Loss of Fancg results in a more severe defect in multiple hematopoietic compartments than loss of Fancc. (A) Total number of progenitors per femur isolated from syngeneic 3-month-old mice with statistically similar weights (25 ± 0.6 g) and bone marrow cellularity (10 × 10⁶ ± 0.7 × 10⁶ cells per femur). *p = 0.0009 by 1-way ANOVA. (B) Number of multipotential colony-forming unit–granulocyte, erythroid, monocyte, megakaryocyte progenitors (CFU-GEMMs), *P < 0.0001 by 1-way ANOVA. (C) colony-forming unit–granulocyte-macrophage progenitors (CFU-GMs), *p = 0.0230 by 1-way ANOVA. (D) burst forming unit–erythroid progenitors (BFU-Es), *p = 0.0030 by 1-way ANOVA. Data represent mean ± SEM of 3 independent experiments each of which was plated in triplicate cultures. (E-F) Genome-wide transcriptomal analysis of marrow cells from 3-month-old mice from each of the 4 indicated FA genotypes was performed. A total of 12 BMC samples from 12 mice (Fancc^−/−;Fancg^−/−, Fancc^−/−, Fancg^−/− and WT) were analyzed using 12 AffyMetrix MOE 430 2.0 GeneChip arrays. (E) Unsupervised hierarchical clustering, the scale represents the level of correlation between the groups. (F) Principal component analysis. In this case the multidimensional data were reduced to a new coordinate system such that the greatest variance (20.8%) by any possible projec-tion of the data are plotted on the first coordinate (the first principal component, PCA1). The second greatest variance (14.5%) was plotted on the second coordi-nate (PCA2), and the third greatest (10.8%) plotted on the third coordinate (PCA3). Results from both analyses confirmed that double mutant cells diverged widely from both the WT samples and both of the single mutant samples.