Proc Natl Acad Sci U S A. 31(2): 82–84. (1945)
Communicated January 15, 1945

Several years ago1 there was found in the house mouse a mutation with several striking effects including absence or shortening of the tail, absence or abnormality of one or both kidneys, absence of external anus and genital aperture and abnormalities of other parts of the urogenital system. In the stock in which it occurred this whole syndrome of effects behaved as a unit and showed simple segregation from normal. The mutation acted as a lethal, all homozygotes Sd Sd being tailless and dying shortly after birth, all showing imperforate anus and absence of both kidneys. Heterozygotes showed a lesser expression of these defects, having short tails, and less severe urogenital malformations. The mutation in the original stock therefore acted as a dominant in respect to its effect on the tail, as recessive or nearly so in its lethal effect, and as incompletely dominant in its effect on urogenital development.

When the Sd mutation was removed from the stock in which it originally occurred and was transferred by a series of successive backcrosses to another inbred normal stock, the tail length of heterozygotes progressively decreased until after five backcross generations nearly all were tailless, while the viability of the heterozygotes decreased, due to the greater effect of Sd on the urogenital system.1 About 90 per cent of all Sd+ animals at birth had abnormal kidneys.2 The dominance of Sd on tail length appeared to have been increased by the genetic constitution of the new stock while the lethal effect appeared also to have become partially dominant. There was no evidence of necessary connection between the effect upon tail length and upon viability.

Since the above observations were published we have transferred the Sd mutation to two other normal-tailed inbred stocks by repeated backcrossing. In one of these stocks (identified as m) the tail length of heterozygotes increased, and the proportion of tailless animals among the heterozygotes decreased. In the F1, BC, and BC2 generations the cross of Sd+ by normal m produced 142 normal, 90 short-tailed and 25 tailless; while in BC3 and BC4 the comparable figures are 39 normal, 23 short and no tailless. The totals, 181 normal and 138 Sd+(x2=5.26, p=0.02) indicate a lowered viability of Sd+ associated with increasing tail length. This was probably due to increased severity of the urogenital lesions, since of four heterozygotes dissected at birth, three had urogenital abnormalities.

When backcrossed to the other inbred stock, a normal albino stock known as CF (Carworth Farms), changes in the opposite direction occurred in the Sd heterozygotes; that is, the tails became shorter, the proportion of tailless heterozygotes increased, and the relative viability of the heterozygotes increased. The figures are F1-BC2: 59 normal, 14 short, 53 tailless; BC3-BC5: 26 normal, 7 short, 18 tailless. The totals of 85 ++ 92 Sd+ indicate no excess mortality of short and tailless Sd+. Forty-three of these heterozygotes were dissected and in all cases the urogenital system was normal. Thus the CF genetic constitution, which increased the severity of the tail defect, eliminated the deleterious effect of Sd+ on viability and on the urogenital defects chiefly responsible for the viability effects of Sd. The conclusion is obvious that the several effects of Sd upon the heterozygotes are modified by different genetic factors.

Since this is so, the conclusions of Fisher and Holt4 concerning dominance modification of Sd will have to be examined critically. Fisher and Holt similarly outcrossed Sd+ animals obtained from this laboratory to other stocks in the Cambridge University laboratory, and set up selection lines in one of which the tail length of heterozygotes was increased, while in the other, tail length remained short. The viability of heterozygotes in the longer-tailed line improved. Fisher and Holt supposed that the chief factor in viability was tail length, so that natural selection for viability aided the selection for longer tails, while in the negative line it was "always acting in opposition, in that mice most defective in the development of the caudal vertebrae will also, on the whole, be most defective in other respects." This assumption, as our observations show, is entirely unwarranted. They assume also that as the tail length of the heterozygotes increases "so one might expect the chance of homozygotes becoming less abnormal, with a consequent lengthening of life, to increase." This is based on the same reasoning as that above and is supported only by the occurrence in the line selected for tail length, of two animals diagnosed as homozygotes (Sd Sd), one of which lived for 22 days and the other for 72 hours. Since tailless heterozygotes with internal lesions as severe as those of the longer lived assumed homozygote have been described by Gluecksohn-Schoenheimer,2 the diagnosis of homozygosity rests on the occurrence of imperforate anus and cloaca. These defects have been found recently in rare cases among heterozygotes also, so it is possible that Fisher and Holt were dealing with an extreme heterozygote.

There is no doubt that the tail length of heterozygous Sd+ mice is readily modified by other genetic factors, which thus may be said to modify the dominance of Sd in its effect upon the tail. But these other factors, as our results show, may change the dominance of Sd in its other effects in ways opposite to that in which the tail expression is affected.

In order to derive some meaning from such observations for Fisher's general theory of the origin of dominance in evolution, it would be necessary to specify whether modifiers of Sd would be selected because of their effect on tail length or because of their effect on urogenital development. In view of the evident relation between the urogenital expression and viability there can be little doubt that modifiers which tend to make the urogenital effect of Sd recessive would be selected. But these same genetic constitutions may, as we have seen, increase the dominance of Sd in respect to tail length. It can be concluded that the evolutionary significance of factors affecting dominance can be properly assessed only when their physiological effects are known.

*The observations reported were made with the aid of a grant from the Josiah Macy, Jr., Foundation.

  1. Dunn, L. C., Gluecksohn-Schoenheimer, S., and Bryson, V., "A New Mutation in the Mouse Affecting Spinal Column and Urogenital System," Jour. Heredity, 31, 343-348 (1940).
  2. Gluecksohn-Schoenheimer, S., "The Morphological Manifestations of a Dominant Mutation in Mice Affecting Tail and Urogenital System," Genetics, 28, 341-348 (1943).
  3. Dunn, L. C., "Changes in the Degree of Dominance of Factors Affecting Tail-Length in the House Mouse," The American Naturalist, 76, 552-569 (1942).
  4. Fisher, R. A., and Holt, S. B., "The Experimental Modification of Dominance in Danforth's Short-Tailed Mutant Mice," Ann. Eugenics, 12, 102-120 (1944).