Recent discoveries in the molecular pathogenesis of the inherited bone marrow failure syndrome Fanconi anemia

Abstract

Fanconi anemia (FA) is a rare autosomal and X-linked genetic disease characterized by congenital abnormalities, progressive bone marrow failure (BMF), and increased cancer risk during early adulthood. The median lifespan for FA patients is approximately 33 years. The proteins encoded by the FA genes function together in the FA-BRCA pathway to repair DNA damage and to maintain genome stability. Within the past two years, five new FA genes have been identified - RAD51/ FANCR, BRCA1/FANCS, UBE2T/FANCT, XRCC2/FANCU, and REV7/FANCV - bringing the total number of disease-causing genes to 21. This review summarizes the discovery of these new FA genes and describes how these proteins integrate into the FA-BRCA pathway to maintain genome stability and critically prevent early-onset BMF and cancer.

1. Introduction

In the 1920’s, Swiss pediatrician Guido Fanconi first described the disease later to become formally recognized as Fanconi anemia (FA).¹ In a family with five children, three brothers died of a severe condition that resembled pernicious anemia. The three patients’ disease manifested between the ages of five and seven and was associated with congenital microcephaly, café au lait spots, cutaneous hemorrhage, and hypoplasia of the testes, concurrent with a current day diagnosis of FA.²

With an estimated incidence of 1 in 360,000 live births and a carrier frequency of approximately 1 in 181, FA is relatively uncommon and can be difficult to diagnose due to patients presenting with a wide variety of symptoms.^{3, 4} Symptoms that could raise suspicion of FA among physicians include various congenital anomalies including microcephaly, microphthalmia, abnormal thumbs or radii, and slow growth rate. Hematological signs can include early-onset aplastic anemia, myelodysplastic syndrome (MDS), acute myeloid leukemia (AML) at an atypically young age, and one or more unexplained cytopenias of any cell lineage, including thrombocytopenia, neutropenia, and anemia. Additionally, clinicians should be mindful of the presentation of solid tumors at a particularly young age, specifically, head, neck, esophageal, and gynecological squamous cell carcinomas. Familial cancer predisposition or a history of chemotherapeutic hypersensitivity can also raise the suspicion of FA (see Practice points).^{3, 5, 6}

After many decades of work, the FA genes are rapidly being identified. The FANCC gene was identified in 1992.^{7, 8} Subsequent discoveries of the FANCA,^{9, 10} FANCG,¹¹ FANCE,¹² FANCF,¹³ and FANCD2¹⁴ genes followed. Correlating with rapid advances in genetic and biochemical technologies, the rate of FA gene identification has promptly increased. With the most recent additions, there are now 21 confirmed FA genes; FANCA, FANCB, FANCC, BRCA2/ FANCD1, FANCD2, FANCE, FANCF, FANCG, FANCI, BRIP1/FANCJ, FANCL, FANCM, PALB2/FANCN, RAD51C/FANCO, SLX4/FANCP, ERCC4/FANCQ, and most recently, and as highlighted in this review, RAD51/FANCR, BRCA1/FANCS, UBE2T/FANCT, XRCC2/FANCU, and MAD2L2/REV7/FANCV. The previously designated FA-H complementation group was found to be analogous to FA-A, and as such FANCH was removed.^{15, 16}. The addition of five new FA genes over the past two years is testament to the continued resolve of the international FA research community and the Fanconi Anemia Research Fund.

2. The FA-BRCA pathway

One major function of the FA-BRCA pathway is to orchestrate the repair of DNA interstrand crosslinks (ICLs).^{17, 18} Examples of ICL-inducing agents include diepoxybutane (DEB) and mitomycin C (MMC). ICLs pose a direct physical block to DNA replication and RNA transcription and result in cellular cytotoxicity and chromosome structural aberrations if not properly repaired. FA patient cells from all complementation groups are characteristically hypersensitive to ICLs, and this phenotype forms the basis of the clinical FA diagnostic test.^{3, 19} Specifically, FA patient cells exhibit increased radial chromosome formations following ICL induction, a consequence of a molecular roadblock in ICL repair. ICL repair mediated by the FA proteins can be described as a tri-phasic process.

In the first phase, the upstream FA proteins FANCA, FANCB, FANCC, FANCE, FANCF, FANCG, and FANCL, as well as the Fanconi anemia-associated proteins FAAP20 and FAAP100, assemble to form the FA core complex (Fig. 1). Upon DNA damage, the FA core complex is recruited to chromatin where it interacts with UBE2T/FANCT, which is constitutively chromatin localized.^{20, 21} Recent studies have identified three distinct modules within the FA core complex: the FANCB-FANCL-FAAP100 module, which provides the essential monoubiquitination catalytic activity, and the FANCA-FANCG-FAAP20 and FANCC-FANCE-FANCF modules, the exact functions of which have yet to be determined.^{22, 23} The FANCM anchor complex, comprising FANCM, FAAP24, FAAP16/MHF1 and FAAP10/MHF2, promotes the chromatin recruitment of the FA core complex.^24–27 Together, the FA core complex and UBE2T/FANCT constitute an active multi-subunit E2/E3 ubiquitination enzyme complex. The RING domain-containing FANCL subunit functions as the E3 ubiquitin ligase while UBE2T/FANCT functions as the E2 ubiquitin-conjugating enzyme.^28–30 This complex catalyzes the second phase of the pathway, the conjugation of a single ubiquitin moiety (monoubiquitin) to K561 of FANCD2 and K523 of FANCI.^31–33 FANCD2 and FANCI form a heterodimer known as ID2.³⁴ Ubiquitin is a 76-amino acid protein that is covalently posttranslationally attached to target proteins. Monoubiquitin functions as a molecular signal that regulates diverse cellular processes including the targeting of proteins to distinct subcellular locations and the promotion of protein-protein interactions.³⁵ As evidence of the critical role the FA core complex plays in this process, FA patient cells harboring deleterious mutations in FANCA, FANCB, FANCC, FANCE, FANCF, FANCG, FANCL, or FANCT lack the capacity to monoubiquitinate the ID2 heterodimer.^31–33 Monoubiquitinated ID2 localizes to chromatin where it associates with several established DNA repair proteins, including BRCA1/FANCS, BRCA2/FANCD1, and RAD51/ FANCR^{31, 36, 37} Monoubiquitinated ID2 is thought to function in the recruitment of several DNA repair proteins, including CtIP, Fanconi anemia associated nuclease 1 (FAN1), SLX4/FANCP, and ERCC4/FANCQ.^38–44 SLX4/FANCP and the ERCC4/FANCQ endonuclease catalyze the unhooking of the ICL,³⁹ enabling translesion DNA synthesis (TLS) beyond the ICL by the multi-subunit TLS polymerase Polζ, one subunit of which is REV7/FANCV (Fig. 2).^{45, 46}

Figure 1.

Schematic of FANCD2 and FANCI monoubiquitination. Following exposure to DNA damaging agents and during S-phase of the cell cycle, the FANCM anchor complex, comprising FANCM, FAAP24, FAAP16/MHF1 and FAAP10/MHF2, recognizes the damage, remodels the fork, and promotes the recruitment of the FA core complex. The FA core complex, which is comprised of three sub-complexes - FANCB/FANCL/FAAP100, FANCC/FANCE/FANCF, and FANCA/FANCG/FAAP100 - together with the E2 ubiquitin-conjugating enzyme UBE2T/FANCT, constitutes an active multisubunit E2/E3 ubiquitination enzyme complex. This E2/E3 enzyme complex catalyzes the site-specific monoubiquitination of FANCD2 K561 and FANCI K523.

Figure 2.

Schematic of downstream steps of the FA pathway. During the later stages of ICL repair, SLX4/FANCP and ERCC4/FANCQ catalyze the unhooking of the ICL (Step 1). The unhooked ICL is then bypassed by Polζ (Step 2). REV7/FANCV is a structural component of Polζ that enhances the polymerase activity of the catalytic subunit, REV3. Downstream of FANCD2 and FANCI monoubiquitination, BRCA2/FANCD1, BRIP1/FANCJ, PALB2/FANCN, RAD51C/FANCO, RAD51/FANCR, BRCA1/FANCS, and XRCC2/FANCU all function cooperatively to repair the remaining broken duplex by homologous recombination (HR). Following 5′–3′ DNA strand resection, RAD51/FANCR forms nucleoprotein filaments on single-stranded DNA and catalyzes homologous pairing and DNA strand invasion. BRCA2/FANCD1, PALB2/FANCN, RAD51C/FANCO, and XRCC2/FANCU facilitate these RAD51/FANCR-mediated processes.

During the final phase of ICL repair, the downstream FA proteins - BRCA2/FANCD1, BRIP1/FANCJ, PALB2/FANCN, RAD51C/FANCO, BRCA1/FANCS, RAD51/ FANCR, and XRCC2/FANCU - function cooperatively to repair the remaining broken duplex via homologous recombination (HR) (Fig. 2). HR is predominantly a conservative and error-free process whereby DNA damage is repaired using a homologous DNA template, typically the sister chromatid.^{47, 48} RAD51/FANCR is the major eukaryotic HR repair protein. RAD51/FANCR forms nucleoprotein filaments on 5′–3′ resected single-stranded DNA and catalyzes homologous pairing and DNA strand invasion and exchange. Many of the downstream FA proteins, e.g. BRCA2/FANCD1, PALB2/FANCN, and RAD51C/FANCO, are known to facilitate RAD51 function.^49–51. Disruption of the FA-BRCA pathway leads to defective HR and an increased dependence on the typically error-prone nonhomologous DNA end joining (NHEJ) repair pathway.^52–54

3. UBE2T/FANCT is responsible for a new FA subtype

The covalent attachment of ubiquitin to proteins regulates a variety of cellular pathways. This transfer is completed through a cascade of ubiquitin-related enzymes: an ubiquitin-activating enzyme (E1), an ubiquitin-conjugating enzyme (E2), and an ubiquitin ligase (E3) (Fig. 1).⁵⁵ Monoubiquitination of FANCD2/FANCI is the central step of the FA-BRCA pathway, and it is estimated that this step is defective in >90% FA patients, and intricately tied to patients’ increased BMF and cancer risk. UBE2T, newly termed FANCT, is the E2 ubiquitin-conjugating enzyme for FANCD2 and FANCI.^{28, 29, 56} In 2000, Zhang et al. cataloged three hundred cDNAs of previously undefined genes expressed in CD34 positive hematopoietic stem cells, one of which, HSPC150, contained a ubiquitin-conjugating motif.⁵⁷ HSPC150 was mapped to 1q31, a region known to be amplified in a range of cancers, including breast cancers, hepatomas, and cervical carcinomas.⁵⁸ This protein, later renamed UBE2T, was demonstrated in biochemical assays to interact with FANCL, the E3 ubiquitin ligase of the FA pathway.²⁹ UBE2T was subsequently shown to be required for FANCD2 monoubiquitination. Consequently, similar to other FA patient cells, loss of UBE2T leads to increased sensitivity to ICL-induced chromosome damage.²⁹ These data clearly implicated UBE2T as the ubiquitin-conjugating enzyme for the FA-BRCA pathway, yet stopped short of classification of a new subtype.

In 2015, Hira et al. from the DNA damage signaling laboratory of Kyoto University reported two unrelated FA patients who, through WES and Sanger sequencing, were found to have biallelic UBE2T mutations.⁵⁹ Both patients presented with congenital malformations, hematological abnormalities, and early onset BMF at ages 8 and 13. A heterozygous missense mutation leading to p.Q2E was uncovered in both patients.⁵⁹ This highly conserved glutamine residue is found in an amino-terminal helix, which is part of the E2–E3 interacting interface.⁶⁰ Disruption of this interface results in less efficient FANCD2 ubiquitination.⁵⁹ Additionally, both patients harbored unique mutations in their second alleles, a 23 kilobase genomic deletion in one patient and skipped exon resulting in a frameshift and premature stop codon in the other. Complementation of patient cells with wild-type UBE2T restored efficient FANCD2 monoubiquitination and nuclear foci formation, and rescued the increased sensitivity to MMC-induced chromosome breakage.⁵⁹ Hira et al. suggested that this new FA complementation group be named FA-T.

Coinciding with this study, the Laboratory of Genome Maintenance at Rockefeller University described an individual with biallelic mutations in UBE2T, born with microcephaly and bilateral thumb malformations, who tested positive in the DEB chromosome breakage test.⁶¹ However, this patient had normal bone marrow cellularity, normal leukocyte and thrombocyte counts, and only mild anemia, preventing a firm FA diagnosis for the first 16 years of life. Proband cells lacked the ability to monoubiquitinate FANCD2 and FANCI, failed to form FANCD2 nuclear foci, and were hypersensitive to ICL-inducing agents. Just months later, Virts and Jankowska et al. at Indiana University School of Medicine reported on the same 16 year old FA patient with biallelic mutations in UBE2T.⁶² In both cases, Sanger sequencing of genomic DNA identified two germline mutations, a paternal deletion and a maternal duplication of exons 2–6, both caused by aluY-mediated recombination events. Just as with the patients reported by Hira et al., retroviral complementation with wild-type UBE2T rescued the ICL-induced cell cycle arrest and chromosome breakage phenotypes.^{59, 61, 62} With these three announcements of biallelic UBE2T mutations leading to FA, the new FA-T complementation group has been clearly established and demonstrates that the FANCT alias is warranted.

4. Mutations in MAD2L2/REV7 underlie FA complementation group V

The most recent addition to the family of FA-BRCA genes is MAD2L2/REV7, a subunit of DNA polymerase ζ (Polζ). Involved in translesion DNA synthesis (TLS), Polζ is capable of synthesizing directly across template DNA lesions; specifically inserting nucleotides opposite DNA adducts, abasic sites, and, of particular interest in FA, DNA crosslinks. Through this process, the cell can tolerate the effects of DNA lesions during replication, effectively guarding against genomic instability. Structurally, Polζ is comprised of REV1, REV3, REV7, POLD2, and POLD3.⁶³ REV7 has been shown to stabilize and enhance the polymerase activity of the catalytic subunit, REV3 by a factor of 20–30 demonstrating its role as a critical subunit.⁶³ Additionally, the interaction between REV7 and REV3 has been shown to be necessary to confer DNA damage resistance.⁶⁴ In 2005, Okada et al. demonstrated that Rev1, Rev3, and Rev7 depletion in chicken lymphocytes lead to hypersensitivity to various DNA damaging agents, including crosslink-inducing UV irradiation.⁶⁵ These data suggested a possible phenotypic link to the FA-BRCA pathway.

In August 2016, Bluteau et al. reported a patient presenting with clinical and cellular phenotypes of FA.⁶⁶ The patient exhibited various physical anomalies including short stature, microcephaly, and renal abnormalities, and presented with severe multi-lineage BMF. Patient-derived cells exhibited increased susceptibility to ICL-induced chromosome breakage and cell cycle arrest, concurrent with an FA diagnosis. Sequencing of known FA genes failed to identify pathogenic mutations. Whole-exome sequencing analysis revealed a homozygous mutation in the MAD2L2/REV7 gene, c.354T>A, resulting in a nonsynonymous p.V85E change in the translated protein. This substitution was predicted to be pathogenic, as it lies in a highly conserved domain of the protein that is known to facilitate interaction between catalytic subunits REV1 and REV3.⁶⁴ No REV7 protein could be detected in patient cells despite the presence of normal transcript levels.⁶⁶ Retroviral complementation with wild type REV7 rescued the increased susceptibility of patient cells to ICL-induced chromosome breakage and cell cycle arrest, establishing REV7 as FANCV. Confirming a role for REV7 in the FA-BRCA pathway and ICL repair, homozygous mutation of MAD2L2/REV7 using CRISPR/Cas9 recapitulated the hallmark FA cellular phenotypes.⁶⁶ Furthermore, in support of an important role for REV7 in hematopoiesis, murine hematopoietic stems cells (HSC) deficient in Rev7 exhibited impaired colony forming ability and increased differentiation, similar to Fancg-deficient HSCs.⁶⁶ Taken together, these findings implicate REV7/FANCV as a necessary downstream effector of the FA-BRCA pathway, most likely functioning in the translesion DNA synthesis step of ICL repair (Fig. 2). It remains to be seen if mutations in other Polζ subunit genes are causative for FA.

5. Discovery of BRCA1 as an FA gene

BRCA1 is one of the most well known breast and ovarian cancer susceptibility genes, along with BRCA2/FANCD1. In 1990, Mary-Claire King’s group at University of California, Berkeley mapped a gene responsible for inherited breast cancer to chromosome 17q21 and, four years later, BRCA1 was identified by positional cloning.^{67, 68} In 2001, a biochemical link between the FA pathway and BRCA1 was uncovered, leading to the coining of the phrase the FA-BRCA pathway. Monoubiquitinated FANCD2 was shown to co-localize with BRCA1 in discrete sub-nuclear foci following exposure to DNA damage, as well as in synaptonemal complexes of meiotic chromosomes.³¹ In late 2012, at the Perelman School of Medicine at the University of Pennsylvania, the laboratory of Roger A. Greenberg identified a patient with one BRCA1 allele with a deleterious mutation and a second allelic variant of unknown clinical significance (VUS): the first report of a patient with biallelic BRCA1 mutations.⁶⁹ While never formally diagnosed with FA, the patient presented with classical FA-like phenotypes, including physical anomalies, microcephaly, developmental delay, and cellular sensitivity to chemotherapeutic agents, which is not typically displayed with single allelic mutations. The proband also exhibited an increased susceptibility to cancer, evident in the development of stage IV papillary serous ovarian carcinoma by 28 years of age (Table 1). The patient’s VUS was predicted to result in a valine to alanine change at amino acid 1736 (p.V1736A) in the carboxy-terminal BRCT domain.⁶⁹ This domain plays a critical role in BRCA1 function by mediating interaction with phosphorylated proteins.⁷⁰ The V1736A change led to a diminished interaction with the RAP80 protein and reduced localization of BRCA1 to DSBs. The second allele contained a truncating mutation, 2576delC, leading to a premature stop codon in exon 11.⁶⁹

Table 1.

Summary of the clinical phenotypes of FA-R, FA-S, FA-T, FA-U, and FA-V patients described in this review article.

Patient	Microcephaly	Growth retardation	Skin pigmentation	Radial- ray defects	Microphthalmia	Renal anomalies	Brain anomalies	Genital anomalies	Hipdysplasia	Cancer	BMF (y)
FA- R57	+	+	+	+	+	+	+	NR	NR	NR	NR
FA- R56	+	+	NR	+	NR	NR	+	+	NR	−	−
FA-S63	+	+	NR	NR	NR	NR	NR	NR	NR	+a	−
FA-S65	+	+	+	+	NR	+	NR	NR	+	+b	−
FA- T71 (PNG S-252)	NR	+	NR	+	NR	NR	NR	+	NR	NR	13
FA- T71 (PNG S-255)	NR	+	NR	+	NR	NR	NR	NR	NR	+	8
FA- T73,74 (10016 6/1)	+	+	+	+	NR	+	NR	NR	NR	NR	−
FA- U75	+	+	+	+	NR	+	NR	+	NR	NR	−
FA- V66 (EGF- 123)	+	+	NR	NR	NR	+	NR	NR	NR	NR	8

^aStage IV papillary serous ovarian cancer

^bDuctal breast carcinoma

NR, Not reported

Early in 2015, the Greenberg laboratory reported the presence of biallelic BRCA1 mutations in another female patient exhibiting phenotypes concordant with FA and presenting with ductal breast carcinoma at the age of 23.⁷¹ Patient cells were deficient in localization of BRCA1 to DNA damage sites, demonstrated increased sensitivity to ICL-inducing agents, and had their disease state phenotypes ameliorated by expression of wild-type BRCA1. WES of the proband revealed biallelic compound heterozygous variants in BRCA1. One allele contained a 4 base pair deletion at exon 10 leading to a frame shift in exon 11 (p.S198Rfs*35) and the second harbored a point mutation in exon18 (p.A1699T), predicted to result in misfolding of the BRCT domain and protein instability.^{71, 72} After chromosome breakage results were found to be in the range for a FA diagnosis, the patient was formally diagnosed with FA subtype S (FA-S), finally establishing BRCA1 as an FA gene.⁷¹

Numerous biochemical connections between BRCA1/FANCS and the FA-BRCA pathway have been established. For example, BRCA1/FANCS associates with BRCA2/FANCD1 and RAD51/FANCR (see below) and phosphorylated BRIP1/FANCJ interacts with the carboxy-terminal BRCT domains of BRCA1/FANCS.^73–75 Nevertheless, the exact role(s) of BRCA1/FANCS in ICL repair remains to be clearly determined. Recent studies have shown that BRCA1/FANCS is necessary for the removal of the CMG replicative helicase during ICL repair.⁷⁶ And previous studies indicate that BRCA1 competes with 53BP1 early in the repair process to promote HR and restrict error-prone NHEJ.^{77, 78} Therefore, it seems likely that BRCA1/FANCS may play several distinct roles during the process of ICL repair.

6. Institution of a new alias for RAD51: FANCR

The bacterial recA protein catalyzes strand exchange between homologous regions of two double-stranded DNA molecules in E. coli. The human homolog, RAD51, plays a parallel role in HR during the repair of DNA double-strand breaks (DSBs).^{79, 80} Multiple biochemical connections between the FA proteins and RAD51 have been established over the past two decades. For example, RAD51 is necessary for resistance to ICL-inducing agents,^{81, 82} and FANCD2, BRCA1/FANCS, BRCA2/FANCD1 and RAD51 function in a common DNA damage response pathway.^{31, 36, 37, 75, 83} Nevertheless, a causative genetic link between FA and RAD51 was unexpected given the early embryonic lethality exhibited by rad51 nullizygous mice.⁸⁴

Within the past year, however, two groups have described heterozygous dominant-negative RAD51 mutations in patients with an FA-like disorder.^{85, 86} Wang et al. from the Smogorzewska laboratory at Rockefeller University described a de novo heterozygous mutation c.391A>C in RAD51, resulting in a threonine to proline amino acid change (p.T131P).⁸⁶ Patient cells exhibited ICL hypersensitivity, yet remained proficient for HR, most likely because of a low ratio of mutant to wild-type RAD51.⁸⁶ These results indicate that RAD51 has additional roles in ICL repair distinct from its recombinase function. For example, several studies have shown that RAD51 plays an important role in replication fork protection prior to DSB formation.^87–89 Mutant RAD51-T131P acts antagonistically to wild-type RAD51 causing a disease-state in the context of a single mutant allele, representing the first report of an FA being caused by a de novo dominant-negative heterozygous mutation.⁸⁶ In the second report, Ameziane et al., described an individual with growth retardation, microcephaly, skeletal abnormalities and other physical phenotypes consistent with FA (Table 1).⁸⁵ A DEB chromosome breakage test provided grounds for an FA diagnosis; patient cells displayed increased levels of chromosomal aberrations and extensive accrual in late S-G2 phase of the cell cycle. Whole-genome sequencing, whole-exome sequencing (WES), and Sanger sequencing of the patient implicated a de novo g.41022153G>A mutation in RAD51 as the cause of disease. This mutation, leading to alanine 293 being substituted for threonine (p.A293T), is found in a highly conserved region of RAD51 affecting its ability to undergo oligomerization. RAD51 oligomerization is essential for nucleoprotein filament formation, underlying the dominant-negative nature of this mutation. ^{79, 80} Biochemical studies of mutant RAD51-A293T also revealed impaired D-loop formation and decreased DNA binding.⁸⁵ Taking into account these two reports, the role of RAD51 in the FA pathway as a downstream FA protein has been clearly established and RAD51 assumes the alias of FANCR.

7. XRCC2/FANCU and establishment of FA-U

The most recent and twentieth FA gene is XRCC2/FANCU (X-ray cross-complementing gene 2).⁹⁰ XRCC2 is a member of the RAD51 family of proteins, encoding for products integrally involved in HR and critical for the maintenance of genome integrity.⁵¹ XRCC2 is a member of the BCDX2 protein complex - comprising RAD51B, RAD51C, RAD51D, and XRCC2 - which is thought to have several functions in HR, including stabilization of RAD51 nucleoprotein filaments and the promotion of RAD51-dependent homologous pairing and strand invasion.^{51, 91, 92} Underscoring the critical functions of XRCC2, homozygous disruption of murine Xrcc2 results in embryonic lethality occurring from mid-gestation, with Xrcc2^−/− embryonic cells exhibiting high levels of chromosome aberrations.⁹³

In 2012, Shamseldin et al. described a 2 year old Saudi Arabian patient presenting with classical FA phenotypes, including microcephaly, bilaterally absent thumbs, and kidney malformations. Positional mapping and WES uncovered a homozygous biallelic stop mutation in XRCC2.⁹⁴ This mutation, c.643C>T, is predicted to encode p.R215* resulting in the production of a carboxy-terminal truncated protein. However, due to the presence of mutations in other genes, XRCC2 could not be clearly established as the causative gene in the absence of complementation studies.⁹⁴ In 2016, Park et al. of the Andreassen laboratory at Cincinnati Children’s Hospital Medical Center further examined this patient.⁹⁰ On a cellular level, the patient exhibited a heightened sensitivity to DEB and accumulation in G2-M phase of cell cycle, in line with FA. Reintroduction of wild-type XRCC2 corrected all pathological cellular phenotypes, establishing that the XRCC2 mutations were responsible for the observed cellular defects. Re-expression of XRCC2 also restored RAD51 nuclear foci to wild-type levels and corrected the increased ionizing radiation and ICL sensitivity. Consistent with XRCC2 functioning at a later stage of the FA-BRCA pathway, FANCD2/FANCI monoubiquitination was not affected in patient cells.⁹⁰ Heterozygous mutations in XRCC2 have previously been shown to be associated with an increased predisposition to breast cancer.⁹⁵ This is consistent with other downstream FA genes - BRCA1/FANCS, BRIP1/FANCJ, PALB2/FANCN, and RAD51C/ FANCO - that are also associated with an increased risk of developing breast cancer.^96–102

With this new classification of XRCC2 as an FA gene, and in conjunction with the establishment of RAD51C as FANCO, it is clear that the RAD51 paralog complex BCDX2 plays critical functions in the FA-BRCA pathway.^{90, 103} This leads to speculation that mutations in other RAD51 paralog-encoding genes could lead to FA. Further experiments and genetic screening will be needed to determine the status of other RAD51 paralogs as potential FA genes.

8. Summary and final remarks

The discovery of these five new bona fide FA genes further strengthens the molecular ties between FA and DNA repair. These discoveries not only further elucidate key facets of the FA-BRCA repair pathway, but also have important implications for FA genetic testing. As there are many FA patients who currently cannot be assigned to any of the existing complementation groups, it is certain that more FA genes will be discovered. The identification of new genes using the powerful combination of next-generation sequencing and classical biochemistry approaches represents one critical aspect of the urgent quest in improve therapeutic options for this devastating disease, and to provide much needed hope for FA patients and their families.

Practice points.

Any of the following clinical symptoms should raise suspicion for FA. If FA is suspected, physicians are advised to refer the patient to a hematologist/oncologist and/or geneticist for FA genetic testing, preferably at a center experienced in FA diagnostic testing (see fanconi.org).

• Atypically early-onset myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML), in addition to any of the following

• Familial cancer predisposition and/or a history of excessive toxicity to radiation or chemotherapy

• Unexplained cytopenias of any cell lineage due to bone marrow failure, e.g. thrombocytopenia, neutropenia, or anemia

• Idiopathic macrocytsosis in the absence of identifiable cause, e.g. vitamin B12 or folate deficiency

• Congenital anomalies including short stature, microcephaly, microphthalmia, hypo- or hyper-pigmentation of the skin, radial ray defects, male hypogonadism, and structural renal defects

• Atypically early-onset tumors, particularly head and neck squamous cell carcinomas in the absence of a history of risk factors such as alcohol or tobacco use

Research agenda.

• Continue to combine next-generation sequencing and classical biochemistry approaches to identify new FA genes

• While a role for the FA proteins in the repair of ICLs generated by exogenous agents in vitro has been clearly established, the endogenous source(s) of genomic instability in the physiological setting still remains to be determined

• Determine the mechanisms underlying the tissue-specificity of FA cancers. For example, why are FA patients at increased risk for head and neck squamous cell carcinomas?

• Discover novel strategies to effectively prevent and treat FA BMF and cancers