Fig 1.
Contributions of Col17a1 and other genetics to Lamc2jeb/jeb modifier effects.
(A-C) ear scores weekly, from 0 (‘unaffected’) to 6 (‘very affected’). (D-F) tail scores using same method. (G-K) tension at various ages. All mice tested are male, Lamc2jeb/jeb homozygotes. Congenic names are for Col17a1 (e.g., MRL.FVB is MRL-Lamc2jeb/jeb Col17a1FVB/FVB). Ear and tail score statistics are based on age when scores first reach ‘4’ (‘moderately affected’) calculated as survival in PRISM by both Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon, with the less significant of the two indicated. G, MRL required euthanization before 15 wks due to their severity of JEB. Tension statistics are PRISM 1-way ANOVA. Tension bars indicate mean with SD. ns = not significant, * <0.05, ** <0.01, *** <0.001, **** <0.0001. Includes some B6, 129, FVB and MRL data previously published [16].
Fig 2.
Direct comparisons of B6 and MRL Col17a1 congenics on FVB background.
(A) Tail tension test at 20 weeks of age. (B-C) Average ear and tail scores from 0 ‘not affected’ to 6 ‘very affected’ for mice scored weekly. (D) Cumulative censored survival data. (E) Amino acid sequence for Col17a1 AA 1275–1292 for tested inbred strains, from Sproule et al 2014. All (A-D) data shown is for FVB males homozygous Lamc2jeb/jeb and for B6 or MRL Col17a1 congenic segments as indicated. (E) MRL differs from FVB at Col17a1 AA 1277. B6 differs from FVB at both AA 1277 and 1292. Tension bars indicate mean with SD. Tension p-values are 1-way ANOVA. Ear and tail score p-values are less significant of Log-rank (Mantel-Cox) or Gehan-Breslow-Wilcoxon, based on age when each mouse first reaches a score of ‘4’ treated as survival.
Fig 3.
(B6 x 129X1)F2-Lamc2jeb/jeb tail tension R/QTL.
(A) Tension test results for parental B6 and 129X1, F1 and F2 male mice at 10 weeks of age (P-values: * < 0.05, ** <0.01, **** <0.0001). (B) LOD score of tension x marker analysis for the B6X1F2 tension set with LOD score of 0–5 expanded in (C) to show greater detail. Horizontal lines bottom to top are 63, 10, 5 and 1% cutoffs. (D) Effect plots for SSLP markers nearest the peaks. X-axis is genotype: BB = B6, BX = het, XX = 129X1. (E) 95% confidence intervals for the four most significant QTL. Positions of SSLP markers tested with Mb position names are indicated (i.e. 1_22 is chr1, 22Mb). Blue lines indicate 1.5 LOD drop interval. Green lines indicate Bayesian confidence intervals. 95% CIs converted to Mb are: chr1:22-39Mb (peak 34Mb), chr5:40-111Mb (peak 58Mb), chr11:90-122Mb (peak 119Mb) and chr13:20-104Mb (peak 59Mb). (F) Tail tension comparison of ‘long’ and ‘short’ B6.129X1-Lamc2jeb/jeb congenics. (G) Mb map of same. 129 is black, B6 is white. All mice are male, Lamc2jeb/jeb homozygotes. Tension bars indicate mean with SD.
Fig 4.
(MRL.FVB x FVB)F2-Lamc2jeb/jeb Col17a1FVB/FVB tail tension R/QTL.
(A) Male 15 week old tail tension test results for MRL.FVB x FVB F2 (n = 176) with F1 and parental FVB and MRL.FVB controls (all homozygous Lamc2jeb/jeb and Col17a1FVB/FVB). Bars indicate mean with SD. (B) R/QTL LOD scores showing peaks on chr6, 10 and 13. (C) Inclusion of body weight as a covariant increased significance of chr6 and 13 peaks. (D) Effect plots for SSLP markers nearest the QTL peaks. (E) 95% confidence intervals for the three most significant QTL. Converted to Mb are: 6:66-127Mb (peak 110Mb), 10:66-117Mb (peak 104Mb) and 13:11-74Mb (peak 41Mb). Positions of SSLP markers tested with Mb position names are indicated (i.e. 6_26 is chr6, 26Mb). Blue lines indicate 1.5 LOD drop interval. Green lines indicate Bayesian confidence intervals. (F) Linear regression of tension vs body weight for F2 mice is not statistically significant.
Table 1.
Phenotypes of QTLs and percent of variation explained.
Fig 5.
(A-F) Mouse Phenome Database (MPD) CGD-MDA1 percent polymorphism plots (PPP) across 1:22-39Mb QTL 95% confidence interval: (A) B6 vs 129X1 percent polymorphic, (B) FVB/MRL and (C) B6/129X1 polymorphic while FVB/MRL matched. Each peak/bar = 0.1 Mb. Map beneath indicates extent of congenic segments tested. Red box indicates Dst position. (G) MPD PPP for Dst. (H) B6 chr1 congenic 10 wk old male tail tension test. All homozygous Lamc2jeb/jeb. B6.129 and B6.AJ homozygous for chr1 congenic intervals 1:22-37Mb and 1:25-44Mb respectively. Bar indicates mean with SD. (D) Male ear and tail scores for the same congenics. (J) B6/129 missense SNPs mapping to Dst-e. A/J matches B6 at 2 of 5. Statistics in (H) are PRISM 1-way ANOVA all compared to B6-Lamc2jeb/jeb control (B6) except as indicated. Statistics in (I): both Log-rank (Mantel-Cox) or Gehan-Breslow-Wilcoxon tests were performed, based on age when each mouse first reaches a score of ‘4’ treated as survival. Values shown are the less significant of the two. **** is <0.0001, *** is <0.001, ** is <0.01, * is <0.05, ns is not significant.
Fig 6.
Col17a1 and 1/22-37Mb (Dst) double congenic effects on the Lamc2jeb/jeb disease.
(A) 10-week-old male tension comparisons of B6, B6-Dst129/129 congenic, R03Q B6-Col17a1PWD/PWD congenic, B6-Col17a1PWD/PWD Dst129/129 double congenic and B6 wild-type (B6-Lamc2wt/wt) (all Lamc2jeb/jeb except the last). Red symbols indicate clamp slipped off without removing substantial skin. (B) Tension expressed as survival of same with red values censored gives statistical difference between B6-Col17a1PWD/PWD and double congenic. (C) Photos showing the typical amount of skin removed during the tail tension test from (top to bottom): B6-Lamc2jeb/jeb, B6-Lamc2jeb/jeb Dst129/129, B6-Lamc2jeb/jeb Col17a1PWD/PWD, B6-Lamc2jeb/jeb Dst129/129 Col17a1PWD/PWD double congenic and B6 wild-type, all at 10 wks of age. (D-E) Ear and tail scores for the same excluding B6 wt control and adding FVB-Lamc2jeb/jeb for comparison. All males. All Lamc2jeb/jeb except ‘B6-Lamc2wt/wt’. Dst is contained within the 1/22-37Mb congenic and used here to represent it. Ear and tail statistics are survival based on age when each mouse first received a score of ‘4’ (moderately affected). **** p < .0001, *** p < .001, ** p < .01, * p < .05, ns = not significant.
Fig 7.
Mouse Phenome Database CGD-MDA1 percent polymorphism plots of 129X1/SvJ (129X1) vs 129S1/SvImJ (129S1) throughout the genome by 0.1Mb increments, limited to SNPs with allele data for both strains. B6 x 129X1 F2s give QTL. B6 vs 129S1, but not 129X1, full genome SNP polymorphism data is available in Sanger. Percent polymorphism plots reveal regions of the genome where 129S1 can fairly be used as a substitute for 129X1 in Sanger B6 vs 129 SNP analysis, including all regions of interest to us in this study: 1:22-39Mb/Dst, 5:50-53Mb/Ppargc1a, 11:90-122Mb/Itgb4 and 13:38-39/Dsp (red underlines). Col17a1 on chr19, which does genetically differ between 129X1 and 129S1, is also underlined for reference.
Fig 8.
(A) Percent MPD CGD-MDA1 SNP polymorphism map (each bar = 0.1 Mb) for the chr5 QTL interval. Clustered bars indicate regions of disparate strain heritage. Comparison of B6/129 (top, differ at QTL) to FVB/MRL (middle, predicted to match at QTL) does little to reduce candidate interval (bottom, B6/129 and FVB = MRL). (B) Chr5 congenic 10 wk old male tail tension test. * after name indicates cross made using a reduced B6-Lamc2jeb/jeb congenic (<10 Mb) instead of originally published (>80 Mb). Statistical significance for each congenic is based on comparison to its parental short or long congenic strain. (C) Focused chr5 congenic map of protective (red) and non-protective (black) lines limits the main candidate interval to 50.5–52.366 Mb. (D) Ensembl view of 5:50.5–52.366Mb showing genes present. (E) Percent polymorphism plots for reduced congenic interval and candidate gene Ppargc1a from the MPD CGD-MDA1 SNP data set. (F) List of B6 vs 129X1 polymorphisms in Ppargc1a excluding synonymous and intronic from GenomeMUSter GRCm38 (mpd.jax.org/genotypes). All are predicted MRL = FVB.
Fig 9.
(A) Strain heritage comparisons across QTLs: B6 v 129 top, MRL v FVB middle and appropriate prediction-based combined plot bottom for each. Bars indicate percent polymorphism among MPD CGD-MDA1 SNP markers in 0.1Mb increments, only looking at markers which contained genotype information for all 4 strains. Labels indicate full congenic intervals. Red underline indicates location of genes of interest. (B) MPD percent polymorphism for strain pairs for specific candidate intervals and genes. (C) Tail tension test of 10 week old male chr11 QTL congenic vs B6 control. Both groups Lamc2jeb/jeb. Bars indicate mean with SD.
Table 2.
Candidate SNPs in Itgb4 and Dsp.
Fig 10.
Comparison of ear (A, C) and tail (B, D) scores for Col17a1 allele matched B6 and 129X1 versus MRL and FVB matched for Col17a1FVB/FVB (A, B) or Col17a1MRL/MRL (C, D). All scores are from 0 (‘unaffected’) to 6 (‘very affected’). (E) Tension comparison of B6, 129, MRL and FVB 10 week old males where B6 and 129 are naturally Col17a1-matched and MRL and FVB are congenically matched for Col17a1MRL/MRL. All (A-E) mice are male, homozygous Lamc2jeb/jeb and homozygous for the indicated alleles of Col17a1. Ear and tail score statistics (A-D) are less significant of Log-rank (Mantel-Cox) or Gehan-Breslow-Wilcoxon based on age each mouse first to score ‘4’. Tension statistics (E) are 1-way ANOVA. Tension bars indicate mean with SD.