We have developed a highly effective marker set for assigning medaka mutants to specific linkage groups (LG) in bulk segregation analysis by PCR product length polymorphisms (PLPs). The marker set permits analysis using an automated microchip electrophoresis system as well as conventional agarose gel electrophoresis.

Medaka is a small, egg-laying, freshwater teleost fish found throughout eastern Asia (1). Its small size, high fecundity, short life cycle, and transparency of eggs and embryos make this species an ideal model organism for studies of vertebrate developmental genetics. In addition, 10 highly polymorphic, inbred lines are readily available (2). Indeed, extensive N-ethyl-N-nitrosourea mutagenesis screening has been conducted in medaka, resulting in >500 mutants, which are maintained and available from the National Bio Resource Project (NBRP Medaka; www.shigen.nig.ac.jp/medaka).

To identify and clone the genes responsible for mutant phenotypes by candidate gene approaches or chromosome walking (3,4), the mutants must first be assigned to specific linkage groups (LGs). Bulk segregation analysis is the most effective method for assigning mutant loci to specific LGs. With this method, mutant model organisms are segregated according to phenotype and allele frequencies within each pool are measured. The allele frequencies of F2 wild-type pools derived from different strains are expected to be the same throughout the genome, except for the specific mutant allele responsible for the phenotype. Therefore, mutants may be assigned to specific LGs by comparison of allele frequencies between pools.

The M-marker 2003 primer set (5), which consists of 48 PCR length polymorphism (PLP) markers (two per LG) has been used to map mutants in medaka by bulk segregation analysis. Because the amplicon size of some of the markers is >1 kb, PCR amplification is affected by template quality. Additionally, this method requires relatively labor-intensive acryl-amide gel electrophoresis because the range of amplicon size is too wide to analyze with agarose gel electrophoresis. Recent advances in microchip technology have resulted in the development of cost-effective, automated electrophoresis systems, such as the MultiNA (Shimadzu Corporation, Kyoto, Japan) and the LabChip 90 (Caliper Life Science, Hopkinton, USA) platforms. Although these systems can load samples automatically and effectively analyze size differences among PCR fragments, the maximum resolution is relatively small (~300 bp). We therefore undertook the design of a new primer set that could be applied to mapping analyses using both automated electrophoresis systems and 4% agarose gel electrophoresis. We refer to this primer set as M-marker 2009.

M-marker 2009 markers were designed to amplify genomic regions containing insertion/deletion (indel) polymorphisms between the Southern and Northern Japanese medaka strains, which are most frequently used for position-based cloning of causal genes in mutants. In this study, we report on the feasibility and robustness of the M-marker 2009 primer set in crosses of inbred medaka lines and specific mutants using an automated electrophoresis system and agarose gel electrophoresis.

A previous genome-wide SNP analysis showed that the southern Japanese population is highly polymorphic (6). To construct a marker set capable of detecting polymorphisms between the Southern and Northern Japanese medaka populations, we designed PLP markers based on indel polymorphisms between the HNI-II and Hd-rR-II strains. Indel polymorphisms are not compared as frequently as SNPs, but once identified, may be more robustly detected in Southern and Northern Japanese medaka populations. Therefore, indel polymorphisms are suitable for the first rough mapping step, while SNPs are preferable for subsequent fine mapping. A total of 169 primer pairs located at 10-centi-morgan (cM) intervals throughout the medaka genome were designed for use as polymorphic PLP markers. All of the markers were confirmed using genomic DNA from HNI-II, Hd-rR-II, and their F1 progeny. We also tested the markers using genomic DNA extracted from 10 other inbred strains obtained from the NBRP to identify polymorphisms between Southern and Northern Japanese medaka populations. Finally, we selected 48 PLP markers (two markers per LG) for the M-marker 2009 kit (Supplementary Table S1). When resolved on the MultiNA system, the sizes of amplicons among the Southern Japanese strains were indistinguishable (Supplementary Table S2). However, amplicons derived from the three Northern Japanese strains could be distinguished from the Southern Japanese strains when resolved using either the MultiNA or agarose gel electrophoresis. The amplicons derived from the non-Japanese populations were similar in size to those of the Japanese populations. Specifically, in the Nilan and HSOK strains, 31 and 32 and 6 and 8 amplicons were the same size as amplicons from the Southern and Northern Japanese strains, respectively. In the same non-Japanese strains, 1 and 2 markers did not amplify, and 10 and 6 amplicons were polymorphic compared with the Japanese populations, respectively. Together, these findings indicate that this marker set may be applied to evaluate all crosses between Southern and Northern Japanese medaka populations.

To evaluate the applicability of the M-marker 2009 markers to bulk segregation analysis, we mapped the leucophorefree (lf) mutant. The lf mutant, which has been mapped previously on LG01 (7), completely lacks differentiated leucophores (8,9). We crossed lf male and HNI-II female fish to produce two pools consisting of 96 mutant and 36 wild-type fish, which we assayed using the M-marker 2009 kit. The results revealed that a pair of LG01 markers produced intense mutant-derived bands (Figure 1, A and B). None of the other LG markers showed this trend. Notably, MID0117 had only a mutant-derived band in the lf pool. This suggested that MID0117 was closely located to the lf causal mutation. Individual genotyping showed that 3 out of 192 carried the nonmutant allele, enabling us to map the lf locus to 1.6 cM apart from MID0117. These findings confirmed that the lf mutant was correctly mapped on LG01.

Figure 1. Results of bulk segregation analysis for two mutants,lfandpl2with MultiNA and 4% agarose gel electrophoresis. (Click to enlarge)

After demonstrating the ability of the method to map the lf mutant, we attempted to map the pectoral finless 2 (pl2) mutation, which had not previously been mapped. The pl2 mutant lacks pectoral fins throughout life, but is viable when homozygous. We crossed the pl2 and Kaga strains to generate DNA pools from mutants and their siblings. We then performed bulk segregation analysis of pl2 using the M-marker 2009 markers. As a result, both LG10 markers showed an increase of mutant-derived product in the pl2 pool (Figure 1, C and D). Other markers did not show this trend. Comparison between those two markers showed that MID1012 had a more intense mutant-derived band than MID1014. This indicated that MID1012 was closer to the pl2 locus than MID1014. To confirm these observations, we performed individual genotyping using these markers. The pl2 mapped to 12 cM from the MID1012 and 30 cM from MID1014 (data not shown). We therefore concluded that pl2 mapped to LG10.

We specifically designed the new marker set for bulk segregation analysis on automated electrophoresis platforms such as MultiNA. Given that the markers were able to detect polymorphisms between Southern and Northern Japanese inbred strains, it is likely that the M-marker 2009 markers will work on almost all crosses between Northern and Southern Japanese populations. This is particularly relevant if one considers that lf and pl2 mutants are not inbred with the Northern inbred strains, HNI-II and Kaga. Notably, the M-marker 2009 markers were able to map pl2 outside the region of the two LG10 markers. MID1014, which was 30 cM apart from the pl2 locus, showed linkage. As all markers were within 30 cM of the terminal marker of each LG (Supplementary Table S1), this marker set covered the entire medaka genome. Although we performed bulk segregation analysis using 96 mutants, it may be possible to map mutants using fewer samples (~32 mutants and 32 wild-type siblings) as described previously (5). Since the M-marker 2009 markers can be used with agarose gel electrophoresis (Figure 1, B and D), we believe that it has the potential to be widely applied to study genome sequencing and gene annotation, and that it will simplify mapping of medaka mutants in many laboratories.