Identification and High-Density Mapping of Gene-Rich Regions in Chromosome Group 5 of Wheat

Kulvinder S. Gill,* Bikram S. Gill,* Takashi R. Endot and Elena V. Boyko* *Wheat Genetics Resource Center and Department of Plant Pathology, Kansas State University, Manhattan, Kansas 66506 and tLaboratory of Genetics, Faculty of Agnculture, Kyoto University, Kyoto, 606 Japan Manuscript received August 17, 1995 Accepted for publication March 13, 1996

ABSTRACT The distribution of genes and recombination in the wheat genome was studied by comparing physical maps with the genetic linkage maps. The physical maps were generated by mapping 80 DNA and two phenotypic markers on an array of 65 deletion lines for homoeologous group 5 chromosomes. The genetic maps were constructed for chromosome 5B in wheat and 50 in Triticum tauschii. No marker mapped in the proximal 20% chromosome region surrounding the centromere.More than 60% of the long arm markers were present in three major clusters that physically encompassed <18% of the arm. Because 48% of the markers were cDNA clones and the distributions of the cDNA and genomic clones were similar, the marker distribution may represent the distribution of genes. The gene clusters were identified and allocated to very small chromosome regions because of a higher number of deletions in their surrounding regions. The recombination was suppressed in the centromeric regions and mainly occurred in the gene-rich regions. The bp/cM estimates varied from 118 kb for gene-rich regions to 22 Mb for gene-poor regions. The wheat genes present in these clusters are, therefore, amenable tomolecu- lar manipulations parallel to the plants with smaller genomes like rice.

HE bread wheat (Triticum aestivum L. em. Thell., resultant physical maps are then compared with the T 2n = 6X = 42) possesses a large genome (16 bil- corresponding genetic linkage maps to analyze the lion bp per haploid genome), which is about six times physical distribution of recombination andorder of the size of maize and 35 times that of rice (BENNETT markers within each chromosome region. The genetic and SMITH1976). The three crop plants most probably mapping may be performed in wheat or in any of its originated from a common ancestor -60 million years relative species. A physical map is compared with a ge- ago (BENNETZENand FREELING1993). Besidespoly- netic linkage map by drawing lines to join the common ploidy in wheat, a key step in the evolution of these markers. The resultant composite map is called a cyto- three crops was differential amplification of DNA, to a genetic ladder map (CLM) (GILL andGILL 1994b). The greaterextent in wheat than in maize or rice. The previously constructed CLMs have established that the amount of actively transcribing DNA is probably not distributions of markers and recombination are uneven much different among the three genomes. The genes along the wheat chromosomes (WERNERet al. 1992; in wheat may be present in uninterrupted clusters, indi- GILLet al. 1993a; KOTA et al. 1993; HOHMANNet al. 1994; vidually interspersed by repetitive DNA blocks, or in a DELANEYet al. 1995a,b; MICKELSON-YOUNGet al. 1995). combination of the two arrangements. Thedistribution In the present study, we illustrate the effectiveness of pattern of genes will greatly influence the choice and the CLM strategy to identify and preferentially map success of techniques for molecular manipulation of gene-rich regions in homoeologous group 5 chromo- the genome. Map-based cloning of the genes individu- somes of wheat. ally interspersed by noncoding repetitive DNA would be difficult. Conversely, clustered genes will be amena- MATERIALS AND METHODS ble to such manipulations, parallel to the plants with smaller genomes like rice. Genetic stocks: Genetic markers were physically mapped We proposed a mapping strategy to target gene-rich using 65 deletion lines for wheat group 5 chromosomes (5A, 54 and 50).Twenty of these deletion lines were for the short regions of the wheat genome (GILL and GILL1994b). arms and 45 were for thelong arms. There were eight deletion Briefly, single-break deletion lines are used to divide lines each for the short arm of chromosomes 5A and 54 and each wheat chromosome into small regions marked by four for 5D. For the long arm, there were 22, 14 and nine chromosome bands, protein and DNA markers. The deletion lines for chromosomes 54 54 and 50, respectively. The deletion lines were generated using the gametocidal Corresponding author: Kulvinder S. Gill, Wheat Genetics Resource chromosome of Aegilops qlindn'ca (ENDO1988; ENDOand GILL Center, Department of Plant Pathology, 4307 Throckrnorton Hall, 1996). Nullisomic-tetrasomic (NT) lines (missing apair of Kansas State University, Manhattan, KS 66506. chromosomes, the deficiency of which is compensated by a E-mail: [email protected] pair of homoeologous chromosomes) and ditelosomic lines

Genetics 143 1001-1012 (June, 1996) 1002 K. S. Gill et al.

TABLE 1 The clones used for genetic and physical mappingof wheat homoeologous group 5 chromosomes

No. of bands Chromosome locationChromosome bands cDNA/ Total no. Clone"genomic of bands Enzyme 5A5D 5B T. tauschii Wheat PgsP C 3 EcoRI 1115s pTa71 C 8 EcoRI 0015 lBS, 5DS dhn2 C 12 EcoRI 2115 5L pHvabc309 C 5 HindIII 0115 51, pHvabg705 G 3 HindIII 111 5s pHvcnlbcdl57 C 4 HindIII 1215 5L pHudcd204 C 6 EcoRI 2125 5L pHvmlbcd351 C 7 EcoRI 2225 5L pHbcnlbcd450 C 3 HindIII 1115 5L pHvmlbcd508 C 10 HindIII 1111, 5 5L, IL pHvmlbcdlO87 C 2 HindIII 0015 5L pAsmlcdo57 C 7 HindIII 0127s 5L, 2A,7AS, 7BL, 7DL pAsnlcdo213 C 3 HindIII 1115 5L pAsdcdo388 C 13 HindIII 103IS, 2L, 3L, 4S, 5, 6 5L, lBS, IDS, 2DL PAsdcdo400 C 5 EcoRI 1125 5L, 1L pAsmlcdo412 C 3 HindIII 1115 5L PAsdCdo484 C 4 EcoRI 011 5L, 4L, 2L pAscnlcdo677 C 3 EcoRI 1115 5s pAsmlcdo687 C 8 HindIII 0117s 5s, 7L pAsmlcdo786 C 4 HindIII 0017s 5L pAscnlcdol049 C 3 EcoRI 1115 5L pAsmlcdo1312 C 4 EcoRI 1004, 5 5AL, 4BL,4DL pAsmlcdo1333 C 6 EcoRI 1004, 5 5AL, 4BL,4DL pAsmlcdol335 C 4 HindIII 1115 5s pHvh8 C 12 EcoRI 1625 5L pHvksu24 C 8 EcoRI 3225 5L pHvksu26 C 3 HindIII 1115 5L pHvksu5S C 5 HindIII 2125 5L pTtkma3 G 7 EcoRI 2125 5L pTtksudl6 G 4 HindIII 1115 5L pTtksud30 G 7 EcoRI 2225 5L pTtksud42 G 3 HindIII 111- 5L PTtksufl G 5 EcoRI 1115 5L pTtksug7 G 13 HindIII 1025, 7 5AL, 5DL,7L pTtksugl2 G 6 EcoRI 0015, 7L 5DL,7AL pThgl4 G 3 EcoRI 1115 5L pTtksug44 G 3 HindIII 1115 5L pTtksug5 7 G 3 HindIII 1115, 2 5L pTtksug60 G 5 HindIII 2105 5s pTtksuh8 G 15 HindIII 0102, 3, 5, 7 5BS,7L pTtksui26 G 7 EcoRI 1125 5s pTtksum2 G 4 BamHl 1125 5L pTtksus1 G 4 HindIII 1115 5L pTamwg522 G 4 EcoRI 1215 5L pTamwg602 G 3 HindIII 0115 5L pTapsr 79 C 5 EcoRI 1115 5L pTapsrll5 C 7 EcoRI 0215 4AL, 5BL, 5DL pTapsrll8 C 5 HindIII 1115 5s pTapsrl20 C 8 EcoRI 2235 5L pTapsrl28 C 3 EcoRI 1115 5L pTapsrl45 C 12 EcoRI 3445 5L p Tapsrl70 C 8 EcoRI 1003 54 3L pTapsr3 70 C 4 EcoRI 2115 51 pTapsr580 C 3 EcoRI 0115 4AL, 5BL, 5DL pTapsr628 C 4 EcoRI 1115 51 pTapsr63 7 C 3 EcoRI 1115 51 pTagl65 G -3 HindIII 0125 5 Mapping Gene-Rich Regions of Wheat 1003

TABLE 1 Continued

~~ ~~~ No. of bands Chromosome location cDNA/ Total no. Clone“genomic of bands Enzyme5B 5A 50 T. tauschii Wheat pTag222 G 1 HindIII 0 102 WL, 5BL pTag251 G 6 HindIII 1105 5L pTag317 G 10 HindIII1 1 0 3,4,5 5, 6A pTag319 G 7 HindIII 0 2 25 5s pTag354 G 4 HindIII 0 115 4A, 4B, 5BL, 5DL pTag614 G 6 HindIII 2 115 5L, 6B, 6D pTag621 G 1 HindIII 00 15 5L pTag644 G 3 HindIII 11 15 5L pTag651 G 3 HindIII 0 115 5, 7AS pTag695 G 7 EcoRI 2 125 5L pTag754 G 6 EcoRI 0 105 5, 7B pTacnlwgl14 G 4 HindIII 10 04 4BL, 4DL, 5AL pTantlwg363 G 3 HindIII 111 5s pTacnlwg341 G 4 EcoRI 11 15 5L, 6L, 7L pTantlwg419 G 3 EcoRI 11 15 5L pTanlwg530 G 3 EcoRI 0 115 5L pTacnlwg564 G 10 HindIII 14 35 5L pTanlwg583 G 3 EcoRI 11 15 5L pTacnlwg444 G 3 EcoRI 11 15 5L pTanlwg889 G 6 EcoRI 011 5L, 3L pTacnlwg908 G 5 EcoRI 12 15 5L pTacnlwg909 G 5 EcoRI 11 15 5L, 2BS, 7BS pTacnEwglO26 G 5 EcoRI 13 15 5L “Total no. of bands” and “No. of bands for5A, 5B, and 50” are from mappingon nulli-tetra lines of wheat cultivar Chinese Spring. The type of probe is indicated as C for the cDNA clones and G for the genomic. The restriction enzyme used for genomic DNA digestion of the aneuploid lines is mentioned under “Enzyme.” The probes shown boldface were mapped on the 5B population of RSLs. a bcd, cdo, and wg were barley cDNA, oats cDNA, and wheat genomic clones, respectively, from Dr. MARK SORRELLS;mwgwere barley genomic from Dr. A. GRANER;psrwere wheat cDNA from Dr. MIKEGALE; Tagwere wheat ”genomic from Dr. KOICHIROTSUNEWAKI: Huksu and Ttlzsu were barley cDNA and T. tauschii PstI genomic clones, respectively, from our laboratory.

(SEARS1954) were used to assign DNA restriction fragment were only interchromosomally mapped to group 5 chromo- bands to theirrespective chromosome arms. The populations somes using NT lines. The “dhn2” probe is a barley dehydrin used for linkage analyses are described in the later sections. gene that mapped to chromosome5H (TIMCLOSE, personal All the deletion lines, genetic and aneuploidstocks are main- communication). The rest of the 20 probes were genetically tainedat the Wheat Genetics Resource Center (WGRC), mapped to chromosome 5D of T. tauschii (272 = 14, DD) Throckmorton Hall, Kansas State University, Manhattan, KS. by restriction fragment lengthpolymorphism (RFLP) linkage Arm ratio/-fraction length (FL)measurements: The analysis (GILLet al. 1991b). mean of 10 chromosomes was used for all measurements of Physical mapping: Of the 65 deletion lines, 43 were homo- the deletion lines. Standard errors were calculated for com- zygous (the deletion chromosome present as a pair), nine parison of the measurements. Arm ratios were calculated by were hemizygous (only one deletion chromosome), and 13 dividing the short arm measurements with that of the long were heterozygous (one deletion and one normal chromo- arm. The breakpoint FL values of the short armdeletion some). Each group Sspecific DNA fragmentband was lines were calculated by dividing the armratio of the deletion mapped to a chromosome region flanked by breakpoints of chromosome with that of the normal homologue. Fraction the largest deletion possessing the fragment band and the length values for the long arm deletions were calculated by smallest deletion lacking it. For homozygous and hemizygous dividing the arm ratio of the normal homologue (GILLet al. deletion lines, the DNA probes were scored for the presence 1991a) with that of the deletion chromosome. The standard or absence of the DNA fragment(s). For the heterozygous errorsfor the FL measurementsranged from 20.005 to deletion lines, however, the DNA fragment bandswere scored 20.021. for theirintensities. The intensity of a DNA fragment detected DNA probes: Eighty wheat homoeologous group Sspecific in a deletion line was compared with that of Chinese Spring probes were used for physical and genetic mapping. Chromo- (CS) and/or other lines possessing the normal chromosome. some locations,the numbersof DNA fragments detected, and For example, one or moreof chromosome 5B or 50 deletion the sources of these probes are given in Table 1. Thirty-eight lines were used as a control for comparing intensities of 5A- (48%) probes were cDNA clones and 42 were genomic. Sixty specific fragment bands. As an internal check, the intensities (75%)probes [“bcd”, “cdo”, and “wg” (ANDERSON etal. 1992), of noncritical fragment bandswere also compared. For exam- “Tag” (NASUDA1992), and “psr” probes (SHARP et al. 1989)] ple, the intensities of 5A- and 5specific fragment bandswere 1004 K. S. Gill et al.

Chinese Spring (CS) X 5B disomic substitution from Triticum dicoccoides in ‘CS’ were counted. Based on random distribution, 4.5 dele- 40+ 40+ I. tions are expected for each region (total 45 deletions). 58 \ 58 Eight deletions were observed in the proximal 30% of monotelo 5BL F1 the long arm and no deletionwas observed in the distal “+Y 40+ 10% (Figure 4). Theexpected numbers of deletions for 1 5B these regions are 13.5 and 4.5, respectively. Conversely, five deletions were observed for a chromosome region 136 plants, each possessing 40 normal wheat chromosomes and a 5B chromosome (lacking the 5BL telo), were selected and allowed to self from FL 0.3 to 0.4, and seven each for the FL 0.5-0.6, 40+1-()orlor---- or- and FL 0.7-0.8 regions. For these three regions, dele- or + parental type 5B + tions were observed in all threehomoeologues. The From the progeny of each plant, a 42 chromosome plant, possessing recombinant three regions may be more susceptible to breaks and chromosome in disomic condition, was selected and used for mapping. this property is conserved among homoeologues. FIGURE1.-The crossing scheme used for generating the Physical maps: The physical maps of chromosomes population of5B RSLs. The two parents differed only for 5A, 5B, and 5D are given in Figure 2. The 80 probes chromosome 5B and only one cycle of meiosis was allowed detected 209 loci on the three homoeologues (240 ex- before the recombinant products were isolated individually pected). Sixty-three loci were detected on chromosome as separate RSLs. 5A and 73 each on5B and 5D. Fifty-one probes detected loci on all three chromosomes (marked bold in Figure also compared while comparing the intensities of 5Bspecific fragment bands. Ethidium bromide-stained pictureof the gel 3), 18 detected loci on two, and 11 on only one chromo- was also used for intensity comparisonsto make sure that an some. Several deletions that could not be orderedcyto- equal amount of DNA was loaded in the lanes being com- logically because of similar FL values were differenti- pared. If the DNA fragment band being compared wasof ated by DNA markers. For example, the FL value for approximately half the intensity ofthe CS band andthe other deletions 5AL-10, -22, and -3 was 0.56. However, homoeologous chromosome fragment band intensities were the same in the two lines, the probe was scored as “missing” markers cd0388 and Tag644 mapped in the chromo- in the deletion chromosome. For the probes that detected some region distal to the breakpointsof deletions 5AL- more than one DNA fragment band for a chromosome, the 10 and 5AL-22 but proximal to thatof 5AL-3. Similarly, DNA fragments were scored to be nonallelic (different alleles the submicroscopic chromosome region flankedby de- shown by a letter at the end of probe names) if they map to letions 5AL-15 and 5AL-6 possessed nine markers. Sig- different chromosome regionsor if the probe detected more nificantly, the submicroscopicchromosome regions than two DNA fragments bigger than its own size. Genetic mapping: The 50 RFLP genetic linkage map was bracketed by deletions 5BL-14 and 5BL-9, and 5DL-8 developed using an F2 population (56 plants) of a cross be- and 5DL-9 were homoeologous to that flankedby dele- tween two accessions of T. tauschii (TA1691 X TA1704) (GILL tions 5AL-15 and 5AL-6 (revealed by the presence of et al. 1991b).All 80 probes were concurrently mapped physi- the same markers) (Figure 2). cally and in T. tauschii, exceptfor probes cdo484,Krud42, Physical locations of two agronomically important wg363, and wg889. These probes failed to detect polymor- phismusing restriction enzymesDruI, HindIII, EcoRI, and genes were studied andboth mapped to submicro- EcoRV. The mapping data were analyzed using the computer scopicregions, eachbracketed by two deletion program MAPMAKER (LANDER et al. 1987). Twenty-seven of breakpoints with the same FL value. The Qgene,which the 80 probes were genetically mapped in wheat usinga chro- makes wheat heads square and free thrashing, mapped mosome 5B population of 136 recombinant substitution lines to chromosome region at FL 0.87 bracketed by (RSLs) (marked bold in Table 1).The population was devel- 5AL oped as described in Figure 1. The cultivar CS was crossed deletions 5AL-7 and 5AL-23. Similarly, the Phl, a chro- with a CS substitution line containing chromosome5B from mosome-pairing regulator gene of wheat, mapped to a an emmer wheat (T. dicoccoides). The F, wascrossed with chromosome region at FL 0.55 bracketed by deletions mono telo 5BLto recover each recombinant chromosome 5BL-1 and phlb. individually as an RSL. For chromosome 5B, each RSL repre- Except for thetelomeric region of chromosome 5AL, sents a gamete. The mapping dataof 136 recombinant substitution lines were analyzed using MAPMAKER. The gel-blot the location and the orderof the markers were the same DNA hybridization procedures were as previously described among the three group5chromosomes, suggesting that (GILLet al. 1993a). the gene synteny is conserved among wheat homoeologues. The distal region of chromosome 5AL was in- RESULTS volved ina complex translocation involving chromosomes 4A and 7B (NARANJOet al. 1987). A part of 5AL Nature of chromosome breaks: The breakpoints of region distal to FL 0.87 is homoeologous to chromo- all 65 deletion lines are marked by arrows on the left somes 4B and 40. The markers cdo1333, cdo1312, and of thechromosome drawings (Figure 2). To test the wgl14 are wheat homoeologous group4 probes (Figure randomness of the deletion breakpointdistribution, the 3, see also chromosome 5A in Figure 2). Correspond- group 5 long arm was divided into 10 regions (FL 0.1, ingly, group 5 probes cdo484, psrll5, and psr580 did not 0.2, . . . , 1.0). For each region, the number of deletion map on5AL because the chromosome regionpossessing breakpointsin all three homoeologouschromosomes these probes is present on 4A (Figures 2 and 3). Mapping Gene-Rich Regions of Wheat 1005 22 b .P M vi0 .eOh .z 0 2E 2g co " E .5 0 uh 55 +c05 Me c 'C $2 :$ ma.e w 52 &!bg& da ~a&d & 4 44 j A jjjj 'CI mq 8 882 4 a $883 uc m r-m 0 m 0 O\N*W P- '4'9 ? $%% c, '5 0 03 0 oo3 2 2 zszs E ? 35 9 cu r\' 3E E2 50 seoE 3.c ' $2mug .e '5 .=, .s EO2 5,r 0 g;;ua,u 35D ggg 3 2; @I: 2.2 &E@-g 50 -22 5 -- VI0 O\N x 2% "5 $$ $a%$; Oi w -09 $2 0- 0 00 oooqa=a co 00 s25 9 DEU m 0 x Y0 b 25 6% LPId ;z g. $9 58; .B a; 1.c 82 $1 5 -2 .? '0 cm -3 kQ f$E$ gs . ?? .k c :g :@ quoc c1.$4 2% 23 c;r: m5c 3: 3: a? 0 B?J 68 $$ is: '" g 'Eo 32 3%* $ L $.g '5 a9 26 %, B .h v; g: 2: f g g 92 2 $$ 4 3s g 9 F 558Y 6% 223 @x? 5 6.5 % $& m .s % E%>Ru u 5, -mo s 8 I 'J .- 0 a, g.e z s Sa LL++a, Io3 cj3, ea 2 '8 5 2" * 5 2," 9- - iQ$ 0 Pa 1006

I31 7

- abg705 - Tag319 - wg363

-Tag1 65

- wg4 19 -psr370 -CdO388b

-psr115

"cd0484 7psr580 -Tag3 54

FICX'KF.%-A CLM of wheat homoeologous group 5. Consensus physical map of wheat group 5 (center) was compared with the genetic linkage maps of 5D of 7.. In7rsrlrii (right) and 5B of wheat (left). The common markers across maps are joined by lines. The probes marked in bold on the consensus physical map detected loci on all three chromosomes of group 5. The region marked with dotted pattern at the tip of the long arm of the consensus physical map marks the region of chromosome 511 Mapping Gene-Rich Regions of Wheat of Regions Gene-Rich Mapping 1007

Consensus physical map: Because of the conservation of gene svnteny, the physical mapping information of the three homoeologues can be combinedto gener- ate a higher resolution consensus physical map. The group 5 consensus physical map was constructed as fol- lows: a hypothetical chromosome was drawn that was divided into 67 chromosome intervals flanked by 65 deletion breakpoints (Figure3). Each probe was placed on the consensus map to the shortest possible interval. For example, the probes ksuu3, rug909, rug41 9, ksum2 and ksud42 mapped to the chromosome regions between FL 0.16 and 0.32 in 5A, and proximal to FL 0.29 in 5D. However, in chromosome5B the probes mapped 4 4-4 to the region spanned by FL 0.26 and 0.29. On the consensus map, the five markers were placed between FL 0.26 and 0.29. The FL values and the size of each marked region on the consensus map were the same as that of the chromosome where thesize of the region 0 0.1 02 0.3 04 OS 06 07 OX 119 I was the smallest. For example, the probes ksusl and Fraction length of 5BL hu8 mapped to chromosome region between FL 0.46 FIGURE4.-The distribution patterns of markers and delc- and 0.56 in 5A and 0.48 and 0.60 in 50. In chromosome tion breakpoints on the long arm of wheat homoeologous 5B and the consensus map, the two probes mapped to group 5 chromosome. The long arm was divided into 20 re- a submicroscopic chromosome region at FL 0.55. gions, each equivalent to FL 0.0.5. The number of deletions, with their breakpoint..in each region, were counted from Although the relative order and the FL location of chromosomes 5A, 5B, and 51). The number of markers was most of the markers were the same among three ho- counted from the consensus map.0, number of markers; moeologues, a few discrepancies were observed for FL the number of deletions. ., location. The markers wg1026,cd0400, psr637, ksu57, hu24, p.v-120, Tag251, and rug583 mapped to submicro- mapped in submicroscopic chromosome regions.T~us, scopic regions at FL 0.75 in 5B and 5D. A submicro- 30% of the long arm markers mappedin the submicroscopic region of 5A containing these markers was pres- scopic chromosomal regions. entat FL 0.67. Onthe consensus map,the region Genetic Iiikage maps: Figure 3 showsRFLP-based containing these markers was placed at FL 0.75. genetic linkage maps of chromosome 511 of 7.. lnusrhii Distribution of markers: The distribution of the (right) and5B of wheat (left). All 80 markers (mapped markers was highly uneven alongthe chromosome physically) were also mapped in T. tnzwchii, except for length (Figures 3 and 4). Uneven distribution of the cdo484, ksud42, zug363, and rug889, which did not detect markers was more apparent from the consensus map polymorphism between the two parents of the popula- than was from the individual physical maps. With the tion with restriction enzymes Drd, HindIII, EcoRI, and centromere as the reference point, more markerswere EcoRV.Fifty-six of the 80 probes were also tested for present in the distal region compared to the proximal. polymorphism on CS and the 5R substitution line (the No marker was observed in the chromosome region two parents of the 5B mapping population). Thirty- surroundingthe centromere (20% of thechromo- seven (66%) probes detected polymorphicfragments some). About 46% (32/69) of the long arm markers between the two parents with one or more of the restric- were present in the distal 25% of the arm. This general tion enzymes Drd, HindIII, IkoRI, EroRV, IJnrnHI, Kid, marker distribution pattern was superimposed by the and, XhoI. The 5B genetic linkage map consist5 27 of presence of submicroscopic regions of high probe den- the 37 probes along with two Cband markers (Figure sity. On the long arm, three markerclusters were obvi- 3). The orderof the probes is identical between the 5R ous, surrounding FL 0.30, 0.55, and 0.75 (marked on and the 5D maps. Relative genetic distances among the Figure 3). Forty-two of the 69 (61%) long armmarkers markers were similar between the two maps, however, mapped to the three regions that encompass <18% of the overall recombination was reduced in the 5B map the arm. Ten markers mapped in the submicroscopic compared to the 5D map. region at FL 0.75 and eightin the one atFL 0.55. Three Cytogenetic ladder map (CLM): Both 5R and 51) ge- other markers (cdo388n, Tng644, and bcd508n) also netic linkage maps gave a misleading appearance that

involved in translocation with chromosome 4A. The group 5 probes present on chromosomes4A, 5B, and 51) are shown on the right of the region, whereas, the group 4 probes present on chromosome 5A, 4B, and 41) are shown on the left. The probes marked with * mapped to different chromosomal regions behveen wheat and 7'. /mtschii. Centromeres are marked bv C's. The two Gband markers on the 5B linkage map are FiRL1.42 and 5RL2.62. 1008 IC S. Gill et al.

toides, A. sharonasis, A. lonpnima, A. cylindrica, or Secab cerealp, when present in wheat background, cause chromosome breakages (TSUJIMOTO and TSUNEWAKI1985; 20 ENDO 1988; KOTA and DVORAK 1988;LUKASZEWSKI 1995). For the purpose of deletion stock isolation, A. cylindricctbased system is the most ideal among them. 15 Chromosome breaks occur in the gametes lacking the gametocidal chromosome of A. cylindrica, therefore, the deletions are recovered in a pure wheat background. 10 In all the other Aegilops systems, breaks occur in the gametes carrying the gametocidal gene. Therefore, it is difficult to recover deletions in a pure wheat back- 5 ground. Second, the complex cytological aberrations or rearrangements that are frequentin the presence of A. speltoida or S. cereak chromosomes (KOTA and DVO- 0 RAK 1988; LUKASZEWSKI1995) were not observed in our 0 0.1 0.2 0.3 0.4 0.5 06 0.7 0 8 09 I deletion lines. More than 400 deletion lines for all 21 Fraction length of 5BL wheat chromosomes have been isolated and character- FIGURE5.-The distribution patterns of markers and re- ized using Gbanding andRFLP markers (WERNERet dl. combination on the long armof wheat homoeologous group 1992; GILLet al. 1993a,b; KOTA et al. 1993; MICKELSON- 5 chromosome. The long arm was divided into 20 regions, YOUNG et al. 1995; ENDO and GILL 1996). Almostall each equivalent to FL 0.05. The number of markers were deletions resulted from single breaks followed by the counted from the consensus mapas in Figure 4. The recombi- healing of telomeres and subsequent loss of the regions nation for each region was estimated from the 5D genetic linkage map. The recombination between markers mapping distal to each break. It may partly be due to the fact in adjoining regions was divided equally between the two re- that care was taken to select for deletions caused by gions. 0,number of markers; .,recombination in 10 cM. single breaks (ENDOand GILL1996). Third, theA. qlin- dricebased system breaks all wheat chromosomes and markers are concentrated around thecentromeres. The arms compared to the other systems where breakages physical maps clearly showed that centromeric regions occur only at certain sites on a few specific wheat chro- are devoid of markers that are distally located in clus- mosomes (KOTA and DVORAK1988; LUKASZEWSKI1995). ters. However, physical maps alone cannot reveal ge- The deletions were equally frequent both in hetero- netic distances and marker order within each region. chromatin and in euchromatin (GILLet al. 1993a; ENDO It is only when information on the physical and genetic and GILL1996). maps is combined in a CLM (GILL and GILL1994b) that Sensitivity to breaks us. marker density: In group 5 the precise distribution of recombination and genes in chromosomes, breakages were semi-random andthe specific regions become apparent. marker-rich regions were the preferredsites. As a result, Figure 3 shows a CLM of wheat group 5 chromosomes marker clusters were identified and localized to submi- constructed by comparing the consensus physical map croscopic chromosome regions bracketed by deletion with the genetic linkage maps of 5Band 51). The recom- breakpoints (Figure 4). This correlated pattern of dele- bination distribution pattern coincided with that of the tion breakpoints distribution with that of markers was markers (Figure 5). Therecombination was suppressed also observed for other wheat chromosomes (WERNER in the marker-poor chromosome region surrounding et al. 1992; GILLet al. 1993a,b; KOTA et al. 1993; MICKEL the centromere. Contrarily, 58% of the total long arm SON-YOUNG et al. 1995). The deletions always occur in recombination occurred in the distal 25%. Most of the marker-rich regions, although, factors other than gene long arm recombination occurred in the regions density also influence breakages. For chromosome 5A, around the three marker clusters (Figure 5). The ge- six deletions were observed at FL 0.98 of the shortarm netic length of the submicroscopic region at FL 0.55 and three atFL 0.11 of the long arm. The two regions was 44 cM and atFL 0.75 was 114 cM. Similarly for the were devoid ofmarkers. An apparent exception was for short arm, recombination was very low in the proximal the chromosome region distal to FL 0.87. No deletion 20% compared to the distal region where it was very was observed in anyof thethree chromosomes, al- high (Figures 3 and 5). The markers gyp, abg705, psrl70, though at least 10 markers mapped in the region (Fig- and Tag31 7 physically mapped in the distal 25% of the ures 2 and 4). This is probably because of the lack of short arm. The four loci were not even linked to the telomeric Gband in the long arm. The deletion chro- 51) linkage map. mosomes were identified by theGbanding analysis. Without a diagnostic telomeric Gband, loss of the distal DISCUSSION 13% of the arm would not be easily detected. The six Uniqueness of the gametocidal systemused: It is well deletions at FL 0.98- 1.O of chromosome 5ASwere iden- established that some chromosomes from Aegzlops spel- tified because of the presence of the telomeric Gband. M apping Gene-RichMapping Regions of Wheat 1009

Apparently, the susceptibilityof the marker-rich chromosome regions to the gametocidal gene is conserved among homoeologues. Among the regions that possess breakpoints of morethan one deletion at the same FL, the corresponding deletionsin homoeologues were detected only for the marker-rich regions. The marker cluster at FL 0.75 is the best example for which deletions bracketing the submicroscopic region were recovered for all three homoeologues. One reason could be that loosely packed transcriptionally active chromosome regions are more accessible to breaks compared to tightly packed inactive regions. Moreover, gene-rich regions are also high in recombination (discussed later) that may also be related to high breakage frequency. Irrespective of the reason for the differential sensitivity to chromosome breaks, it would be interesting to see the extent of their conservation across the grass family. Conversely, deletion clusters in marker-poor regions 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 hadbn length of 5RL were not conserved across homoeologues. Three deletions were recovered at FL 0.11 of 5AL, but none was FIGURE6.-Distribution of cDNA us. genomic clones on observed in the corresponding regions of 5BL or 5DL. the long arm of wheat homoeologous group 5. The long arm was divided into 20 equal regions as in Figures 4 and 5. The The deletion cluster at FL 0.98 in 5AS is another exam- number of cDNA or genomic clones was counted from the ple. No deletion was observed in thecorresponding consensus map. W, the cDNA clones; 0,the genomic clones. regions of 5BS or 5DS. Both these chromosome arms have telomeric Gbands. Deletions around FL 0.98, if ing nulli-tetra lines and were not selected for theirabil- occurred, wouldhave beendetected. Similarly, the ity to detect polymorphism. The distribution pattern of breakage hotspot at FL 0.42 of 5BS was not observed at ksud42, hcd450, 7ug363, and lug889 that did not detect the correspondingregions of 5ASor 5DS. In these cases, polymorphism in T. tazlschii was not unusual; ksud42 marker density cannot be the cause of high breakages. and bcd450 mapped in the major gene clusters, rug889 Except for 5AS region at FL 0.98, other regions lie in in an interstitial chromosome region and wg363 in the recombination-poor areas and rule out thepossible role short arm along with three other markers. The markers of recombination in breakage frequency. Besides, the were from 10 different libraries from wheat, barley, oats, distribution pattern of recombination is conserved or T.tawchii.Therefore, the probes used in the present among homoeologues (discussed later). Thus, in cer- study represent a random sample of wheat genes. tain cases chromosome-specific intrinsic elements may The results from the present study lead to the conclu- be related to high breakage frequency. sion that genes in wheat are present in clusters inter- Distribution of genes: Of the 80 markers used in the spersed by blocks of repetitive sequences visualized as present study, 39 were cDNA clones, 29 were PstI geno- regions of low gene density. Three major gene clusters mic, and 12 were random genomic clones. No obvious were observed in the long arm interspersed in gene- difference was observed between the distribution pat- poor regions (Figures 3 and 4). The precision with terns of cDNAand the genomicclones (Figure 6).Many which the gene-rich and gene-poor blocks can be differ- J?$d genomic clones probably also represent genes be- entiated, depends upon the number of deletions for cause P$tI libraries are enriched with undennethylated the chromosome. The uneven distribution of markers actively transcribing regions of the genome (BURRet in the presentstudy was not as obviousfrom the individ- al. 1988, for review). The distribution pattern of the ual maps as it was from the consensus map where more markers may, therefore, reflect that of wheat genes. It deletion data points were available.These geneclusters has been suggested that the highly conserved genes may be localized to even smaller chromosome regions (housekeeping genes) may be present in recombina- as moredeletions become available.Eleven markers tion-poor regions of the genome (around centromeres) mapped in the region just distal to the submicroscopic and less conserved genes map in the recombination- region at FL 0.75 with 10 markers. Some of these adja- rich regions (AISSANI and BERNARDI1991). If this hy- cent markers may turn out to be part of the FL 0.75 pothesis is true forwheat, the probes thatdo not detect cluster with the availabilityof fine physical mapping polymorphism among various accession/parents should tools. map in the centromeric regions, whereas, the probes Gene distribution has been studied in animal systems thatdetect polymorphism shouldmap towardte- and it seems that all animal genomes and chromo- lomeres. Sixty probes (75%) used in the present study somes, to some degree, are divided into gene-rich and were only interchromosomally mapped to group 5 us- gene-poor compartments (SUMNERet al. 1993, for re- 1010 K. S. Gill et al. view). Gene distribution in thehuman genome was the chromosome location. Figure 5 shows the compari- studied by several methods including in situ hybridiza- son of the distribution of markers with that of recombi- tion of a random pool of mRNA (YUNISet al. 1977), in nation. The recombination is the highest at the marker situ of isochore H3 (SACCONEet al. 1992), which com- clusters. In fact, recombination around the marker clus- prises the highest gene concentration of any fraction ters accounts for most of the long arm genetic length. of DNA (BERNARDI1989; MOUCHIROUDet al. 1991), or Recombination in thelargest marker cluster (at FL 0.75) by DNAase hypersensitivity (WEINTRAUBand GROUDINE was the highest among the three clusters. 1976; ELGIN1988). Theunanimous conclusion was that It is apparent that thestructure of each chromosome the genes in the human genome are localized in R- region determines the extent of recombination as the bands and are more concentrated in the terminal R- relative genetic distances between markers were similar bands called T-bands (SUMNERet al. 1993). In general, in the genetic linkage maps of polyploid and diploid more R-bands are present toward the telomeric regions wheat (Figure 3). It implies that this chromosomal orga- compared to the proximal regions. The gene distribu- nization is conserved among homoeologues. tion in wheat is similar to that in humans, although R- The correlation between marker density and recom- bands are not detected in wheat. bination suggests that recombination occurs only in the Differences in gene densities among chromosomal gene-rich regions. However, if recombination does oc- regions were observed in yeast, although, large regions cur in the gene-poor regions also, it would not be fully devoid of genes were not detected (DUJONet al. 1994). detected because of the lack of markers. A positive cor- Compared to that of wheat or humans, the yeast ge- relation was observed between the number of markers nome is very small. Correspondingly, the proportion of and observed recombination in wheat as well asin other repetitive DNA in the yeast genome is also small. The crop plants (NILSSONet al. 1993, for review). Previously differentiation of the genome intogene-rich and gene- undetectedrecombination events were revealed in poor compartments is probably universal to all eukary- maize, barley (NILSSONet al. 1993), rice, and wheat by otes. The size of the gene-poor compartments and the mapping more markers. In rice an RFLP genetic link- extent of unevenness in the gene distribution depend age map with 1383 markers (KURATAet al. 1994) is 30% upon the size of the genome. In smaller genomes of longer than the map with 600 markers (TANKSLEYet al. yeast or Arabidopsis, the gene distribution would ap- 1993). Based on the data on two to four polymorphic pear moreuniform compared to that of a large genome C-bands, an average of about two crossovers/arm were of wheat where it is highly uneven. The gene density observed for chromosome 5BL of wheat (DVORAKand in gene-rich compartments is probably comparable CHEN1984; CURTISand LUKASZEWSKI1991). However, among various eukaryotes. when 36 markers were mapped on the arm, theaverage Distribution of recombination: Recombination in number ofcrossovers increased to about four (GILL the telomeric region (distal 25%) of the group 5 long and GILL 1994a; the present study). As many as five arm was more than 20 times the recombination in the crossovers/arm were observed during a single cycle of centromeric region (proximal 25%) (Figures 3 and 5). meiosis. Double crossovers spanning < 10% of the arm These data supportprevious reports of uneven distribu- were frequent. The gene-poor region block may be as tion of recombination in wheat (DVORAKand CHEN large as 25% of the arm. Double crossovers occurring 1984; CURTISand LUKASZEWSKI1991; WERNERet al. in these blocks would not be detected. 1992; GILLet al. 1993a; KOTAet al. 1993),animal systems Because of the nonrandom distribution of recombi- (STEINMETZet al. 1987; BOLLAGet al. 1989), andin other nation along the chromosome length, bp/cM would plants (RICK1971; GANALet al. 1989). Proximity to cen- differ among regions. The wheat CLMs revealfairly tromeres tends to suppress crossing over in other organ- precise estimates of bp/cM especially for the gene-rich isms also (GAUDETand FITZGERALD-HAYES1990). In to- regions. The current study shows that bp/cM estimates mato,reduced recombination in thecentromeric for wheat chromosomes may vary from 118 kb in gene- regions was attributed to the presenceof heterochroma- rich regions to 22 Mb in gene-poor regions. Our prelim- tin (RICK 1971). In wheat, recombination in the proxi- inary results from the long-range mapping experiments mal 20% of all wheat chromosomes is suppressed, even indicate that one cM for the gene-rich cluster at FL though these region mayvary 10- to 20-fold in the 0.55 may translate into <54 kb (K. S. GILLand B. S. amount of heterochromatin. GILL,unpublished data). The above bp/cM estimates Earlier conclusions of suppressed recombination are comparable to that of similar regions in other crop around centromeres and increased recombination to- plants. Based on the genome size and RFLP linkage ward telomeres in wheat were based on either a few C- map length, onecM in tomato is estimated to be 550 kb bands (DVORAKand CHEN1984; CURTISand LUKASZEW- (GANALet al. 1989). However, the estimates for regions SKI 1991) or low density CLMs (WERNERet al. 1992; GILL around various disease-resistancegenes ranged from 43 et al. 1993a; KOTA et al. 1993). The high density maps to 90 kb (CANALet al. 1989; SEGALet al. 1992; ZHANG et presented here reveal that the distribution of recombi- al. 1994). Similarly, rice contiguous maps have revealed nation may depend upon markerdensity rather than on that 1 cM may vary from 120 kb to 1 Mb compared to Mapping Gene-Rich Regions of Wheat 1011 the predicted 273 kb (UMEHARAet al. 1995). The bp/ BENNETZEN,J. L., and M. FREELING,1993 Grasses as a single genetic system. Trends Genet. 9: 259-261. cM estimates for the regions around thegenes arecom- BERNARDI,G., 1989 The isochoreorganization of thehuman ge- parable among various crop plants and may be any- nome. Ann. Rev. Genet. 23: 637-661. where from 40 to 120 kb/cM. However, the upperlimit BOI.IAG, R. J., A. S. WALDMANand R. M. LISKAY,1989 Homologous recombination in mammalian cells. Annu. Rev. Genet. 23: for such estimates depends upon the genomesize. The 199-225. upper limit for bp/cM estimates is determined by the BURR,B., R. A. BURR,K. H. THOMPSON,M. C. ALBERTSEN and C. W. size ofgene-poor compartments. Inrice the upperlimit STUBER,1988 Genemapping with recombinant inbreads in maize. Genetics 118: 519-526. is 1 Mb compared to 22 Mb in wheat. The wheat ge- CURTIS,C. A,, and A. J. LUKASZEWSKI,1991 Genetic linkage between nome is -35 times larger than that of rice. C-bands and storage protein genes in chromosome 1B of tetra- Conclusions: Because of its large size (16 billion bp) , ploid wheat. Theor. Appl. Genet. 81: 245-252. DEIANEY,D., S. NASUDA,T. R. ENDO,B. S. GILI. andS. H. HUI.BEm, the wheat genome is commonly believed not to be suit- 1995a Cytologically based physical maps of the group2 chro- able for molecular manipulations. Our results showed mosomes of wheat. Theor. Appl. Genet. 91: 568-573. that a high proportion of actively transcribing wheat DELA~JEY,D., S. NASUIIA,T. R. ENDO,B. S. GILLand S. H. HULBERT, 1995b Cytologically based physical maps of the group3 chrc- genes are present inclusters. The commonly used RFLP mosomes of wheat. Theor. Appl. Genet. 91: 780-782. genetic analysis failsto detect these gene clusters; there- DUJON,B., D. ALEXANDRAKI, B. ANDRE,W. ANSORGE, V. BAIAIIRONet fore, they cannot be efficiently used for molecular ma- al., 1994 Complete DNA sequence ofyeast chromosome XI. Nature 369: 371-378. nipulations. The CLM strategy not only efficientlyiden- DVORAK,J,, and K. C. CHEN,1984 Distribution of nonstructural tifies thegene clusters but also preferentially maps variation between wheat cultivars along chromosome arm 6Bp: them. Thegametocidal system ofA. cylindrica frequently evidence from the linkage map and physical map of the arm. Genetics 106: 325-333. causes breaks in the gene cluster regions and is there- EI.GIN,S. R., 1988 The formation and function of DNAase I hyper- fore an aid in the identification and localization of gene sensitive sites in the process of gene activation. J. Biol. Chem. clusters to small chromosomal regions. The gene clus- 263: 19259-19262. ENDO,T. R., 1988 Induction of chromosomal structural changes by ters are also preferred sites forrecombination and a chromosome of Aegihps cylindrica L. in common wheat. J. marker order can be resolved genetically. Furthermore, Hered. 79: 366-370. agronomically importantgenes are present in these ENIIO,T. R., and B. S. GILL,1996 The deletion stocks of common wheat. J. Hered. (in press). clusters. For group 5, we mapped two such genes (Phl GALIBA,G., S. A. QUARRIE, J. SUTKA,A. MORGUN~Vand J. W. SNAPE, and 9) using the deletion lines and both mapped in the 1995 RFLP mapping of the vernalization (Vrnl) and frost resis- gene clusters. Two other genes conferring resistance to tance (Frl) genes on chromosome 511 of wheat. Theor. Appl. Genet. 90: 1174-1179. wheat against environmental stress (cold tolerance)and GANAL,M. W., N. D. YOUNGand S. D. TANKSLEY,1989 Pulsed field response to vernalization ( Vml) (GALIBAet al. 1995) gel electrophoresis and physical mapping of large DNA frag- are flanked by two probes mapping in the gene cluster ments in the Tm-Pa region of chromosome 9 in tomato. Mol. Gen. Genet. 215: 395-400. at FL 0.75. This appears to bea general pattern of GAUDET,A. M., and M. FITZGERALDHAWS,1990 The function of distribution for most of the agronomically useful genes centromeres in chromosome segregation, pp. 118-131 in Thp in wheat. Therefore, most wheat genes are amenableto Eukaqotic Nucleus. Molerular Biochemisty and MacromolecularAssem- blies, edited byP. R. STRAUSSand S. H. WIISON. Telford Press, map-based cloning similar to other plants with smaller Caldwell, NJ. genomes. Many wheat genes, especially the ones con- GILL,B. S., B. FRIEBEand T. R. ENDO,1991a Standard karyotype trolling the basic physiological functions, are probably and nomenclature system for description of chromosome bands and structural aberrations in wheat (Triticum uestivum).Genome similar to their counterparts in small genome crops 34: 830-839. like rice. However, some important genes areprobably GILL,K. S., and B. S. GILL,1994a Higher level of recombination in wheat-specific and must be cloned from wheat that is wheat is due to more cross-overs/arm and localized recombination hotspots, p. 64 in Plant Genome II. The Second International the only source of their genetic resources. According ConfM(.ncr on the Plant Genome, San Diego. to our analysis, cloning would be feasible for the genes GII.I., K. S., and B. S. GILL, 199413 Mapping in the realm of poly- that are present in gene-cluster regions. ploidy: the wheat model. BioEssays 16: 841-846. GILI., K. S., E. L. LUBBERS,B. S. GILL,W. J. RAUPP and T. S. Cox, This is contribution no.96-27-5 from the Kansas Agricultural Exper- 1991b A genetic linkage map of Triticum tauschii (DD) and its iments station, Kansas State University, Manhattan, KS 66506. The relationship to the D genome of bread wheat (AABBDD). Ge- nome 34 362-374. research was supported by U.S. Department of Agriculture-National GILL, K. S., B. S. GILL andT. R. ENDO,1993a A chromosome region- Research Initiative Plant Genome Award and U.S.japan National specific mapping strategy reveals gene-rich telomeric ends in Science Foundationjapan Society of Plant Science collaborative re- wheat. Chromosoma 102: 374-381. search program. We thank Mr. DUANEWILSON for greenhouse help, GILL,K. S., B. S. GILL,T. R. ENDOand Y. MUM, 199313 Fine physical mapping of Phl, a chromosome pairing regulator gene in polyploid wheat. Genetics 134: 1231-1236. LITERATURE CITED HOHMANN,U., T. R. ENDO,K. S. GII.I.and B. S. GILL,1994 Compari- son of geneticand physical maps of group 7 chromosomes from AISSANI, B., and G. BERNARDI,1991 CpG islands: features and distri- Triticum aestivum L. Mol. Gen. Genet. 245: 644653. bution in the genomes of vertebrates. Gene 106: 173-183. KOTA, R. S., and J. DVORAK,1988 Genomic instability in wheat in- ANDERSON, J. A., Y. OGIHARA,M. E. SORRELLSand S. D.TANKSLEY, duced by chromosome 6Bs of Triticum speltaides. Genetics 120: 1992 Development of a chromosomal arm map forwheat based 1085-1094. on RFLP markers. Theor. Appl. Genet. 83: 1035-1043. KOTA,R. S., K. S. GILL,B. S. GILLand T.R. ENDO,1993 A cytogeneti- BENNETT, M. D., and J. B. SMITH,1976 Nuclear DNA amounts in cally based physical map of' chromosome 1B in common wheat. angiosperms. Philos. Trans. R. SOC.London, Ser. B 274 227-274. Genome 36: 548-554. 1012 K S. Gill et al.

KURATA,N., Y. NAGAMURA,K. YAMAMOTO, Y. HARUSHIMA, N. SUE et rounding the I2 Fusarium oxysporum resistance locus in tomato. al., 1994 A 300 kilobase intelval genetic map of rice including Mol. Gen. Genet. 231: 179-185. 883 expressed sequences. Nature Genet. 8: 365372. SHARP, P. J., S. CHAO,S. DESAIand M. D. GAL.E,1989 The isolation, LANDER, E. S., P. GREEN,J. ABRAHAMSON, A. BARLOW,M. J. DALY characterization and application in the Triticeae of a setofwheat et al., 1987 MAPMAKER an interactive computer package for RFLP probes identifying each homoeologous chromosome arm. constructing primary genetic linkage maps of experimental and Theor. Appl. Genet. 78: 342-348. natural populations. Genomics 1: 174-181. STEINMETZ,M., Y. UEMATSUand L. K. FISCHER,1987 Hotspot3 of LUKASZEWSKI,A. J., 1995 Chromatid and chromosome type break- homologousrecombination in mammalian genomes.Trends age-fusion-bridge cycles in wheat (Triticum aestivumL.). Genetics Genet. 3: 7-10. 140: 1069-1085. SUMNER,A. T., J. D. L. TORREand L. STUPPIA,1993 The distribution of genes on chromosomes: a cytological approach. J. Mol. Evol. MICKELSON-YOUNG,L., T. R. ENDO and B. S. GILL,1995 A cytoge- 117-122. netic ladder map of the wheat homoeologous group-4 chromo- 37: TANKSLEY,S. T. M. FULTONand S. R. McCouc~,1993 Linkage somes. Theor. Appl. Genet. 90: 1007-1011. D., map of rice (O?yra sativa) (272 = 24), pp. 6.61-6.76 in Gnetic MOUCHIROIJD,D., G. D’ONOFRIO,B. AISSANI,G. MACAYA,C. GAUTIER Maps, edited by S. J. O’BRIEN.Cold Spring Harbor Laboratory et al. 1991 The distribution of genes in the human genome. Press, Cold Spring Harbor, NY. Gene 100: 181-187. TSIJIIMOTO,H., and K. TSUNEWAKI,1985 Hybrid dysgenesis in NARANJO,T., A.ROC& P. G. GOICOECHEAand R. GIM.DEZ, 1987 common wheat caused by gametocidal genes. Jpn. J. Genet. Arm homoeology of wheat and rye chromosomes. Genome 29: 60: 565-578. 873-882. UMEHARA,Y., A. INAGAKI, H. TANOUE,Y. YASUKOCHI,Y. NAGAMURA et NASUDA,S., 1992 Molecular genetical studies of common wheat and al., 1995 Construction and characterization of rice YAC library its relatives-telosomic analysis of the RFLP loci in common for physical mapping. Mol. Breed. 1: 79-89. wheat and a study on chloroplast genome diversity in Aegzlops WEINTRAUB,H., and M. GROUDINE,1976 Chromosomalsubunits caudata L., MSc. Thesis. Kyoto University, Japan. in active genes have analtered conformation. Science 193: NIISSON, N.-O., T. SALLand B. 0. BENGTSSON,1993 Chiasma and 848-856. recombination data in plants: are they compatible. Trends WERNER,J. E., T. R. ENDOand B. S. GILL,1992 Towards a cytogeneti- Genet. 9: 344-348. cally based physical map of the wheat genome. Proc. Natl. Acad. RICK,C. M., 1971 Some cytological features of the genome in dip Sci. USA 89 11307-11311. loid plant species. Stadler Genet. Symp. 1,Z 153-174. YUNIS, J. J., M. T. KUO and G. F. SAUNDERS,1977 Localization of SACCONE,S., A. D. SARIO,G. D. VALE and G. BERNARDI,1992 The sequences specifying messenger RNA to lightstaining Gbands highest gene concentrations in the human genome are in te- of human chromosomes. Chromosoma 61: 335-344. ZHANG, H.B., G. B. MARTIN, S. D. TANKSLEYand R.A. WING, 1994 lomere bands of metaphase chromosomes. Proc. Natl. Acad. Sci. Map-based cloning in crop plants: tomato as a model system. 11. USA 894913-4917. Isolation and characterization of a set of overlapping yeast artifi- SEARS,E. R., 1954 The aneuploidsof common wheat. Research Bul- cial chromosomes encompassing the jointless locus. Mol. Gen. letin 572 University of Missouri Agricultural Experiment Station. Genet. 244: 613-621. SEGAL,G., M. SARFATTI, M. A. SCHAFFER,N. ORI, D. ZAMIR et aL, 1992 Correlation of genetic and physical structure in the region sur- Communicating editor: J. A. BIRCHLER