SEGREGATION ANALYSIS OF THE TESTIS-DETERMINING AUTOSOMAL TRAIT, TDA, THAT DIFFERS BETWEEN THE C57BL 6J AND DBA/2J MOUSE STRAINS SUGGESTS A MULTIGENIC THRESHOLD-MODEL/

The testis-determining autosomal trait (Tda) of the mouse was uncovere d when the Y chromosome of the poschiavinus variety of Mus musculus do mesticus was introduced into the C57BL/6J laboratory strain background . Testis development is normal in the F-1 generation but, in the backc ross and subsequent crosses to C57BL/6J females, XY individuals with t he poschiavinus Y chromosome expressed bilateral ovaries or various co mbinations of an ovotestis with a contralateral ovary or testis or bil ateral ovotestes and few had testes bilaterally. In other strain backg rounds, such as DBA/2J, XY individuals with the poschiavinus Y chromos ome always expressed normal testes bilaterally. The first breeding ana lysis of this difference in the interaction of strain background with the poschiavinus Y chromosome suggested that the Tda trait was due to a single gene, but attempts to map it failed. We constructed two strai ns of C57BL/6J and DBA/2J that are consomic for the poschiavinus Y chr omosome in order to conduct a segregation analysis of the Tda trait. I n the C57BL/6J.Y-POS consomic strain, liability to express incomplete testis development is normally distributed and thresholds in developme nt specify the probability of different classes of ovary, ovotestis, a nd testis combinations. Testis development is complete in the DBA/2J.Y -POS consomic strain. We demonstrated previously that the Tda trait of C57BL/6J is recessive to that of DBA/2J and the segregating first bac kcross generation of embryos rejected the single-gene model. We have e xtended our analysis to a F-2 generation of embryos that also rejects a single-gene model. We also report a test mating analysis of the firs t backcross generation. It was initiated to provide an independent ass essment of the single-gene model, but the analysis of the distribution of test mating results suggests that the difference in the Tda trait between C57BL/6J and DBA/2J may be due to a small number of loci, poss ibly four or five, and that the phenotypic effect between loci may be additive.