Our recent studies have shown that types of lymphomas and leukaemias are determined by host genes in certain experimental models (Yamada et al, 1994a(Yamada et al, , 1994b. Propylnitrosourea (PNU)-induced T-lymphomas are one such example. F344 rats exhibit high susceptibility, whereas many other strains of rats develop predominantly erythro-or myeloid leukaemias (Shisa and Hiai, 1985;Shisa and Suzuki, 1991) We have assumed that a single dominant gene of F344 rats determines the susceptibility to T-lymphomas and that another independently segregating dominant gene determines the length of the latency period (Shisa and Hiai, 1985). Subsequent efforts to map such genes with crosses between F344 and LE rats, however, have met with great difficulty (Shisa et al, unpublished observation). The difficulty seems to arise partly from paucity of rat genetic marker at that time and partly because these traits are quantitative in nature and mutigenic, rather than a single gene inheritance.The recombinant inbred (RI) strains represent a set of stable inbred strains derived from F2 intercross between two parental inbred stains. The RI strains are extremely useful for the genetic analysis of complex traits (Taylor, 1978;Bailey, 1981). To analyse T-lymphoma susceptibility, we have established a new set of rat RI strains, LEXF, from matings between F344 and LE/Stm strains rats. The LEXF RI strains comprise 11 independent strains and 13 of their sublines. The strain distribution pattern (SDP) of LEXF has been extensively analysed (Shisa et al, 1997;Lu et al, 1998).In this paper, we report the PNU-induced carcinogenesis in the LEXF RI strains, and survey the host genes responsible for determination of the disease types and length of the latency period calculated using quantitative trait locus (QTL) analysis with the Map Manager QT software developed by Manly (1993). We found that numerous QTL with opposed functions are involved in determining the types of lymphomas and latency periods. To control type I and type II errors (Belknap et al, 1992(Belknap et al, , 1996, we further evaluated these putative loci in 137 backcross generation of (F344 × LE)F1 × LE rats. Significant linkage with T-lymphomagenesis was confirmed for three loci on chromosome 5, 7 and 10 and suggestive linkage, for a locus on chromosome 1.
MATERIALS AND METHODS
AnimalsThe LE/Stm rats were originally derived from a closed colony of Long-Evans rats at Ben May Laboratory, University of Chicago (Chicago, IL, USA) and maintained at Saitama Cancer Center Research Institute by brother-sister matings for > 50 generations. The F344 rats were derived from a pair of F344/DuCrj rats purchased from Charles River Japan, Inc. (Kanagawa, Japan) and thereafter maintained by brother-sister matings for > 23 generations. The LEXF was a set of 11 independent RI strains and 13 sublines (Shisa et al, 1997) In this study all LEXF RI strains except for LEXF 6B, a poor breeder, were used. These sublines were branched out at the 7th to 11th generations after an attempt to fix coat colour. ...