39. Assignment of RFLP linkage groups to rice chromosomes using primary trisomics

Z. H. Yu¹, G. S. KHUSH² and S. D. TANKSLEY¹

1) Department of Plant Breeding and Biometry, Cornell University, 252 Emerson Hall, Ithaca, NY 14853, U.S.A.

2) Division of Plant Breeding, Genetics and Biochemistry, International Rice Research Institute, P. O. Box 933, 1099 Manila, Philippines

Molecular linkage maps directly using DNA sequences would provide many advantages because such DNA markers normally are codominant, free of epistatic interactions, and have no deleterious effect on phenotype. The number of DNA markers on a linkage map can be large enough for plant breeding and genetics purposes such as gene tagging and cloning (Tanksley et al. 1989). However, one of the important steps in constructing a molecular map is to associate random linkage groups of DNA clones with respective chromosomes in a species. Once such linkage groups have been tentatively identified based on F`2` or BC`1` segregation analysis of restriction fragment length polymorphism (RFLP) markers, they can be assigned to the chromosomes by dosage analysis using primary trisomics (Young et al. 1987). In rice, two primary trisomic series were employed for associating classical linkage groups of rice with respective chromosomes. One series of primary trisomics was developed in an indica variety IR36 (Khush et al. 1984). The extra chromosomes of the IR36 trisomics were identified at pachytene stage of meiosis. The other series of primary trisomics was developed in a japonica variety Nipponbare (Iwata and Omura 1984). The extra chromosomes of this trisomic series were identified at mitotic prophase. Until recently, however, the chromosome numbering of these two trisomic series did not agree with each other. Recently agreement has been reached to unify the chromosome numbering system (Khush 1990). The newly unified system is based on the chromosome length at pachytene stage of meiosis (Shastry et al. 1960).

In this study, IR36 trisomics were used to assign the RFLP linkage groups to rice chromosomes. The origin, morphology and cytology of these trisomics were described previously (Khush et al. 1984). A complete set of inbred primary trisomic plants was grown at IRRI and the lyophilized leaf tissues shipped to Cornell for DNA extraction. Because DNA from all primary trisomics was not available, hybrid trisomics were also employed to assign the linkage groups to rice chromosomes. These hybrid trisomics were developed at IRRI by crossing IR36 primary trisomics to a japonica cultivar, Ma Hae. Ten hybrid trisomics (missing triplo 1 and triplo 3) were grown at IRRI and the lyophilized leaf tissues shipped to Cornell for DNA analysis.

Southern blots for trisomic analysis were prepared similarly to those for the mapping parents and F`2`s, with special care taken to calculate the concentration of DNA for each triplo plant so that exactly 5ug DNA would be aliquoted to each lane in a gel. Further, large digests were prepared, and several filters were made for each digestion batch to assure uniformity among filters. All triplo DNAs were digested with a single restriction enzyme DraI except for those hybrid trisomics used for internal comparisons among polymorphic alleles. In the latter, any enzymes giving polymorphism between IR36 and Ma Hae were used to digest hybrid triplo DNAs. A few filters were made of PstI-digested primary triplos as the first rice genomic library was generated with the PstI (McCouch et al. 1988). The autoradiograms were exposed in a -80 deg C freezer to ensure the proper intensity of the signal.

Dosage comparison of the exposed signal was performed by visual inspection, without densitometric scanning. A high correlation between densitometry and visual inspection was observed in a previous study (Young et al. 1987). Because most rice triplo lines are available, DNA clones could be assigned to specific chromosomes by observing increased hybridization signal at about 1.5: 1 ratio of a given clone to DNA to one of the trisomic series. Several other DNA clones of different sized and known chromosome association were included in single primary trisomic filters during hybridization to give "within-lane" comparisons (Tanksley et al. 1987). The second approach of the "within-lane" comparison was to overlay several autoradiograms of a single trisomic filter probed with different DNA clones of known chromosome association. In few cases, "internal comparison" was carried out between two alleles of a polymorphic clone probed on a hybrid trisomic filter. In such cases, an increased hybridization signal at a ratio of 2: 1 was observed (Fig. 1).

Fig. 1. An autoradiogram showing the dosage effect in IR36 hybrid trisomics. Ten hybrid trisomics [missing triplo 1 and 3, order converted to the new chromosome numbering system (Khush 1990)], along with Indica and Javanica diploid parents (left), were digested with HindIII, a restriction enzyme detecting polymorphism between the parents. A polymorphic DNA clone, RG182, was labeled to this hybrid trisomic filter. The increased signals (see arrows) were detected from Indica alleles in triplo 5 for both copies of RG182. Also note that triplo 10 might undergo a mutation (see triangles). The increased molecular weight (about 100 bp) is probably due to either the seed mixture of a triplo 10 line or the insertion/deletion of a DNA segment.

A total of seventy-three rice genomic (RG) clones, plus the R45S clone, were analyzed on the trisomic filters. Among them, seven clones were probed only on primary trisomic filters, thirty-eight clones were probed only on hybrid trisomic filters, and the rest of twenty-nine clones were probed on both primary and hybrid trisomic filters. Figure 1 shows ten hybrid lines digested with HindIII, a restriction enzyme giving polymorphism between two parents when probed with a rice genomic clone RG182. After probing this Southern blot with RG182, one could easily observe the increased signal dosage for the Indica fragments in DNA from the triplo 5 line. Both loci hybridized with RG182 showed consistent results. Hybrid trisomics automatically provided convenient internal dosage controls where allelic differences between the Indica and Javanica parents were apparent. Sometimes novel bands were observed during the trisomic analysis. The increased molecular weight (about 100 bp) in an Indica allele of triplo 10 probed with RG182 might be due to either seed mixture of the triplo 10 line or possible insertion of a DNA fragment (such as a transposable element) (Fig. 1). Using trisomic dosage analysis, fifty-three DNA clones were unambiguously assigned to rice chromosomes (Table 1). Among them, six were confirmed by the primary

Table 1. Chromosome assignment of RG clones by dosage analysis in IR36
         trisomics^a
_______________________________________________________________________________
  1             2           3           4          5           6
_______________________________________________________________________________
RG109^c        RG139^c      RG100^b     RG169^bc    RG119^c      RG123^c  
RG147^b        RG144^c      RG104^c     RG208^bc    RG182^bc     RG138^c  
RG233^c        RG151^b      RG127^b     RG214^c     RG207^c      RG213^b  
RG246^bc       RG152^c      RG166^bc    RG329^c     RG229^c  
RG317^bc       RG157^c      RG227^c                RG346^c  
RG345^bc       RG171^c      RG391^bc  
RG350^bc       RG365^c
_______________________________________________________________________________
  7            8           9           10         11         12
_______________________________________________________________________________
RG30^bc        RG108^c      R-45S^c      RG134^c     RG167^bc     RG181^c
RG128^bc       RG136^c      RG141^bc                RG303^bc     RG190^c
RG146A^c       RG150^c      RG148^c                 RG304^c      RG241^bc
RG165^bc                   RG386^bc                RG353^c      RG341^b
RG351^bc
_______________________________________________________________________________
a   Out of seventy-four DNA clones analyzed on the IR36 trisomic filters,
    fifty-three clones were assigned unambigously to rice chromosomes.
    Thirty-one clones (in bold) were transfered by linkage analysis to a
    recent rice framework map based on a backcross population derived from an
    interspecific cross (Tanksley et al.  1992)
b   Chromosome number assigned by IR36 primary trisomics.
c   Chromosome number assigned by IR36 hybrid trisomics.

trisomic analysis, twenty-seven were confirmed by the hybrid trisomic analysis, and the rest of twenty were confirmed by both primary and hybrid trisomics. For those DNA clones probed on both primary and hybrid trisomic filters, the results were con sistent in most cases. Thus, every one of twelve rice chromosomes has at least one RFLP marker confirmed with trisomic tests.

Fig. 2 shows a recent RFLP framework map of the rice genome based on a backcross population derived from an interspecific cross (BS125 X WL02) (Tanksley et al. 1992). It is now possible to place on this map via linkage analysis thirty-one trisomic assigned loci (in enlarged bold), distributing on every one of twelve rice chromosomes, to assure the correct chromosome assignment. Such associations of linkage groups with rice chromosomes provide a basic tool for genetic studies. Trisomic dosage analysis is useful in chromosome assignment, not only at protein (isozyme) level, but also at DNA (RFLP) level. In this study, rice hybrid trisomics were more effective than its primary trisomics due to easy control of internal dosage changes between two alleles. Most of chromosomal assignments of RFLP markers were consistent with their linkage analyses during the map construction.

References

Iwata, N. and T. Omura, 1984. Studies on the trisomic in rice plants. VI. An accomplishment of a trisomic series in Japonica rice plants. Jpn. J. Genet. 59(3): 199-204.

Khush, G. S., 1990. Report of meetings to discuss chromosome numbering system in rice, and Announcement-Changes in the designation of trisomics of IR36. RGN 7: 12-15.

Khush, G. S., R. J. Singh, S. C. Sur and A. L. Librojo, 1984. Primary trisomics of rice, origin, morphology, cytology and use in linkage mapping. Genetics 107: 141-163.

McCouch, S. R., G. Kochert, Z. H. Yu, Z. Y. Wang, G. S. Khush, W. R. Coffman and S. D. Tanksley, 1988. Molecular mapping of rice chromosomes. Theor. Appl. Genet. 76: 815-829.

Shastry, S. V. S., D. R. R. Rao and R. N. Misra, 1960. Pachytene analysis in Oryza, I. Chromosome morphology in Oryza sativa. Ind. J. Genet. Plant Breed. 20: 15-21.

Tanksley, S. D., M. Causes, T. Fulton, N. Ahn, Z. Wang, K. Wu, J. Xiao, Z. Yu, G. Second and S. McCouch, 1992. A high density molecular map of the rice genome. RGN 9:111-115.

Tanksley, S. D., J. Miller, A. Paterson and R. Bernatzky, 1987. Molecular mapping of plant chromosomes. In J. P. Gustafson and R. S. Appets(eds.), Chromosome structure and function, p. 157-173. Plenum Press, New York.

Tanksley, S. D., N. D. Young, A. H. Paterson and M. W. Bonierbale, 1989. RFLP mapping in plant breeding-new tools for an old science. Biotechnology 7: 257-264.

Young, N. D., J. C. Miller and S. D. Tanksley, 1987. Rapid chromosomal assignment of multiple genomic clones in tomato using primary trisomics. Nucleic Acids Reseach 15: 9339-9348.