Induction and Characterization of Ph1 Wheat Mutants

Corresponding author: Graham Moore, John Innes Centre, Norwich Research Park, Colney, Norwich NR4 7UH, England., tracie.foote{at}bbsrc.ac.uk (E-mail)

Communicating editor: J. A. BIRCHLER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

The cloning of genes for complex traits in polyploid plants that possess large genomes, such as hexaploid wheat, requires an efficient strategy. We present here one such strategy focusing on the homeologous pairing suppressor (Ph1) locus of wheat. This locus has been shown to affect both premeiotic and meiotic processes, possibly suggesting a complex control. The strategy combined the identification of lines carrying specific deletions using multiplex PCR screening of fast-neutron irradiated wheat populations with the approach of physically mapping the region in the rice genome equivalent to the deletion to reveal its gene content. As a result, we have located the Ph1 factor controlling the euploid-like level of homologous chromosome pairing to the region between two loci (Xrgc846 and Xpsr150A). These loci are located within 400 kb of each other in the rice genome. By sequencing this region of the rice genome, it should now be possible to define the nature of this factor.

HEXAPLOID wheat (Triticum aestivum L., 2n = 42) is composed of three related genomes (A, B, and D), each consisting of seven pairs of chromosomes (SEARS 1952 ). Although each chromosome of hexaploid wheat has the potential to pair with either its homologue or the two pairs of homeologues from the other two genomes, chromosome pairing is largely restricted to homologues. Thus, hexaploid wheat behaves meiotically as a diploid with 21 bivalents observed at meiotic metaphase I. The major locus, Ph1, controlling this behavior is located on chromosome 5B (OKAMOTO 1957 ; RILEY and CHAPMAN 1958 ). Two deletions of this locus have long been known, one in hexaploid wheat (ph1b; SEARS 1977 ) and the other in tetraploid (A and B genomes) wheat (ph1c; GIORGI 1978 ). Association and recombination between the homeologous chromosomes of the related genomes is evident in these deletion lines as observed by multivalent formation at meiotic metaphase I. Over the last 20 years, the two deletion lines have been important for breeding purposes because they permit the wheat chromosomes to pair with homeologues from related species.

In the past, the absence of the Ph1 locus has been defined by the presence of multivalents at meiotic metaphase I. Recent results indicate that the deletion of the Ph1 locus may affect several premeiotic and meiotic processes. In summary, it affects the level of premeiotic association of homologues at their centromeres (ARAGON-ALCAIDE et al. 1997A ; MARTINEZ-PEREZ et al. 1999 ), centromere structure (ARAGON-ALCAIDE et al. 1997B ), the timing of telomere bouquet formation during meiotic prophase and, hence, the timing of intimate association of homologues (MARTINEZ-PEREZ et al. 1999 ), and recombination between homeologous chromosomes or segments (DUBCOVSKY et al. 1995 ; LUO et al. 1996 ). SEARS 1977 also reported that the ph1b line exhibited reduced fertility compared to euploid wheat, implying chromosome pairing/segregation failure and gamete abortion. Recent studies suggest some form of premeiotic control of chromosome pairing in wheat (ARAGON-ALCAIDE et al. 1997A ; VEGA and FELDMAN 1998 ). However, from our current knowledge of chromosome pairing, largely based on yeast studies, it is difficult to envisage how all the effects observed might be controlled by a single gene, although it is possible that the disruption of a premeiotic process could result in a cascade of effects on meiotic processes. It is therefore important to establish whether the Ph1 locus is a single gene, a multigene family, or a complex of linked genes influencing different meiotic processes.

The deleted segment of chromosome 5B in the ph1b line is ~70 Mb in size and, therefore, a number of genes are likely to have been deleted (GILL et al. 1993 , GILL et al. 1996 ). Further analysis of the genetic content of this locus requires identification of new deletions involving the region and an approach for revealing the gene content of this 70-Mb region of the wheat genome. A strategy has been suggested for using rice as the cereal model (MOORE et al. 1993 ). The genome of hexaploid wheat consists of 17,000 Mb per 1C nucleus (BENNETT and SMITH 1991 ) and is composed of ~80% repetitive sequences. The rice genome on the other hand is only 400 Mb and is composed of only 50% repetitive sequences. Comparative mapping has revealed that the genes on rice chromosome 9 are in a similar order to those on a part of wheat chromosome 5B including the region containing the Ph1 locus (FOOTE et al. 1997 ). The present study describes an approach of combining the identification and characterization of wheat lines carrying new deletions of the Ph1 locus with the establishment of a complete yeast artificial chromosome (YAC) contig to the corresponding region in rice.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Plant material:
The following hexaploid wheat (T. aestivum; 2n = 6x = 42; genome AABBDD) lines were used: euploid Chinese Spring; the Chinese Spring mutant line ph1b (SEARS 1977 ); Chinese Spring nullisomic-tetrasomic lines N5AT5B, N5BT5D, and N5DT5A (SEARS 1966 ); and a doubled haploid single chromosome substitution line, Highbury (CS5B) line in which the Highbury 5B chromosome has been substituted by that from Chinese Spring (SNAPE et al. 1988 ). The tetraploid wheat (T. durum; 2n = 4x = 28; genome AABB) Cappelli and its derivative deletion line ph1c (GIORGI 1978 ) were used in deletion mapping experiments. Deletion lines for screening were generated de novo in two separate experiments. First, 3000 seeds from the wheat line Highbury (CS5B) were subjected to a fast-neutron irradiation dose of 3.0 GyNf at the International Atomic Energy Agency's Laboratories (Seibersdorf, Austria; protocol no. FN662). The M₁ generation was grown in a greenhouse, allowed to self-fertilize, and the seed was harvested. A total of 5213 seeds from 1042 M₁ lines were sown in field plots with up to 15 plants per row (each row representing one M₁ line), and progeny from 25 M₁ different lines were selected for further analysis because they exhibited reduced fertility in mature spikelets. The M₃ generation from the 25 selected M₁ lines was grown under greenhouse conditions and screened for deletions of the Ph1 locus by PCR (QU et al. 1998 ). M₄ plants from one M₁ line, H455, were analyzed further.

For the second experiment, 1300 F₁ seeds from ph1b segmental hemizygotes of a cross between Chinese Spring and the ph1b deletion line were subjected to fast-neutron irradiation (as above). Half of the M₁ seeds were grown in a greenhouse and selfed seed was collected. Almost 5000 juvenile M₂ plants (up to 10 from each M₁ parent) were screened using the multiplex PCR ph1b assay, and selected plants were selfed or crossed with Chinese Spring lines and grown further.

PCR-based characterization of the ph1b domain:
Primers PSR2120.f (5'-TTA ACG CCA GGG CAT ACT C) and PSR2120.r (5'-CTG CAG GAG GCG CTG GA) amplify a wheat 232-bp DNA fragment PC3-9 (subsequently designated Xpsr2120) that maps within the ph1b deletion and can be used in a plus/minus PCR assay to detect lines deleted in this region (QU et al. 1998 ). The DNA sequences of two restriction fragment length polymorphism (RFLP) probes PSR128 and PSR574 were known (J. B. SMITH, K. M. DEVOS and M. D. GALE, personal communication) that have been located in the ph1b deletion (DUNFORD et al. 1995 ). Therefore, to obtain additional PCR-based markers, primers were designed to amplify B genome-specific DNA fragments of these two probes. Optimal primer pairs for B genome-specific Xpsr128 and Xpsr574 amplification were designated PSR128.f (5'-ATC GCT CCT CTG CTT GCT TC) and PSR128.r (5'-GAC CGC CTG AAA CCT CCC), and PSR574.f (5'-AGC GTA TAT TCA CGC GCT CC) and PSR574.r (5'-CGT AAG AAC TCC CCA GGG TTT G), respectively. Primers AWJL3.f (5'-TGG CAC CCT CAA TGT AGA C) and AWJL3.r (5'-GCT TGC CCA TTT CAC AAC), derived from the early meiosis wheat cDNA clone pAWJL3 (JI and LANGRIDGE 1994 ) and mapped to wheat chromosome 2A (M. A. ROBERTS, unpublished results), were included in multiplex PCR assays as a positive control.

Fresh intact leaf material was prepared for PCR amplification as described by QU et al. 1998 . Each 50-µl PCR reaction contained one prepared leaf slice (0.5 cm²), 1x reaction buffer (10 mM Tris-HCl, pH 9.0, 1.5 mM MgCl₂, and 50 mM KCl), 2 units of Taq DNA polymerase (Pharmacia Biotech, Piscataway, NJ), 2 mM dNTPs, 50 pmol each primer, and 1x loading buffer (2% Ficoll, 0.1 mM EDTA, and 0.02% Orange G dye). The first PCR reaction cycle contained a denaturating step at 94° for 30 sec, an annealing step at 65° for 30 sec, followed by a product extension step at 72° for 1 min. For the next 12 cycles, the annealing temperature was reduced by 0.7° per cycle, with the final annealing temperature of 56° continued for another 23 cycles (VOS et al. 1995 ). Reaction products were fractionated through 3% agarose gels.

RFLP analysis:
Genomic DNA was extracted from wheat lines (QU et al. 1998 ) and analyzed using standard procedures (FOOTE et al. 1997 ). When hybridizing with either wheat or rice probes, Southern blots of wheat-digested genomic DNA were blocked with 5 µg/ml (final concentration) denatured sonicated wheat genomic DNA to produce optimal hybridization results.

Cytogenetic analysis:
Chromosome pairing of spring sown control and deletion lines was analyzed in pollen mother cells at meiotic metaphase I in Feulgen-stained preparations. Observed differences in pairing frequencies between lines were assessed for statistical significance using Tukey's comparisons for single factor one-way analysis of variation (ANOVA). The single variable analyzed was that of the nature of the chromosome pairing in pollen mother cells derived from the single M₂ and M₄ deletion lines.

YAC contig development:
Forty-seven rice YACs located in the rice chromosome 9 region analogous to the overlapping wheat ph1b and ph1c deletions were selected from our rice chromosome 9 partial YAC physical map (FOOTE et al. 1997 ) for further analysis. Each of the rice YAC clones was plated out from -70° glycerol stocks onto fresh YPDX complete medium (1% yeast extract, 2% peptone, 2% dextrose, 2% agar, 50 µg/ml uracil, 50 µg/ml L-tryptophan, and 30 µg/ml adenine hemisulfate) and then transferred onto SD minimal medium (0.67% yeast nitrogen base without amino acids, 2% D-glucose, and 2% agar) supplemented with 20 µg/ml adenine hemisulfate, 20 µg/ml L-histidine, 30 µg/ml L-isoleucine and 30 µg/ml L-lysine, to select yeast strain AB1380 colonies containing functional left (TRP1) and right (URA3) arms of the pYAC4 vector (BURKE et al. 1987 ). At least five separate colonies from each YAC clone were picked for high molecular weight mini-prep DNA isolation (adapted from SCHWARTZ and CANTOR 1984 ). DNA was fractionated by pulsed-field gel electrophoresis (PFGE) using the Pharmacia LBK 2015 Pulsaphor System, typically with a 50- to 120-sec pulse at 170 V over 48 hr to separate fragments of 100–800 kb. Following Southern transfer to Hybond N⁺ nylon filters, YACs were sized by probing with the 1.4-kb PvuII/SalI pBR322 fragment (pYAC4 right arm probe) and the 2.3-kb PvuII/EcoRI pBR322 fragment (pYAC4 left arm probe). Integrity of each YAC clone was assessed by comparing YAC sizes from five or more independent colonies and, if fragmentation or deletion was evident, then the largest YAC clone was used further.

For isolation of YAC ends using the inverse PCR method (SILVERMAN 1993 ), individual YACs were obtained after PFGE of high-concentration total DNA plugs. YAC DNA was digested with each of the restriction endonucleases AccI, EcoRV, HaeIII, HhaI, HincII, NlaIII, NlaIV, RsaI, SalI, and SphI. DNA fragments were circularized by ligation and subjected to PCR using the following primer pairs with the indicated enzyme combinations for each arm: A5 (5'-ATA CAA TTG AAA AAG AGA TTC C) and S8 (5'-GTA GCC AAG TTG GTT TAA GG) with NlaIII for left arms; A13 (5'-GGA CGG GTG TGG TCG CCA TGA TCG CG) and S8 with EcoRV, HaeIII, NlaIV, and RsaI for left arms; S2 (5'-GAC TTG CAA GTT GAA ATA TTT CTT TCA AGC) and A3 (5'-AGT CGA ACG CCC GAT CTC AA) with HaeIII, HhaI, NlaIII, NlaIV, and SphI for right arms; and S11 (5'-AGG AGT CGC ATA AGG GAG AG) and A3 with AccI, HincII, and SalI for right arms. Selected PCR products were cloned into the pGEM-T Easy vector (Promega, Madison, WI). Putative YAC end clones were hybridized to Hybond N⁺ filters gridded with the 47 rice YACs to confirm clone identities and to screen for overlapping YACs. DNA sequencing of YAC end clones was performed using the Thermo Sequenase dye terminator cycle sequencing kit (Amersham, Arlington Heights, IL).

	RESULTS AND DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Development of multiplex PCR ph1b assay:
The wheat ph1b deletion generated by SEARS 1977 spans ~70 Mb (GILL et al. 1993 ). To screen for novel deletion lines with different breakpoints and thereby define more closely the region containing the Ph1 locus, it was necessary to develop PCR markers for loci dispersed within the ph1b deletion. On the long arm of wheat chromosome 5B, wheat RFLP loci Xpsr100 (at 3.2 cM from the centromere) and XKsu8 (at 22.7 cM from the centromere) broadly delimit the ph1b deletion (GALE et al. 1995 ; FOOTE et al. 1997 ). Three PCR-generated loci already reported are all located distally near the Xksu8 breakpoint (GILL and GILL 1996 ; SEGAL et al. 1997 ; QU et al. 1998 ), so the proximal RFLP loci Xpsr128 (at 4 cM) and Xpsr574 (at 10.9 cM) were selected for development into PCR-based assays. Two sets of primer pairs tested, PSR128.f/PSR128.r and PSR574.f/PSR574.r, were found to amplify the expected wheat chromosome 5B sequences located within the ph1b deletion (Figure 1). In combination with the primers PSR2120.f/PSR2120.r (QU et al. 1998 ) and positive control primers AWJL3.f/AWJL3.r, it proved feasible to undertake a single multiplex reaction, which provided a plus/minus assay for the determining the presence/absence of these three ph1b-linked loci.

View larger version (76K): In this window In a new window Download PPT slide   Figure 1. (a) PCR amplification products from alkaline-treated wheat leaves using primers PSR128f, PSR128r, PSR574f, PSR574r, PSR2120f, PSR2120r, AWJL3f, and AWJL3r multiplexed together (see MATERIALS AND METHODS). (b) PCR amplification products from extracted genomic DNA using the same primers multiplexed together. The absence of the product in the nullisomic-tetrasomic line N5BT5D indicates the mapping of the product to the missing chromosome. Products corresponding to each of the loci are indicated.

The multiplex PCR reaction worked efficiently on both purified genomic wheat DNA and wheat leaf slices prepared for PCR using a simple NaOH lysis/boiling technique (QU et al. 1998 ; Figure 1, a and b). In our hands, neither the KSU8-derived (GILL and GILL 1996 ) nor the WPG90-derived (SEGAL et al. 1997 ) primers amplified efficiently with PCR-prepared leaf slices as template. Both assays require time-consuming DNA extraction and the KSU8-derived assay requires restriction enzyme digestion of the PCR products to detect Chinese Spring euploid vs. ph1b polymorphisms. The PCR assay developed here has neither of these requirements and therefore allows higher throughput screening of mutagenized lines. However, the KSU8-derived assay still provides a useful marker for this region.

Identification of Highbury (CS5B) ph1 deletion lines:
When initially describing the ph1b deletion line, SEARS 1977 noted that its average seed yield was only 34% that of euploid Chinese Spring. This reduction in fecundity was attributed to an interstititial deficiency that included the Ph1 locus. In the present study, M₂ plants from the progenies of selfed fast-neutron irradiated Highbury (CS5B) M₁ lines were screened for reduced seed set in mature spikes. Approximately 2.3% of the lines appeared to be segregating for reduced fertility, and these were selected for further anaylsis. A total of 295 low-fertility segregants (M₂) obtained from the selected self-fertilized 25 M₁ plants were also self-fertilized. Leaf material from 295 M₃ progeny (a single progeny from each M₂ plant) was tested using the PSR2120 PCR-based assay for the ph1b deletion (QU et al. 1998 ). Only 3 M₃ Highbury (CS5B) plants, H455-2, H455-9, and H455-11, descended from M₁ line H455, were found to be missing the ph1b-located PSR2120 PCR marker. None of the 12 remaining M₃ plants descended from H455, nor any other of the M₃ plants tested, were found to be deficient for Xpsr2120. The PSR2120-based PCR assay produced clear results from 269 (91.2%) of the M₃ plants, including 93 (31.5%) that were screened twice, with the balance unscorable due to lack of or unclear PCR amplification products.

The 14 M₃ plants (including H455-2 and H455-9) descended from H455 were allowed to self-fertilize, and M₄ seeds were harvested and germinated. Leaf material from juvenile M₄ plants was screened by the multiplex PCR ph1b assay developed in this study for the presence or absence of Xpsr128, Xpsr574, and Xpsr2120 loci (Table 1). Two plants of the three M₄ progeny of H455-2, viz. H455-2-2 and H455-2-3, were deficient for all three loci. The third plant, H455-2-1, lacked none of the markers and is likely to be the result of spurious out-pollination. H455-9-1, the only progeny of H455-9, lacked only Xpsr2120, while H455-11-1, one of the four progeny of H455-11, also lacked the Xpsr2120 locus. The remaining 43 M₄ plants were not homozygous for deletions within the ph1b region.

Southern and PCR analysis of phenol/chloroform-extracted genomic DNA indicated that H455-2-2 was deficient for all markers tested within the ph1b region, and therefore the deleted region is of the same size or larger than the ph1b deletion (Table 1). H455-2-2 and H455-2-3 possibly carry the same deficiency. However, H455-9-1 was deficient only for loci Xpsr150A, Xpsr2120, and Xksu8, and the plant's interstitial deletion was delimited proximally by Xrgc846 and distally by Xpsr150B. The H455-9-1 deletion is therefore smaller than the ph1b deletion. It can be inferred that plants H455-9-1 and H455-11-1 probably possess a similar deficiency. Identification of M₄ plants carrying two different-sized deletions indicates that their common irradiated progenitor line H455 was chimeric.

Identification of Chinese Spring ph1 deletion lines:
In a second experiment to identify novel Ph1 deficiencies in wheat, seeds hemizygous with respect to the ph1b deletion (the F₁ generation from a Chinese Spring euploid x ph1b/ph1b cross, thus carrying a complete chromosome 5B and a ph1b deletion chromosome 5B) were irradiated by fast-neutron bombardment. Assuming transmission of both 5B chromosomes from a Ph1/ph1 heterozygote, three classes of progeny were anticipated from screening M₂ plants with the multiplex PCR ph1b assay: (1) all three ph1b PCR markers present, in nondeleted homozygous Ph1 plants (expected frequency F 0.25) and from hemizygous ph1b deletion plants (F 0.5); (2) all three ph1b PCR markers absent, in homozygous ph1b plants (F 0.25) and in hemizygous plants with the ph1b deletion and novel 5BL deletions of similar size or completely spanning the ph1b region (unknown F); and (3) one or two ph1b PCR markers absent, in plants homozygous or hemizygous for a novel deficiency within the ph1b region (unknown F). For the purposes of this experiment, only class 3 progeny were studied further.

Up to five M₂ progeny from each of two ears chosen at random from 450 M₁ plants were screened using the multiplex PCR ph1b assay. Five novel deletion lines were identified among the 4283 juvenile plants screened: CS239, CS1010, and CS3793 lacked the Xpsr2120 PCR locus but possessed Xpsr128 and Xpsr574 loci, while CS977 and CS1002 lacked both Xpsr128 and Xpsr574 but possessed Xpsr2120 (Table 1; Figure 1). The multiplex PCR ph1b assay generated scorable results in 98.9% of plants tested, including 13.5% that were repeated because the initial results were unclear.

Southern RFLP and PCR analysis of phenol/chloroform-extracted genomic DNA from the novel deletion lines confirmed the results of the leaf-based assays and allowed for further characterization (Table 1; Figure 2). The deficiency in CS239 could be delimited by two loci of the probe PSR150 located on the distal portion of chromosome 5BL. Xpsr150A had been previously located within the ph1b and ph1c deletions while Xpsr150B was located outside both deletions (FOOTE et al. 1997 ). In contrast, the deficiency in CS1010 was delimited outside these two Xpsr150 loci, providing another deletion breakpoint within the ph1b region. Lines CS977 and CS1002 harbored deficiencies that encompassed the proximal region of the ph1b deletion, and both the deficiencies extended distally from the locus Xrgc846 (identified using a rice probe) up to breakpoints before Xpsr150A (Table 1; Figure 2). The deficiency in line CS3793, which lacks the PCR locus Xpsr2120 located in the distal region of the ph1b deletion, also encompasses Xksu8 (Table 1).

View larger version (99K): In this window In a new window Download PPT slide   Figure 2. PhosphorImage of a Southern blot of genomic DNA digested with HindIII and hybridized with the rice DNA probe R2790. The arrow indicates the band missing in N5BT5D, ph1b, ph1c, CS977, and CS1002 lines. The absence of the band in the N5BT5D indicates the mapping of the band to the missing chromosome 5B.

The screening and partial genotyping of deleted M₂ juvenile plants has resulted in the identification of five new lines that contain deficiencies with different breakpoints within the ph1b region. It is important to remember, however, that the experiment was designed to allow detection of deficiencies overlapping with the ph1b deletion in both homozygous and heterozygous conditions. Only one line, CS1010, deficient for the locus Xpsr150B that is present in the ph1b line (Table 1), can be identified as homozygous for its deficiency, while the remaining four lines could be either homozygous or ph1b hemizygous for their respective deficiencies. It was still feasible to analyze the chromosome pairing behavior of both hemizygous and homozygous M₂ lines to attempt to correlate the different deficiencies with pairing phenotype, because only one copy of the Ph1 locus is necessary to prevent wheat meiotic homeologous pairing and maintain diploid-like homologous pairing (OKAMATO 1957; RILEY and CHAPMAN 1958 ).

Cytological analysis of ph1 deletion lines:
Meiotic pairing in wheat lines is commonly assessed using F₁ hybrids produced from crosses between the study line and related Triticeae species, e.g., Secale cereale (rye; WALL et al. 1971 ), Aegilops variabilis (SEARS 1977 ; CEOLONI and DONINI 1993 ) and A. sharonensis (MELLO-SAMPAYO 1972 ). In these F₁ hybrids homeologous chromosome associations are clear due to the absence of any homologous chromosome pairs, and the effects of genetic alterations can therefore be more easily diagnosed. However, such F₁ hybrids are sterile, and their production is time consuming. Moreover, interpretation of the effect of the Ph1 locus can be complicated by the presence of additional pairing loci on related Triticeae chromosomes. Therefore in this study meiotic pairing was only assessed in the control and deletion lines themselves. The feasibility of differentiating Chinese Spring euploid and Chinese Spring N5BT5D lines using selfed material has been demonstrated, albeit that differences in univalent, bivalent, and higher-order chromosome associations between lines are subtle (RILEY and KEMPANNA 1963 ).

Highbury (CS 5B) M₄ deletions lines, Chinese Spring M₂ deletion lines, and appropriate controls were scored in blind trials at meiotic metaphase I for univalent, rod and ring bivalent, and higher-order association (tri-, quadri-, penta-, and hexavalent) chromosome formations (Table 2). The 17 controls contained progeny derived from the same ears as the deletion lines, including those identified as having a ph1/ph1 genotype (i.e., loci Xpsr2120, Xpsr574, and Xpsr128 absent). Lines such as CS178 and CS1219 (both 2n = 42), which appeared from the multiplex PCR ph1b assay to be homozygous Ph1/Ph1 or hemizygous Ph1/ph1 (i.e., loci Xpsr2120, Xpsr574, and Xpsr128 present), formed predominantly 21 bivalents at MI and exhibited low levels of univalent and higher-order chromosome formation, as expected for Chinese Spring lines carrying the Ph1 locus. In contrast, lines CS270 and CS276 (also both 2n = 42), which were diagnosed by PCR as having a ph1/ph1 genotype, formed slightly fewer bivalents and more higher-order chromosome associations, similar in phenotype to Sears' ph1b/ph1b strain (Table 2). Although the total number of bivalents formed in Ph1 euploid and ph1 deficient lines was not statistically distinct (P = 0.01), it proved informative enough to differentiate between rod (mainly homeologous) and ring (mainly homologous) bivalent formation: lines CS(ph1b/ph1b), CS270, and CS276 that are deficient for the Ph1 locus produced more rod bivalents and less ring bivalents than lines CS178 and CS1219 that contain the Ph1 locus (each character statistically significant at P = 0.01). Data from cellular and cytogenetical studies suggest that the Ph1 locus must in part function to actively promote homologous pairing and cannot simply be preventing homeologous pairing (GILLIES 1987 ; ARAGON-ALCAIDE et al. 1997A ; MARTINEZ-PEREZ et al. 1999 ). This suggests that the level of homologous chromosome pairing is an appropiate measure for the effect of the Ph1 locus. Therefore, to assess whether novel deletion lines were normal or abnormal for homologous chromosome pairing behavior, the number of ring bivalents in control and experimental lines was compared.

Three classes of pairing phenotype could be distinguished on the basis of ring bivalent formation during meiotic metaphase I (statistically significant at P = 0.01): Chinese Spring euploid frequencies of ring bivalent formation were seen in CS239: H455-2-3, CS1002, and CS1010 produced similar numbers of ring bivalents to the CS(ph1b/ph1b) deletion, and very low levels of ring bivalent associations were observed in H455-9-1 (Table 2). Line H455-2-2 had only 41 chromosomes and was not included in the statistical analyses; however, it appears to have a similar level of homologous pairing to the CS(ph1b/ph1b) deletion.

Only three cells could be scored in line H455-11-1 (2n = 40), and these limited data indicated that the line had homologous chromosome pairing behavior similar to H455-9-1 (Table 2). However, the line was completely sterile and has perished. The lines CS3793 and CS977 were also not robust. Only eight pollen mother cells could be scored in the M₂ line CS977 (2n = 42) and the line had an apparent ph1b-like phenotype, with 0.50 ± 0.53 univalents, 3.88 ± 2.10 rod bivalents, and 14.38 ± 1.77 ring bivalents. Extensive backcrossing of these lines is being undertaken to improve their viability. If improved viability is achieved, the pairing frequencies of their progeny will then be assessed. All the lines identified by the screening procedures except for the CS239 (which possessed euploid-like pairing behavior) exhibited reductions in fertility and viability. Lines H455-9-1, CS977, and CS3793 exhibit the most severe levels of sterility. SEARS 1977 observed a low transmission rate for the chromosomes carrying the ph1b deletion. The transmission rate for chromosomes carrying the new deletions appears even lower than that of the ph1b chromosome. This may account for the high levels of sterility observed in these lines.

Irradiation can induce nonreciprocal chromosome translocations resulting in multivalents at metaphase I and apparent "homeologous pairing." The level of multivalent formation did not accurately predict the presence or absence of the Ph1 locus (Table 2). Its scoring may be complicated by the occurrence of such translocations. The presence of multivalents suggests that both homeologous pairing and recombination has occurred, while the presence of univalents, as GILLIES 1987 indicated, could result from homeologous pairing without homeologous recombination. Pollen mother cells derived from the ph1b line often do not possess any multivalents but do possess univalents. Therefore it is not surprising that the frequency of multivalents does not accurately predict the absence of the Ph1 locus. However, by scoring the frequency of ring bivalents, the number of chromosomes engaged in homologous pairing can be estimated. Conversely, the number of chromosomes not present as ring bivalents provides an estimate for those chromosomes engaged in both homeologous pairing with and without homeologous recombination. The use of ring bivalent frequencies at metaphase I in blind scoring of the lines proved sufficiently accurate to identify the lines previously determined by PCR to be homozygous for the ph1b deletion, including progeny from the same ear as the deletion lines. If there were nonreciprocal translocations in the M₁ plants, then they did not interfere with the characterization of these M₂ progeny. Therefore it is reasonable to suppose that the determination of the pairing types of the progeny carrying the new Ph1 deletions was also accurate. For the Highbury (CS 5B) deletion lines, however, it was not possible to undertake a similar study. Therefore the H455-9-1 line could carry nonreciprocal translocations causing the apparently higher level of homeologous pairing. An alternative explanation is that the deletion breakpoint in this line is actually located within the Ph1 locus, resulting in a new phenotype.

Characterizing the Ph1 locus:
The cytological analyses have shown the meiotic chromosome pairing behavior of novel wheat lines with deletions within the ph1b domain. The data presented in Table 1 and Table 2 allow for the construction of a map of deletion breakpoints that minimally define the part of the Ph1 region on chromosome 5B that controls the level of homologous/homeologous pairing (Figure 3, top). Deletion line CS1010 is deleted distal to the locus Xrgc846, and line CS1002 is deleted proximal to Xpsr150A, yet both lines show ph1b-like level of homologous pairing behavior. Line CS977 is deleted for the same markers as CS1002, and at the M₂ generation both lines exhibited ph1b-like behavior. Therefore, the euploid-like level of homologous chromosome pairing is being controlled by a factor that must lie between loci Xrgc846 and Xpsr150A. Line CS239, deleted distal to locus Xpsr150A, exhibits euploid-like pairing, confirming that this factor cannot be located distal to Xpsr150A.

View larger version (32K): In this window In a new window Download PPT slide   Figure 3. A comparison of the physical map of rice chromosome 9 with the presence or absence of markers on chromosome 5B in the new deletion lines. All the lines except CS239 have high homeologous pairing. The marker names are coded according to the species of origin and are designated as follows: C (or Xrgc in wheat), cDNA clones derived from rice callus library; R (or Xrgr in wheat), cDNA clones derived from rice root library; S, cDNA clones derived from rice shoot tissue; G (or Xrgg in wheat), rice genomic clones; psr (or Xpsr in wheat), wheat DNA clones; bcd, barley cDNA clones; cdo, oat cDNA clones. YACs are designated Y****; probes derived from YAC end clones are designated as Y****R (right arm) or L (left arm). The physical map shows only those markers hybridizing to overlapping YACs.

Line H455-9-1, with the same Ph1 marker deficiencies as CS1010 (but possesses the locus Xpsr150B, which CS1010 does not), exhibited higher levels of homeologous pairing than lines CS1010, CS1002, CS977 and Sears' ph1b/ph1b deletion line (Figure 3 and Table 2). As indicated previously, it is possible that the breakpoint of the deletion in H455-9-1 disrupts the Ph1 factor, causing an abnormal and more pronounced phenotype than complete loss of the gene or genes, or possesses nonreciprocal translocations.

Comparative physical mapping in rice:
Previous analyses indicated that the ph1b line carries a deletion of a 70-Mb segment containing the Ph1 locus on the long arm of chromosome 5B (GILL et al. 1993 ). Comparative mapping of markers derived from this segment revealed that they were in the same order as related markers on rice chromosome 9, suggesting a region of gene synteny (FOOTE et al. 1997 ). A partial YAC contig of this region of rice chromosome 9 was therefore established (FOOTE et al. 1997 ). Sizing of the rice YAC inserts indicated that the equivalent region to the 70-Mb wheat segment on rice chromosome 9 was at least 4 Mb in size. The analysis of the new deletion lines identified in the present study indicates that the position of the gene(s) controlling the euploid-like pairing lies in the interval defined by the loci Xrgc846 and Xpsr150A. Completion of the YAC contig for the region encompassing these markers would provide a source of more markers and sequence information with which to further characterize the new deletion lines and define the position of the Ph1 locus.

The rice YACs previously identified by C846, PSR150, and their flanking markers (FOOTE et al. 1997 ) formed the basis for further analysis in this study. Ends of the left and/or right arms from 10 YACs were cloned using an inverse PCR method (see MATERIALS AND METHODS) to generate probes for identifying any overlaps between the YACs. Comparison of the sequences from the resulting YAC ends revealed that 39% matched sequences in the GenBank/EMBL databases. Filters dotted with 47 YACs derived from the part of rice chromosome 9 with apparent gene synteny to the ph1b segment of chromosome 5B were probed with the YAC end clones. Left and right arms of three rice YACs (Y142, Y4603, and Y3960) proved useful for completing a contig from markers psr2B to S1057, including the region defined by markers C846 and psr150A (Figure 3). The right end arm of Y142 had a significant match (Blastn score, 202; probability, P = 4e - 50) with the rice heat shock protein 82 (accession no. Z11920) while its left arm had a significant match (Blastn score, 248; probability P = 1.9e - 10) with Arabidopsis ethylene-insensitive-like 3 (accession no. AF004215). The end sequences of the YACs Y4603 and Y3960 had no significant matches in the GenBank/EMBL databases. Southern blot analysis reveals that, in wheat and rice, the probes CDO786, PSR150, and CDO57 all detect a banding pattern similar to that detected by the rice heat-shock protein 82 probes C985 and R1562 (KURATA et al. 1994 ). Sequencing of the probe PSR150 revealed a significant match (Blastn score, 379; probability P = e - 103) with the rice heat-shock protein 82. Moreover, two heat-shock protein 82 loci are located on rice chromosome 9 in similar arrangement with other markers to the two corresponding heat-shock protein loci (Xpsr150A and Xpsr150B) on wheat chromosome group 5 (FOOTE et al. 1997 ). As described above, one of these loci (Xpsr150A) is located within the ph1b and ph1c deletions and the other locus (Xpsr150B) is outside both deletions. The interval between the rice heat-shock protein locus corresponding to Xpsr150A and C846 is no more than 400 kb in the rice genome (Figure 3). Shotgun sequencing of the entire region of the rice genome defined by these two markers is currently underway using subclones from rice bacterial artificial chromosomes. The sequence data will enable the deletion breakpoints (lines CS977, CS1002, CS1010, and H455-9-1) that occur between loci Xrgc846 and Xpsr150A to be characterized, thereby further delimiting the Ph1 locus.

In this study, the effect of the Ph1 locus on homologous/homeologous pairing as observed at metaphase I has been scored. However, the Ph1 locus has been shown to affect the association of homologues via centromere interactions during premeiotic floral development, the structure of the centromeres, the timing of telomere bouquet formation and telomere pairing, as well as recombination between homeologous chromosomes (LUO et al. 1996 ; ARAGON-ALCAIDE et al. 1997A ; MARTINEZ-PEREZ et al. 1999 ). Chromosome pairing of the new deletion lines of the Ph1 locus will be analyzed in detail using anther sectioning and confocal microscopy to determine whether all these effects can be defined to the same region or different regions lying between the Xrgc846 and Xpsr150A loci. Combining this information with deletion breakpoint analysis of the Ph1 mutants will reveal whether the Ph1 locus is a single gene, a multigene family, or a complex of linked genes affecting different premeiotic and meiotic processes.

	ACKNOWLEDGMENTS

We thank Mark Sorrells and Steven Tanksley (Cornell University, Ithaca, NY), Bikram Gill (Kansas State University, Manhattan, KS), and Takuji Sasaki (Rice Genome Program, Japan) for providing probes and YACs for this study.

Manuscript received April 30, 1999; Accepted for publication August 12, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

ARAGON-ALCAIDE, L., S. M. READER, P. J. SHAW, A. F. BEVEN, and T. E. MILLER et al., 1997a  Association of homologous chromosomes during floral development. Curr. Biol. 7:905-908[Medline].

ARAGON-ALCAIDE, L., S. READER, T. MILLER, and G. MOORE, 1997b  Centromeric behaviour in wheat with high and low homeologous chromosomal pairing. Chromosoma 106:327-333[Medline].

BENNETT, M. D. and J. B. SMITH, 1991  Nuclear DNA amounts in angiosperms. Philos. Trans. R. Soc. Lond. B Biol. Sci. 334:309-345.

BURKE, D. T., G. F. CARLE, and M. V. OLSON, 1987  Cloning of large segments of exogenous DNA into yeast by means of artificial chromosome vectors. Science 236:806-812[Abstract/Free Full Text].

CEOLONI, C. and P. DONINI, 1993  Combining mutations for the two homeologous pairing suppressor genes Ph1 and Ph2 in common wheat and in hybrids with the Triticeae. Genome 36:377-386.

DUBCOVSKY, J., M. C. LUO, and J. DVORAK, 1995  Differentiation between homeologous chromosomes 1A of wheat and 1A (M) of Triticum monococcum and its recognition by the wheat Ph1 locus. Proc. Natl. Acad. Sci. USA 92:6645-6647[Abstract/Free Full Text].

DUNFORD, R. P., N. KURATA, D. A. LAURIE, T. A. MONEY, and Y. MINOBE et al., 1995  Conservation of fine-scale DNA marker order in the genomes of rice and the Triticeae. Nucleic Acids Res. 23:2724-2728[Abstract/Free Full Text].

FOOTE, T., M. ROBERTS, N. KURATA, T. SASAKI, and G. MOORE, 1997  Detailed comparative mapping of cereal chromosome regions corresponding to the Ph1 locus in wheat. Genetics 147:801-807[Abstract].

GALE, M. D., M. D. ATKINSON, C. N. CHINOY, R. L. HARCOURT, J. JIA et al., 1995 Genetic maps of hexaploid wheat, pp. 29–40 in Proceedings of the 8th International Wheat Genetics Symposium, edited by Z. S. LI and Z. Y. XIN. China Agricultural Scientech Press, Beijing.

GILL, K. S. and B. S. GILL, 1996  A PCR-based screening assay of Ph1, the chromosome pairing regulator gene of wheat. Crop Sci. 36:719-722[Abstract/Free Full Text].

GILL, K. S., B. S. GILL, T. R. ENDO, and Y. MUKAI, 1993  Fine physical mapping of Ph1, a chromosome pairing regulator gene in polyploid wheat. Genetics 134:1231-1235[Abstract].

GILL, K. S., B. S. GILL, T. R. ENDO, and E. V. BOYKO, 1996  Identification and high-density mapping of gene-rich regions in chromosome group 5 in wheat. Genetics 143:1001-1012[Abstract].

GILLIES, C. B., 1987  The effect of the Ph gene alleles on synaptonemal complex formation in Triticum aestivum x Triticum kotshyi hybrids. Theor. Appl. Genet. 74:430-438.

GIORGI, B., 1978  A homeologous pairing mutant isolated in Triticum durum cv. Cappelli. Mutat. Breed. Newsl. 11:4-5.

JI, L.-H. and P. LANGRIDGE, 1994  An early meiotic cDNA from wheat. Mol. Gen. Genet. 243:17-23[Medline].

KURATA, N., Y. NAGAMURA, K. YAMAMOTO, Y. HARUSHIMA, and N. SUE et al., 1994  A 300 kilobase interval genetic map of rice including 883 expressed sequences. Nat. Genet. 8:365-372[Medline].

LUO, M. C., J. DUBCOVSKY, and J. DVORAK, 1996  Recognition of homeology by wheat Ph1 locus. Genetics 144:1195-1203[Abstract].

MARTINEZ-PEREZ, E., P. SHAW, S. READER, L. ARAGON-ALCAIDE, and T. MILLER et al., 1999  Homologous chromosome pairing in wheat. J. Cell Sci. 112:1761-1769[Abstract].

MELLO-SAMPAYO, T., 1972  Compensated monosomic 5B-trisomic 5A plants in tetraploid wheat. Can. J. Genet. Cytol. 14:463-475.

MOORE, G., M. D. GALE, N. KURATA, and R. B. FLAVELL, 1993  Molecular analysis of small grain cereal genomes: current status and prospects. Nat. Biotech 11:584-589.

OKAMOTO, M., 1957  Asynaptic effect of chromosome V. Wheat Inf. Serv. 5:6.

QU, L.-J., T. N. FOOTE, M. A. ROBERTS, T. A. MONEY, and L. ARAGON-ALCAIDE et al., 1998  A simple PCR-based method for scoring the ph1b deletion in wheat. Theor. Appl. Genet. 96:371-375.

RILEY, R. and V. CHAPMAN, 1958  Genetic control of the cytologically diploid behaviour of hexaploid wheat. Nature 182:713-715.

RILEY, R. and C. KEMPANNA, 1963  The homeologous nature of the non-homologous meiotic pairing in Triticum aestivum deficient for chromosome V (5B). Heredity 18:287-306.

SCHWARTZ, D. C. and C. R. CANTOR, 1984  Separation of yeast chromosome-sized DNAs by pulse-field gradient gel electrophoresis. Cell 37:67-75[Medline].

SEARS, E. R., 1952  Homeologous chromosomes in Triticum aestivum.. Genetics 37:624.

SEARS, E. R., 1966 Nullisomic-tetrasomic combinations in hexaploid wheat, pp. 22–45 in Chromosome Manipulations and Plant Genetics, edited by R. RILEY and K. R. LEWIS. Oliver and Boyd, Edinburgh.

SEARS, E. R., 1977  An induced mutant with homeologous pairing in common wheat. Can. J. Genet. Cytol. 19:585-593.

SEGAL, G., B. LIU, J. M. VEGA, S. ABBO, and M. RODOVA et al., 1997  Identification of a chromosome-specific probe that maps within the Ph1 deletions in common and durum wheat. Theor. Appl. Genet. 94:968-970.

SILVERMAN, G. A., 1993  Isolating vector-insert junctions from yeast artificial chromosomes. PCR Methods Appl. 3:141-150.

SNAPE, J. W., L. A. SITCH, E. SIMPSON, and B. B. PARKER, 1988  Tests for the presence of gametoclonal variation in barley and wheat doubled haploids produced using the Hordeum bulbosum system. Theor. Appl. Genet. 75:509-513.

VEGA, J. M. and M. FELDMAN, 1998  Effect of the pairing gene Ph1 and premeiotic colchicine treatment on intra and interchromosome pairing of isochromosomes in common wheat. Genetics 150:1199-1208[Abstract/Free Full Text].

VOS, P., R. HOGERS, M. BLEEKER, M. REIJANS, and T. VAN DE LEE et al., 1995  AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23:4407-4414[Abstract/Free Full Text].

WALL, A. M., R. RILEY, and V. CHAPMAN, 1971  Wheat mutants permitting homeologous meiotic chromosome pairing. Genet. Res. 18:311-328.

This article has been cited by other articles:

				N. Al-Kaff, E. Knight, I. Bertin, T. Foote, N. Hart, S. Griffiths, and G. Moore Detailed Dissection of the Chromosomal Region Containing the Ph1 Locus in Wheat Triticum aestivum: With Deletion Mutants and Expression Profiling Ann. Bot., April 1, 2008; 101(6): 863 - 872. [Abstract] [Full Text] [PDF]

				A. M. Linkiewicz, L. L. Qi, B. S. Gill, A. Ratnasiri, B. Echalier, S. Chao, G. R. Lazo, D. D. Hummel, O. D. Anderson, E. D. Akhunov, et al. A 2500-Locus Bin Map of Wheat Homoeologous Group 5 Provides Insights on Gene Distribution and Colinearity With Rice Genetics, October 1, 2004; 168(2): 665 - 676. [Abstract] [Full Text] [PDF]

				K. S. Caldwell, P. Langridge, and W. Powell Comparative Sequence Analysis of the Region Harboring the Hardness Locus in Barley and Its Colinear Region in Rice Plant Physiology, October 1, 2004; 136(2): 3177 - 3190. [Abstract] [Full Text] [PDF]

				E. Jenczewski, F. Eber, A. Grimaud, S. Huet, M. O. Lucas, H. Monod, and A. M. Chevre PrBn, a Major Gene Controlling Homeologous Pairing in Oilseed Rape (Brassica napus) Haploids Genetics, June 1, 2003; 164(2): 645 - 653. [Abstract] [Full Text] [PDF]

				C. FEUILLET and B. KELLER Comparative Genomics in the Grass Family: Molecular Characterization of Grass Genome Structure and Evolution Ann. Bot., January 1, 2002; 89(1): 3 - 10. [Abstract] [Full Text] [PDF]

				K. M. Devos and M. D. Gale Genome Relationships: The Grass Model in Current Research PLANT CELL, May 1, 2000; 12(5): 637 - 646. [Abstract] [Full Text]

				E. Martinez-Perez, P. J. Shaw, and G. Moore Polyploidy Induces Centromere Association J. Cell Biol., January 24, 2000; 148(2): 233 - 238. [Abstract] [Full Text] [PDF]