Silkworm, Bombyx mori L., domesticated for silk production is an agriculturally important insect and comprises a large number of geographical races and inbred lines that show substantial variation in their qualitative and quantitative traits (Mirhoseini et al., 2007). Currently, it is the major economic resource for nearly 30 million families in countries such as China, India, Vietnam, and Thailand (Miao et al., 2005). With the establishment of stable transformation (Tamura et al., 2000; Yamao et al., 1999), the silkworm has shown the potential to produce pharmaceutically important proteins in high yield (Tomita et al., 2003); opening up new applications for sericulture in medical, agricultural, and industrial fields (Yamamoto et al., 2006). Analysis of the silkworm genome began a few years ago because of its importance for breeding and genetic studies, for isolating valuable genes and promoters and for comparative genomics (Goldsmith et al., 2005). Mita et al. (2003) first initiated intensive sequencing of the silkworm genome using expressed sequence tags (ESTs). Recently, this group (Mita et al., 2004) and a second group (Xia et al., 2004) reported the results of whole-genome shotgun sequencing and provided public access to the assembled silkworm genome data (http://www.dna.affrc.go.jp/genome/; Wang et al. 2005; http:// silkworm. genomics.org.cn/). Genetic linkage mapping has proven to be a powerful tool in genetic studies of many organisms. A complete linkage map is necessary to efficiently carry out molecular-based analyses such as molecular marker-assisted selection, QTL mapping of agronomically important traits, prediction of heterosis and comprehensive investigations of genomic evolution between lineages (Tan et al., 2001). Genetic linkage mapping of the silkworm (B. mori L.) as an important insect using molecular markers is essential for genetic studies and for breeding purposes.
Presently, genome studies in B. mori have generated genetic linkage maps based on morphological markers (Doira et al., 1992) and molecular markers including the Restriction Fragment Length Polymorphism (RFLP) (Nguu et al., 2005; Shi et al., 1995; Goldsmith, 1991), Random Amplified Polymorphic DNA (RAPD) (Li et al., 2000; Yasukochi, 1998; Promboon et al., 1995), Selective Amplification of DNA Fragments (SADF) and RAPD (He et al., 2001), amplified fragment length polymerphism (AFLP) (Sima et al., 2006; Lu et al., 2004; Tan et al., 2001), the microsatellite (Miao et al., 2005) and Single Nucleotide Polymorphism (SNP) (Yamamoto et al., 2006). The AFLP technique (Vos et al., 1995; Zabeau and Vos, 1992) has demonstrated to be a convenient and reliable tool to generate highly polymorphic molecular markers that greatly facilitate building linkage maps (Qi et al., 1998; Waugh et al., 1997; Becker et al., 1995). AFLP markers do allow one to construct linkage maps with wide genome coverage without engaging in extensive sequencing or marker development programmes. The AFLP technique is also faster than individual codominant marker types because a single polymerase chain reaction (PCR) can derive multiple loci simultaneously (Erickson et al., 2004). Because of these features, AFLP has been widely employed for genetic mapping in various organisms.
Since the AFLP technique enables the generation of many polymorphic markers in a single PCR, it can be used to generate high-resolution genetic maps. This study reports a high-resolution AFLP-based genetic linkage map of silkworm (B. mori L.). Development of the linkage map lays an important foundation for future genomics research on the silkworm and provides valuable tools for determining the genetic basis of economically important traits, such as silk production and resistance to diseases.
MATERIALS AND METHODS
Insect materials and crosses: One F2 segregating family that had resulted from mating between a Japanese inbred line (P107) as female parent and an Iranian native strain (Khorasan Lemon) as male parent was used in the study. These two inbred lines ands strain exhibit high phenotype diversity for economically important characters such as whole cocoon weight, cocoon shell weight and cocoon shell percentage suggesting that considerable polymorphism exists at the DNA level (Dalirsefat and Mirhoseini, 2007). Indeed, the highest and the least quantities of mentioned traits corresponded to P107 and Khorasan Lemon, respectively. These inbred lines and strains have undergone a high degree of inbreeding and are relatively homozygous. A number of 78 individuals (39 females and 39 males) from the F2 population were used to construct a genetic linkage map. The parents and the F1 progeny were used to establish the segregation pattern of the molecular markers. The crossing experiments were established in the Iran Silkworm Research Center (ISRC) located in Rasht, center of Guilan province.
AFLP analysis: Genomic DNAs were isolated individually from all the parents, F1 and F2 populations at the moth stage by using the phenol/chloroform method (Suzuki et al., 1972) and as modified by Nagaraja and Nagaraju (1995). DNAs were quantified using a known standard (λDNA) (DNA Lambda, Roche, Germany) on agarose gels.
All individuals were subjected to genotyping with AFLP markers according to Vos et al. (1995) with some modifications. Briefly, genomic DNA was double digested with PstI and TaqI restriction enzymes which can produce polymorphic DNA fragments in the silkworm (Mirhoseini et al., 2007; Tan et al., 2001) The DNA fragments were ligated with PstI and TaqI adaptors, generating template DNA for PCR amplification. Two primers were designed on the basis of the adaptor sequences and restriction site sequences for use in Polymerase Chain Reaction (PCR) amplification. Selective nucleotide sequences were added to the 3’-end of each primer. PCR amplification was conducted in two steps: a pre-amplification and a selective amplification. For the selective amplification, a total of 81 primer combinations obtained from two sets of PstI and TaqI selective primers (Table 1) were screened. Among them, 20 primer pairs that produced fragments with clear dominant inheritance patterns and were reproducible were used for linkage analysis. Polymorphism screening of AFLP products was conducted on a 6% polyacrylamide gel using a SequiGen 38×30 cm gel apparatus (BioRad Laboratories Inc., Hercules, CA, USA). Bands were detected by the silver staining procedure (Promega, Technical manual No.023) and gel images were scanned and saved as jpeg files for scoring and further analysis.
Linkage analysis and map construction: Using genotype information of 81 AFLP primer combinations, 20 primer combinations which produced clearly readable and polymorphic fragments among parents were employed to analyze linkage mapping. The AFLP fragments were scored based on 0 and 1 and then converted to A, B, C and D letters according to Map manager QTX (Manly et al., 2001) instruction manual. The data were analysed by using the Kosambi’s map function (Kosambi, 1944) of Map manager QTX (Manly et al., 2001) to develop a linkage map for population. By genotyping 78 F2 progenies using 204 polymorphic bands, a genotypic data matrix in a dimension of 78×204 was constructed and used for linkage mapping. Recombination rates among markers were first evaluated and were then converted to the map distance based on centiMorgan using Kosambi’s map function (Kosambi, 1944).
Among the 81 AFLP primer combinations screened, approximately one-third of the primer combinations (28) produced polymorphic fragments between the P107 inbred line and Khorasan Lemon native strain. Twenty pairs of AFLP primer combinations for segregation analysis of the F2 populations based on reproducibility and the degree of polymorphism were selected. Only the polymorphic fragments that segregated in a dominant manner and could be scored unambiguously were used for linkage map construction. An example of AFLP gel electrophoresis and polymorphism screening corresponding to the TP13 primer (Ptat-Ttac) is shown in Figure 1.
Twenty PstI/TaqI primer combinations produced a total of 845 clearly detected bands of which 204 qualified polymorphic fragments showing good agreement of 3:1 segregation were analyzed for linkage mapping. The frequency of polymorphic AFLP markers derived from the clearly detected bands of the P107×Khorasan Lemon cross was 24.14% (Table 2). This frequency was close to that (25.7%) obtained in the Dazao×C100 silkworm cross (Lu et al., 2004) as well as 27.2% in the eastern oyster, Crassostrea virginica Gmelin, (Yu and Guo, 2003) but significantly higher than that (14.0%) in the Proctor×Nudinka cross reported in barley (Castiglioni et al., 1998) and also 11.2% in the Guppy fish, Poecilia reticulata (Shen et al., 2007). However it was dramatically lower than (60.7%) that in the no. 782×od100 cross for the silkworm (Tan et al., 2001).
Aproximatly, 104 fragments (51%) of 204 polymorphic fragments were detected in the male parent, Khorasan Lemon strain, and 100 fragments (49%) were observed in the female parent, the P107 inbred line. On average, each primer combination generated 10.2 polymorphic fragments that could be used for linkage mapping. The number of polymorphic bands produced using the 20 primer combinations ranged from 7 bands (18.42% and 15.55%) corresponding to TP5 and TP12, to 16 bands (38.09%) for TP16. The different levels of polymorphism observed for each primer combination are illustrated in Table 2.
The linkage map generated from the P107× Khorasan Lemon cross contained 204 AFLP markers that were assigned to 12 linkage groups at the LOD threshold of 2 (Fig. 2). The largest and the smallest linkage groups belonged to LG10 with 53 markers and LG11 with 2 markers covering 938.4 cM and 12.3 cM of silkworm genome, respectively. The average distance between markers was 20.89 cM. Seven major linkage groups consisting of LG1, LG2, LG3, LG8, LG9, LG10 and LG12 contained 13-53 markers whereas five small linkage groups including LG4, LG5, LG6, LG7 and LG11 had 2-6 markers.
In present study, we developed an AFLP-based linkage map for the silkworm (B. mori) was developed. In the early nineteenth century, the silkworm developed into a model for scientific discovery in microbiology, physiology, and genetics at a period when enormous pattern alterations had effected our perception of biology (Willis et al., 1995). Consequently the availability of molecular linkage maps is very valuable in the improvement of research in such disciplines. The map generated in this study consists of 204 AFLP markers. The current map has a total length of 4262 cM and an average marker resolution of 20.89 cM.
Different molecular marker techniques which have their advantages and disadvantages are currently being used to construct genetic linkage maps. Simple sequence Repeats (SSRs) are highly prized as molecular markers due to their codominance and high levels of polymorphism, but a significant effort is required to develop SSR-based maps. The SNP-based genetic markers have attracted significant attention when creating dense genetic linkage maps. SNPs are the most abundant class of polymorphisms and they also provide gene-based markers that may prove useful in identifying candidate genes of interest to be associated with quantitative trait loci (Rafalski, 2002). However the main disadvantage of SNPs is the small number of alleles typically present. AFLP markers are easy to use and reveal large sets of genetic loci, but their transferability between detection platforms (for instance, polyacrylamide gel electrophoresis, gel-based sequencers, and capillary sequencers) can sometimes be difficult and cumbersome (Papa et al., 2005). AFLP not only has higher reproducibility, resolution, and sensitivity at the whole genome level compared to other techniques, but it also has the capability to amplify between 50 and 100 fragments simultaneously (Vos et al., 1995).
In this study, polymorphic AFLP fragments with a clear dominant inheritance pattern were employed to construct a linkage map; that is, the suitable fragments must show complete dominance expression in one parent and complete recessive expression in the other, and all F1 individuals must be heterozygous. Several studies have demonstrated segregation of some AFLP fragments in a codominant manner (Yin et al., 2002; Piepho and Koch, 2000; Castiglioni et al., 1999). However, it is extremely difficult to identify codominantly segregating fragments from the polyacrylamide amongst several hundred AFLP fragments suggesting the lack of interest in employing the codominant AFLP markers (Zhong et al., 2004).
A total of 24.14% of clearly readable and qualified AFLP bands were polymorphic between the P107 inbred line and Khorasan Lemon native strain of the silkworm. A higher level (61%) of polymorphic AFLP marker has been reported by Tan et al. (2001) in a single backcross (no. 782 and od100) family of silkworm (Table 2). To explain this approach, they accounted for several factors: First, employing two distinct silkworm strains as in the present study, P107 and Khorasan Lemon are two examples of distinct silkworm inbred lines and strains. The former is from the Japanese bivoltine system and the latter is from the Iranian native monovoltine system. Second, detecting high levels of polymorphisms by the AFLP technique (Wan et al., 1999; Huys et al., 1996; Latorra and Schanfield, 1996; Mackill et al., 1996) and lastly the fact that a large fraction of the silkworm genome consists of families of transposable elements such as Bm1, BMC1 (a member of the LINE1 a family of transposable elements), mariner, mariner-like elements (Bmmar1), long terminal repeat transposons (LTRs), non-long terminal transposons (nonLTRs) and others (Shimizu et al., 2000; Wang et al., 2000; Tomita et al., 1997; Robertson and Asplund, 1996; Xiong et al., 1993; Xiong and Eickbush, 1993; Herrer and Wang, 1991; Ueda et al., 1986). A relatively high level (24.14%) of polymorphic AFLP markers in this study compared to the results obtained in barley (Castiglioni et al., 1998) and Guppy fish (Shen et al., 2007) may be due to the previously mentioned three factors.
Among the 20 pairs of AFLP primer combinations applied in this study, an average of 10.2 polymorphic AFLP markers per primer combination for linkage mapping was recognized. This rate was close to that (7.1) of barley (Castiglioni et al., 1998) and significantly higher than those obtained by AFLP linkage map studies in tef (Eragrostis tef (Zucc.) Trotter) (Bai et al., 1999) and red flour beetle (Zhong et al., 2004) which produced on average 4.5 and 4.8 polymorphic fragments per primer combination, respectively. However it was considerably lower than two other AFLP linkage and QTL mapping studies on silkworm with 35.7 (Tan et al., 2001) and 36.4 (Lu et al., 2004) fragments per primer. This may be due to the degree of differences between parental lines and strains and primer combinations.
In this study the total recombination distance over 12 linkage groups was 4262 cM which was longer than previous estimates in silkworm, i.e., 1800 cM for the dense RAPD map (Yasukochi, 1998), 3676.7 cM for the AFLP map in a single backcross family (Lu et al., 2004), 3431.9 cM for the SSR linkage map (Miao et al., 2005), 1868.10 cM and 2677.50 cM for the AFLP maps in two F2 subgroups (Sima et al., 2006) and 1305 cM for SNP-based linkage map (Yamamoto et al., 2006). However it was shorter than the total length of AFLP linkage map (6512 cM) reported in silkworm (Tan et al., 2001). Miao et al. (2005) have suggested that although many conditions influence map length, including differences in mating strategy and strains used, the distribution of markers is a possible causative aspect and increased marker density should converge on a more realistic map length value. It has been demonstrated theoretically that, with additional markers typed, the map length may increase when marker density is not saturated or may decrease when marker density is in a saturation state (Tan et al., 2001). For instance, Causse et al. (1994) have constructed a rice map with 762 markers covering 4026.3 cM, whereas Harushima et al. (1998) have obtained a 2275 marker genetic map of rice covering 1521.6 cM. This may explain why the length of the AFLP map of this study is larger than that of the silkworm linkage map studies mentioned above except for Tan et al.’s (2001) AFLP map. Considering that the estimated genome size of B. mori is 530 Mbp (Gage, 1974), the average physical distance per recombination distance is about 124 kb/cM. It seems that the AFLP markers do not exhibit significant clustering near centromeres or the distal region of chromosomes, suggesting that they provide good coverage of the genome (Zhong et al., 2006) (Fig. 2).
Several publications reported that AFLP markers generated from EcoRI/MseI restriction enzymes tend to cluster around centromere regions (Haanstra et al., 1999; Vuylsteke et al., 1999; Young et al., 1999; Qi et al., 1998). In the present study, strict clustering of AFLP markers generated from PstI/TaqI restriction enzymes was not observed. It is important to note that in this investigation, some AFLP markers generated by the same primer combinations have been mapped to similar positions (Table. 3). It has also been reported that some EcoRI/MseI based AFLP markers generated by the same primer combinations are mapped to similar positions in the linkage groups of red flour beetle (T. castaneum). It may be that these AFLP primer combinations have amplified gene clusters or repeated sequences that are physically linked (Zhong et al., 2004).
The AFLP map of this research consisted of 12 linkage groups whereas the haploid genome of the silkworm has 28 chromosomes. As reported in previous studies, it may be due to nonequivalence between the number of linkage groups and the number of chromosomes (He, 1998; Young et al., 1998; Promboon et al., 1995). In the RFLP based linkage map by Goldsmith (1991), 15 linkage groups were reported. However by using morphological (Doira, 1992), RAPD (Yasukochi 1998), RFLP (Nguu et al., 2005), AFLP (Sima et al., 2006) and SNP (Yamamoto et al., 2006) markers, 28 linkage groups and by using SSR markers (Miao et al., 2005), 29 linkage groups have been recognized in silkworm. It has also been indicated that the large number of chromosomes in the haploid silkworm genome (n=28), typical of lepidoptera, makes it difficult to construct maps without missing some chromosomes (Yasukochi, 1998).
In summary, 204 AFLP markers were employed to construct a linkage map for B. mori, with an average marker resolution of 10.2 cM. Since AFLP amplification is highly reproducible, the development of an AFLP linkage map provides an invaluable tool for studying silkworm genetics, such as identification of strain-specific markers for tracking allele frequency changes and QTL analysis for economically important traits.
This work was supported by funds from the Iran Ministry of Agriculture and the Biotechnology Research Institute-Northern region of Iran. We thank Mr. Moeineddin Mavvajpour (The Director) and Dr. Alireza Seydavi (adviser) at the Iran Silkworm Research Centre (ISRC) for establishing silkworm crosses and phenotypic information. The technical assistance of the staff of ISRC is gratefully acknowledged. We are also grateful to Mohammad Naserani for contribution of data it means typing of phenotypic records in Excel to edit and input data in linkage map and QTLs softwares for further analysis.