Several properties of endogenous retroviruses ( 1) make them useful reagents for the study of evolutionary relationships among species. In a variety of mammalian and avian systems it has been demonstrated that endogenous viral gene sequences are transmitted through the germline and are present in the normal cellular DNA in multiple copies ( 2, 3). As such, they are subjected to the same evolutionary pressures as are cellular DNA sequences. Because, in some species, these viruses can be isolated and transmitted in cell culture, it is possible to readily purify their genomes as well as the proteins produced by the viral RNA. We have previously taken advantage of this property of endogenous primate type C viruses to study the evolutionary relationship between primates ( 4, 5); similar studies have also been performed with other endogenous mammalian type C and type B viruses ( 6, 7, 8). One surprising result of some of the above studies was that transmission of type C viruses does not occur between the germlines of distantly related species ( 9, 10). In other cases, type C viruses are found only in certain animals and in certain tissues of a species and is the result of infection.
In the latter category are the viruses isolated from one woolley monkey with a fibrosarcoma ( 11, 12) and from several gibbons, both normal and leukemic ( 13, 14, 15). DNA transcripts of the viral RNA show homology to rodent and not to primate cellular DNA; in particular, transcripts would hybridize most extensively to the cellular DNAs of various Mus species ( 6). To define more precisely the origin of the woolly monkey-gibbon group of infectious type C viruses, various species of Southeast Asian mice were surveyed for their endogenous type C viruses. Since the gibbons were the primate species that showed widespread infection by this group of viruses, Southeast Asian members of the family Muridae were studied for viruses which may be present that might best account for the transmissible leukemia virus of gibbons ( 16).
In this report, we describe the properties of the various endogenous retroviruses of Southeast Asian species of Mus. These animals, we have found, readily release distinct groups of retroviridae. On the basis of biological and biochemical criteria, two subclasses of endogenous type C viruses could be distinguished. One subclass (C-I) is related to the woolly monkey-gibbon ape infectious primate type C viruses; the other subclass (C-II) is more closely related to laboratory murine leukemia viruses (MuLVs) that are commonly isolated from inbred strains and feral populations of M. musculus ( 6, 17). In addition, a novel class of endogenous retroviruses (designated M432) could be distinguished which is unrelated by immunological and molecular hybridization criteria to all other classes of retroviruses ( 8, 18) and type B particles have been identified in the milk of M. cervicolor mice which are morphologically and antigenically related to M. musculus mammary tumor virus (MMTV) ( 19).
Using viral reverse transcripts we have determined the number of endogenous retrovirus-related sequences in various members of the genus Mus. Taken together, the total complement of retroviral sequences represents a significant portion of the cellular genome, approximately 0.04%. The three classes of viral sequences we have studied (C-I, C-II, and M432) appear to be conserved to different extents, suggesting differing evolutionary pressures on each.
Table 1 lists the species of Mus which have been studied as well as their origin. The mice were initially trapped by Dr. J. Marshall in different provinces of Thailand, except M. dunni from India and M. musculus molossinus from Japan ( 20). Recently, random-bred colonies of certain of these species have been established in laboratories in the United States.
Using M. cervicolor as the point of reference, the evolutionary relationships of the different species were determined. Radioactive nonrepetitive cellular DNA was prepared from a tissue culture cell line derived from M. cervicolor lung tissue ( 6). The labeled DNA was used as a probe to measure nucleic acid sequence homology between the nonrepetitive DNA of M. cervicolor and cellular DNA extracted from pooled organs of different rodents. Table 2 lists the final extent of hybridization (at a C0t of 2 x 104) and the ΔTmR of the hybrid. The ΔTmR is a measure of the thermal stability of the hybrids formed and is the index of the base-pair mismatch between the DNAs of the species being tested. The results reveal a broad range of homology between the M. cervicolor cellular genome and those of the other murine species. By these criteria, M. cervicolor cervicolor, the subspecies M. cervicolor popaeus, and M. cookii are all closely related. Of the other species tested, M. caroli, M. musculus, and M. dunni are more distantly related to M. cervicolor. The two least related species are M. pahari and M. shortridgei. Labeled nonrepetitive cellular DNA from M. shortridgei, however, when hybridized to mouse, rat, and hamster cellular DNAs, showed more homology to mouse DNA than to either rat or hamster, consistent with it being a member of the genus Mus (data not shown). The evolutionary relationship of various Mus species with Mus cervicolor as the point of reference is summarized in the phylogenetic tree of the Muridae family shown in Figure 1. This is based on the final extent of hybridization of nonrepetitive cellular DNA, since it is difficult to assign time scales to the divergence points in the absence of paleontological references.
Three different classes of endogenous retroviruses have been isolated from Asian species of Mus. These viruses were obtained either from tissue culture cell lines which spontaneously release virus or by induction of tissue culture cell lines with the pyrimidine analog, BUdR, followed by mixed cell cultivation with sensitive cell lines chosen for their permissiveness for the replication of retroviruses.
On the basis of biological and biochemical criteria, two subclasses of type C virus could be identified. Subclass C-I replicates only in nonmurine tissue culture cell lines, whereas subclass C-II can replicate in cell lines derived from different species of the genus Mus. Table 3 shows the host ranges of representatives of each subclass, isolated from M. cervicolor tissue culture cells. By immunological criteria, the viral-associated proteins of the two subclasses are readily distinguishable. The major structural protein p30 of the C-I subclass competes in a woolly monkey-gibbon ape group-specific radioimmunoassay ( 17), whereas the respective proteins of the C-II subclass show no competition ( Figure 2A). In this assay, the slope of the competition curve can be used as a measure of relatedness between the labeled test antigen and the competing protein. Thus, by this criterion, the p30 proteins of the type C-I viruses exhibit varying degrees of relatedness to the test protein derived from the woolly monkey virus, with the M. shortridgei virus being the most distantly related. Other than the various isolates of MuLV from feral and inbred strains of M. musculus, the only other members of the C-II subclass are virus isolates we have obtained from M. cervicolor and M. cookii cell cultures. The viral-associated p30 proteins of this subclass compete in a m. musculus MuLV group-specific radioimmunoassay, albeit the M. cervicolor and M. cookii viral proteins compete with a reduced slope ( Figure 2B). Although the p30 proteins of these two subclasses of viruses can be distinguished by their unique antigenic determinants, they do share other determinants in common. This can be demonstrated by using a rodent interspecies assay in which the p30 proteins of both subclasses of murine type C viruses compete with the test antigen, while the rat type C viral protein does not ( Figure 2C). A similar pattern between the two subclasses of viruses can be demonstrated by the ability of antisera against the viral-associated reverse transcriptase of the woolly monkey virus or R-MuLV to inhibit enzymatic activity of the respective subclass of viruses ( 6, 21).
Approximately half of the viral genome of type C viruses is believed to contain the coding sequences for the viral structural proteins (the gag proteins) and reverse transcriptase (the pol gene product) ( 22). Using labeled viral cDNA probes representing the entire viral genome, the relationship between the M. cervicolor C-I and C-II subclass virogenes was measured (data not shown). In agreement with the immunological data, the C-I and C-II virogenes share very little nucleic acid homology (approximately 10% under the conditions used). Moreover, the M. cervicolor C-I probe hybridizes partially with the M. cookii and M. caroli C-I viral RNAs and to a lesser extent with M. dunni and M. shortridgei as well as SSAV and GALV-1 viral RNAs. In contrast to these results, the M. cervicolor C-II probe did not hybridize appreciably with either of the viruses from M. cookii or M. musculus; this suggests that the C-II subclass of virogenes has been less evolutionarily conserved than the virogenes of the C-I subclass.
A third class of endogenous retroviruses has been isolated from two Asian rodents, M. cervicolor and M. caroli. Members of this class so far have only replicated in cell culture lines derived from M. musculus ( Table 3). They share certain morphological and biochemical properties in common with M. musculus type B mammary tumor virus. Cells producing these viruses contain cytoplasmic A particles and budding forms contain doughnut-shaped nucleoids ( Figure 3). However, the mature particles lack the characteristic surface spikes of MMTV and have a centrally located nucleoid. This group of viruses contains two envelope glycoproteins (gp65, gp32), a major internal protein (p24), as well as p16 and a phosphoprotein (p12). The viral-associated reverse transcriptase, with a molecular weight of 70,000 daltons, is smaller than the M. musculus type B viral enzyme ( 23) although both share a preference for magnesium to fulfill their divalent cation requirement for activity ( 8).
Antisera raised against the M. cervicolor p24 protein show a partial cross-reactivity with the corresponding protein of the M. caroli virus (M832) ( Figure 4). In this assay, MMTV, bovine leukemia virus (BLV), Rauscher MuLV (R-MuLV), the Mason-Pfizer monkey virus (MPMV), and both M. cervicolor type C viruses all fail to compete with the test antigen. Similar results were obtained with antisera prepared against the M. cervicolor M432 viral-associated reverse transcriptase. In addition, by molecular hybridization criteria, the M. cervicolor M432 viral genes are unrelated to M. musculus MMTV, both murine type C viral subclasses, as well as MPMV and BLV ( 8). Table 4 summarizes the classes of viral isolates obtained from each species. With the one exception, M. musculus, viruses of the type C subclass C-I have been isolated from every species tested. Type C subclass C-II viruses have been isolated from only M. musculus, M. cervicolor, and M. cookii. The M432-related viruses have been isolated from M. cervicolor and M. caroli, while the type B class contains viral isolates from M. musculus. A virus morphologically and antigenically related to the MTV subclass of viruses has recently been observed in the milk of M. cervicolor mice ( 19).
The cellular DNA of Asian species of Mus were examined for the presence of nucleic acid sequences related to the genomes of the M. cervicolor C-I, C-II, and M432 subclasses of viruses. Table 5 summarizes the final extents of hybridization and the thermal stabilities of the hybrids formed with DNA from the listed species and labeled M. cervicolor viral DNA transcripts synthesized in the endogenous polymerase reaction. As shown, viral-related nucleic acid sequences corresponding to the three subclasses could be detected in all of the species of Mus as well as more distantly related rodents. All three viral DNA probes hybridized most extensively and with the highest thermal stability to cellular DNA from Mus cervicolor tissues. The extent to which the viral sequences were detected in other related Mus species depended both on the genetic distance from M. cervicolor DNA, as determined by the nonrepetitive cellular DNA hybridization studies (see Table 2) and on the viral subclass. Thus the M. cervicolor C-I and M432 subclasses of virogenes are conserved with a pattern similar to the nonrepetitive cellular DNA of this species. The fact that the final extents of hybridization and thermal stabilities of the hybrids formed decrease in parallel indicates that the partial homologies detected are not the result of the conservation of discrete, contiguous regions of the viral genome. Rather, it suggests that the differences are due to base-pair substitution throughout much of the viral genome. The virogene sequences related to the M. cervicolor C-II subclass appear much more species-specific. In species more distantly related to M. cervicolor than M. cookii, only low levels of homology can be detected. A similar result has been found using viral DNA probe from M. musculus C-II viruses (either BALB/c ecotropic or BALB/c xenotropic viruses) ( 6). Thus, there seems to be less rigorous evolutionary conservation of virogene sequences for this subclass of type C viruses.
It has been shown in a number of mammalian and avian species that the genomes of endogenous retroviruses are represented in the cellular DNA in multiple copies. Using labeled viral DNA, the frequency of viral-related sequences of each M. cervicolor viral subclass was measured. Both the type C-I and C-II virogenes are reiterated approximately six times per haploid cellular genome, while the type M432 virogenes are reiterated approximately 25 times ( Figure 5). Recently, Morris et al. have estimated that there are at least 25 copies of type B virogenes (MMTV) per haploid genome in M. cervicolor ( 7). The summaries of our estimates of the reiteration frequency of each subclass of virogenes as well as mouse ribosomal RNA genes ( 24) are compared in Table 6. The approximate genetic complexity of a single viral genome is 10,000 base pairs and that of the cellular genome is 1.5 x 109 base pairs/haploid genome ( 25). Thus at least 0.04% of the M. cervicolor cellular genome contains the genetic information for endogenous retroviral-associated proteins. This value is comparable to the genetic complexity of the structural genes coding for 28S and 18S ribosomal RNAs. It is therefore fair to conclude that the retroviral genes represent a significant portion of the cellular genome. However, these represent minimal estimates as additional subclasses of retroviruses may yet be isolated from one or more of these Mus species.
On the basis of nucleic acid homology, it has been possible to establish the evolutionary relationships between a variety of Southeast Asian species of Mus ( 6, 26). We have placed the species of Mus so far examined into three groups based on the nucleic acid homology they share with the point of reference, the M. cervicolor cellular genome. On this basis, M. cervicolor cervicolor, M. cervicolor popaeus, and M. cookii are highly related, whereas M. carolii, M. dunni, and M. musculus are a somewhat more distantly related group. The M. pahari and M. shortridgei cellular genomes share the least homology with that of M. cervicolor.
The Mus cellular genome contains the genetic information for type B, at least two subclasses of type C, and a novel class (M432) of endogenous retroviral genes. Each of the four classes is present in the cellular genome in multiple copies. In M. cervicolor the total complement of endogenous retroviral sequences accounts for a minimum of 0.04% of the cellular genome. This calculation takes into account only the genetic information corresponding to the viral-associated structural proteins. Additional genetic information would be required to account for the various mechanisms the cell has evolved to regulate the expression of these virogenes. Although the absolute copy number of a given subclass of virogenes may vary with the species of Mus, it is apparent that they represent a significant portion of the cellular genome.
The evolutionary conservation of these classes of endogenous retroviral genes is not uniform. The most highly conserved is subclass C-I which has evolved at a rate comparable to the nonrepetitive cellular DNA sequences. The M432 class of virogenes, although highly conserved, seems to have evolved at a faster rate than the C-I subclass. The C-II subclass is the least conserved. By nucleic hybridization there is little homology between the M. cervicolor type C-II virogenes within members of this species and even less in more distantly related species. Similarly, the various viral isolates of type C-II from M. musculus also share varying degrees of nucleic acid homology between their genomes ( 27, 28). The relationship between members of this subclass is more easily demonstrated by the shared unique antigenic determinants of the p30 and reverse transcriptase proteins.
The existence of two subclasses of endogenous type C virogenes in the murine cellular genome is not unique to the genus Mus. By the criteria of nucleic acid hybridization, the rat cellular genome contains two subclasses of unrelated endogenous type C virogenes ( 29, 30). The domestic cat and the Old World monkey genomes ( 5, 31) also have been shown to harbor multiple copies of two or more distinctly related retroviral genomes. The extensive genetic and immunologic data available in the mouse (Mus musculus) and its close relatives ( 20, 32) make it especially attractive for studies on the evolution and function of these viral gene sequences.
1. Dalton, A.J., Melnick, J.L., Bauer, H., Beaudreau, G., Bentvelzen, P., Bolognesi, D., Gallo, R., Graffi, A., Haguenau, F., Heston, W., Huebner, R., Rowe, W., and Todaro, G.J. (1974). Intervirology 4: 201.
2. Todaro, G.J., Benveniste, R.E., Callahan, R., Lieber, M.M., and Sherr, C.J. (1974).
Cold Spring Harbor Symp. Quant. Biol. 39: 1159.
See also
PubMed.
3. Aaronson, S.A., and Stephenson, J.R. (1976). Biochim. Biophys. Acta 458: 323.
See also
PubMed.
4. Benveniste, R.E., and Todaro, G.J. (1976). Nature 261: 101.
See also
PubMed.
5. Todaro, G.J., Benveniste, R.E., Sherwin, S.A., and Sherr, C.J. (1978) Cell 13: 775.
See also
PubMed.
6. Benveniste, R.E., Callahan, R. Sherr, C.J., Chapman, V. , and Todaro, G.J. (1977). J. Virol. 21: 849.
See also
PubMed.
7. Morris, V.L., Medeiros, E., Ringold, G.M., Bishop, J.M., and Varmus, H.E. (1977). J. Mol. Biol. 114: 73.
See also
PubMed.
8. Callahan, R., Sherr, C.J., and Todaro, G.J. (1977). Virology 80: 401.
See also
PubMed.
9. Benveniste, R.E., and Todaro, G.J. (1974). Nature 252: 456.
See also
PubMed.
10. Benveniste, R.E., Sherr, C.J., and Todaro, G.J. (1975). Science 190: 886.
See also
PubMed.
11. Theilen, G.H., Gould, D., Fowler, M., and Dungworth, D.L. (1971). J. Natl. Cancer Inst. 47: 881.
See also
PubMed.
12. Wolfe, L.G., Deinhart, F., Theilen, G.H., Rabin, H., Kawakami, T., and Bustad, L.K. (1971). J. Natl. Cancer Inst. 47: 1115.
See also
PubMed.
13. Kawakami, T., Huff, S.D., Buckley, P.M., Dungworth, D.L., Snyder, S.P., and Gilden, R.V. (1972). Nature New Biol. 235: 170.
See also
PubMed.
14. Kawakami, T.G., and Buckley, P.M. (1974). Transplant. Proc. 6: 193.
See also
PubMed.
15. Lieber, M.M., Sherr, C.J., Benveniste, R.E., and Todaro, G.J. (1975). Virology 66: 616.
See also
PubMed.
16. Charman, H.P., Kim, N., White, M., Marquardt, H., Gilden, R.V., and Kawakami, T. (1975). J. Natl. Cancer Inst. 55: 1419.
See also
PubMed.
17. Lieber, M.M., Sherr, C.J., Todaro, G.J., Benveniste, R.E., Callahan, R., and Coon, H.G. (1975). Proc. Natl. Acad. Sci. USA 72: 2315.
See also
PubMed.
18. Callahan, R., Benveniste, R.E., Sherr, C.J., Schidlovsky, G., and Todaro, G.J. (1976). Proc. Natl. Acad. Sci. USA 73: 3579.
See also
PubMed.
19. Schlom, J., Hand, P., Teramoto, Y.A., Callahan, R., and Todaro, G., and Schidlovsky, G. (1978). J. Natl. Cancer Inst. 61: 1509.
See also
PubMed.
20. Marshall, J.T. (1977). Bull. Amer. Mus. Nat. Hist. 158: 173.
21. Sherr, C.J., Fedele, L.A., Benveniste, R.E., and Todaro, G.J. (1975). J. Virol. 15: 1440.
See also
PubMed.
22. Baltimore, D. (1974). Cold Spring Harbor Symp. Quant. Biol. 39: 1187.
See also
PubMed.
23. Dion, A., Vaidya, A., Font, G.S., and Moore, D.H. (1974). J. Virol. 14: 40.
See also
PubMed.
24. Gaubatz, J., and Cutler, R.G. (1975). Biochemistry 14: 760.
See also
PubMed.
25. Laird, C. (1971) Chromosoma 32: 378.
See also
PubMed.
26. Rice, N.R., and Strauss, N.A., (1973). Proc. Natl. Acad. Sci. USA 70: 3546.
See also
PubMed.
27. Callahan, R., Benveniste, R.E., Lieber, M.M., and Todaro, G.J. (1974). J. Virol. 14: 1394.
See also
PubMed.
28. Callahan, R., Lieber, M.M, and Todaro, G.J. (1975). J. Virol. 15: 1378.
See also
PubMed.
29. Scolnick, E.M., Goldberg, R.J., and Williams, D. (1976). J. Virol. 18: 559.
See also
PubMed.
30. Scolnick, E.M., Williams, D., Maryak, J., Vass, W., Goldberg, R.J., and Parks, W.P. (1976). J. Virol. 20: 570.
See also
PubMed.
31. Benveniste, R.E., and Todaro, G.J. Primate and Human Cancer Meeting, San Antonio, Texas, in press.
32. Chapman, V.M., and Shows, T.B. (1976). Nature 259: 665.
See also
MGI.
33. Benveniste, R.E., Heinemann, R., Wilson, G.L., Callahan, R., and Todaro, G.J. (1974). J. Virol. 14: 56.
See also
PubMed.
34. Benveniste, R.E., and Todaro, G.J. (1974). Proc. Natl. Acad. Sci. USA 71: 4513.
See also
PubMed.
35. Ando, T. (1966). Biochim. Biophys. Acta 114: 158.
See also
PubMed.
36. Benveniste, R.E., and Scolnick, E.M. (1973). Virology 51: 370.
See also
PubMed.
37. Britten, R.J., and Kohne, D.E. (1968). Science 161: 529.
See also
PubMed.
38. Britten, R.J., and Smith, J. (1970). Carnegie Inst. Washington Yearbook 68: 378.
39. Kohne, D.E. (1970). Q. Rev. Biophys. 3: 327.
See also
PubMed.
40. Jainchill, J.L., Aaronson, S.A., and Todaro, G.J. (1969). J. Virol. 4: 549-553.
See also
PubMed.
41. Benveniste, R.E., and Todaro, G.J. (1977). Proc. Natl. Acad. Sci. USA 74: 4557.
42. Hartley, J.W., and Rowe, W.P. (1975). Virology 65: 128.
See also
PubMed.
43. Birnstiel, M., Chipchase, M., and Speirs, J. (1971). Progr. Nucleic Acids Res. Mol. Biol. 11: 351.
See also
PubMed.