36. Comparison between PCR-based RFLP and Southern-based RFLP as DNA markers for germplasm classification in rice

Behzad GHAREYAZIE, Ning HUANG, Gerard SECOND, John BENNETT and Gurdev S. KHUSH

International Rice Research Institute, P.O. Box 933, Manila, Philippines

Among DNA markers, restriction fragment length polymorphism (RFLP) has proved to be a powerful tool in germplasm classification and genome mapping of rice (Tanksley et al. 1992; Second 1991). The potential of RFLP markers is being exploited in various aspects of rice breeding. However, RFLP analysis requires large quantities of hiah quality DNA and involves detection systems that use either radioisotopes or complex biochemistry. As an alternative, sequence tagged sites (STS) generated through the polymerase chain reaction (PCR) offer a simple, high-speed, non-radioisotopic approach to the detection of DNA polymorphism (Tragoonrung et al. 1992). Williams et al. (1991) converted some rice RFLP markers into STS. Combined with restriction digestion of PCR products, some DNA polymorphisms were revealed as PCR-based RFLP. Southern-based RFLP detects DNA variation present within as much as 30 kb of the marker locus while PCR-based RFLP can detect polymorphism occurring only within the DNA segment delimited by the primers. However, PCR-based RFLP offers higher resolution in the detection of variation. It remains therefore unclear in the level of DNA polymorphism of PCR-based RFLP is comparable with that of Southern-based RFLP.

In an attempt to classify a collection of Iranian rice (O. sativa L.) germplasm which is well known for its excellent grain quality (e.g. strong aroma, intermediate amylose content, gel consistency), the Southern-based RFLP was compared to PCR-based RFLP in order to determine a suitable approach for our study. Thirty five Iranian rice accessions along with four Indica or Japonica varieties as control were used in this comparison study. A high level of variation in these selected accessions can be assumed, because the same type and number of alleles identified in a world collection of rice germplasm (Gerard Second, unpublished data) were also identified in the 35 rice accessions with the markers used in this study. Pairs of primers for 15 loci (Table 1) were either synthesized based on sequence information of RFLP markers (Williams et al. 1991) or obtained from Raymond L. Rodriguez (University of California, Davis) and Susan McCouch (Cornell University, Ithaca). The primers were used for PCR amplification of rice genomic DNA isolated from the 39 selected rice accessions. PCR products were separated in 2% agarose gel to examine Amplified Fragment Length Polymorphism (RFLP) or digested with eight different restriction endonucleases (4 or 5 base cutters,) to detect point mutations or small addition/deletion events causing PCR-based RFLP.

Table I shows the results of the two approaches. With the application of only a single restriction enzyme (EcoRV or SCAI) 11 out of the 12 RFLP markers studied (except RG386) identified allelic variation among the 39 rice lines. The degree of polymorphism was relatively high using all probes. By contrast, PCR-based RFLP revealed lower level of DNA polymorphism. Two types of polymorphism were observed with the PCR-based approach. The first type comprised addition/deletion events that could be read directly from agarose gels (Fig. 1C and Table 1, column 4). Six of the 15 primer pairs showed this type of polymorphism. The second type of polymorphism comprised point mutation events and/or small additions/deletions that were identified from agarose gel electrophoresis of RCR products following digestion with 4-5 base cutters (Fig.

Table 1. Polymorphism of 15 DNA markers across 39 rice accessions

=============================================================================
                                                    PCR-based
Marker    Southern-based     ================================================
Loci           RFLP          AFLP (-digestion)  RFLP (+digestion)  Increased
          ================   =================  ================ polymorphism
           Pattern   %DP     Pattern      %DP   Pattern     %DP*
=============================================================================
PTA 248    poly       30     poly          38    poly       46       yes
RG   13    poly       47     poly          47    poly       50       yes
RG   64    poly       22     poly          22    poly        4       yes
RG  100    ND         ND     mono           0    mono        0       no
RG  118    poly        9     mono           0    poly       49       yes
RG  120    poly       33     mono           0    mono        0       no
RG  173    poly       19     mono           0    mono        0       no
RG  214    poly       25     poly          22    poly       23       yes
RG  235    poly       50     poly          13    poly       13       no
RG  241    poly       20     mono           0    poly        5       yes
RG  257    ND         ND     mono           0    poly        5       yes
RG  329    ND         ND     mono           0    mono        0       no
RG  365    poly       17     poly           8    poly        8       no
RG  386    mono        0     mono           0    mono        0       no
Waxy       poly       31     mono           0    poly       20       yes
=============================================================================
ND=no data.
DP=Degree of polymorphism (percentage of accessions differing from majority
pattern of RFLP).
*The highest degree of polymorphism among enzyme/marker combination.

1B and 1D). Eight pairs of primers showed increased polymorphism on digestion, either producting polymorphism from monomorphic PCR products (Fig. 1A and 1B) or giving additional (new) polymorphisms as compared to the ones observed prior to digestion (Fig. 1C and 1D, Table 1 last column). However in some cases, digestion of polymorphic PCR products did not reveal additional polymorphism (Fig. 1C and 1E). Out of 9 pairs of primers which amplified monomorphic bands prior to digestion, 4 pairs showed polymorphism after digestion with restriction enzymes. In total, 10 loci showed polymorphism with PCR-RFLP, i.e., 67% of the 15 loci examined.

In general, the degree of polymorphism and the number of detected alleles were lower with the PCR approach than with Southern-based RFLP. Of the 15 markers used, PTA248 and RG64 were the only markers giving the same polymorphism for both the PCR-based approach and Southern-based RFLP. Nevertheless, PCR-based RFLP offers the advantage of being fast, highly repeatable and not based on the use of radioisotopes. If more probes are sequenced and/or informative primers from known sequences are made available, one may expect to accumulate sufficient number of polymorphic markers to use PCR-based RFLP for classification purposes. Furthermore, PCR-based RFLP can be applied in marker-aided selection and gene mapping projects (see Res. note 33).

References

Second, G. 1991. Molecular markers in rice systematics and the evaluation of the genetic biotechnology in agriculture and forestry. Vol. 14, Rice. (ed. by Y.P.S. Bajaj). Spinger-Verlag, Berlin, Heidelberg.

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

Tragoonrung, S., V. Kanazin, P.M. Hayer and T.K. Blake, 1992. Sequence tagged site facilitated PCR for barley genome mapping. Theor. Appl. Genet. 84: 1002-1008.

Williams, M.N.V., N. Pande, S. Nair, M. Mohan and J. Bennett. 1991. Restriction fragment length polymorphism analysis of polymerase chain reaction products amplified from mapped loci of rice (Oryza sativa L.). Theor. Appl. Genet. 82: 489-498.