Involvement of Fanconi Anemia (FA) Genes FANCG and FANCA in Human DNA Replication, Mitosis and Chromosome Segregation: A Gene Expression Profiling Study.

FA is an autosomal recessive disorder with diverse clinical symptoms, including developmental anomalies, progressive bone marrow failure, and a predisposition to the development of malignancies. FA cells are hypersensitive to DNA cross-linking agents, and display chromosomal instability and defective DNA repair. FANCG-deficient pts accounts for 10% of all FA pts, and are a high risk FA population for acute myeloblastic leukemia. FANCG protein is part of the FA nuclear core complex, interacting with FANCA and FANCF. It can also interact with FANCD1, RAD51, and CYP2E1, suggesting roles in homologous recombination DNA repair and oxidative DNA-damage protection. However, FANCG lack apparent homology to other proteins involved in such processes. To address whether FANCG mRNA expression level was associated with some microarray-based functional signatures using Affymetrix HG-U133A Genechips, we first used a dataset including 111 T-acute lymphoblastic leukemia (T-ALL) samples. This large series was selected because FA pts do not usually develop T-ALL (risk of FANC gene mutation/inactivation low), FANCG is highly expressed in thymus and lymphoblasts, FANCG gene expression was significantly varying across these samples. We first selected samples having the lowest (FANCGlow) and highest (FANCGhigh) FANCG expression levels (n=15 per group, median expression values: 36 vs. 329), and identified genes differentially expressed between these 2 groups using SAM (fold change>3; qvalue<1%). Among the 368 probesets (308 known genes) significantly up-regulated in FANCGhigh group (likely to be co-expressed/regulated with FANCG), 55 were correlated to FANCG expression profile (>.75) when analyzing all 111 T-ALL samples. These genes were mainly involved in cell proliferation, mitotic cell cycle and its regulation (DNA replication, traversing Start control point, G2/M transition, chromosome segregation and cytokinesis), and were also representing cellular components involved in such processes (kinesin complex, spindle, and kinetochore). FANCA, RAD51, RAD51C, and PIR51 were also among the genes having expression patterns close to FANCG one. In addition, 11 of these 55 genes, all involved in DNA replication, were also correlated with FANCA (>.75). A robust regression analysis identified CKS1B, CCNB2, RFC5, MELK, STK18, and CENPF as the genes most significantly associated with FANCG expression profile (p<.0001, power=100%). Finally, we compared normal fibroblasts (CCL153) to primary fibroblasts from 2 FANCG-deficient pts grown under normal conditions or with MMC 100 nM for 36 hours (PD829, PD352). Fifty-eight of the 585 probesets down-regulated in FANCG-deficient fibroblasts (fold change>2; qvalue<1%) were also present in the FANCGhigh list of genes, and were mainly involved in DNA replication, mitotic chromosomal positioning and segregation, as well as in cytokinesis. Among these genes, GMNN (DNA replication inhibition by preventing MCM complex incorporation), MCM10 (DNA replication initiation), POLE2 (DNA replication and repair), and RRM2 (DNA synthesis) were also correlated with FANCA expression pattern. Of note, FANCC, FANCE, FANCF, FANCD1/BRCA2, and FANCL expression patterns were not significantly correlated with the one of FANCG or FANCA.

Conclusion: these results suggest that the FA core complex, and especially FANCG and FANCA, is involved in DNA replication and mitosis.