Experiment
Strains of mice show large differences in acute lung injury (ALI) mortality. A mouse model of differential hyperoxic ALI (HALI) survival was previously established to tease out genetic factors affecting mortality [J:127224]. Separate genetic analyses of backcross and F2 populations generated from sensitive C57BL/6J (B) and resistant 129X1/SvJ (X1) progenitor strains identified two quantitative trait loci (QTLs Shali1 and Shali2) with strong, equal but opposite, within-strain effects on survival. Reanalysis of the original F2 data gave analogous age-related findings, and also supported sex-specific linkage for Shali1 and Shali2.
The long-term goal of the current project was to identify specific genes or pathways beneficial for increased survival to oxidant-induced lung injury. Accordingly, survival time was selected as the trait of interest.
Consomic, congenic and subcongenic lines described in this study were all maintained on the 129X1/SvJ (X1) background. The consomic lines for Shali1 and Shali2 were 129X1/SvJ-Chr 1C57BL/6J/J and 129X1/SvJ/-Chr 4C57BL/6J/J respectively. Congenics and subcongenics were produced using a backcross-intercross breeding scheme. Each crossover was reproduced by backcrossing to the X1 strain to generate a breeder mate, and the congenic regions were fixed to homozygosity for testing and maintenance of lines. Panels of nested congenic lines were made using microsatellites to screen the QTL intervals.
Initial congenic lines were tested in hyperoxia at larger numbers and sorted into groups of mice at </= 6 weeks, 7-9 weeks, and >/= 10 weeks old for analysis of an age effect. To determine whether the age effect was associated with survival time in the original F2 population (n = 840), the dataset was sorted into the same 3 age groups for comparisons.
A sex-effect was also assessed in the newly defined age-specific groups for controls, congenics and F2 cohorts.
The F2 dataset [J:127224] was reanalyzed in several ways. First, the total population (n = 840) was split by sex and age group (</=6 weeks, 7-9 weeks, and >/= 10 weeks old), and mean survival times (MSTs) calculated and evaluated for differences using t-tests with multi-test corrections. Linear regression was used to evaluate age by sex interactions. Next, the total dataset of exposed F2 mice was stratified by genotypes at the peak markers for Shali1 and Shali2, with interest only in the two cohorts of mice carrying homozygosity for both resistance alleles (X1-B, n = 50) or both sensitivity alleles (B-X1, n = 58) at D1Mit303 and D4Mit308, respectively.
Each of these genotype groups was then parsed by age (younger or older) using 56 d.o. as the discriminator. To assess a possible sex effect, each of these four age groups was then divided into males and females. Groups were assessed for significant differences by t-tests with the appropriate Bonferroni corrections for multiple comparisons. In a third analysis, the F2 dataset was divided into younger versus older aged cohorts using a single-age threshold of 56 d.o. and separate QTL analyses were performed for both sexes in each age group using R/QTL.
Fig 3A,3B-
The QTL analysis plots of all F2 mice </= 56 d.o. at exposure (n = 546) were similar to the full F2 dataset of 6-12 week old mice [J:127224], and reaffirmed Shali1 and Shali2 as important putative QTLs for survival time differences.
Next, the two age-specific F2 cohorts were stratified by sex for separate QTL analyses.
Younger males (n = 256) yielded a significant linkage peak at Shali2, whereas younger females (n = 290) were highly significant for Shali1 and Shali2 . Of note, the male and female linkage peaks for Shali2 did not represent the same mapping locations on Chr 4. No other significant QTLs were found in these younger cohorts. For older males (n = 161), only the previously identified male-specific QTL, Shali5 on distal Chr1, was identified.
Next, whether the cohorts of F2 mice also showed differential effects based on age and/or sex was examined. Mice carrying reciprocal allelic pairs (i.e., X1X1BB vs.BBX1X1 for Shali1 Shali2, respectively) demonstrated significant survival differences(p < 0.0001), with MSTs differing by 50% [Fig 4]. Younger mice (MST = 177 h; n = 36) with resistance alleles at Shali1 and Shali2 were 33% more resistant (p = 0.0015) than older mice (MST = 133 h; n = 14) carrying the same alleles.
To determine whether congenic lines carrying Shali1 and Shali2 also demonstrated the age and sex-related effects, X1.B-1EA.a (sensitive) and X1.B-4BB (resistant) congenic lines of mice were exposed to hyperoxia and MSTs calculated (Table 2). Both sexes in all three age groups (</= 6 weeks, 7-9 weeks, >/=10 weeks) spanning the 4-week trait transition period were tested and compared.
For both sexes, the MSTs of 6-week congenic X1.B-1EA.a mice were significantly less than MSTs of young X1 mice (males: 142 h vs. 183 h; females: 94 h vs. 171 h), thus validating the capture of Shali1 sensitivity in the X1.B-1EA.a congenic line.
With the identified age effect, the focus for Shali1 was on younger highly sensitive mice, using a </= 56 d.o. age threshold to sort the data. Fig 5 displays the genetic regions captured in the Shali1 congenic lines, along with separate MSTs for males and females. The original validated Shali1 interval was 102.8 Mb which encompassed slightly more than the proximal half of Chr1. The five subcongenics of the 1A interval have reduced this interval to ~8.5 Mb (60.05-68.55 Mb), as represented by the overlapping regions of the highly sensitive line 1EA.aa at 60-70.4 Mb and the congenic line 1EA.ab at ~68.574.1 Mb, which had a MST similar to control X1 mice.
The most refined single congenic that still carries a significant Shali2 resistance effect (i.e., X1.B-4BC.b) has an interval of ~28 Mb (>126.7-154.9 Mb). This congenic has lost some of the overall resistance (especially in females) that was seen in the X1.B-4BB line, suggesting that additional sub-QTLs are present in the larger 4BB interval.
Analysis Tools