gave discreet phenotypic interpretations, 48 probes were polymorphic for at least one of four restriction enzymes tested, with EcoRI and EcoRV giving a higher frequency of polymorphism than XbaI or HindIII (data not shown). Progeny blots were subsequently made with only EcoRI- and EcoRV-digested DNA.

More than 114 RAPD primers from Operon primer sets A, B, C, D, F, H, and M were tested for parental and [F.sub.1] DNA polymorphism. Of the 96 primers that successfully amplified DNA, 325 polymorphic bands were detected, resulting in a mean of 3.4 polymorphic bands per primer. The three ISSR primers identified 58 polymorphic bands between the two parental lines (19.3 polymorphic bands per primer). Five enzyme systems were scored for isozyme polymorphism, and three displayed polymorphism (Me, 6pgd, and Pgi). The GRSTS was polymorphic when undigested and after digestion with DpnII or HaeIII restriction enzymes.

Marker Dosage and Segregation

Twenty-three RFLP probes were selected for segregation analysis based on their potential to identify 3 or 4 alleles at a distinct locus in the [F.sub.1] hybrid parent. Fifteen RAPD primers were selected for segregation analysis based on their ability to produce several distinct polymorphic markers. The RFLP probes, RAPD primers, and three ISSR primers used in scoring [F.sub.2] progeny segregation revealed 101, 103, and 43 segregating marker bands, respectively. Figure 1 displays a composite gel image of dominantly scored ISSR markers. One polymorphic marker band was displayed for each of the undigested, DpnII digested, and HaeIII digested GRSTS as well as the Me, 6pgd, and Pgi isozyme scored gels.

[FIGURE 1 OMITTED]

Of the 246 marker bands analyzed for goodness of fit to a single dominant marker, a total of 85 bands gave segregation ratios significantly different (P < 0.05) from 3:1 (presence:absence) expectations. However, 29 of these bands were most likely present in two doses in the [F.sub.1] parent giving rise to the [F.sub.2] population, being either in the duplex alleles at the same locus or two identical single-dose alleles at duplicated unlinked loci. Therefore, 56 of the 217 single-dose marker bands (25.8%) were skewed from the expected 3:1 segregation ratios. Ten (17.9%) of these skewed loci arose from the Moroccan accession used as the male parent and 46 skewed loci (82.1%) came from the autogamous accession used as the female parent of the [F.sub.1] plant that produced the [F.sub.2] population (Tables 1 and 2).

On the basis of autoradiographic signal intensity (e.g., Fig. 2 and 3), 63 RFLP bands from 17 RFLP probes were also tested for goodness of fit to 1:2:1 segregation ratios when scored as codominant markers. Twenty-nine of these markers gave segregation ratios significantly different than those expected for single-dose co-dominantly scored markers, with five of these most likely being present in two doses in the [F.sub.1] plant. Thus, 24 of the 58 (41.4%) single-dose RFLP markers had segregation ratios significantly different from values expected for codominantly scored single-dose alleles. Sixteen of these single-dose alleles had segregation ratios significantly different from 3:1 when scored as dominant (3:1) alleles, while 14 additional alleles appeared to segregate significantly different from 1:2:1 when scored codominantly due to a high proportion of heterozygotes identified by codominant scoring (Table 2). Five of the 24 skewed codominantly scored alleles (20.8%) arose from the Moroccan parent, while the majority (19 of 24, 79.2%) of the skewed alleles arose from the autogamous parent (Table 2).

[FIGURES 2-3 OMITTED]

On the basis of linkage analysis and [chi square] goodness of fit ratios, the five double-dose codominantly scored markers appeared to be duplex alleles at single loci (Table 3). One marker at first presumed to be double-dose (2023b) did not fit inheritance patterns expected for one duplex or two simplex marker alleles (Table 2), suggesting unusual segregation distortion in the inheritance of this marker.

Double reduction was observed for three RFLP markers, 309d, 3037d, and 30117a. Double reduction is observed in chromatid-type segregation after a crossover between an allele and the centromere, nondisjunction of sister alleles during the first meiotic division, and passage of both sister alleles to the same gamete in the second meiotic division. Double reduction at these loci was indicated by the presence of three copies of one allele in [F.sub.2] progeny when only one allelic copy was present in the [F.sub.1] parent.

Linkage Analysis

Linkage analysis of 51 RFLP markers at 14 loci scored codominantly and 166 isozyme, PCR, and RFLP markers scored as single-dose dominant alleles was performed using Map Manager and MapMaker. Although isozymes can be scored as codominant markers, poor band resolution and complex banding patterns allowed only dominant scoring of the three isozyme markers analyzed. The RFLP markers that had more than one allele significantly departing from 1:2:1 ratios in comparison with 3:1 ratios (alleles identified by RFLP probes 2041 and 2075), and those that were not reliably scored for allele dosage (alleles identified by RFLP probes 204, 206, 2012, 2021, 308, 309B, 3079, and 30107), were analyzed as dominant alleles. Both codominantly and dominantly scored alleles identified by RFLP probe 2023 were analyzed to aid in its map placement. The majority of duplex dominant PCR markers were not mapped (excluding possible duplex markers noted below), due to low informativeness on map position.

When a LOD value of 4.0 was used as a criterion for linkage between the 217 markers, 32 linkage groups were identified, having from two to 18 markers and spanning from 1.4 cM on Linkage Group 7 up to 110.2 cM on Linkage Group 5-C/6-C (all references to linkage groups, loci, and map distances are depicted in Fig. 4). Homologous linkage groups were identified and anchored by the presence of shared RFLP loci. Since most of the RFLP probes analyzed were preselected to identify four RFLP alleles (Fjellstrom et al., 2001), these tetra-allelic RFLP markers allowed a somewhat straight-forward assignment of homologous linkage groups. Dropping the LOD score criterion to 3.0 joined four of these linkage groups, while dropping the LOD criterion to 2.5 allowed placement of several anchoring loci that were not linked at higher criteria and resulted in the formation of 26 linkage groups.

[FIGURE 4 OMITTED]

Several steps were taken to map the duplex alleles of anchoring RFLP loci. Instead of calculating recombination values with every simplex marker, only a subset of markers were mapped relative to these duplex markers. The subset used were those mapping to regions where the duplex markers were expected to be based on the location of simplex alleles from anchoring RFLP on homologous linkage groups. Since dominantly-scored duplex markers were poorly placed on genetic maps, codominantly-scored duplex alleles were evaluated for linkage with markers expected to be near them. Duplex RFLP markers were coded for the presence of 0 to 4 copies of the allele when scored codominantly. Equations developed to provide maximum likelihood estimations of recombination between codominantly-scored duplex loci and dominantly and codominantly scored loci in coupling (see Appendix) were used to map and calculate LOD scores for duplex markers.

Composite Linkage Group Assignment

Six composite linkage groups were developed from the aligned homologous linkage groups, using anchoring RFLP loci to orient dominantly scored simplex marker loci. Composite linkage groups were numbered 1 to 6 roughly in order of size, going from longest (118.1 cM) to shortest (72.7 cM). Distances between anchoring loci on composite linkage groups were averaged, and positions of simplex markers were adjusted according to their proportional distance between anchoring loci. This adjustment could not be done for markers found outside anchoring loci, so their map distances from anchoring loci were not recalculated. One linkage group comprised of only two markers, was not associated with markers found on anchored linkage groups and was separately designated as Linkage Group 7. Out of the 217 single-dose markers analyzed, 15 markers were unlinked to any other marker.

Composite Groups 2, 3, and 4 were well defined and the easiest to delineate. Homologous linkage groups of Groups 3 and 4 were readily aligned since the relative order of RFLP loci was retained for each of the homologous linkage groups. The homologous linkage groups of Group 2 were more difficult to align, since one linkage group (2-D) carried only one of the four shared RFLP loci (locus 308). An allele for RFLP locus 2012 on Group 2-D could not be scored and was most likely a null allele, as suggested by the presence of numerous progeny with only one scorable allele for this probe and the presence of only one allele (that segregated 3:1 in the [F.sub.2] progeny) in the autogamous seed parent. Furthermore, RFLP locus 30107C had markers of low autoradiographic intensity, with only two alleles being reliably scored.

Apparent allele duplication made the mapping and division of Groups 5 and 6 problematic. One and four alleles of RFLP markers 204 and 30107, respectively, had two copies of alleles that were nearly identical in size. Allele intensity was used to assign the presence or absence of alleles 30107b and 30107c. The duplicated alleles of 30107b and 30107c were of low autoradiographic intensity and could not be reliably scored. Allele 30107d was readily scored for presence or absence, and was treated as a single-dose marker in linkage analysis, even though it segregated closer to expectations for two loci rather than a single locus (Table 1). Alleles 204b and 204d shared an identical RFLP band (segregating 82:0), but 204d also had a weak intensity, cosegregating band that allowed scoring of its presence or absence. The dominantly-scored presence or absence of markers 204b and 30107a (both having bands segregating 82:0 when scored by presence:absence) were less accurately assigned by signal intensity and taking into account the presence of alternative alleles. Two linkage groups (5-B and 5-D) appeared to be imperfectly divided between RFLP loci 309 and 30107, with loose linkages (average = 25.8 cM, LOD = 2.7) seen between the linkage groups. However, alleles 30% and 30107b were linked on Group 5-C. Instead of combining linkage groups together for the two loosely linked homologs, it was decided to divide Groups 5-C and 6-C at locus 30107. As a result, RFLP locus 30107 was mapped onto the terminal ends of both Groups 5 and 6 and listed as locus 30107A/B, without discrimination between the two loci 30107A and 30107B.

Composite Linkage Group 1 was difficult to align because of loose linkages, skewed segregation ratios, and possible rearrangements between homologous linkage groups. Although alleles for RFLP loci identified by probes 202, 2043, and 3117B remained linked in each group, markers identified by probes 2019, 2023, 306, and 3037 showed inconsistent linkage. The RFLP markers for 2019 and 306 were unlinked on Group 1-A, even though they were closely linked on the other three homologs. Markers 202 and 2019 were closely linked on Groups 1-A and 1-D, only loosely linked on Group 1-C, and unlinked on Group 1-D. Markers 2023 and 3037 were closely linked on Group 1-D, loosely linked on Groups 1-A and 1-C, and unlinked on Group 1-B. Markers 306 and 2023 were not linked on 1-A and only loosely linked on the other three homologs. Except for markers 3037c, 306d, and GAT-520, the markers on Group 1-C displayed high segregation distortion with few band absent genotypes or phenotypes. The only two homologs that had all anchoring loci in linkage at LOD = 2.5 or higher were Groups 1-C and 1-D, and these groups differed in their gene order. Whereas locus 30117 was terminal and locus 306 was internal for group 1-C, these loci were flipped around such that the former became internal and the latter became terminal in Group 1-D. This order could not be resolved by groups 1-A and 1-B, since these loci were unlinked in those groups. Maximum likelihood estimations of simultaneous recombination between all four homologs in tetrasomic inheritance (available at the internet site http:// usda-ars-beaumont.tamu.edu/biotech.html) supported the gene order displayed by Group 1-C, which was used to order the composite map of Group 1. Linkage Groups 1-A and 1-B are both combinations of two unlinked groups, each displayed with the likely order of loci on these co-aligned groups.

The six composite linkage groups constructed from these marker data ranged from 72.7 cM to 118.1 cM in length, spanning a total of 572.1 cM. While Groups 1 to 4 were readily discriminated from each other, Groups 5 and 6 were not easily separated. With both groups including Loci 30107A/B and opa10-460, Group 5 spans 96.3 cM and Group 6 spans 72.7 cM. If these groups were combined, they would span 165.8 cM, becoming the largest linkage group and result in the formation of only five linkage groups. For reasons given in the following discussion section, the division of Groups 5 and 6 as presented was considered the most likely genetic map for this species. A total of 143 loci were mapped from the 217 marker bands scored. The number of loci per composite linkage group ranged from 15 loci for Group 5 to 39 loci for Group 1. Although shown as independent loci on the composite maps, it is likely that ISSR markers from the same primer, of similar molecular weight, and located near each other on composite maps were actually alleles from the same locus. Gel images suggest this is true for GAT-520 and GAT-540 on Group 1, GAT-1180 and GAT-1260 on Group 2, and GAT-880 and GAT-920 on Group 6.

DISCUSSION

Ample polymorphism was available for linkage analysis of the DNA markers surveyed. The isozyme markers had rather limited polymorphism along with complex banding patterns that made codominant scoring of bands difficult, making these protein markers less desirable for genetic mapping. The use of RFLP markers allowed the efficient identification and anchoring of homologous linkage groups. It would have been extremely difficult to assign dominantly scored PCR markers to homologous linkage groups, particularly in a tetrasomic species, without the use of anchoring RFLP loci.

Scoring duplex markers codominantly allowed the mapping of these markers, which could not have been done if the markers were scored dominantly for presence or absence. Not surprisingly, all the codominantly scored duplex markers mapped to similar positions where their homologous simplex alleles were mapped. Estimating recombination between codominantly scored RFLP loci of all four homologous linkage groups simultaneously also allowed the gene order of composite Linkage Group 1 to be more confidently determined. Scoring genotype x band intensity can lead to false results, since subjectivity is used to assign how many doses of each band are present. False results have been avoided in this research by scoring genotypes with those loci that clearly have four allelic doses in the [F.sub.1] parent, scoring when a loss of one allele dose can be obviously correlated with the addition of another allele dose in the [F.sub.2] progeny, and omitting progeny genotype data when allelic dosage was not readily identified. It can be noted that barely half of the RFLP markers scored (51 of 101) in this research fulfilled these criteria well enough to be scored codominantly. Greater mapping accuracy will result when markers can be scored codominantly, particularly when determining recombination values with duplex markers. Because there has been a lack of simplified formulas to estimate recombination involving codominantly scored duplex markers, it is hoped that the presentation of these formulas by this research will help others working with marker inheritance in polysomic species.

The molecular marker analysis reported in this research has allowed the construction of the first genetic linkage map in L. corniculatus. With 139 loci providing a genome coverage of 572.1 cM, this map is fairly comparable to maps constructed for alfalfa, a related 2n = 4x = 32 forage legume. Genetic maps in diploid alfalfa have spanned from 234 to 659 cM (Brummer et al., 1993; Kiss et al., 1993; Echt et al., 1994; Tavoletti et al., 1996; Barcaccia et al., 1999; Kalo et al., 2000) and the only map of tetraploid alfalfa (Brouwer and Osborn, 1999) spans 443 cM. Genetic maps for L. japonicus (Regel) K. Larsen (= L. corniculatus var. japonicus Regel) are actively being developed, with this species being internationally studied as a model legume (Handberg and Stougaard, 1992; Stougaard, 2001). Preliminary results indicate that this diploid Lotus species has a genome size similar to that found in this study of tetraploid L. corniculatus (Hayashi et al., 2001; P.M. Gresshoff, 2002, personal communication).

The 25.8% and 41.4% segregation distortion seen among the dominantly and codominantly scored simplex markers, respectively, falls well within the range of 9 to 58% distortion seen in diploid alfalfa mapping studies (Kiss et al., 1993; Tavoletti et al., 1996). In this study, segregation distortion was higher for loci scored codominantly than dominantly, predominantly because of a unexpectedly high number of heterozygotes being identified after codominant scoring. Unusually high proportions of heterozygotes also accounted for most of the segregation distortion seen in alfalfa genetic studies, particularly in [F.sub.2] progeny populations of selfed [F.sub.1] hybrids (Brummer et al., 1993; Kiss et al., 1993). As in alfalfa mapping studies, the most likely reason for segregation distortion resulting from high heterozygosity would be the detrimental effects of inbreeding. Lotus corniculatus is predominantly an outcrossed species, with noticeable inbreeding depression seen after selfing that reduces heterozygosity and increases the proportion of potentially lethal or detrimental alleles. Although one of the parents was autogamous, molecular marker evidence suggested this autogamous line retained heterozygosity by genome differentiation that caused preferential pairing (Fjellstrom et al., 2001). This resulted in fixed heterozygosity (Soltis and Riesberg, 1986) within its gametes, similar to that seen in disomic polyploids such as wheat. With most of the markers having segregation distortion arising from the autogamous seed parent, it appeared that heterozygosity was particularly enforced in gametes or offspring of that line. Regardless, segregation distortion does not preclude genetic mapping, although it can affect map distance.

Double reduction was observed at three RFLP loci, 309A, 3037, and 30117A, found on Linkage Groups 5, 1, and 5, respectively. Double reduction is expected to be more common near telomeres and not common near centromeres (Mather, 1935b), which appears to agree with our map placement of these loci near the end of their respective linkage groups.

For a species with tetrasomic inheritance, four homologous linkage groups of each chromosome are expected, giving an expectation of L. corniculatus having a total of 24 linkage groups from its six chromosomes. Use of a LOD criterion of 3.0 identified 28 linkage groups, but did not join anchoring RFLP alleles expected to be on the same linkage group. When the LOD criterion was dropped to 2.5, these anchoring RFLPs were joined onto their expected groups, but this also caused the joining of two homologs of Groups 5 and 6. Since six composite groups were expected and the data indicated the presence of five distinct groups, it was thought that either Groups 5 and 6 or Group 1, being the longest continuous groups identified, should be divided.

Groups 5 and 6 were divided based on two points of evidence. With a LOD criterion of 3.0 or greater, only one homolog had Groups 5 and 6 joined (5-C and 6-C), while two homologs of Linkage Group 1 (1-C and 1-D) were joined. At a LOD criterion of 2.5, two homologs of Linkage Groups 5 and 6 became joined at locus 30107A/B, which apparently was a duplicated locus. It was concluded that Group 5-D was joined to Group 6-D, respectively, not through linkage but by association of identical heterologous alleles of the duplicated 30107A/B locus. As noted above, locus 309A also displayed one of the few instances of double reduction seen in this study. For double reduction to occur, it seems likely that locus 309A would be located near the end of a chromosome, which would be more likely if Groups S and 6 were divided, but less likely if these groups were joined to place locus 309A near the center of this linkage group. It is not apparent why Groups 5-C and 6-C were linked at a 3.0 LOD threshold while their homologous groups were not. Perhaps there was selection pressure for these linkage groups to segregate together to the same gamete, which could cause them to appear genetically linked. There is still the possibility that Linkage Groups 5 and 6 are on the same chromosome and that Linkage Group 1 should be divided, which more mapping research should be able to support or reject.

Group-1 was the longest composite linkage group and displayed several interesting features. This composite group had the greatest number of markers and was the only linkage group that showed different orders among the anchoring loci. Certain marker alleles (306b and 3037d) expected to be in Group 1 were unlinked to any markers, even though their homologous alleles displayed strong linkages with other Group 1 markers. With the large number of markers, differing orders, and lost linkages in this group, it appears this composite group is the most differentiated in the L. corniculatus genome.

The genetic map constructed for L. corniculatus provided a beneficial tool for genetic research and plant improvement. Genetic maps can serve as a foundation for the analysis and dissection of important quantitative traits in trefoil such as root- and crown-rotting disease resistance, winter hardiness, and seedling vigor. Genetic maps can aid in the development of markers for marker aided selection of quantitative traits as well as markers for useful, simply inherited traits like rhizome production (Nualsri and Beuselinck, 1998). This genetic map has also provided insight into the nature of L. corniculatus genome organization. We have shown that the most polymorphic linkage group (Group 1) appeared to have structural divergence not seen among the other linkage groups, and that gene duplications were common throughout the L. corniculatus genome and can confound mapping efforts. Future efforts in L. corniculatus mapping should include the incorporation of SSR markers. Although sequence information is required for SSR marker development, such information is becoming available from extensive sequencing being completed in L. japonicus. Because RAPD markers can have poor reproducibility among different laboratories, SSR markers, RAPDs converted into sequence-characterized amplified region markers (Paran and Michelmore, 1993), or other sequence-specific PCR markers should be developed. These markers are readily reproduced in independent laboratories. Using sequence-specific PCR markers mapped to both species will also help extend functional genomics work developed in L. japonicus to future genetic analysis in L. corniculatus as well as other members of the L. corniculatus complex (e.g., L. glaber Mill. and L. uliginosus Schkuhr).

Abbreviations: cM, centimorgan: GR, glutathione reductase; ISSR, inter-simple sequence repeat; PCR, polymerase chain reaction: RAPD, randomly amplified polymorphic DNA; RFLP, restriction fragment length polymorphism; SSR, simple sequence repeat; STS, sequence tagged site; TBE, tris-borate.

APPENDIX

See Tables A1, A2, and A3.
Table A1. Expected genotype frequencies based on the recombination
fraction (p) between a duplex codominant marker (X)
linked to a simplex dominant marker (Y) in coupling and scores
(d ln m/dp) used for maximum likelihood estimation of recombination
fraction based on the observed number of progeny
with a particular genotype combination.

                       Expected progeny
Progeny genotype       proportions (m)            d ln m/dp

XXXX Y-yy          (1 - [p.sup.2])/36         -2p/(1 - [p.sup.2])
XXXX yyyy          [p.sup.2]/36                2/p
XXXx Y-yy          (2 - p)/9                  -1/12 - p)
XXXx yyyy          p/9                         1/p
XXxx Y-yy          (7 - p + [p.sup.2])/18      (2p - 1)/(7 - p +
                                                 [p.sup.2])
XXxx yyyy          (2 - p)(1 + p)/18           (1 - 2p)/(2 + p -
                                                 [p.sup.2])
Xxxx Y-yy          (1 + p)/9                   1/(1 + p)
Xxxx yyyy          (1 - p)/9                  -1/(1 - p)
xxxx Y-yy          [[1 - (1 - p).sup.2]]/36    (2 - 2p)/(2p -
                                                 [p.sup.2])
xxxx yyyy          [(1 - p).sup.2]/36         -2/(1 - p)

Table A2. Expected genotype frequencies based on the recombination
fraction (p) between a duplex codominant marker (X)
linked to a simplex dominant marker (Y) in coupling and scores
(d ln m/dp) used for maximum likelihood estimation of recomnation
fraction based on the observed number of progeny
with a particular genotype combination.

Progeny       Expected progeny
genotype      proportions (m)                d ln m/dp

XXXX YYyy   [(1 - p).sup.2]/36    -2/(1 - p)
XXXX Yyyy   p(1 - p)/18            (1 - 2p)/(p - [p.sup.2])
XXXX yyyy   [p.sup.2]/36           2/p
XXXx YYyy   (1 - p)/9             -1/(1 - p)
XXXx Yyyy   1/9                    0
XXXx yyyy   p/9                    1/p
XXxx YYyy   [2 + p (1 - p)]/18     (1 - 2p)/(2 + p - [p.sup.2])
XXxx Yyyy   [4 + [(1 -p).sup.2]    (4p - 2)/(5 - 2p + 2[p.sup.2])
              + [p.sup.2]/18
XXxx yyyy   [2 + p (1 - p)/18      (1 - 2p)1(2 + p - [p.sup.2])
Xxxx YYyy   p/9                    1/p
Xxxx Yyyy   1/9                    0
Xxxx yyyy   (1 - p)/9             -1/(1 - p)
xxxx YYyy   [p.sup.2]/36           2/p
xxxx Yyyy   p(1 - p)/18            (1 - 2p)/(p - [p.sup.2])
xxxx yyyy   [(1 - p).sup.2]/36     -2/0 - p)

Table A3. Expected genotype frequencies based on the recombination
fraction (p) between a duplex codominant marker (X) linked to a duplex
codominant marker (Y) in coupling and scores (d ln m/dp) used for
maximum likelihood estimation of recombination fraction based on the
observed number of progeny with a particular genotype combination.

Progeny          Expected progeny
genotype         proportions (m)                   d ln m/dp

XXXX YYYY   [(1 - p).sup.4]/36           -4/(1 - p)
XXXX YYYy   p[(1 - p).sup.3]/9            (1 - 4p)l(p - [p.sup.2])
XXXX YYyy   [p.sup.2][(1 - p).sup.2]/6    (2 - 4p)/(p - [p.sup.2])
XXXX Yyyy   [p.sup.3](1 - p)/9            (3 - 4p)/(p - [p.sup.2])
XXXX yyyy   [p.sup.4]/36                  4/p
XXXx YYYY   p[(1 - p).sup.3]/9            (1 - 4p)/(p - [p.sup.2])
XXXx YYYy   2[(1 - p).sup.2](1 - p + 2   -1(3 - 7p + 8[p.sup.2])/[(1
              [p.sup.2])/9                  - p)(1 - p + 2[p.sup.2])]
XXXx YYyy   p(1 - p)(5 - 6p +             (5 - 22p + 36[p.sup.2] - 24
              6[p.sup.2])/9                 [p.sup.3])/[(p - [p.sup.2])
                                            (5 - 6p + 6[p.sup.2)]
XXXx Yyyy   2[p.sup.2](2 - 3p +           (4 - 9p + 8[p.sup.2])/(2p - 3
              2[p.sup.2])/9                 [p.sup.2] + 2[p.sup.3])
XXXx yyyy   [p.sup.3](1 - p)/9            (3 - 4p)/(p - [p.sup.2])
XXxx YYYY   [p.sup.2][(1 - p).sup.2]/6    (2 - 4p)/(p - [p.sup.2])
XXxx YYYy   p(1 - p)(5 - 6p +             (5 - 22p + 36[p.sup.2] - 24
              6[p.sup.2])/9                 [p.sup.3])/[(p - [p.sup.2])
                                            (5 - 6p +  6[p.sup.2])]
XXxx YYyy   (9 - 20p + 38[p.sup.2] -      4(-5 + 19p - 27[p.sup.2] +
              36[p.sup.3] + 18              18/[p.sup.3])/(9 - 20p + 38
              [p.sup.4])/18                 [p.sup.3] - 36[p.sup.3] +
                                            18[p.sup.4])
XXxx Yyyy   p(1 - p)(5 - 6p +             (5 - 22p + 36[p.sup.2] - 24
              6[p.sup.2])/9                 [p.sup.3])/[(p - [p.sup.2])
                                            (5 - 6p + 6[p.sup.2])]
XXxx yyyy   [p.sup.2][(1 - p).sup.2]/6    (2 - 4p)/(p - [p.sup.2])
Xxxx YYYY   [p.sup.3](1 - p)/9            (3 - 4p)/(p - [p.sup.2])
Xxxx YYYy   2[p.sup.2](2 - 3p +           (4 - 9p + 8[p.sup.2])/(2p -
              2[p.sup.2])/9                 3[p.sup.2] + 2[p.sup.3])
Xxxx YYyy   p(1 - p)(5 - 6p +             (5 - 22p + 36[p.sup.2] - 24
              6[p.sup.2])/9                 [p.sup.3])/[(p - [p.sup.2])
                                            (5 - 6p +  6[p.sup.2])]
Xxxx Yyyy   2[(1 - p).sup.2](1 - p +     -(3 - 7p + 8[p.sup.2])/[(1 -
              2[p.sup.2])/9                 p)(1 - p + 2[p.sup.2])]
Xxxx yyyy   p[(1 - p).sup.3]/9            (1 - 4p)/(p - [p.sup.2])
xxxx YYYY   [p.sup.4]/36                  4/p
xxxx YYYy   [p.sup.3](1 - p)/9            (3 - 4p)/(p - [p.sup.2])
xxxx YYyy   [p.sup.2][(1 - p).sup.2]/6    (2 - 4p)/(p - [p.sup.2])
xxxx Yyyy   p[(1 - p).sup.3]/9            (1 - 4p)/(p - [p.sup.2])
xxxx yyyy   [(1 - p).sup.4]/36           -4/(1 - p)

Table 1. Segregation counts of Lotus corniculatus markers having
segregation ratios significantly deviating from expectations,
mapped as dominantly scored single-dose markers.

                 Segregation

                                                Donor
                                     [chi       parent
                                   square]     ([double   Linkage
Marker       Presence   Absence   ([dagger])   dagger])   group(s)

GAT-520         70        11        5.63 *        A       1-C
opA5-480        71         9        8.07 **       A       1-C
CTG-1200        72         9        8.33 **       A       1-C
GAA-1500        72         8        9.60 **       A       1-C
3079a           73         8        9.88 **       A       1-C
opC10-1450      75         7       11.85 **       A       1-C
opC9-280        75         7       11.85 **       A       1-C
GAA-520         76         6       13.67 **       A       1-C
opA5-1000       76         5       15.31 **       A       1-C
opC16-720       76         5       15.31 **       A       1-C
opA6-780        75         7       11.85 **       A       1-D
2012c           49        33       10.16 **       M       2-A
GAT-1260        69        12        4.48 *        M       2-B
308b            70         9        7.80 **       A       2-D
opA5-220        72         9        8.33 **       A       2-D
opA9-300        73         9        8.60 **       A       2-D
opA19-220       74         8       10.16 **       A       2-D
opC10-790       74         8       10.16 **       A       2-D
opM3-650        77         5       15.63 **       A       2-D
CTG-1250        42        39       23.15 **       M       3-A
GAA-350         42        39       23.15 **       M       3-A
opM12-550       74         8       10.16 **       A       3-C
opC10-180       62         6        9.49 **       A       3-D
opa19-1010      74         8       10.16 **       A       3-D
opA5-270        49        31        8.07 **       M       4-A
GAT-430         69        12        4.48 *        A       4-D
opC7-350        68        11        5.17 *        A       4-D
30107a          76         6       13.67 **       M       5-A-6-A
GAT-940         70        11        5.63 *        A       5-C
CTG-700         71        11        5.87 *        A       5-C
30107d          73         9        8.60 **       A       5-D-6-D
opa10-460       73         9        8.60 **       A       5-D-6-D
opC4-380        71         8        9.32 **       A       6-C
204b            73         9        8.60 **       A       6-D
opC10-430       70        12        4.70 *        A       6-D
GAT-880         70        11        5.63 *        A       6-D
opC7-530        70        10        6.67 **       A       6-D
30107d          73         9        8.60 **       A       6-D
204b            76         6       13.67 **       A       6-D
opC16-520       70        12        4.70 *        A       7
GAA-700         71        11        5.87 *        A       7
opM12-780       71        11        5.87 *        A       unlinked
opH11-550       71        10        6.92 **       M       unlinked

* Significant at the 0.05 level of probability.

** Significant at the 0.01 level of probability.

([dagger]) Expectations for dominantly scored markers segregating
3:1 for presence:absence, respectively (df = 1).

([double dagger]) A designates autogamous parent, M designates
Moroccan parent.

Table 2. Segregation counts of Lotus corniculatus restriction
fragment length polymorphism markers scored as single-dose
alleles significantly deviating from expected segregation
ratios when scored as codominant or dominant markers.

          Allele dose       [chi square]
          segregation        ([dagger])

                                                Donor
                                                parent
                                               ([double   Linkage
Marker   2    1    0      1:2:1       3:1      dagger])    group

2023d     4   54   24   18.00 **     0.80         M         1-A
2023c    14   54   14    8.24 **     2.75         M         1-B
3037c    18   52   12    6.78 *      4.70 *       A         1-C
306d     16   53   13    7.24 *      3.66         A         1-C
202c     27   47    8   10.56 **    10.16 **      A         1-C
2019a    28   46    8   10.98 **    10.16 **      A         1-C
2043c    27   48    6   13.67 **    13.37 **      A         1-C
2023b    32   50    0   28.93 **    27.33 **      A         1-C
202d     10   48   23    6.95 *      0.50         A         1-D
30117d    9   52   19    9.70 **     0.07         A         1-D
2023a     5   54   23   16.15 **     0.41         A         1-D
30100b    8   41   33   15.24 **    10.16 **      M         3-A
3026c     9   34   38   22.857 **   20.74 **      M         3-A
3058d    12   29   38   22.70 **    22.49 **      M         3-A
3058a    28   42    9    9.46 **     7.80 **      A         3-D
3026a    30   44    7   13.67 **    11.56 **      A         3-D
30100a   30   46    6   15.27 **    13.67 **      A         3-D
303c     26   43   12    5.15        4.48 *       A         4-C
2037d    20   48   12    4.80        4.27 *       A         4-D
2027b    15   54   13    8.34 **     3.66         A         4-D
303d     10   60   12   17.71 **     4.70 *       A         4-D
309a      8   57   16   15.02 **     1.19         A         5-C
2041b    10   57   15   13.10 **     1.97         A         6-C
2075a    10   58   14   14.49 **     2.75         A         6-C
2075c    17   53   12    7.63 *      4.70 *       A         6-D
2041a    14   57   11   12.71 **     5.87 *       A         6-D

* Significant at the 0.05 level of probability.

** Significant at the 0.01 level of probability.

([dagger]) Expectations for codominantly scored (segregating 1:2:1
for 2 doses:l dose:0 doses, respectively, df = 2) or dominantly
scored (3:1 for presence:absence, respectively, df = 1) markers.

([double dagger]) A designates autogamous parent, M designates
Moroccan parent.

Table 3. Segregation counts of Lotus corniculatus markers
considered to be duplex alleles.

                                  Tetrasomic
              Allele dose        [chi square]
              segregation         ([dagger])

             2 +
Marker   ([section])    1   0   27:8:1   35:1

2043a        63        16   2   0.33     0.03
2019b        63        19   0   2.35     2.34
30117e       71         9   1   6.92 *   0.71
2027c        63        17   2   0.15     0.03
30117c       60        22   0   3.10     2.34

              Disomic
            [chi square]
          (double dagger])

                             Donor
                             parent(s)       Linkage
Marker    11:4:1     15:1    ([paragraph])   groups

2043a     3.71      1.98           M          1-AB
2019b     6.01 *    5.47 *         M          1-AB
30117e   13.72 **   3.48          A/M         1-BC
2027c     3.28      2.03           M          4-AB
30117c    5.47 *    5.47 *         M          5-AB

* Significant at the 0.05 level of probability.

** Significant at the 0.01 level of probability.

([dagger]) Expectations in chromosomal-type tetrasomic inheritance
for codominantly (segregating 27:8:1 for 2+ doses:1 dose:0 doses,
respectively, df = 2) or dominantly (35:1 for presence:absence,
respectively, df = 1) scored markers.

([double dagger]) Expectations in disomic inheritance (indicating
simplex alleles at two independent loci) for codominantly (11:4:1
for 2+ doses:1 dose:0 doses, respectively, df = 2) or dominantly
(15:1 for presence:absence, respectively, df = 1) scored markers.

([section]) 2+ designates 2, 3, or 4 doses.

([paragraph]) A/M designates one allele originated from the
autogamous parent and one originated from the Moroccan parent; A
or M designates that both alleles originated from either the
autogamous or Moroccan parent, respectively.


ACKNOWLEDGMENTS

The authors thank P.M. Frank and C.J. Poklemba for technical assistance with this research and P.M. Frank for graphics preparation. The senior author thanks Dr. Guowei Wu for checking and providing maximum likelihood formula derivations, Dr. Peter M. Gresshoff for providing progress information on the international L. japonicus genome program, and Dr. Perry Cregan for valuable review comments. The mention of a trademark or a proprietary product does not constitute a guarantee or warranty of the product by USDA and does not imply its approval to the exclusion of other products may also be suitable.

(1) The use of trade names in this publication does not imply endorsement of the products named nor criticism of similar ones not mentioned.

REFERENCES

Al-Janabi, S.M., R.J. Honeycutt, M. McClelland, and B.W.S. Sobral. 1993. A genetic linkage map of Saccharum spontaneum L. 'SES 208'. Genetics 134:1249-1260.

Barcaccia, G., E. Albertini, S. Tavoletti, M. Falcinelli, and F. Veronesi. 1999. AFLP fingerprinting in Medicago spp.: Its development and application in linkage mapping. Plant Breed. 118:335-340.

Beuselinck, P.R., and W.F. Grant. 1995. Birdsfoot trefoil, p. 237-248. In R.F. Barnes et al. (ed.) Forages. Vol. I: An introduction to grassland agriculture. Iowa State Univ. Press, Ames, IA.

Beuselinck, P.R., B. Li, and J.J. Steiner. 1996. Rhizomatous Lotus corniculatus L.: Taxonomic and cytological study. Crop Sci. 36: 197-185.

Bonierbale, M.W., R.L. Plaisted, and S.D. Tanksley. 1988. RFLP maps based on a common set of clones reveal modes of chromosomal evolution in potato and tomato. Genetics 120:1095-1103.

Brouwer, D.J, and T.C. Osborn. 1999. A molecular marker linkage map of tetraploid alfalfa (Medicago sativa L.). Theor. Appl. Genet. 99:1194-1200.

Brummer, E.C., J.H. Bouton, and G. Kochert. 1993. Development of an RFLP map in diploid alfalfa. Theor. Appl. Genet. 86:329-332.

Bubar, J.S., and R.K. Miri. 1965. Inheritance of self-incompatibility and brown keel tip in Lotus corniculatus L. Nature (London) 205: 1035-1036.

Buzzell, R.I., and C.P. Wilsie. 1963. Genetic investigations of brown keel tip color in Lotus corniculatus L. Crop Sci. 3:128-130.

Creissen, G., E.A. Edwards, C. Enard, A. Wellburn, and P. Mullineaux. 1992. Molecular characterization of glutathione reductase cDNAs from pea (Pisum sativum L.). Plant J. 2:129-131.

Dalrymple, E.J., B.P. Goplen, and R.E. Howarth. 1984. Inheritance of tannins in birdsfoot trefoil. Crop Sci. 24:921-923.

Da Silva, J., M.E. Sorrells, W.L. Burnquist, and S.D. Tanksley. 1993. RFLP linkage map and genome analysis of Saccharum spontaneum. Genome 36:782-791.

Dawson, C.D.R. 1941. Tetrasomic inheritance in Lotus corniculatus L. J. Genet. 42:49-72.

De Winton, D., and J.B.S. Haldane. 1931. Linkage in the tetraploid Primula sinensis. J. Genet. 24:121-144.

Diwan, N., J.H. Bouton, G. Kochert, and P.B. Cregan. 2000. Mapping of simple sequence repeat (SSR) DNA markers in diploid and tetraploid alfalfa. Theor. Appl. Genet. 101:165-172.

Donovan, L.S. 1957. Inheritance studies in broadleaf birdsfoot trefoil, Lotus corniculatus L. Ph.D. diss. Univ. Nebraska, Lincoln, NE.

Echt, C.S., K.K. Kidwell, S.J. Knapp, T.C. Osborn, and T.J. McCoy. 1994. Linkage mapping in diploid alfalfa (Medicago sativa). Genome 37:61-71.

Fjellstrom, R.G., P.R. Beuselinck, and J.J. Steiner. 2001. RFLP marker analysis supports tetrasomic inheritance in Lotus corniculatus L. Theor. Appl. Genet. 102:718-725.

Freyre, R., S. Warnke, B. Sosinski, and D.S. Douches. 1994. Quantitative trait locus analysis of tuber dormancy in diploid potato (Solanum spp.). Theor. Appl. Genet. 89:474-480.

Gebhardt, C., E. Ritter, T. Debener, U. Schachtschabelk, B. Walkemeier, H. Uhrig, and F. Salamini. 1989. RFLP analysis and linkage mapping in Solanum tuberosum. Theor. Appl. Genet. 78:65-75.

Grant, W.F., and E. Small. 1996. The origin of the Lotus corniculatus (Fabaceae) complex: A synthesis of diverse evidence. Can. J. Bot. 74:975-989.

Hackett, C.A., J.E. Bradshaw, R.C. Meyer, J.W. McNicol, D. Milbourne, and R. Waugh. 1998. Linkage analysis in tetraploid species: A simulation study. Genet. Res. 71:143-154.

Haldane, J.B.S. 1930. Theoretical genetics in autopolyploids. J. Genet. 22:359-372.

Handberg, K., and J. Stougaard. 1992. Lotus japonicus, an autogamous, diploid legume species for classical and molecular genetics. Plant J. 2:487-496.

Hart, L.R.H., and C.P. Wilsie. 1959. Inheritance of a flower character, brown keel tip, in Lotus corniculatus L. Agron. J. 51:379-380.

Hayashi, M., A. Miyahara, S. Sato, T. Kato, M. Yoshikawa, M. Taketa, M. Hayashi, A. Pedrosa, R. Onda, H. Imazumi-Anraku, A. Bachmair, N. Sandal, J. Stougaard, Y. Murooka, S. Tabata, S. Kawasaki, M. Kawaguchi, and K. Harada. 2001. Construction of a genetic linkage map of the model legume Lotusjaponicus using an intraspecific [F.sub.2] population. DNA Res. 8:301-310.

Kalo, P., G. Endre, L. Zimanyi, G. Csanadi, and G.B. Kiss. 2000. Construction of an improved linkage map of diploid alfalfa (Medicago sativa). Theor. Appl. Genet. 100:641-657.

Kiss, G.B., G. Csanadi, K. Kalman, P. Kalo, and L. Okresz. 1993. Construction of a basic genetic map for alfalfa using RFLP, RAPD, isozyme, and morphological markers. Mol. Gen. Genet. 238:129-137.

Lander, E., P. Green, J. Abrahamson, A. Barlow, M. Daly, S. Lincoln, and L. Newburg. 1987. MAPMAKER: An interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1:174-181.

Landry, B.S., and R.W. Michelmore. 1985. Selection of probes for restriction fragment length analysis from plant genomic clones. Plant Mol. Biol. Rep. 3:174-179.

Luo, Z.W., C.A. Hackett, J.E. Bradshaw, J.W. McNicol, and D. Milbourne. 2001. Construction of a genetic linkage map in tetraploid species using molecular markers. Genetics 157:1369-1385.

Mather, K. 1935a. The combination of data. Ann. Eugenics 6:399-410.

Mather, K. 1935b. Reductional and equational separation of the chromosomes in bivalents and multivalents. J. Genet. 30:53-78.

Mather, K. 1936. Segregation and linkage in autotetraploids. J. Genet. 32:287-314.

Meyer, R.C., D. Milbourne, C.A. Hackett, J.E. Bradshaw, J.W. McNichol, and R. Waugh. 1998. Linkage analysis in tetraploid potato and association of markers with quantitative resistance to late blight (Phytophthora infestans). Mol. Gen. Genet. 259:150-160.

Nualsri, C., and P.R. Beuselinck. 1998. Rhizomatous Lotus corniculatus L.: IV. Inheritance of rhizomes. Crop Sci. 38:1175-1179.

Paran, I., and R.W. Michelmore. 1993. Development of reliable PCR-based markers linked to downy mildew resistance genes in lettuce. Theor. Appl. Genet. 85:985-993.

Qu, L., and J.F. Hancock. 2001. Detecting and mapping repulsion-phase linkage in polyploids with polysomic inheritance. Theor. Appl. Genet. 103:136-143.

Raelson, J.V., P.C. Lemaitre, K.M. Starkie, and W.F. Grant. 1989. An isozyme study in the genus Lotus (Fabaceae). Theor. Appl. Genet. 77:360-368.

Ramnani, A.D., and D.A. Jones. 1984. Inheritance of the brown keel tip character in Lotus corniculatus L. Bangladesh J. Bot. 13:52-59.

Ripol, M.I., G.A. Churchill, J.A.G. da Silva, and M. Sorrells. 1999. Statistical aspects of genetic mapping in autopolyploids. Gene 235: 31-41.

Sambrook, J., E.F. Fritsch, and T.A. Maniatis. 1989. Molecular cloning: A laboratory manual. 2nd ed. Cold Spring Harbor Laboratory, Woodbury, NY.

Soltis, D.S., and L.H. Riesberg. 1986. Autopolyploidy in Tolmiea menziesii (Saxifragaceae): Genetic insights from enzyme electrophoresis. Am. J. Bot. 73:310-318.

Steiner, J.J. 1993. Registration of AG-S4 autogamous broad-leaf birds-foot trefoil germplasm. Crop Sci. 33:1424-1425.

Steiner, J.J., C.J. Poklemba, R.G. Fjellstrom, and L.F. Elliott. 1995. A rapid one-tube genomic extraction process for PCR and RAPD analyses. Nucleic Acids Res. 23:2569-2570.

Stougaard, J. 2001. Genetics and genomics of root symbiosis. Curr. Opin. Plant Biol. 4:328-335.

Tavoletti, S., F. Veronesi, and T.C. Osborn. 1996. RFLP linkage map of an alfalfa meiotic mutant based on an [F.sub.1] population. J. Hered. 87:167-170.

Van Eck, H.J., J.M. Jacobs, P. Stam, J. Ton, W.J. Stiekema, and E. Jacobsen. 1994. Multiple alleles for tuber shape in diploid potato detected by qualitative and quantitative genetic analysis using RFLPs. Genetics 137:303-309.

Wu, K.K., W. Burnquist, M.E. Sorrels, T.L. Tew, P.H. Moore, and S.D. Tanksley. 1992. The detection and estimation of linkage in polyploids using single-dose restriction fragments. Theor. Appl. Genet. 83:294-300.

Wu, S.S., R. Wu, C.-X. Ma, Z.-B. Zeng, M.C.K. Yang, and G. Casella. 2001. A multivalent pairing model of linkage analysis in autotetraploids. Genetics 159:1339-1350.

Yu, K.F., and K.P. Pauls. 1993. Segregation of random amplified polymorphic DNA markers and strategies for molecular mapping in tetraploid alfalfa. Genome 36:844-851.

Zietkiewicz, E., A. Rifalski, and D. Labuda. 1994. Genome fingerprinting by simple sequence repeat (SSR)-anchored polymerase chain reaction amplification. Genomics 20:176-183.

Robert G. Fjellstrom, * Jeffrey J. Steiner, and Paul R. Beuselinck

R.G. Fjellstrom, USDA-ARS, Rice Research Unit, 1509 Aggie Dr., Beaumont, TX 77713; J.J. Steiner, National Forage Seed Production Res. Cntr., USDA-ARS, 3450 SW Campus Way, Corvallis, OR 97331; P.R. Beuselinck, USDA-ARS, Plant Genetics Research Unit, Univ. Missouri, Columbia, MO 65211. Received 2 Jan. 2002. * Corresponding author (r-fjellstrom@tamu.edu).
COPYRIGHT 2003 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Fjellstrom, Robert G.; Steiner, Jeffrey J.; Beuselinck, Paul R.
Publication:Crop Science
Date:May 1, 2003
Words:10060
Previous Article:Herbage mass, nutritive value, and ergovaline concentration of stockpiled tall fescue. (Forage & Grazing Lands).
Next Article:Genetic mapping of the [Or.sub.5] gene for resistance to Orobanche Race E in sunflower. (Cell Biology & Molecular Genetics).
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters