Sci. Aging Knowl. Environ., 25 September 2002
Murine Chromosomal Regions Correlated With Longevity
Rebecca Gelman, Ada Watson, Roderick Bronson, and Edmond Yunishttp://sageke.sciencemag.org/cgi/content/abstract/sageke;2002/38/cp24
Abstract: In this longevity analysis of 360 BXD recombinant inbred female mice (20 different strains), 2 strains had very significantly shorter survival and 1 strain had very significantly longer survival than the other 17 strains; 4 other strains had less significant lengthening of survival compared to the other 13 strains in a proportional hazards model of survival. Mean survival on the shortest lived strain was 479 days; on the longest lived strain the mean survival was almost double (904 days). Ranges of survival within strain were very large (averaging 642 days), and strain accounted for only 29% of the variation in survival, showing that there are important environmental and/or special developmental effects on longevity even in this colony housed in a single room. Each strain had been typed for markers of 141 regions on 15 chromosomes; 101 of these markers had distinguishable distributions on the 20 strains. The two shortest lived strains had the same alleles for 63% of the markers. The single region most significantly correlated with survival (marked by P450, Coh, Xmmv-35 on chromosome 7) divided the mice into two groups with survival medians which differed by 153 days (755 days for mice with a B genotype; 602 days for mice with a D genotype). Evaluated individually, 44% of the genetic markers (including some markers on 11 of 15 chromosomes with any markers typed) were found to be significantly correlated with survival (P > 0.05) although one would only expect 5% of the markers to be significant by chance. While studies of many markers should adjust for the multiple comparisons problem, one interpretation of these crude P values is that any experiment with only one of these "significant" markers typed would be likely to conclude that the marker was a significant predictor of survival. Two types of multiple regression models were used to examine the correlation with survival of groups of genes. When a proportional hazards model for survival was done in terms of genotype regions, a six genetic region model best correlated with survival: that marked by P450, Coh, Xmmv-35 on chromosome 7 (B allele lives longer), Ly-24 on chromosome 2 (B allele lives longer), 2M and H-3 on chromosome 2 (D allele lives longer). Lamb-2 on chromosome 1 (D allele lives longer), Ltw-4 on chromosome 1 (B allele lives longer), and the Igh area of chromosome 12 (Igh-Sa4, Igh-Sa2, Igh-Bgl, Igh-Nbp, Igh-Npid, Igh-Gte, Odc-8, and Ox-1; D allele lives longer). A linear model that regressed mean survival (per strain) on genetic markers found a similar six region model to be best, but replaced Coh by D12Nyu1 on chromosome 12. It should be noted that in both types of regression, there were many other models almost as good as the best one. The total number of chromosomal regions marked by the genotype of the longer lived B parent (out of a possible 141) was not, in general, correlated with survival, although the two shortest lived strains had the most B genes. It appears that BXD recombinant inbred strains can vary widely in survival both within and between strains, that no single genetic marker which has yet been identified can account for much of this variance, (although groups of six or more markers may do so), and that it is not always those strains which inherit the most genes from the long-lived parent B that live longest. The large number of genetic markers found to be significantly correlated with survival raises questions of the reliability of conclusions based on survival studies of only one or two genetic regions.
Reproduced by permission.
Rebecca Gelman, Ada Watson, Roderick Bronson, Edmond Yunis, Murine Chromosomal Regions Correlated With Longevity. Genetics 118, 693-704 (1988).
Citation: R. Gelman, A. Watson, R. Bronson, E. Yunis, Murine Chromosomal Regions Correlated With Longevity. Science's SAGE KE (25 September 2002), http://sageke.sciencemag.org/cgi/content/abstract/sageke;2002/38/cp24
Science of Aging Knowledge Environment. ISSN 1539-6150