Methods, technology, and resources
High-speed system to generate congenic strains in medaka
2023 年 98 巻 5 号 p. 267-275
詳細
2023 年 98 巻 5 号 p. 267-275
The congenic strain, an inbred strain containing a small genomic region from another strain, is a powerful tool to assess the phenotypic effect of polymorphisms and/or mutations in the substituted genomic region. Recent substantial progress in the genetic studies of complex traits increases the necessity of congenic strains and, therefore, a quick breeding system for congenic strains has become increasingly important in model organisms such as mouse and medaka. Traditionally, more than ten generations are necessary to produce a congenic strain. In contrast, a quick method has been reported previously for the mouse, in which the use of genetic markers reduces the required number of backcross generations to about a half that of the traditional method, so that it would take around six generations to obtain a congenic strain. Here, we present an even quicker congenic production system, which takes only about four generations. The system can produce medaka congenic strains having part of the HNI-II (an inbred medaka strain derived from the northern Japanese population, Oryzias sakaizumii) genome in the HdrR-II1 (another inbred strain from the southern Japanese population, O. latipes) background. In this system, the availability of frozen sperm and genotype data of the BC1 male population makes it possible to start marker-assisted congenic production after obtaining the BC2 population. Our evaluation revealed that the system could work well to increase the percentage of recipient genome as expected, so that a congenic strain may be obtained in about one year.
The teleost fish medaka is a widely used model animal in genetic studies. The availability of many inbred strains makes medaka a good tool for studying complex traits such as vertebral number (Kimura et al., 2012) and startle behavior (Tsuboko et al., 2014). To dissect complex traits and/or to identify the underlying genetic factors, congenic strains are helpful. As a congenic strain carries a genomic region from an inbred strain in the background of another inbred strain, we can assess the effect of the genomic region easily.
In the traditional method, a congenic strain is produced by repeated backcrossing of a donor strain carrying an allele of interest to a recipient inbred strain. Starting from the mating to obtain F1 progeny, backcrossing over nine subsequent generations is usually performed and gives rise to individuals heterozygous for the locus of interest with about 99.9% of the recipient genome (Falconer, 1989). The locus is then made homozygous by intercrossing the heterozygotes. Thus, more than ten generations are necessary to obtain a congenic strain. Even though medaka has a relatively short generation time, it still takes about three years to complete production of a congenic strain.
To obtain a congenic strain in a shorter time, a marker-assisted congenic breeding strategy has been proposed and shown to work in mice (Markel et al., 1997). In this strategy, individuals are genotyped using a genetic marker set covering the mouse genome, and the carrier with the highest percentage of recipient genome is then selectively backcrossed. This genomic scanning and selective mating in each generation reduces the required number of backcross generations to four. Based on this strategy, there may be a way to further reduce the number of generations. To create a congenic strain, we first cross two inbred strains to obtain F1 and then backcross an F1 individual to one of the strains. Selection of the best individual for further mating starts from the BC1 generation. That is, as long as the combination of donor and recipient strains is the same, the BC1 population can be shared to produce a congenic strain of any targeted region. Once a set comprising frozen sperm and genotype data from the BC1 male population is prepared, researchers can then start to produce any congenic strains of the donor–recipient combination from the obtained BC2 population: selecting the BC1 male of best fit by looking at the genotype data and inseminating eggs with the frozen sperm. Thus, the use of the set above can eliminate two more generations in the congenic production procedure, so that a congenic strain will be obtained within about one year in the case of medaka, which mature about 2.5–3 months after fertilization.
Here, we aimed to develop a high-speed congenic (HSC) system using two medaka inbred strains, the HNI-II strain for the donor and the HdrR-II1 strain for the recipient (derived from Oryzias sakaizumii and O. latipes, respectively). Its speed would be achieved as described above, by applying sperm cryopreservation to the marker-assisted congenic breeding strategy developed in mice.
The system, outlined schematically in Figure 1, was constructed to generate congenic strains in a shorter period. The recipient and donor strains were HdrR-II1 and HNI-II, respectively. To develop the system, we generated ‘male BC1 stock’, which consisted of frozen sperm and the genomic information from each male BC1 fish.
This system works as follows. First, looking at the genomic information, we select the best BC1 male fish. The best fish carries the targeted region originating from HNI-II and the least amount of donor genome outside that region. Frozen sperm from the best BC1 male is used to artificially inseminate mature unfertilized eggs obtained from HdrR-II1 females. The BC2 progeny are bred and the best BC2 fish is determined by examining a PCR product length polymorphism (PLP) marker set which covers the medaka genome. Next, the best BC2 fish is backcrossed to obtain BC3 progeny. By repeating the selection and backcrossing, BC4 progeny are prepared. Finally, the targeted region is made homozygous by intercrossing the best pair among the BC4 population, resulting in a congenic strain. If few progeny are obtained from the best fish, the second-best and/or third-best fish can additionally be mated.
Construction of the PLP marker setTo assess the amount and region of the genome originating from the donor strain, we prepared a PLP marker set covering the medaka genome at approximately 10-cM intervals (Table 1). This marker set consists of 187 PLP markers, including 130 known markers that were found in the MLBase database (http://mbase.nig.ac.jp/mbase/ml_base.html) or have been reported elsewhere (Kimura et al., 2004, 2005, 2012; Naruse et al., 2004). New markers were designed to fill the large gaps. Eighty primer sets were tested, and 69 were found to be co-dominant PLP markers. Based on their mapped positions on the chromosomes, 57 markers were selected to add to the initial PLP marker set for the HSC system. The 187 PLP markers covered the medaka genome with an average interval of 8.86 ± 2.49 cM. The primer sequences are shown in Supplementary Table S1.
| Chromosome | Genetic length (cM) | No. of markers | Interval distance (cM) | ||
|---|---|---|---|---|---|
| Average | Minimum | Maximum | |||
| 1 | 72.4 | 9 | 9.0 | 5.3 | 12.3 |
| 2 | 64.2 | 8 | 9.2 | 5 | 14.7 |
| 3 | 66.9 | 8 | 9.6 | 8.5 | 11.5 |
| 4 | 57.8 | 7 | 9.6 | 4 | 12.7 |
| 5 | 61.9 | 8 | 8.8 | 6.2 | 10.8 |
| 6 | 52.2 | 7 | 8.7 | 5 | 12.6 |
| 7 | 54.2 | 7 | 9.0 | 5.9 | 12.5 |
| 8 | 58.7 | 8 | 8.4 | 5.6 | 12.9 |
| 9 | 56.6 | 7 | 9.4 | 4.7 | 12.9 |
| 10 | 62.8 | 7 | 10.5 | 8.3 | 12.7 |
| 11 | 57.7 | 8 | 8.2 | 2.6 | 13 |
| 12 | 59.8 | 8 | 8.5 | 4.7 | 12.6 |
| 13 | 68.9 | 7 | 11.5 | 10 | 12.8 |
| 14 | 57.3 | 8 | 8.2 | 2.8 | 12.7 |
| 15 | 53.7 | 7 | 9.0 | 7.3 | 11.3 |
| 16 | 60.1 | 8 | 8.6 | 5.6 | 11.6 |
| 17 | 66.8 | 9 | 8.4 | 3.4 | 11.9 |
| 18 | 70 | 9 | 8.8 | 6 | 12.7 |
| 19 | 55.9 | 8 | 8.0 | 5.3 | 9.2 |
| 20 | 56.2 | 8 | 8.0 | 4.7 | 10.4 |
| 21 | 60.8 | 8 | 8.7 | 5.1 | 11.2 |
| 22 | 54.9 | 8 | 7.8 | 6.3 | 10.5 |
| 23 | 60.1 | 7 | 10.0 | 5 | 12.7 |
| 24 | 53.6 | 8 | 7.7 | 2.6 | 11.2 |
| Total | 1446.6 | 187 | 8.9 | 2.6 | 14.7 |
A total of six families were prepared to generate the male BC1 stock. F1 fish from an HdrR-II1 and HNI-II cross were backcrossed to HdrR-II1 fish. Three of the six families (A1 to A3) were obtained by backcrossing each of three F1 female fish to HdrR-II1 male fish, and the rest (B1 to B3) were obtained by backcrossing each of three F1 male fish to HdrR-II1 female fish. The male BC1 stock was derived from 93 BC1 males: 24 males from the A1 family, 10 from A2, 10 from A3, 23 from B1, 14 from B2 and 12 from B3.
Sperm of each of the 93 BC1 males was cryopreserved, and each genome was scanned using the PLP marker set. The genotype data of each BC1 male are shown in Supplementary Table S2. The percentage of recipient genome in the male BC1 stock followed a normal distribution with a mean of 73.11 ± 4.10% (Supplementary Fig. S1).
Evaluation of the HSC systemTo test the system, we generated four congenic strains using the high-speed congenic system. The four strains were named HdrR-II1-Chr5^HNI-II, HdrR-II1-Chr6^HNI-II, HdrR-II1-Chr15^HNI-II and HdrR-II1-Chr22^HNI-II, in which HdrR-II1 chromosome 5, 6, 15 and 22, respectively, is replaced with its HNI-II equivalent. These are actually consomic strains, as a single entire chromosome was transferred from the HdrR-II1 origin to HNI-II in each strain. We selected these regions as targets because QTLs for craniofacial morphology had been found on these chromosomes (Kimura et al., 2007), and we aimed to use the consomic strains for further analyses. Family trees for each strain are presented in Supplementary Figure S2 to S5.
First, the best BC1 male fish was selected for each targeted region from the BC1 male stock, by looking at the genotype data shown in Supplementary Table S2. A total of 25 BC1 males were heterozygous for all markers on chromosome 5, and of those, B1-1 was selected as the best BC1 male with the highest percentage of recipient genome (79.61%) in the whole genome other than chromosome 5 (Table 2 and Supplementary Table S3). For the targeted region of chromosome 6, 31 fish were carriers, and B1-22 was the best with 82.78% recipient genome (Table 2 and Supplementary Table S4). To generate HdrR-II1-Chr15^HNI-II, we chose B3-4, which showed 82.22%, the highest proportion of recipient genome out of 29 carriers (Table 2 and Supplementary Table S5). For HdrR-II1-Chr22^HNI-II, B3-13 was selected from 25 carriers (Table 2) as the fish showing the highest percentage of recipient genome (81.01%) (Supplementary Table S6).
| Congenic strain | Backcross generation | No. of carriers | Observed % recipient genomea | |||||
|---|---|---|---|---|---|---|---|---|
| Average | ± | SD | Best | Second best | Third best | |||
| HdrR-II1-Chr5^HNI-II | BC1 | 25 | 73.51 | ± | 3.01 | 79.61 | 78.77 | 77.93 |
| BC2 | 12 | 89.41 | ± | 3.14 | 94.69 | 92.46 | 92.18 | |
| BC3 | 8 | 97.42 | ± | 0.96 | 98.60 | 98.32 | 97.77 | |
| BC4 | 7 | 99.20 | ± | 0.59 | 100.00 | 99.72 | 99.44 | |
| HdrR-II1-Chr6^HNI-II | BC1 | 31 | 73.84 | ± | 4.40 | 82.78 | 82.22 | 79.44 |
| BC2 | 22 | 92.56 | ± | 2.81 | 96.67 | 95.83 | 95.83 | |
| BC3 | 10 | 97.75 | ± | 1.02 | 99.44 | 99.17 | 98.61 | |
| BC4 | 7 | 99.84 | ± | 0.22 | 100.00 | 100.00 | 100.00 | |
| HdrR-II1-Chr15^HNI-II | BC1 | 29 | 73.57 | ± | 3.84 | 82.22 | 80.56 | 78.61 |
| BC2 | 15 | 91.59 | ± | 1.46 | 93.85 | 93.06 | 93.06 | |
| BC3 | 5 | 96.94 | ± | 1.48 | 99.44 | 96.94 | 96.39 | |
| BC4 | 7 | 99.68 | ± | 0.30 | 100.00 | 100.00 | 100.00 | |
| HdrR-II1-Chr22^HNI-II | BC1 | 25 | 72.92 | ± | 4.16 | 81.01 | 78.77 | 78.77 |
| BC2 | 5 | 90.11 | ± | 2.06 | 92.46 | 91.62 | 90.50 | |
| BC3 | 14 | 96.03 | ± | 1.35 | 98.60 | 97.49 | 96.93 | |
| BC4 | 5 | 99.44 | ± | 0.52 | 100.00 | 99.72 | 99.44 | |
Thereafter, following the work flow described in Figure 1, each step was undertaken for all four congenic strains. Changes in the percentage of recipient genome in the carrier, which was heterozygous for the targeted region, are summarized in Table 2. In HdrR-II1-Chr15^HNI-II and HdrR-II1-Chr22^HNI-II strain development, enough progeny were obtained from the backcross of the best fish at each generation, and 5 to 16 carriers could be examined for the PLP markers. From the results, we obtained three BC4 fish with the HdrR-II1 allele homozygous for all markers except for the targeted region for the HdrR-II1-Chr15^HNI-II strain (Supplementary Table S5), and one such BC4 fish for the HdrR-II1-Chr22^HNI-II strain (Supplementary Table S6). For the HdrR-II1-Chr5^HNI-II strain, the best fish in the BC3 generation did not yield enough progeny. Thus, additional crosses were made using the two individuals of the third best, as the second-best fish died before maturation (Supplementary Table S3). A total of seven carriers were obtained from the BC3 backcrosses, resulting in a BC4 fish with the percentage of recipient genome of 100%. In the process generating the HdrR-II1-Chr6^HNI-II strain, because the best fish in the BC2 generation did not lay enough eggs, the two second-best fish were additionally backcrossed to HdrR-II1 (Supplementary Table S4). Those three backcrosses raised 10 carriers, and the backcross of the best fish in the BC3 generation resulted in four fish with 100% recipient genome except for chromosome 6 in the BC4 generation. Starting from the insemination with the best BC1 male sperm, it took 11 to 14 months to obtain the best fish in the BC4 generation for each strain.
As the last step of the HSC system, the targeted region was made homozygous for the HNI-II allele (Fig. 1). To obtain the homozygote from the selected BC4 fish, three to six months were necessary, and it took at least three months more to obtain a homozygous pair for maintaining a homozygous strain. Only one out of seven carriers in the BC4 generation was male in the HdrR-II1-Chr5^HNI-II development. Thus, the male with 99.16% of the recipient genome was crossed to the best female with 100% of the recipient genome except for the targeted region (Supplementary Table S3). Of 65 siblings from the cross, a pair were HNI-II-homozygous in the entire chromosome 5, but both of them showed 99.44% of the recipient genome in the rest of the genomic regions. By intercrossing their offspring, we could obtain pairs that were HNI-II-homozygous for chromosome 5 and HdrR-II1-homozygous for all other chromosomes (Supplementary Fig. S2). To generate the HdrR-II1-Chr6^HNI-II strain, a pair of the best fish (percentage of recipient genome of 100%) in the BC4 generation were intercrossed (Supplementary Table S4). Of 45 obtained fish, only one female (LG6-16x13-19) was HNI-II-homozygous for the entire chromosome 6. The female was crossed to a sibling male fish (LG6-16x13-35) that was heterozygous for one marker on chromosome 6 (Supplementary Fig. S3). From this cross, several pairs were obtained in which the entire chromosome 6 was HNI-II-homozygous. For the HdrR-II1-Chr15^HNI-II strain, a pair of the best fish (percentage of recipient genome of 100%) in the BC4 generation were intercrossed (Supplementary Table S5), and 24 fish were obtained; however, none were HNI-II-homozygous for the entire chromosome 15. Thus, a sibling (LG15-17x27-8) that was heterozygous for one marker on chromosome 15 was mated with one of the best fish in the BC4 generation (B3-4-21-4-29, Supplementary Fig. S4). The pair yielded three homozygotes for the entire chromosome 15; however, all fish grew as males, even though they were genetically female. We tried to obtain a female fish that was HNI-II-homozygous for the entire chromosome 15 by intercrossing several pairs in which one chromosome 15 was derived from the HNI-II and all the others from HdrR-II1. One such female fish was obtained, with two male fish that were genetically female. In HdrR-II1-Chr22^HNI-II strain development, the best (female) and the second best (male) in the BC4 generation were intercrossed (Supplementary Table S6), and 39 offspring were examined for HNI-II homozygosity in the entire chromosome 22. Two such fish with 100% of the recipient genome in other chromosomes were found, but both of them were male. To make a pair that were HNI-II-homozygous for the entire chromosome 22, one of the males (LG22-25x32-2) was crossed with a sibling female fish (LG22-25x32-16) that was heterozygous for one marker on chromosome 22 (Supplementary Fig. S5). The cross yielded the desired pairs with the percentage of recipient genome of 100%.
Data and resource availabilityThe system described here starts from selecting the best BC1 male based on the genotype data for the PLP marker set. The data are shown in Supplementary Table S2. Frozen sperm of the BC1 males is stored in NBRP Medaka (https://shigen.nig.ac.jp/medaka/). Upon request, NBRP Medaka can inseminate Hd-rRII1 unfertilized eggs with this sperm and provide the resulting eggs. To assess genotype in each generation, a PLP marker set has been prepared in this study. Information about the PLP marker set such as primer sequences is shown in Supplementary Table S1. Furthermore, we can share the primers upon request as long as we have them. If a researcher wants to use other methods to select the best BC1 male, we will provide the DNA extracted from the BC1 males upon request. The consomic strains HdrR-II1-Chr6^HNI-II and HdrR-II1-Chr22^HNI-II are available from NBRP Medaka. HdrR-II1-Chr5^HNI-II and HdrR-II1-Chr15^HNI-II strains were lost before frozen sperm stocks were made, so we cannot offer those strains.
In this study, we have developed a HSC system to generate congenic strains in medaka. The speed was achieved by preparation of a male BC1 stock in addition to the marker-assisted congenic breeding strategy reported for mice (Markel et al., 1997). The male BC1 stock prepared here displayed an average percentage of recipient genome of 73.11%, which was close to 75%, the expected value based on Mendel’s law. The genomic regions derived from the donor were steadily reduced at each backcross, resulting in the best BC4 fish with no HNI-II alleles except for the introgressed chromosome of interest for all four strains tested (Table 2). Markel et al. (1997) reported the empirical results of their marker-assisted congenic strain production in which the Apoe null allele had been introgressed into each of six inbred mouse strains. According to their data, the observed percentage of recipient genome for the best male was 81.04–86.36% in the BC1 generation and increased as backcrossing proceeded. Because they calculated the values including the Apoe locus, the percentage of recipient genome in their study was underestimated as about 0.3% less than that in our experiments. In our study, the percentage of recipient genome for the best male was 79.61–82.78% in the BC1 generation, so that more of the donor genome remained compared to Markel et al.’s production in this generation. Why the values in our study were lower is unknown, even though we tested more than twice the number of carriers. After the BC2 generation, however, the value in both studies became similar: 90.00–96.15% in theirs and 92.46–96.67% in ours in the BC2 generation, 95.54–98.96% and 98.60–99.44% in the BC3 generation, and 95.88–99.73% and 100% in the BC4 generation, respectively. In these generations, the number of carriers was similar in the two studies, and one difference was that we used not only the best but also the second- and third-best individuals in the production of the HdrR-II1-Chr5^HNI-II and HdrR-II1-Chr6^HNI-II strains, as Markel et al. (1997) recommended in their paper. Taking all these observations together, the reduction of the donor genome in this study progressed comparably to that observed in the marker-assisted congenic strain production in mice. It took Markel et al. (1997) five generations to obtain a heterozygote for the targeted region with 99.90% of the recipient genome, whereas only three generations are necessary in our system because congenic production starts from generating the BC2 population.
Neither we nor Markel et al. (1997) have set any standard value of the percentage of recipient genome for the ‘best’ individual in each generation. Instead, the ‘best’ individual has been selected relatively, by selecting the individual with the highest percentage of recipient genome among those genotyped. Thus, the percentage of the recipient genome of the ‘best’ fish depends on the number of tested individuals. We aimed to genotype about 50 medaka in order to find carriers, which resulted in 5–31 carriers. We could have fixed the standard value of the percentage of recipient genome for the ‘best’ in a generation, and continued to sample offspring from their parental generation until an individual met the criteria of the ‘best’. However, we prefer the relative selection, at least in the case of medaka, because we have to wait about two months after fertilization to make fin-clips for genotyping, and obtaining many eggs from medaka becomes harder as the fish gets older. It may be better to select the ‘best’ fish among individuals obtained within a few weeks, and, if the ‘best’ fish in the BC4 generation has too low a value of the percentage of recipient genome to make homozygotes for a congenic strain, one more backcross will give a good individual. The strategy above probably takes less time than keeping searching for a fish that fulfils the criteria in each generation.
In our test operation of the system, it took one or a few more generations than expected to make homozygotes for the targeted region in some strains after obtaining the BC4 generation. This is because each of the targeted regions was so wide (i.e., a whole chromosome) that homologous recombination occurred in the majority of chromosomes during meiosis. As a result, it was hard to obtain fish that were homozygous for the targeted chromosome in both sexes: one male and one female out of 65 tested fish for HdrR-II1-Chr5^HNI-II, one female out of 45 tested fish for HdrR-II1-Chr6^HNI-II, no homozygotes out of 24 tested fish for HdrR-II1-Chr15^HNI-II, and two males out of 39 tested fish for HdrR-II1-Chr22^HNI-II. In many cases, the genomic region of interest is narrower, so that the user can obtain a pair that are homozygous for the targeted region more efficiently. In fact, we could smoothly generate two congenic strains, in each of which about half of chromosome 15 was targeted, as follows. By intercrossing a pair of heterozygous carriers, six homozygotes out of 32 fish were obtained for one strain, and nine were homozygotes when 49 fish were examined for the other strain (data not shown). In both cases, a few pairs of homozygotes could be generated to maintain as a congenic strain. Thus, the main cause of more generations being required in our evaluation is not in our congenic breeding system, but in the size of the targeted region. As long as the targeted region is the same, there should basically be no difference in efficiency of the step making homozygotes among the traditional method, the marker-assisted congenic breeding strategy, and our system.
One unexpected phenomenon observed in our congenic breeding was the sex reversal. Not all, but most of the genetic females grew up as males in the step of making chromosome 15 homozygous for the donor allele. For this reason, the congenic strain, HdrR-II1-Chr15^HNI-II, could be produced, but soon collapsed. Medaka (O. latipes and O. sakaizumii) employ the XX/XY genetic sex determination system (Aida, 1921). However, both XX males and XY females have been observed in the wild (Shinomiya et al., 2004) and in the laboratory (Nanda et al., 2003; Sato et al., 2005; M. S., unpublished data). Female-to-male sex reversals have been reported to occur in response to environmental factors such as temperature, starvation and green light irradiation (Sato et al., 2005; Hayasaka et al., 2019; Sakae et al., 2020). On the other hand, the frequency of XX males has been shown to differ in each inbred strain (Nanda et al., 2003; Sato et al., 2005), indicating that genetic factor(s) are involved in the sex reversal. In fact, one of the XX males found in a wild population transmitted the female-to-male sex reversal to its progeny as a recessive trait controlled by one or a few major genes (Shinomiya et al., 2010). Their linkage analysis identified one of the loci on chromosome 8 but not 15. Our study suggests that genomic region(s) on chromosome 15 induce female-to-male sex reversal when the region(s) is/are HNI-II-homozygous under an HdrR-II1 background. In addition, two congenic strains have suggested that there are at least two regions on chromosome 15 that affect the female-to-male sex reversal. Each of these two congenic strains carries about half of the HNI-II chromosome 15 in homozygous form in an HdrR-II1 background, and the two HNI-II genomic regions cover the whole of chromosome 15. The frequency of XX males in each strain is much lower than that observed in HdrR-II1-Chr15^HNI-II, so that we can maintain the strains (data not shown). Thus, there seem to be two or more loci involved in sex reversal, and they are divided into the two strains. Alternatively, some phenotypes in HdrR-II1-Chr15^HNI-II may secondarily affect the sex reversal, e.g., a smaller mouth might make the larvae starve, resulting in female-to-male sex reversal.
Using the system shown in Figure 1, a congenic strain can be made through crossings over four generations. As the generation time of medaka is approximately 2.5–3 months, simple calculation indicates that 10–12 months would be enough to obtain a congenic strain. However, it actually took 20 months at the fastest in our evaluation experiments: about 12 months for obtaining the BC4 medaka and about eight months for making a homozygous pair to maintain as a strain. As described above, the reason it took longer to make the homozygous pair was that a few more generations were necessary due to the wide targeted regions (a whole chromosome) and the frequent occurrence of sex reversal. About three more months were spent for the steps to obtain the BC4 medaka. The main cause of this is considered the lack of staff and space for breeding, as we proceeded with the production of four strains simultaneously.
For the genome-wide scan, we prepared and used a PLP marker set that covers the medaka genome at approximately 10-cM intervals (Table 1). A previous report suggested that medaka showed high chiasma interference within the same chromosome arm (Naruse et al., 1988), and double crossing-over in a chromosome arm seems rare in medaka (Naruse et al., 2004). Therefore, we consider that our PLP marker set can work well for monitoring the genome, although we cannot rule out the possibility that unexpected changes in genomic structure occur between the markers. As many genetic markers have been reported in medaka (Naruse et al., 2000, 2004; Kimura et al., 2005; Kimura and Naruse, 2010), genotyping by additional markers should reduce such concerns. Alternatively, whole-genome sequencing would be a reliable monitoring strategy.
In conclusion, we have established a quick system to generate congenic strains in medaka carrying part of the HNI-II genome in the HdrR-II1 background. Only three generations are required to obtain fish that are heterozygous in a genomic region of interest, with the remaining 99.9% being the recipient (HdrR-II1) genome. These strains are well-known inbred strains derived from the northern (HNI-II) and southern (HdrR-II1) Japanese populations, which diverged approximately 18 million years ago (Setiamarga et al., 2009). Both genetic and phenotypic variation has been observed between the two strains. Single-nucleotide polymorphisms were found at a high rate, 3.42% (Kasahara et al., 2007). In addition, a recent study using the multiple arbitrary amplicon sequencing approach showed that these two populations are genetically well differentiated (Fujimoto et al., 2022). As an example of a phenotypic difference, 379 out of 444 craniofacial traits differed between HdrR-II1 and HNI-II (Kimura et al., 2007). Furthermore, HNI-II has a more elongated body shape (Hyodo-Taguchi, 1990) and shows higher sensitivity in the startle response to a visual stimulus (Tsuboko et al., 2014). For these reasons, the two strains have often been used in genetic studies of complex traits (Kimura et al., 2007; Tsuboko et al., 2014). Usually, several or more congenic strains are necessary to define the genomic region that controls such traits, so this system should be a helpful tool for these studies. Furthermore, the strategy adopted in our system can be applied to other combinations of medaka strains as well as to other species in which the sperm can be cryopreserved.
Fish were maintained in an in-house facility at 27 ℃ in a constant recirculating system on a 14 h light/10 h dark cycle. We used two medaka strains, HdrR-II1 and HNI-II, which are inbred strains established from a southern and a northern Japanese population, respectively (Hyodo-Taguchi, 1996). These strains were supplied by NBRP Medaka (https://shigen.nig.ac.jp/medaka/). Cryopreservation of medaka sperm was performed following a method described elsewhere (Sasado, 2009a). The frozen sperm were used to artificially inseminate mature unfertilized HdrR-II1 eggs obtained as previously described (Sasado, 2009b).
Primer design for PLP markersScaffolds located in the target regions for new PLP markers were identified using UTGB (http://utgenome.org/UTGBMedaka/). Downloaded scaffold sequence data were analyzed with Sputnik (not currently available) to identify microsatellites. As the scaffolds were derived from HdrR-II1, the sequences of surrounding microsatellites were compared with the HNI scaffold by BLASTn at the NIG DNA Sequencing Center -medaka site- (http://dolphin.nig.ac.jp/medaka/srch_db/search_medaka_blastdb2.php?lang=jp&lmode=all). If the microsatellites were polymorphic between the two strains, forward and reverse primers were designed by Genetyx software (Genetyx, Tokyo, Japan). Using BLASTn on Ensembl (http://asia.ensembl.org/index.html), both forward and reverse primers were confirmed not to map to any other region of the medaka genome. Each primer pair was experimentally tested with HNI-II fish, HdrR-II1 fish and F1 progeny to determine whether it would work as a co-dominant genetic marker. The new co-dominant PLP markers were then genetically mapped with the known PLP markers, using the same method and samples (184 F2 progeny of HNI-II and HdrR-II1 strains) reported previously (Kimura et al., 2005).
DNA extraction and genotypingGenomic DNA of male BC1 stock was extracted from each animal, which was kept frozen after cryopreservation of sperm; the DNA extraction method is described elsewhere (Kimura et al., 2005). For testing the HSC system, further backcrossed fish were genotyped. Their DNA samples were prepared from fin-clips using a KAPA Express Extract kit (KAPA Biosystems, Wilmington, MA, USA), according to the manufacturer’s protocol with some modifications. The genotype of each fish for PLP marker sets was determined using methods described in Kimura et al. (2005) with some modifications.
We thank Ms. Tomoko Hoshikawa and Ms. Asami Takagi for their valuable assistance with this study; Ms. Yukari Koike for fish maintenance; and all other laboratory members for their help feeding fish. The work presented here was supported by a Grant-in-Aid for Scientific Research on Priority Areas ‘‘Systems Genomics’’ and ‘‘Comparative Genomics’’ from the Ministry of Education, Culture, Sports, Science and Technology in Japan, and by grants from the NIBB Collaborative Research Program (08-202, 09-203, 10-202 and 11-318).
