FBL 2'-O-methylates FANCD2 mRNA to sustain interstrand crosslink repair

FANCD2 control moves upstream of monoubiquitination

The canonical FA model places FANCD2 monoubiquitination at the center of ICL repair: the FA core complex modifies the ID2 heterodimer, enabling chromatin retention, foci formation, and recruitment of downstream effectors such as SLX4 and homologous recombination factors [1][2][4]. Renaud et al. established that monoubiquitinated FANCD2 is required to relocalize Tip60 to damaged chromatin, where H4K16 acetylation blocks 53BP1 binding and permits resection [3]. Huang et al. showed that ICL-proximal replisome remodeling depends on ATR, FANCM, and FANCD2, reinforcing FANCD2 as a hub for coordinating replication-coupled repair [2]. What remained unresolved was how FANCD2 abundance itself is set before monoubiquitination occurs.

The anchor paper addresses that gap by identifying FBL as a 2'-O-methyltransferase acting on FANCD2 mRNA rather than on FANCD2 protein [1]. FBL depletion lowered FANCD2 mRNA and total FANCD2 protein, and the decrease in monoubiquitinated FANCD2 tracked with reduced total protein rather than a specific block in ubiquitination [1]. This reframes FANCD2 activation as a two-tier process: transcript stabilization sets the ceiling, and monoubiquitination determines the fraction recruited. The interpretation is that epitranscriptional regulation is not a parallel pathway but a supply-side input into the canonical cascade.

Nm at position 2837 extends FANCD2 mRNA half-life

Using RTL-P, the authors mapped 2'-O-methylation to position 2837 within the FANCD2 coding region and showed that FBL knockdown reduced Nm at this site [1]. FANCD2 mRNA half-life was shortened in shFBL cells, and reconstitution with wild-type FBL, but not the catalytically inactive E191A/D236A mutant, restored Nm and stability [1]. A biotin-labeled probe centered on U2837 pulled down FLAG-FBL, supporting direct interaction with the methylated region [1]. The functional consequence was increased FANCD2 protein, more FANCD2 foci after MMC, and restored Tip60 and BRCA2 recruitment to damage sites [1].

This mechanism is distinct from previously described RNA modification roles in the DNA damage response. METTL3-dependent m6A influences R-loop accumulation and RAD51/BRCA1 recruitment, and FANCD2 itself has been linked to R-loop management through SRSF1-mediated mRNA export and through hnRNP U/DDX47 interactions [7][8]. Those pathways act on R-loop metabolism or mRNA export, whereas FBL-mediated Nm acts directly on FANCD2 transcript stability [1]. The distinction matters because it predicts that FBL loss should phenocopy FANCD2 insufficiency specifically, rather than broadly disrupting R-loop homeostasis. The supplied data support the stability claim in cell lines; whether Nm at 2837 is the dominant modification controlling FANCD2 in primary hematopoietic cells remains untested.

Tip60 and BRCA2 recruitment depend on FBL-supported FANCD2 levels

The anchor paper places FBL upstream of the Tip60 arm defined by Renaud et al. [3]. In FBL-knockdown HCT116 cells, MMC-induced Tip60 chromatin loading and FANCD2-Tip60 co-localization were reduced, and BRCA2 foci were diminished while gammaH2AX foci accumulated [1]. Reconstitution with wild-type FBL restored BRCA2 foci, whereas the E191A/D236A mutant rescued BRCA1 expression but not BRCA2 foci, arguing that FANCD2 stability, not BRCA1 deficiency, drives the BRCA2 recruitment defect [1]. This is consistent with the precursor model in which Ub-FANCD2 relocalizes Tip60 to chromatin to promote H4K16 acetylation and HR [3].

The comparison with competing FA-pathway regulators clarifies the boundary. Toosendanin inhibits FA signaling by suppressing JAK/STAT1-mediated transcription of FA core subunits such as FANCA, FANCC, FANCF, and FANCM, reducing FANCI/FANCD2 monoubiquitination and ID2-SLX4 interaction in bladder cancer cells [4]. That mechanism operates at the level of core-complex abundance and would be expected to impair monoubiquitination even when FANCD2 mRNA is stable. Hypoxia provides another competing layer: acute hypoxia activates FANCD2/FANCI ubiquitination via ATR, while prolonged hypoxia transcriptionally represses FANCD2 [6]. FBL-mediated Nm could modulate the set point for FANCD2 mRNA available during these stresses, but the supplied evidence does not test whether FBL activity changes under hypoxia or whether FBL loss alters the hypoxic response.

R207S links FBL catalytic loss to AML/MDS-associated FA complications

The paper identifies recurrent FBL mutations in AML/MDS through cBioPortal analysis, including R207S and T173M, both predicted damaging by SIFT and PolyPhen2 [1]. R207S lies in the MTase domain; structural modeling on the SAM-bound FBL structure (PDB 7SE7) suggested that the shorter serine side chain disrupts hydrogen bonds and electrostatic interactions stabilizing the alpha1 helix, weakening SAM binding [1]. Experimentally, R207S reduced methylation at canonical rRNA sites 5.8S 124U, 18S 1713C, and 28S 1858A, confirming compromised catalytic activity [1]. RTL-P showed reduced Nm on FANCD2 mRNA, actinomycin D chase showed faster FANCD2 mRNA decay, and R207S-reconstituted FBL-knockdown cells were more MMC-sensitive than wild-type reconstituted cells [1].

This connects an epitranscriptional defect to the hematologic complications that dominate FA morbidity. However, the evidence is correlative at the patient level: FBL was downregulated in FA patient datasets GSE16334 and GSE157591, and R207S was found in AML/MDS cohorts, but the paper does not show that R207S carriers have reduced FANCD2 mRNA or protein in primary patient samples [1]. Testa et al. note that MDS and AML with chromosome 7 aberrations involve complex mutational events, so FBL variants may be one of several contributors rather than a sole driver [5]. The R207S finding is best read as a mechanistic hypothesis with strong in vitro support, not as an established clinical biomarker.

What the model does not yet establish

The mechanistic work is performed in HCT116, A549, and HEK293T cell lines, with patient datasets used for expression correlation and mutation frequency rather than for functional validation [1]. No animal model is presented, so the requirement for FBL-mediated Nm in organismal ICL repair, hematopoiesis, or tumor suppression is untested. The therapeutic implication, that FBL or Nm could be targeted to sensitize cancers to platinum agents, remains speculative; the paper shows that FBL overexpression increases cisplatin resistance and FBL loss increases sensitivity, but does not test whether inhibiting FBL in vivo improves chemotherapy response [1].

The broader FA literature also cautions against overgeneralizing. FANCD2 has transcriptional and R-loop-related functions beyond ICL repair, including binding at common fragile sites and interacting with RNA processing factors [8], and E2F7 represses FANCD2 among other genomic stability genes, with E2F7 depletion reducing FANCD2 foci and increasing resistance to ICL agents [10]. Intra-tumor heterogeneity can affect FANCD2 at the genomic level in ovarian cancer [9]. These findings do not contradict the FBL-FANCD2 axis, but they mean that FANCD2 abundance in a given tumor or patient sample reflects multiple inputs. The open question is whether FBL-mediated Nm is a dominant, rate-limiting input in FA-relevant tissues or one of several layers that converge on FANCD2.

About These Sources

This research page is built on 10 peer-reviewed studies — published from 2014 to 2026, 3 from 2024 or later, collectively cited 348 times — selected as the most relevant from 12 studies that passed quality screening, drawn from 79 papers retrieved from a database of over 500 million.

Sources used in this answer