Monday, 8 October 2007

The Fanconi anaemia DNA repair pathway

Fanconi anaemia (FA) is an autosomal recessive chromosomal instability, characterised by congenital abnormalities, defective haemopoiesis, a high risk of leukaemia and development of solid tumours. FA patients have a 10% incidence of leukaemia and are 50 times more likely to develop solid tumours (particularly hepatic, oesophageal, oropharyngeal and vulval) by early adulthood. The disease consists of 13 complementation groups, each connected to mutations in 13 corresponding FANC genes (A, B, C, D1, D2, E, F, G, I, J, L, M, and N) which make up the FA pathway. Two thirds of FA cases are caused by inherited mutations in FANCA.

Interstrand crosslink damage

FA cells are characterised by hypersensitivity to damage by DNA Interstrand CrossLinking (ICL) agents such as mitomycin C and diepoxybutane. Cells fail to repair ICL damage, resulting in polysomies, radial chromosomal structures and unrepaired breaks (Figure 1). ICL agents tether two DNA strands together, creating a physical barrier to replication and transcription, which causes arrest of DNA replication. In normal cells, several mechanisms come in place to minimize damage and maintain the replication event. This necessitates removal of the physical barrier, repair of the gap and re-establishment of the replication fork. This complex repair mechanism involves the Fanconi anaemia proteins, translesion synthesis (TLS) polymerases, and the homologous recombination (HR) machinery. The FA pathway has also been implicated in spontaneous DNA damage repair, despite early speculation that the pathway was specific to ICL damage. In addition, there is some uncertainty as to whether the FA pathway also functions in intra-strand crosslink repair.

The FA pathway

The FA proteins are part of a complex DNA repair pathway, which is not yet fully understood. They are commonly arranged in three distinct groups according to their function. The first group is the FA core complex, which consists of FANC A, B, C, E, F, G, L and M. This complex stabilises stalled replication forks, processes the ICL site and activates a second group, which consists of FANCD2 and FANCI. Finally, group 3 consists of the remaining FANC proteins, BRCA2 (which is the same gene as FANCD1), FANCJ and FANCN, which have accessory or uncertain function in the pathway (Figure 2). The most recent model of this pathway is described in Wang (2007), Kennedy & D’Andrea (2005) and Niedernhofer (2007), and a detailed description of the model follows.

An ICL physically blocks replication by preventing separation of the strands by the advancing DNA helicase, which causes the enzyme to arrest. The replication fork arrest activates the ATR kinase, which subsequently phosphorylates FANCD2. A double strand break (DSB) upstream of the ICL is carried out by XPF-ERCC1, MUS81-EME1 or MUS81-EME2. Simultaneously, the monoubiquitin ligase FANCL, part of the FA core complex, monoubiquitinates FANCD2 and FANCI. This stage is regulated by the deubiquitinating enzyme USP1, and the monoubiquitin acts as a chromatin localisation signal. UPS1 is itself regulated by cell cycle controls, and self-cleaves in response to DNA damage. The integrity of the whole of the core complex is essential for this stage, as proven by the lack of chromatin localisation in FA core complex mutants. The downstream function of FANCI-Ub is uncertain, but FANCD2-Ub is arguably the most important component of this pathway, acting as a “fire captain” to recruit the proteins needed for the rest of the pathway (Figure 3).

As group 2 is activated, the FA core complex is also carrying out DNA processing in preparation for the next step in the pathway. FANCM acts as a DNA translocase. It anchors the rest of the FA core complex onto DNA, which stabilises the broken replication fork and facilitates removal of the ICL. A nucleotide excision repair (NER) endonuclease “unhooks” the ICL (which remains attached to the one strand), and translesion synthesis (TLS) polymerases recruited by the FA core complex fill in the gap. The TLS polymerases’ ability to pass through the damaged site has the trade-off of being error-prone, introducing random mutations in the repair site. The ICL adduct is then completely removed by a NER exonuclease, and the result is a replication fork broken by a DSB, with the ICL removed and the gap filled in (Figure 4). The next step is to repair the DSB, which is carried out by homologous recombination (HR) proteins.

HR-mediated replication fork re-establishment

FANCD2-Ub forms a complex with BRCA2 and chromatin. All the FANC proteins, along with BRCA1, NBS1, PCNA, RAD51 and other related proteins are recruited by the FANCD2-Ub/BRCA2-chromatin complex into nuclear foci at the damage site. BRCA1 forms a complex with BRCA2 and is essential for the formation of nuclear foci, even though its exact function is unknown. FANCJ, also identified as the helicase BRIP1 (BRCA1-interacting protein), depends on BRCA1 to translocate to the DNA damage site and unwind the replication fork in order to promote HR. The MRN complex, which consists of MRE11, RAD50 and NBS1 and has 3’-5’ exonuclease activity, processes either side of the DSB to prepare it for HR.

DSB repair can occur in two ways. Firstly by single-strand annealing, an error-prone non-homologous end joining (NHEJ) pathway, where the two strands are trimmed until homologous repetitive sequences are reached, and the strands are then annealed at that region of homology. This is mutagenic since the non-homologous intermediate region is deleted, but this is preferable to the arrest of replication. Secondly by strand invasion HR, where RAD51 is loaded onto chromatin by BRCA2 to form a nucleoprotein filament, within which Sister Chromatid Exchange (SCE) occurs and a Holliday junction forms between the strand being repaired and a homologous strand (Figure 5). Holliday junctions are intersections of four strands of DNA, which appear most commonly during meiosis when “crossing-over” occurs, and are resolved by endonucleases such as MUS81-EME1. FA cells show high levels of SCE in FANCC mutants, while other FANC mutant cells have normal SCE levels, suggesting a regulatory role for FANCC over SCE formation and a generally significant role in HR. HR thus re-establishes the replication fork and at the end of the S-phase, USP1 deubiquitinates FANCD2 to turn the pathway off.

In summary, ICL damage causes replication arrest, one of the strands is separated by a DSB, the FA core complex stabilises the broken fork and TLS repairs the ICL. At the same time, the structural integrity of the FA core complex is essential for the monoubiquitination of the “fire captain” FANCD2, which forms a complex with BRCA2 and chromatin and recruits HR and accessory proteins into nuclear foci. Finally, the MRN complex processes the ends of the DSB for HR to repair it and allow replication to restart.

There has been increased interest in the FA pathway in the past decade or so due to the discovery of its connection with cancer and complex relationship with the HR pathway. The pathway model is still in its early stages but already there are important questions raised about cell cycle regulation and its connection with DNA repair.


Kennedy, R.D. and D'Andrea, A.D. (2005) The Fanconi Anemia/BRCA pathway: new faces in the crowd. Genes Dev. 19: 2925-2940.

Niedernhofer,L.J., Lalai,A.S., and Hoeijmakers,J.H.J. (2005) Fanconi Anemia (Cross)linked to DNA Repair. Cell 123: 1191-1198

Patel,K.J. and Joenje,H. (2007) Fanconi anemia and DNA replication repair. DNA Repair 6: 885-890

Tischkowitz,M. and Dokal,I. (2004) Fanconi anaemia and leukaemia - clinical and molecular aspects. British Journal of Haematology 126: 176-191

Wang,W. (2007) Emergence of a DNA-damage response network consisting of Fanconi anaemia and BRCA proteins. Nat Rev Genet 8: 735-748


Catarina Vicente said...

I had never heard of this... Does any of the pappers that you mention have a nice diagram of the process? I need a picture...

Menelaos Symeonides said...

Yes, check the Niedernhofer one for an early model and the Wang one for a newer but more difficult one. I am still trying to get my head around it, especially the connection between the FA proteins and the HR proteins, it seems to be very complex.

James Lloyd said...

just to clearthings up i read it yesterday and commented but i didnt go through and lunch was over.

here is the question - i have only come across ubiquitinated proteins when it is a signle for breakdown by a proteasome. does FANCD2 avoid breakdown because only one ubiquitin is added and is it USP1 that stops more being added thereofre protecting it from breakdown?

Menelaos Symeonides said...

I will get back to you on that, in the meantime i made some edits today that cleared up some things I got wrong so you might want to have another quick look at it.

Menelaos Symeonides said...

Sorry it took a while but I've been quite busy. After reading a bit more about this (well actually a LOT more), I now understand it a lot more. The monoubiquitin works as a chromatin localisation signal. FANCD2 acts as a "fire captain" to recruit all the other proteins of the pathway (especially the HR proteins) to carry out DSB repair. USP1 is a deubiquitinase whose activity is controlled by the S-phase checkpoint. It acts against the FANCL ubiquitin ligase activity to regulate FANCD2 monoubiquitination. As far as I know FANCD2 only takes one ubiquitin so proteasome breakdown doesn't figure into this... but keep in mind that the pathway model is still very new and we know very little about all these proteins. The importance of FANCD2 is just beginning to come into the foreground and it is beginning to show in the literature, there are several large teams working on FA around the world.

Since I now know a lot more about the pathway, I think I might completely re-edit the post, I will let you know when I've done it.

Menelaos Symeonides said...

Ok, I found this paper that actually confirms that the proteasome is not involved, seems that FANCD2 is a special case of the ubiquitination pathway.

And this paper also says that FANCL is a monoubiquitin ligase, which means that it can only add one ubiquitin and its activity is FANCD2-specific, so you were along the right lines, it's because not enough ubiquitin is added, but it's not because USP1 regulates that, it's because that's all FANCL can do.

This will be really useful, thanks for pointing it out!

Anonymous said...

I saw you mentioned ubiquitin so I thought you might be interested in Science's current webinar:
The Ubiquitin-Proteasome Pathway