The divergency of endogenous mouse mammary tumor virus (MMTV) in European mouse strains indicates acquisition of exogenous MMTV into their genomes during evolution. In the present study, we selected two strains of Asian wild mice, Cas-Hmi (Mus musculus castaneus) and Sub-Kjr (M. m. subspecies) to investigate a possible divergency of endogenous MMTVs among Asian mouse strains. Southern blot analysis and polymerase chain reaction (PCR) demonstrated that Cas-Hmi and Sub-Kjr have the full structure of integrated endogenous MMTVs, suggesting that these mice were diverged during evolution after MMTV endemics.
Mouse mammary tumor virus (MMTV), a replication-competent B-type retrovirus, is known to induce mammary carcinomas in a number of mouse strains. Like other retroviruses, the MMTV genome contains gag (encodes the core protein), pol (RNA dependent DNA polymerase), env (the envelope protein) and long terminal repeats (LTRs) at both 5∣ and 3∣ ends (Salmons and Günzburg, 1987). It is the characteristics of MMTV proviruses to have a single EcoRI restriction site. Cleavage pattern is, thus, utilized as a basis of nomenclature of the MMTV proviruses in the mouse genome (Cohen and Varmus, 1979; Kozak et al., 1987).
MMTV can be transmitted 1) exogenously as a virus particle contained in milk and saliva from mother to the off-spring, and 2) endogenously as an integrated provirus via germ line (Salmons and Günzburg, 1987; Held et al., 1994). The whole life cycle of exogenous MMTV, however, from transmission through the gut epithelium in infancy to produc-tion of mature MMTV particles in lactating mammary glands, remains to be elucidated, though 3∣ LTR encodes superantigen (Held et al., 1994) and thus the immune system is known to play a central role during its life cycle. These studies have been conducted in European mouse strains. On the other hand, endogenous MMTV loci in the genomes of Asian wild mice remain fragmentary.
Imai et al. (1994) recently reported that distribution of endogenous MMTV in several groups of Asian wild mice (Mus musculus) free of exogenous MMTV endemics. In this study, we obtained two strains of Asian wild mice, Cas-Hmi and Sub-Kjr. Cas-Hmi mice, trapped from He-mei, Taiwan, are classified as M. m. castaneus. Sub-Kjr mice, from Kojuri, Republic of Korea, remain unclassified from M. musculus (M. m. subspecies).
The aim of the present study is to investigate an evolutionary event of acquisition of endogenous MMTV. We considered the evolutionary divergency of MMTV, comparing the reports on Asian wild mice using other genetic markers.
MATERIALS AND METHODS
Male and female mice of two strains of Asian wild mice, Cas-Hmi and Sub-Kjr, were provided by Dr. T. Shiroishi (National Institute of Genetics). Cas-Hmi is divided into two strains. Strain 1 has full sequences of MMTV genome (size of EcoRI fragment: MMTV-LTR probe 7.9 kb or 7.9 and 28 kb, -env probe 7.9 kb, -gag-pol probe 7.9 kb or 7.9 and 28 kb) but strain 2 is completely free from MMTV integration (Imai et al., 1994). Mice obtained were not identifiable as either of the two. Sub-Kjr is reported to have a defective MMTV locus containing the 1.0 kb env portion and is reported to lack the rest of MMTV portions (LTR, gag-pol) (Imai et al., 1994). These mice and female C57BL/6 (purchased from Nippon SLC, Shizuoka, Japan) were kpt in plastic cages with wood shavings at 23°C, a relative humidity of 50 % and an LD 12:12 photoperiod in our animal facility.
DNA isolation from tissue
Mice were sacrificed by cervical dislocation. A part of the liver was homogenized in homogenizing buffer (0.1 M NaCI, 0.2 M sucrose, 0.01 M EDTA and 0.3 M Tris-HCI [pH 8.0]) and incubated at 65°C for 1 hr in the presence of 10% SDS. Then the liver DNA was extracted by the phenol-chloroform extraction method as described by Davis et al. (1986). The total amount of DNA was quantified spectrophotometrically as described Sambrook et al. (1989). TheDNA was stored at −20°C until used for Southern blotting and Polymerase chain reaction (PCR).
The pBR322 derived plasmids, VG027 and VG037, containing the LTR (Majors and Varmus, 1981) and a part of the MMTV genome (Ucker et al., 1983) respectively, were provided by Japanese Cancer Research Resources Bank (Tokyo, Japan). Plasmids were isolated from an Escherichia coli host according to Sambrook et al. (1989). A 1 kb fragment of the MMTV LTR portion obtained by digestion of plasmid VG027 with BamHI (Takara, Tokyo; Toyobo, Tokyo, Japan) and Clal (Takara, Tokyo, Japan), and a 1 kb BamHI fragment of the MMTV env portion from plasmid VG037 were used as hybridization probes. These probes were 32P-labeled with a random primer DNA labeling kit (Takara, Tokyo, Japan) to a specific activity of at least 5 × 107 cpm/μg.
Southern blot analysis
Isolated mouse genomic DNA was digested with 10-fold excess amount of restriction enzymes (EcoRI or BamHI), using conditions recommended by the manufacturer. Digested DNA samples (ca. 15 μg) were subjected to 0.8% agarose gel electrophoresis containing 89 mM Tris-borate, 89 mM boric acid and 2 mM EDTA [pH 8.0] at 32 V for 16 hr in an electrophoresis buffer (TBE buffer: 89 mM Tris-borate, 89 mM boric acid and 2 mM EDTA [pH 8.0]). λDNA digested with HindIII (Toyobo, Tokyo, Japan) was used as a molecular size marker. After electrophoresis, the gel was stained with 0.5 μg/ml ethidium bromide solution and transferred to a nylon membrane (Gene Screen Plus, Du Pont, Boston, USA) overnight by a capillary blot procedure (Sambrook et al., 1989) using a transfer buffer (1.5 M NaCl and 0.25 M NaOH) at room temperature. The membrane was UV-linked and pre-hybridized in 15 ml of hybridization solution (1% SDS, 1 M NaCl, 10% dextran sulfate, 8% Denhardt's Solution and 133 μg/ml salmon sperm DNA) at 65°C for 6 hr. Then excess buffer was removed and replaced by hybridization solution containing heat denatured 32P-labeled probe (>107 cpm), and the membrane was incubated at 65°C overnight. The membrane was washed by the method recommended by the membrane supplier (Protocol for electrophoresis and capillary transfer of DNA, Du Pont).
Amplification by polymerase chain reaction (PCR)
A pair of forward (5∣) and reverse (3∣) primers (20 mer each) for the MMTV gag and pol portion were designed with paying attention to Tm value, CG contents and avoidance of misannealing and so on. They were synthesized by a commercial supplier (Nippon Bio Service, Saitama, Japan). Isolated mouse liver DNAs of Cas-Hmi, Sub-Kjr and C57BL/6 were diluted into 100 ng/μl and used as template DNAs. PCR buffer (10 mM Tris-HCI [pH 8.3], 50 mM KCl, 1.5 mM MgCI2 and 0.001% galatin), 200 μM dNTP mixture, 30 μM of 5∣ and 3∣ primers, 50 ng of template DNA and 1.25 U of recombinant Taq DNA polymerase (Takara, Tokyo, Japan) were mixed into 50 μl. The PCR profile was as follows: 94°C for 1 min for denaturation of DNA, 56°C (for the MMTV gag portion) or 60°C (for the MMTV pol portion) for 1 min for primer annealing and 72°C for 1 min for DNA extension. The PCR profile was repeated for 40 cycles.
Detection of PCR product
Primers were designed to amplify an approximately 300 base-pair DNA fragment of the MMTV gag and pol portions. One third of PCR products of the MMTV gag and pol portions were electrophoresed on a 1% agarose gel at 100 V for 1.5 hr. HaeIII-digested ?×174 DNA (Toyobo, Tokyo, Japan) was used as a molecular size marker. After electrophoresis, the gel was stained with ethidium bromide solution for 30 min and checked under UV light.
RESULTS AND DISCUSSION
We detected endogenous MMTV proviruses in genomic DNA of liver from Cas-Hmi and Sub-Kjr. In Southern blot analysis, using MMTV-LTR and -env probe, only a 6.7 kb fragment was detected in EcoRI-digested genomic DNA from both of Cas-Hmi and Sub-Kjr (Fig. 1). This 6.7 kb frag-ment corresponds in size to EcoRI-digested 3∣ fragment of Mtv-8 (6.7 kb) and was found in genomic DNA from C57BL/6 as expected (Fig. 1). This may suggest the similarity of en-dogenous MMTV from Cas-Hmi and Sub-Kjr with Mtv-8.
We next digested genomic DNA from Cas-Hmi and Sub-Kjr with BamHI to characterize this locus. The Mtv-8 endogenous provirus has a single BamHI site close downstream to a single EcoRI site (Salmons and Günzburg, 1987) (see Fig. 2). The MMTV-LTR probe hybridized to 7.2 and 4.3 kb fragment and the latter was also hybridized with the MMTV-env probe (Fig. 1). Thus, 7.2 and 4.3 kb fragments were considered as 5∣ and 3∣ products, respectively.
The results of Southern blotting of genomic DNA after EcoRI- or BamHI-digestion (Fig. 1) suggest that Cas-Hmi and Sub-Kjr have an akin MMTV provirus. They apparently appeared to have a single EcoRI fragment of 6.7 kb. However, as shown in Fig. 2, this 6.7 kb EcoRI fragment was found to be a mixture of a similar size of two fragments, because this MMTV provirus was digested into 7.2 and 4.3 kb fragments with BamHI. Therefore, the endogenous MMTV provirus in the genomes of Cas-Hmi and Sub-Kjr is complete MMTV(s) and is a novel MMTV locus which has never been reported. This result, however, is not in agreement with the reported results (Imai et al., 1994). Their results are indicative of the absence of endogenous MMTV in these strains. Possibilities to explain the discrepancy of the results are to assume individual and/or geographical variations. As suggested by Imai et al. (1994) that one of two strains of Cas-Hmi captured in the same geographic area had endogenous MMTV but another did not, individual variations may likely explain the discrepancy.
In order to analyze the structures of MMTV gag and pol portion in two strains of Asian wild mice, we amplified an approximately. 300 bp of MMTV gag or pol portion by PCR, using template of Cas-Hmi or Sub-Kjr liver DNA. Using Cas-Hmi DNA, both MMTV gag and pol portions were amplified but the former was not produced from Sub-Kjr DNA (Fig. 3), suggesting that deletion and/or mutation may occur in the gag portion of the Sub-Kjr DNA.
Classification of various subspecies of M. musculus in and around Asia has recently been actively in progress. The various molecular criteria using mitochondrial DNA (mtDNA) (Yonekawa et al., 1982, 1988), murine leukemia virus (MuLV) (Inaguma et al., 1991), ribosomal DNA (Suzuki et al., 1986), and H-2 (Moriwaki et al., 1990) have been employed. From the data of restriction patterns of mtDNA, Sub-Kjr are likely assigned to M. m. musculus (Yonekawa et al., 1982, 1988). Yonekawa et al. (1982) also estimated the time of divergence of M. m. castaneus and M. m. musculus (Bulgarian race) as 102 × 104 years by analyzing polymorphism of mtDNA as an evolutionary marker. It is of note that endog-enous MMTVs from Cas-Hmi and Sub-Kjr appeared to be close, though a minor difference is shown in the sequence of gag portion. If so, the timing of acquisition of endogenous MMTV by Cas-Hmi and Sub-Kjr is considered to be close. Although more direct evidence on this issue is obviously needed, we are tempted to hypothesized that segregation of these two strains likely occurred after the acquisition of endogenous MMTV. A research plan to confirm our speculation with various genetic markers other than MMTV are currently under way in our laboratory.
We are grateful to Dr. Toshihiko Shiroishi, National Institute of Genetics, and Dr. Hiromichi Yonekawa, Tokyo Metropolitan Instituteof Medical Science, for providing us Cas-Hmi and Sub-Kjr mice and their good advices.