Previous Next

Biochemical Polymorphisms in Wild Mice

Verne M. Chapman

Department of Molecular Biology
Roswell Park Memorial Institute
Buffalo, New York

The development of electrophoretic techniques coupled with histochemical staining on gels for specific enzyme reaction has provided a powerful tool for evaluating qualitative differences in primary gene products. These techniques have been widely used in mouse genetics to describe new loci and to define the linkage of the structural genes for numerous proteins and enzymes ( 1). This work has resulted in a substantial growth and refinement of the genetic map of the mouse over the past several years.

The ability to qualitatively identify the products of structural genes has also provided a means of examining wild or feral mice for genetic variation. In this regard, several workers interested in the genetic structure of natural populations and the issues of population genetics have surveyed mouse populations in various parts of the world [reviewed by Chapman and Selander ( 2)]. Several generalizations about genetic variation in mice can be drawn from these surveys. First, most mouse populations sampled are highly polymorphic. Typically, about 30 to 40% of the loci examined are variable. Second, while a few loci show a striking tendency to be polymorphic in many of the samples, there is a pronounced difference between populations in the loci that are polymorphic and in the alleles that are present. Third, there are many allelic forms found in wild mice that have not been observed among laboratory strains and stocks of mice. Furthermore, in many cases, genetic variation for a particular gene has been found only in wild mice or in mice recently derived from the wild. That is, the standard inbred strains are monomorphic. The contribution of wild or wild-derived mice to genetic variation at specific loci can be determined from various listings of biochemical genetic variation in the mouse ( 3, 4).

The purpose of this paper is to focus on the wild mouse as a valuable genetic resource that can be used in conjunction with the laboratory strains for many aspects of mouse genetics. In addition to the characters that can be seen qualitatively determined such as electrophoretic mobility differences, it is also possible to use wild mouse populations to look for quantitative variation in enzyme activity levels. In this regard, wild mice have been studied for quantitative variation in α-amylase ( 5), β-galactosidase (Nielson and Chapman, unpublished), and androgen inducibility of kidney β-glucuronidase (Nielson and Chapman, in preparation). Work on these systems is of particular interest because it extends the use of wild mouse populations from a source of structural variants to asking questions concerning variation in gene regulation in natural populations. It also involves using wild populations to look for recombination events between closely linked elements which might be difficult to realize in laboratory experiments. Finally, because wild mice can serve a variety of purposes in biochemical genetics, I would also like to briefly consider different breeding programs for maintaining and using wild-derived mice in the laboratory.

Biochemical Genetic Variation in Wild Mice

Extensive surveys of wild mouse populations for biochemical genetic variation have been conducted in Great Britain, the Faroe, Shetland and Orkney archipelagoes ( 6), in Denmark ( 7) and in North America ( 8, 9, 10). Estimates of genetic variation vary from 5 to 17 loci examined (29%) ( 10) to 17 of 41 (41%) ( 7). These findings and the results of limited samplings of the Asian subspecies M. m. castaneus ( 11) from Thailand and M. m. molossinus from Japan clearly demonstrate that substantial amounts of genetic variation can be found within and between wild populations of M. musculus.

A more striking result is that the potential array of genetic variation in the mouse gene pool is so large. This variation is partially observed among inbred strains and stocks of mice ( 1, 3, 4, 12). To date, about 90-100 biochemical loci have been characterized in the mouse. A partial list of those loci which are characterized by variation from wild mice is shown in Table 1, part 1. A second group of loci are listed where uncommon alleles have been observed in wild mice.

Many of the biochemical loci in the mouse are characterized by quantitative variation and primarily in surveys of inbred mice. Thus, of the total loci characterized, wild-derived variation accounts for over a third of the total reported. However, if only those loci are considered where wild mice have been examined, the contribution of wild-derived alleles to the total genetic variation is more than 60%. These results clearly indicate that the continued growth of the biochemical genetic map can be significantly aided by a systematic sampling of wild mouse populations.

The heterogeneous nature of wild mice is both an advantage and a limitation. In general, we have been confined to genetic characters that can be examined qualitatively such as the electrophoretic mobility of proteins. In the past, much of the electrophoretic variation has been detected by means of starch gel techniques. Most of the phenotypes have been codominant and presumably represent variation in the structural gene. Recent work on two enzymes, acid phosphatase ( 13, 14) and α-mannosidase ( 15, 16), shows that electrophoretic variation can also be present as a recessive-dominant trait. These mutants presumably represent variation in post-translational modification of proteins.

These observations indicate that certain precautions are warranted in surveys of wild mice for new electrophoretic variation. To distinguish between structural loci and processing mutants it is essential to know the kinds of phenotypes present and their distribution in a sample. Furthermore, conventional genetic crosses with standard inbred strains are essential to establish the mode of inheritance and linkage.

In the past, we have been primarily concerned with structural gene variation. However, the development of more sophisticated electrophoretic techniques involving systematic alterations of acrylamide concentrations holds considerable promise for identifying a wide array of processing mutants that might not have been detected in conventional starch gel electrophoretic systems ( 17, 18). A reexamination of both laboratory and wild-derived mice using this technology should substantially increase the number of biochemical markers known in the mouse genome.

Sampling Wild Mice for Variation in β-Glucuronidase Regulation and Structure

The molecular biology and genetics of β-glucuronidase have been extensively studied in the laboratory mouse ( 19). An important feature of mouse β-glucuronidase is that it is androgen inducible in the proximal tubule cells of adult mice. Female mice show a 10- to 30-fold increase in kidney β-glucuronidase within seven days following androgen treatment. This increased activity is an increase in the amount of β-glucuronidase which is the result of an increased rate of synthesis.

Three different induction phenotypes are observed among inbred mice. Two of these phenotypes characterized in the strains C57BL/6J and A/J have been studied in detail ( 20). C57BL/6J is a low inducibility strain (Gurb) and A/J is a high inducibility strain (Gura). These strains differ in induced β-glucuronidase activity by almost threefold at seven days ( Table 2). The F1 values are intermediate. The difference in β-glucuronidase induction segregates as a single gene in backcrosses and the F2 generation, and it is closely linked to the structural locus, Gus, on chromosome 5. The electrophoretic phenotype of the induced F1 kidney is consistent with a cis acting regulatory site closely associated with the structural gene.

In the initial genetic crosses 3/115 presumptive recombinants between the structural gene and the regulatory site were observed. However, the mice were killed as part of the test procedure and it was not possible to verify them as true recombinants by breeding tests. In a subsequent test of 1,042 backcross progeny no recombinants were observed ( 21). These data indicate that the Gur and Gus sites are very closely linked and are probably separated by less than 0.7 recombination units.

Among inbred mice strains those that are Gusa, fast electrophoretic form, are Gura, high inducibility phenotypes. Conversely, nearly all of the strains that are Gusb, slow electrophoretic phenotype, are associated with the low inducibility phenotype, Gurb.

In 1975 J. Tønnes Nielson, the Genetics Institute, University of Aarhus, Aarhus, Denmark, and I examined wild trapped mice from Denmark to determine whether recombinant type chromosomes could be recovered from wild populations. Specifically, we asked the following questions:

  1. Is β-glucuronidase polymorphic in wild mice?
  2. Do recombinant types exist for inducibility and structure, specifically, Gusa-Gurb or Gusb-Gura
  3. What array of Gus-Gur combination exists in wild mice? That is, are there new phenotypes not found in laboratory mice?

Our strategy was to androgen induce wild trapped females directly and test for β-glucuronidase activity levels in kidneys at seven days and the electrophoretic phenotypes. Wild males were mated to a C3H strain (Gusb Gurb) maintained in Aarhus. C3H x wild F1 female progeny were tested to determine androgen inducibility levels at seven days and their electrophoretic mobility. Backcrosses, F2 and F3 generations, were examined for cosegregation of structural gene variation and inducibility levels.

Two inducibility-structural gene combinations were found that are not present among laboratory mice ( Table 3). These include a fast electrophoresis low inducibility type (Gurb Gusa) and a new very high induction phenotype associated with fast electrophoresis (Gura Gusa).

Backcross type matings of the C3H x wild F1 (Gurb/b Gusa/b) with the strain C57BL/6J (Gurb/b Gusb/b) segregated for electrophoretic mobility 16/32 : 16/32 which is consistent with the expected 1 : 1 ratio for a single gene. The electrophoretic patterns of the heterozygote are consistent with codominant expression of subunits for the tetrameric β-glucuronidase.

Seven day induction levels of glucuronidase activity did not segregate ( Table 3). When induction levels of heterozygotes (Gusa/b) and homozygotes (Gusb/b) are compared, there is no significant difference. By contrast, induction levels of Gura/b Gusa/b and Gurb/b Gusb/b females from a congenic backcross are clearly different.

Electrophoretic phenotypes of the heterozygous Gura/b Gusa/b from the congenic backcross and the Gurb/b Gusa/b from the wild-derived mice also differ. The Gura/b phenotype is predominantly GUS-A subunits shifting the distribution of activity to 4A and 3A1B multimeric forms. However, in the wild-derived phenotype the electrophoretic pattern is consistent with an equal synthesis of the Gusa and Gusb type subunits. That is, a binomial expression of (.5 Gusa + .5 Gusb)4.

Additional genetic and structural studies are under way with both of the new wild-derived chromosomes to determine whether the two Gura electrophoretic forms are the same and whether they differ from the Gusa characteristic of the strain A/J. These data are essential for determining whether we are dealing with recombinants of Gur and Gus or three separate structure-induction combinations.

In this regard, we have also observed two additional Gur-Gus combinations from wild-derived mice which are also shown in Table 4. The Gusc allele from M. m. castaneus, trapped in Thailand, produced a β-glucuronidase which differs from any produced by laboratory mice. The second allele, Gusn, was found in wild-derived inbred lines developed by Dr. James Conner, University of Nebraska. The relative electrophoretic mobility of the Gusn β-glucuronidase is slower than any inbred phenotype. It is associated with a very high inducibility at seven days. More complete molecular and genetic characterizations of these two allelic forms are under way.

Our experience with sampling wild mice for β-glucuronidase parallels the findings for many loci. Namely, many allelic forms may be present in wild mice that are not present among inbred strains. Furthermore, we have demonstrated that it is possible to survey wild mice for complex phenotypes such as kidney levels of β-glucuronidase following induction. Even if the presumptive recombinants of Gur-Gus reported here ultimately prove to be different chromosome sets of Gur-Gus with different structural sequences of bet-glucuronidase, the notion of looking for recombination between closely linked genes in wild populations may be useful, particularly when the expected recombination frequency is less than 10-4. This will be especially important as additional cis acting regulatory elements are observed for other systems besides β-glucuronidase induction.

Breeding Programs for Wild-Derived Mice

In most instances, wild populations of mice are fairly heterogeneous with relatively large amounts of polymorphism in most of the samples taken. Developing inbred strains from wild-derived sources has some value in producing a new array of gene combinations that may not exist among inbred mice. For example, M. m. castaneus differs from C57BL/6J for more than 27 biochemical markers on 13 different chromosomes ( Table 5) (Chapman, unpublished). As such, it provides a valuable tool for linkage analysis and as a mutagenesis testing stock. Similar differences may also be present for M. m. molossinus (from Japan) and a strain developed at Roswell Park Memorial Institute called MOR/Cv. However, the development of wild-derived strains is costly in terms of time and money, and there is probably a limit to the number of such strains that fulfill this general need.

As an alternative, I would propose that wild mouse populations with interesting differences from laboratory stocks be kept as random breeding colonies. This will maintain a greater proportion of genetic variation originally present in the wild-derived sample. Reproduction in these stocks will be generally better than in inbreds as well. Where interesting alleles are present for different loci, it is relatively easy to transfer them to a standard inbred strain background. This has the additional advantage of providing a new allele on a known genetic background. This may be of special value in working with quantitative phenotypes.

From a practical standpoint the congenic stocks are easier to handle, i.e., they behave like inbred mice. Also, congenic lines have a special value for linkage tests because they carry a chromosome segment from the donor strain which may contain different alleles for other linked loci.


Feral mouse populations are highly polymorphic and contain genetic variation not present among laboratory stocks of mice. Genetic analysis of wild-derived polymorphisms is possible by crossing wild mice with homozygous inbred strains. Backcross and F2 generations can be readily studied for segregation and linkage analyses.

Complex phenotypes such as quantitative levels of enzyme activity as well as qualitative characteristics such as electrophoretic mobility can be evaluated in crosses with characterized inbred strains. This has been demonstrated for the enzyme loci α-amylase ( 5) and in work on β-glucuronidase.

Wild mouse populations may also serve as an experimental resource. In the case of closely linked elements, sufficient test progeny may not be economically feasible to evaluate recombination distances less than 0.01 map units apart. Cis acting regulatory elements would be reasonably expected to be within that distance from structural loci. Thus, the identification of populations polymorphic for both structure and regulation may provide a possibility of recovering what would otherwise be rare recombination events in the laboratory.


1. Hutton, J.J. (1978). In Origins of Inbred Mice. (H.C.Morse III, ed.). Academic Press, New York.

2. Chapman, V.M., and Selander, R.K. (1978). FASEB Biological Handbook, in press.

3. Chapman, V.M., Paigen, K., Siracusa, L., and Womack, J. (1978). FASEB Biological Handbook, in press.

4. Staats, J. (1976). Cancer Res. 36: 4333.
See also MGI.

5. Nielson, J.T., and Sick, K. (1978). Hereditas 79: 279.

6. Barry, R.J., and Peters, J. (1977). Proc. Roy. Soc. London. B. 197: 485.

7. Selander, R.K., Hunt, W.G., and Yang, S.Y. (1969). Evolution 23: 379.
See also MGI.

8. Selander, R.K., Hunt, W.G., and Yang, S.Y. (1969). Genetics 63: 653.

9. Petras, M.L., Reimer, J.P., Biddle, F.G., Martin, J.E., and Laughton, R.S. (1969). Canada J. Genet. Cytol. 11: 497.
See also PubMed.

10. Ruddle, F.H., Roderick, T.H., Shows, T.B., Weigle, P.G., Chipman, R. K., and Anderson, P.K. (1969). J. Hered. 60: 321.
See also PubMed.

11. Chapman, V.M., Felton, J., Meisler, M., Paigen, K., Swank, R., Tomino, S., Wudl, L. (1973). Mouse News Letter 48: 44.
See also MGI.

12. Roderick, T.H., Ruddle, F.H., Chapman, V.M., and Shows, T.B. (1971). Biochem. Genet. 5: 457.
See also PubMed.

13. Womack, J.E., and Belsky, S. (1975). Mouse News Letter 52: 37.

14. Lalley, P.A., and Shows, T.B. (1977). Genetics 87: 305.
See also MGI.

15. Dizik, M., and Elliott, R.W. (1977). Biochem. Genet. 15: 31.
See also PubMed.

16. Dizik, M., and Elliott, R.W. (1978). Biochem. Genet. 16: 247.
See also PubMed.

17. Johnson, G. (1977). Biochem. Genet. 15: 665.
See also PubMed.

18. Johnson, G. (1978). Proc. Natl. Acad. Sci. USA 75: 395.
See also PubMed.

19. Paigen, K. (1978). In Origins of Inbred Mice. (H.C. Morse III, ed.). Academic Press, New York.

20. Swank, R.T., Paigen, K., and Ganschow, R.E. (1973). J. Mol. Biol. 81: 225.
See also MGI.

21. Swank, R.T., Paigen, K., Davey, R., Chapman, V., Labarca, C., Watson, G., Ganschow, R., Brandt, E.J., and Novack, E. (1978). Progress in Hormone Res., 34: 401.
See also PubMed.

22. Quarantillo, B. (1977). M.S. Thesis, Roswell Park Memorial Institute Division, SUNY at Buffalo.

Previous Next