RETRACTED: A subset of PARP inhibitors induces lethal telomere fusion in ALT-dependent tumor cells

Abstract

About 10% of all tumors, including most lower-grade astrocytoma, rely on the alternative lengthening of telomere (ALT) mechanism to resolve telomeric shortening and avoid limitations on their growth. Here, we found that dependence on the ALT mechanism made cells hypersensitive to a subset of poly(ADP-ribose) polymerase inhibitors (PARPi). We found that this hypersensitivity was not associated with PARPi-created genomic DNA damage as in most PARPi-sensitive populations but rather with PARPi-induced telomere fusion. Mechanistically, we determined that PARP1 was recruited to the telomeres of ALT-dependent cells as part of a DNA damage response. By recruiting MRE11 and BRCC3 to stabilize TRF2 at the ends of telomeres, PARP1 blocked chromosomal fusion. Exposure of ALT-dependent tumor cells to a subset of PARPi induced a conformational change in PARP1 that limited binding to MRE11 and BRCC3 and delayed release of the TRF2-mediated block on lethal telomeric fusion. These results therefore provide a basis for PARPi treatment of ALT-dependent tumors, as well as establish chromosome fusion as a biomarker of their activity.

MRE11 and UBR5 Co-Operate to Suppress RNF168-Mediated Fusion of Dysfunctional Telomeres

“…Previous studies by our lab and others showed that suppression of MRE11, the MRE11 partner BRCC3, or UBR5 alone were sufficient to induced chromosomal fusion in a manner comparable to that seen by combined suppression of both MRE11 and UBR5, or by loss of TRF2 itself (9,12). On the surface these studies therefore suggest that MRE11 or UBR5 independently limit the entire RNF168-mediated fusion process.…”

“…The MGG119 and MGG152 PDX xenograft cells were a gift from Dan Cahill (Massachusetts General, Boston, MA). These cells differ in ATRX status and ALT dependency: MGG119 are ATRX deficient and ALT dependent, whereas MGG152 are ATRX proficient and ALT independent ( 12 , 13 ). SF8628 is an ATRX proficient, K27M mutated, and ALT dependent pediatric diffuse intrinsic pontine glioma (DIPG) cell line ( 14 ) obtained from the Brain Tumor Center (BTC) Tissue Bank at the University of California, San Francisco (UCSF).…”

“…IF-FISH analysis was performed as described previously ( 12 ). In brief, paraformaldehyde-fixed cells were blocked at room temperature for 30 mins and then incubated with 53BP1 (1:200) primary antibodies for 18 hours at 4°C.…”

“…Co-localization of green and red fluorescence from the overlap of PML and TRF2 signal results in yellow foci was considered as APBs and was counted on >200 cells per group. Leading (TelC-Cy3-red) and lagging (TelG-FITC-green) strand probes (PNA Bio) as described previously ( 12 ) was used to quantify telomeric sister-chromatid exchange (T-SCE) by chromosome orientation fluorescence in situ hybridization (CO-FISH). Sister chromatid exchange results in overlapping of the leading and lagging strand probes generating yellow signal were considered T-SCE events.…”

“…The mechanism is used by (5-10%) of many cancer sub-types, and nearly all lower-grade astrocytomas. We previously showed that these cells lines, by virtue of expression of mutant IDH1 and loss of ATRX, have dysfunctional telomeres that trigger a DNA damage response involving telomeric recruitment of both γH2AX and PARP1 ( 11 , 12 ). The telomeres of these ALT-dependent cells, however, retain TRF2, which in turn is sufficient to prevent telomeric fusion.…”