letters - John Innes Centre

Vol 439|9 February 2006|doi:10.1038/nature04434

LETTERS Molecular characterization of Ph1 as a major chromosome pairing locus in polyploid wheat Simon Griffiths1, Rebecca Sharp1, Tracie N. Foote1, Isabelle Bertin1, Michael Wanous2, Steve Reader1, Isabelle Colas1 & Graham Moore1

The foundation of western civilization owes much to the high fertility of bread wheat, which results from the stability of its polyploid genome. Despite possessing multiple sets of related chromosomes, hexaploid (bread) and tetraploid (pasta) wheat both behave as diploids at meiosis. Correct pairing of homologous chromosomes is controlled by the Ph1 locus1. In wheat hybrids, Ph1 prevents pairing between related chromosomes2. Lack of Ph1 activity in diploid relatives of wheat suggests that Ph1 arose on polyploidization3. Absence of phenotypic variation, apart from dosage effects, and the failure of ethylmethane sulphonate treatment to yield mutants, indicates that Ph1 has a complex structure4,5. Here we have localized Ph1 to a 2.5-megabase interstitial region of wheat chromosome 5B containing a structure consisting of a segment of subtelomeric heterochromatin that inserted into a cluster of cdc2-related genes after polyploidization. The correlation of the presence of this structure with Ph1 activity in related species, and the involvement of heterochromatin with Ph1 (ref. 6) and cdc2 genes with meiosis, makes the structure a good candidate for the Ph1 locus. The Ph1 locus lies on chromosome 5B of hexaploid (AABBDD) and tetraploid (AABB) wheats, originally delimited by two overlapping deletions (ph1b and ph1c, respectively)7,8. Gene order on chromosomes 5A, 5D and on the chromosomes of diploid relatives of polyploid wheat is very similar to that on chromosome 5B, but Ph1 activity is unique to 5B. This suggests that Ph1 arose through a structural change on 5B after polyploidization. Evidence indicating that this change was complex includes the failure of ethylmethane sulphonate (EMS) treatment to yield Ph1 mutants4, and the fact that supernumerary chromosomes (heterochromatin) can compensate for the absence of Ph1 (ref. 6). Moreover, duplication of the whole locus gives a stronger Ph1 phenotype5, which would have been preferentially selected during wheat evolution if Ph1 was a single gene effect. Ph1 has pronounced effects both on the condensation of heterochromatin in hybrids of wheat with other Triticeae species, and on the condensation of euchromatin in wheat9,10. Thus, the Ph1 locus is likely to comprise a multigene family, heterochromatin or both, and affect condensation. Mapping Ph1 has been complicated by the absence of allelic variation, making it only possible to score for its presence or absence. This led us to propose a two-part strategy for the molecular characterization of the Ph1 locus11–13. The first part revealed the gene content of the wheat Ph1 region using the conservation of gene order between the highly repetitive, 16,000-megabase (Mb) hexaploid wheat genome and the smaller genomes of rice and Brachypodium sylvaticum (460 Mb and 470 Mb, respectively), thus providing markers with which to saturate the wheat region. The second part of the strategy used deletion lines to dissect physically the Ph1 locus. The original (ph1b) deletion in hexaploid wheat is some 70 Mb in 1

size, with the equivalent region in rice being approximately 4 Mb of chromosome 9 (ref. 14). Previously, fast-neutron irradiation of hexaploid wheat produced five deletions overlapping the ph1b deletion15. Using markers derived from the sequence of rice chromosome 9, we show that the breakpoints for these five wheat Ph1 deletions cluster within a 250-kilobase (kb) section of the 4-Mb equivalent region of rice chromosome 9. These breakpoints delimit Ph1 to a location equivalent to a 140-kb section of the rice chromosome (Supplementary Fig. 1). The nonrandom clustering of the breakpoints suggests that the whole segment is in a different conformation to its flanking regions. All markers derived from the 140-kb rice region were mapped against the wheat Ph1 deletion lines to confirm their location. Half of the rice markers failed to yield clear hybridizing patterns on wheat Southern blots. Therefore, the rice markers were used to screen a B. sylvaticum bacterial artificial chromosome (BAC) library to identify homologous markers from a closer relative of wheat16. Screening this library also identified additional markers with which to analyse the wheat region. The rice markers identified two regions in Brachypodium: region 1 is co-linear with the Ph1 region on rice chromosome 9, and region 2 with rice chromosome 8 (Fig. 1 and Supplementary Fig. 2). This is consistent with known internal duplications in the rice genome17. In contrast to the rice markers, all the Brachypodium markers gave clear hybridizing patterns when mapped against the deletion lines. Forty-two markers, mostly from Brachypodium region 1 genes, were used to screen hexaploid and tetraploid wheat BAC libraries containing 750,000 and 500,000 clones, respectively18,19. Southern blots, prepared from restriction digests of 2,000 positive BACs and a full set of wheat nullisomic/tetrasomic lines and deletion lines, were hybridized with the same probes. The hexaploid wheat BAC library is made from the same variety of wheat as the nullisomic/tetrasomic lines, so BACs could be unambiguously assigned to wheat chromosomes by comparison of hybridizing fragment size. This process eliminated gene family members from other genomic locations, giving 400 BACs from the Ph1 region and the homologous regions of chromosomes 5A and 5D. Overlaps were confirmed by crosshybridization of end clones and comparison of restriction profiles. A minimum tiling path from these contigs was selected for shotgun sequencing. Thirty-six genes were identified, 25 of which were co-linear with their orthologues on rice chromosome 9 and/or Brachypodium region 1. The generation of markers from this sequence enabled some gaps in the contigs to be closed through chromosome walking. After all possibilities for further steps were stopped by large tracts of retroelements, we had assembled five BAC contigs for the region, with approximate sizes of 500 kb, 300 kb, 1.2 Mb, 1 Mb and 200 kb (Fig. 1). More sequencing was undertaken to reveal the gene content of the areas we had walked into, making a

John Innes Centre, Colney, Norwich NR4 7UH, UK. 2Augustana College, Sioux Falls, South Dakota 57197, USA.

© 2006 Nature Publishing Group

749

LETTERS

750

NATURE|Vol 439|9 February 2006


LETTERS


R Figure 1 | BAC tiling path and annotated gene content of the Ph1 region and its equivalent region in rice chromosomes 9 and 8 and Brachypodium regions 1 and 2. Pink vertical lines represent part of rice chromosome 9; orange vertical lines represent Brachypodium chromosome region 1. Co-linearity of rice and Brachypodium markers with markers on part of wheat chromosomes 5B, 5A and 5D is shown with horizontal and diagonal black lines; red dashed lines show the location of fast neutron-induced breakpoints on chromosome 5B. Blue bars represent hexaploid wheat BACs; green bars show tetraploid wheat BACs (T. durum). Pink boxes indicate sequenced BACs. Gene and BAC names are indicated in black, whereas other

total of 33 Triticum aestivum BACs sequenced within this region. So far, this represents the largest study of its kind in wheat. All the genes within the Ph1 5B deletion region (as defined by deletion line breakpoints) were assessed for their expression in different tissues (leaf, spike and seed) and, where expression was detected, none of them showed differences in the presence and absence of Ph1 (Supplementary Table 1 and Supplementary Fig. 3). However, on the basis of their sequence homology to genes of known function, the cdc2 genes are the best candidates for Ph1 function. cdc2-related genes affect chromosome condensation (http://www. ihop-net.org/UniPub/iHOP/). This is the only multigene cluster in the region, and at least one of its members (cdc2-4) is 5B specific. It is striking that this cluster of cdc2 genes is interrupted by a large segment of chromosome 5B (yellow boxed area in Fig. 1) that does not have homologous equivalents on 5A and 5D. Southern analysis shows that the genes in this additional segment are all part of a subtelomeric insertion from chromosome 3AL. It also contains two tandem arrays of a 2.3-kb repeating unit, showing no sequence homology to previously recorded repetitive elements. Our analysis indicates that the tandem array between vmp1 and hyp3 genes is at least 87 kb in size, and between cdc2-3 and the mads pseudogene is at least 20 kb. Some of the BACs carrying this repeat were unstable. Southern analysis reveals that the tandem arrays are not present on wheat chromosomes 5A or 5D, but there are a number of similar repeats on chromosome 3B, also occurring as arrays (Supplementary Fig. 4). These data indicate that an insertion from the subtelomeric region of 3AL has occurred in this part of the newly defined Ph1 region. The novel tandem repeat together with cdc2 genes fulfils all our criteria as a Ph1 candidate structure. In situ hybridization of the tandem repeat on mitotic chromosome spreads from hexaploid wheat root meristems reveals two strongly hybridizing interstitial and two weaker hybridizing subtelomeric locations (Supplementary Fig. 5). The two interstitial sites correspond to the tandem arrays within the Ph1 locus on chromosome 5B, whereas the subtelomeric sites on the long arms represent the 3B arrays. In the absence of Ph1, only the subtelomeric sites can be visualized in meiocytes. The availability of markers for the 3BL telomeric regions enables us to show that these regions pair in all meiocytes in the presence and absence of Ph1 in hexaploid wheat (Supplementary Fig. 5a–f). Thus, telomeres pair correctly independently of Ph1 activity. If the segment containing the tandem repeats and the cdc2-4 (5B specific) pseudogene is responsible for the Ph1 effect, it should be present in relatives of hexaploid wheat possessing Ph1. Chromosome 5B of tetraploid wheat also possesses a Ph1 locus defined by the ph1c deletion8. Southern analysis shows that the additional segment is also within the ph1c deletion (Supplementary Fig. 4). There is a high level of conservation between the restriction profiles of tetraploid wheat BACs carrying the tandem repeat and those derived from the corresponding hexaploid wheat 5B region (Supplementary Fig. 6). Although the S genome of Aegilops speltoides (SS) is the progenitor of the B genome of Triticum dicoccoides (AABB) and the G genomes of Triticum timopheevi (AAGG) and Triticum araraticum (AAGG), it does not possess Ph1 activity, but the B and G genomes of these tetraploids do. Moreover, T. araraticum and T. timopheevi result from

marker names are in red. Black circles indicate the presence of marker from sequence data or by Southern analysis; blue circles show presence of marker by colony hybridization alone; no circle indicates not tested. The yellow box represents a subtelomeric insertion from wheat chromosome 3A. Red diamonds represent tandem repeats. The high level of homology between cdc2-type genes on different BACs means that homologues cannot be distinguished from paralogues; they are therefore connected by black dashed lines. The hashed box indicates overlap between BACs based on fingerprint only. Centromere and telomere directions are indicated with arrows. Scale bar, 100 kb. All wheat BACs are drawn to scale.

a later polyploidization event than their B genome relatives. Despite their separate polyploidization history, polymerase chain reaction (PCR) assays that are 5B specific for the cdc2-4 pseudogene and 2.3-kb repeat (data not shown) in T. aestivum amplified products from all of the T. dicoccoides, T. timopheevi and T. araraticum accessions tested, but failed to amplify products from A. speltoides and its relative Aegilops longissima (S1S1) (Supplementary Fig. 7). Lack of the cdc2-4 pseudogene in the B genome progenitor (the S genome) indicates that it is part of the polyploidization event. The availability of these markers will enable us to understand how Ph1 arose in these species. Fertile hybrids have been generated between accessions of T. timopheevi and T. dicoccoides, so genetic exchange is possible20. Tight coupling of the cdc2/heterochromatin structure with Ph1 activity leads to the conclusion that they are essential components of this locus. The presence of a multigene complex containing heterochromatin, and the potential lethality of cdc2 mutants, means that the dissection of the Ph1 locus by a transgenic approach is difficult. However, the use of drugs such as okadaic acid, which alter cdc2 activity, and a multiplex PCR assay of markers across the locus (Supplementary Fig. 8) to identify deletions of the segment, will allow dissection of Ph1 function and the generation of Ph1 variation essential for breeding. METHODS Plant material. The following hexaploid wheat (Triticum aestivum; 2n ¼ 6 £ ¼ 42; genome AABBDD) lines were used: euploid Chinese Spring, Chinese Spring mutant line ph1b, Chinese Spring nullisomic/tetrasomic lines N5AT5B, N5BT5D and N5DT5A, Chinese Spring 3BL deletion line 4525L7 (ref. 21), Highbury (CS 5B) Ph1 mutant lines 977, 1002, 1010, 3793 and 455-9-1. These genetic stocks have been previously described15. Tetraploid wheat lines used were: Triticum durum (2n ¼ 4 £ ¼ 28 AABB), Cappelli and its derivative deletion line ph1c, T. timopheevi (2n ¼ 4 £ ¼ 28 AAGG), T. araraticum (2n ¼ 4 £ ¼ 28 AAGG) and T. dicoccoides (2n ¼ 4 £ ¼ 28 AABB). Specific T. araraticum and T. dicoccoides accessions used are shown in Supplementary Fig. 7. Diploid wheat species used were: Triticum monococcum (2n ¼ 2 £ ¼ 14 AA), A. longissima (2n ¼ 2 £ ¼ 14 SlSl), A. speltoides (2n ¼ 2 £ ¼ 14 SS) and Triticum tauschii (2n ¼ 2 £ ¼ 14 DD). BAC libraries and their analysis. The hybridization and analysis of the wheat and Brachypodium BAC libraries have been previously described16,18. Sequencing and annotation. Sequencing was performed by GATC Biotech Ltd using standard techniques. Contigs were assembled using PhredPhrap22,23 and viewed using Consed24. Gene models were based on ab initio Genemark25 predictions, sequence conservation between wheat, rice and Brachypodium sylvaticum, and homology to known peptides detected by BLAST26. Intron– exon boundary predictions were refined using Geneseqer27. Sequences were submitted to EMBL/GenBank/DDJB in HTG format with an average of 35 contigs per BAC. Accession numbers, gene content and sequence coverage are shown in Supplementary Table 2. Fluorescence in situ hybridization. Root tips and anthers were collected as previously described9,10. Meiocytes were extruded from individual anthers and spread on slides28. Primers were designed from the 5B Ph1 region tandem repeat and a 2-kb product was amplified by PCR using BAC 218J13 DNA as template. Probes were prepared and in situ hybridization was carried out as previously described9,10. Fluorescence microscopy and image processing. Cells were visualized using a Nikon Eclipse E600 fluorescence microscope connected to a Hamamatsu CCD camera, and digital images were captured using MetaMorph (Universal Imaging


751

LETTERS


Corp.) software. Images were processed using the AutoDeblur (AutoQuant Imaging) and ImageJ (NIH) programs. Analysis of expression by RT–PCR. Total RNA isolation, reverse transcription and measurement of the abundance of products using a Typhoon fluorimeter have all been previously described29. Tissues included wheat leaves as described, immature inflorescences taken at Feekes scale 6–7 and seed imbibed for 24 h (Chinese Spring and ph1b types). PCR was performed on the cDNA using intron-spanning, gene-specific primer pairs. PCR products were cloned using a p-GEM T easy vector kit (Promega) according to the manufacturer’s instructions. Received 29 September; accepted 11 November 2005. 1. 2. 3.

4.

5.

6.

7. 8. 9.

10.

11. 12. 13. 14.

15.

752

Riley, R. & Chapman, V. Genetic control of cytologically diploid behaviour of hexaploid wheat. Nature 182, 713–-715 (1958). Riley, R., Chapman, V. & Kimber, G. Genetic control of chromosome pairing in intergeneric hybrids with wheat. Nature 183, 1244–-1246 (1959). Chapman, V. & Riley, R. Homoeologous meiotic chromosome pairing in Triticum aestivum in which chromosome 5B is replaced by an alien homoeologue. Nature 226, 376–-377 (1970). Wall, A. M., Riley, R. & Gale, M. D. The position of a locus on chromosome 5B of Triticum aestivum affecting homoeologous meiotic pairing. Genet. Res. Camb. 18, 329–-339 (1971). Martinez, M., Naranjo, T., Cuadrado, C. & Romero, C. The synaptic behaviour of Triticum turgidum with variable doses of the Ph1 locus. Theor. Appl. Genet. 102, 751–-758 (2001). Dover, G. A. & Riley, R. Prevention of pairing of homoeologous meiotic chromosomes of wheat by an activity of supernumerary chromosomes of Aegilops. Nature 240, 159–-161 (1972). Sears, E. R. An induced mutant with homoeologous pairing in common wheat. Can. J. Genet. Cytol. 19, 585–-593 (1977). Giorgi, B. A homoeologous pairing mutant isolated in Triticum durum cv Cappelli. Mutat. Breed. Newslett. 11, 4–-5 (1978). Martinez-Perez, E., Shaw, P. & Moore, G. The Ph1 locus is needed to ensure specific somatic and meiotic centromere association. Nature 411, 204–-207 (2001). Prieto, P., Shaw, P. & Moore, G. Homologue recognition during meiosis is associated with a change in chromatin conformation. Nature Cell Biol. 6, 906–-908 (2004). Moore, G., Gale, M. D., Kurata, N. & Flavell, R. B. Molecular analysis of small grain cereal genomes. Nature Biotechnol. 11, 584–-589 (1993). Kurata, N. et al. Conservation of genome structure between rice and wheat. Nature Biotechnol. 12, 276–-278 (1994). Sasaki, T., Matsumoto, T., Antonio, B. A. & Nagamura, Y. Review: from mapping to sequencing and beyond. Plant Cell Physiol. 46, 3–-12 (2005). Foote, T., Roberts, M., Kurata, N., Sasaki, T. & Moore, G. Detailed comparative mapping of cereal chromosome regions corresponding to the Ph1 locus in wheat. Genetics 147, 801–-807 (1997). Roberts, M. A. et al. Induction and characterisation of Ph1 wheat mutants. Genetics 153, 1909–-1918 (1999).

16. Foote, T. N., Griffiths, S., Allouis, S. & Moore, G. Construction and analysis of a BAC library in the grass Brachypodium sylvaticum: its use as a tool to bridge the gap between rice and wheat in elucidating gene content. Funct. Integr. Genomics 4, 26–-33 (2004). 17. Paterson, A. H., Bowers, J. E. & Chapman, B. A. Ancient polyploidization predating divergence of the cereals, and its consequences for comparative genomics. Proc. Natl Acad. Sci. USA 101, 9903–-9908 (2004). 18. Allouis, S. et al. Construction and characterisation of a hexaploid wheat (Triticum aestivum L.) BAC library from the reference germplasm ‘Chinese Spring’. Cereal Res. Commun. 31, 331–-338 (2003). 19. Cenci, A. et al. Construction and characterisation of a half million clone BAC library of durum wheat (Triticum turgidum ssp. durum). Theor. Appl. Genet. 107, 931–-939 (2003). 20. Rao, P. S. & Smith, E. L. Studies with Israeli and Turkish accessions of Triticum L.emend. var. dicoccoides (KORN) BOWDEN. Wheat Inform. Serv. 26, 6–-7 (1968). 21. Endo, T. R. & Gill, B. S. The deletion stocks of common wheat. J. Hered. 87, 295–-307 (1996). 22. Ewing, B., Hillier, L., Wendl, M. & Green, P. Basecalling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res. 8, 175–-185 (1998). 23. Ewing, B. & Green, P. Basecalling of automated sequencer traces using phred. II. Error probabilities. Genome Res. 8, 186–-194 (1998). 24. Gordon, D., Abajian, C. & Green, P. Consed: A graphical tool for sequence finishing. Genome Res. 8, 195–-202 (1998). 25. Borodovsky, M. & McIninch, J. GeneMark: Parallel gene recognition for both DNA strands. Comput. Chem. 17, 123–-133 (1993). 26. Altschul, S. F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–-3402 (1997). 27. Schlueter, S. D., Dong, Q. F. & Brendel, V. GeneSeqer@PlantGDB: gene structure prediction in plant genomes. Nucleic Acids Res. 31, 3597–-3600 (2003). 28. Bass, H. W. et al. Evidence for the coincident initiation of homolog pairing and synapsis during the telomere-clustering (bouquet) stage of meiotic prophase. J. Cell Sci. 113, 1033–-1042 (2000). 29. Dunford, R. P., Griffiths, S., Christodoulou, V. & Laurie, D. A. Characterisation of a barley (Hordeum vulgare L.) homologue of the Arabidopsis flowering time regulator Gigantea. Theor. Appl. Genet. 110, 925–-931 (2005).

Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Acknowledgements This work was funded by the Biotechnology and Biological Sciences Research Council of the UK. Sabbatical support for M.W. was provided by a NIH grant from the NCRR INBRE program. We would like to thank DuPont/ Pioneer Hi-Bred International and P. Prieto for their assistance. Author Information Sequences were submitted to EMBL/GenBank/DDJB and accession numbers are shown in Supplementary Table 2. Reprints and permissions information is available at npg.nature.com/reprintsandpermissions. The authors declare no competing financial interests. Correspondence and requests for materials should be addressed to G.M. ([email protected]).
