Endometrial cancer is the fourth most common cancer in women and the most common gynaecological malignancy in the developed world. No new systemic treatments for endometrial cancer have been developed in recent years and its incidence is expected to double over the next decade. As such, there is a need to gain a better understanding of key molecular pathways that are altered in the disease and could be targeted by novel treatments. The DNA mismatch repair (MMR) pathway is lost in approximately 30% of endometrial cancers. Recently, our lab has shown that MLH1-deficient cells demonstrate a mitochondrial phenotype characterised by reduced oxidative phosphorylation (OXPHOS), reduced mtDNA copy number and Complex I inhibition. OXPHOS-deficient cells have to adapt their metabolism to compensate for energy defects and the inability to efficiently use the tricarboxylic acid cycle to generate energy. We hypothesise that this altered metabolism is driving tumourigenesis by increasing the metastatic potential of the tumour cells and are investigating this using orthotopic models of MLH1-deficient endometrial cancer in vivo. We have performed metabolomic analysis on a panel of MLH1-proficient and deficient paired endometrial cell lines and identified a metabolic map of alterations upon MLH1 loss. Ultimately, we aim to use this knowledge of altered metabolism upon MLH1 loss to identify more targeted treatments for MMR-deficient endometrial cancer patients.

Here we present the characterisation of the so far unstudied NUDIX hydrolase family member NUDT22. We previously identified a unique hydrolase activity of NUDT22 towards UDP-glucose from a family-wide biochemical substrate screen. UDP-glucose hydrolysis results in the production of uridine monophosphate (UMP) and glucose 1-phosphate (G-1-P). We furthermore solved the first crystal structure of NUDT22 in complex with its substrate UDP-glucose [1]. Our mechanistic studies reveal increased replication stress in NUDT22 deficient cells which can be rescued by nucleoside supplementation. We therefore propose the discovery of a novel NUDT22-mediated pyrimidine salvage pathway.
Increased replication rates resulting in replication stress is a hallmark of cancer cells and NUDT22 gene expression alterations are present in several cancer tissues, which makes it an interesting new target for the development of small molecule inhibitors for uses in cancer.
We employed our NUDT22 crystal structure to perform an in silico docking screen on available small molecule libraries to identify starting points for the development of first-in-class NUDT22 inhibitors. Chemically optimised NUDT22 inhibitors are currently being validated in biochemical assays, cellular target engagement assays, and their cellular activity is being assessed in vitro.

[1] M. Carter et al., "Human NUDT22 Is a UDP-Glucose/Galactose Hydrolase Exhibiting a Unique Structural Fold," Structure, vol. 26, no. 2, pp. 295-+, Feb 2018, doi: 10.1016/j.str.2018.01.004.

Non-canonical four-stranded G-quadruplex (G4) structures form through self-recognition of guanines into stacked tetrads within chromatin DNA. Considerable evidences exist for G4s formation in vitro and have linked G4 formation with key biological processes ranging from transcription and translation to genome instability and cancer. However, G4s have not been exploited for colorectal cancer (CRC) treatment before. Our aim was to assess G4s as therapeutic targets in CRC tumorigenesis and characterize the role of G4-mediated genomic instability in CRC. We measured the presence of G4s and DNA damage at G0/G1 and S phase of cell cycle by immunofluorescence and Fiji quantification (N > 200) using antibodies against G4s (termed BG4) and gH2AX respectively in a cellular progression model of CRC. In addition, antitumoral activity of well-known G4-ligands was assessed by Resazurin method. Genomic instability driven by these G4-ligands was also evaluated by gH2AX immunofluorescence. Non-tumoral epithelial cells showed the lowest G4 and gH2AX levels, which significantly increased along CRC progression, suggesting a role in CRC tumorigenesis. Both G4s and gH2AX levels were positively correlated and the typical genome instability in CRC was partially explained by G4s presence. G4-ligands showed cytotoxic activity and significantly induced DNA damage. Subsequent studies are required to further evaluate the therapeutic potential of G4s and associated genomic instability in CRC.

In recent years, the infiltration of tumors by axons or nerves has been increasingly reported and has been linked to poor prognosis. This includes, among others, publications by Magnon et al (Science 341, 2013); and our own work (Faulkner et al., FASEB BioAdvances 2, 2020). Thus, recent investigations have been defying the old mechanistic, non-participating, view of the role of the nervous system in the tumor microenvironment. The “nerve-cancer connection” now is believed to encompass novel therapeutic targets already reported for breast, prostate and gastric cancers. However, the role of the autonomic nervous system in ovarian cancer development and progression remains unclear. We aimed to characterize this new component in ovarian tumors microenvironment. We identified the infiltration of peripheral axons in some ovarian tumors. In addition, ovarian tumors expressed neurotrophins, including nerve growth factor (NGF), in particular in the initial onset of the tumor. Our work expose the need to further comprehend the role of the nervous system in female cancers, namely in the unique microenvironments of ovarian tumors.

PSPC1 is a member of the Drosophila behaviour/human splicing protein family (DBHS), which were previously found to have a role in DNA repair, however its precise function in the DNA damage response was unclear. To further understand the role of PSPC1 in response to DNA damage, we silenced PSPC1 and treated cells with several DNA damaging agents including cisplatin, mitomycin C (MMC), hydroxyurea (HU), olaparib and irradiation. Interestingly, we observed that silencing PSPC1 promoted sensitivity to DNA interstrand crosslinking agents such as cisplatin and MMC but not the remaining DNA damaging agents. Given that cells with deficiency in the Fanconi Anaemia (FA) pathway are also sensitive to DNA interstrand crosslinking agents, we next investigated the role of PSPC1 in the FA pathway. Our results demonstrate for the first time that activation and mono-ubiquitination of FANCD2 was significantly reduced, upon MMC treatment in PSPC1-depleted cells. In addition, the expression of RPA32, FANCD2 and ATR were significantly reduced in PSPC1 knockout cells. Significantly, following RNA sequencing of the PSPC1 KO cells, expression of a large proportion of the FA core complex and FA-associated proteins (FAAPs) were reduced at the RNA level. Moreover, our data showed increased levels of DNA damage was induced and the repair of MMC-induced DNA damage was delayed upon PSPC1 loss. Taken together, our results suggest that PSPC1 is required to activate FANCD2-mediated DNA repair upon treatment with DNA interstrand crosslinking agents. Our results also suggest that loss of PSPC1 is a biomarker of sensitivity to DNA interstrand crosslinking agents. Given that PSPC1 loss is a feature of many tumour types, our data suggests that screening patient tumours for PSPC1 expression may represent an important biomarker for therapeutic selection.

Cancer cells respond to increases in DNA damage by upregulating their DNA damage response (DDR). The base excision repair (BER) pathway corrects damage to single DNA bases through the action of multiple enzymes, including the central protagonist, apurinic/apyrimidinic endonuclease 1 (APE1). Numerous studies have shown association between increased APE1 levels and enhanced growth, migration, and drug resistance in human tumor cells, as well as with decreased patient survival. APE1 has been implicated in over 20 human cancers, making this an attractive target for developing anticancer therapies. Despite intensive effort, there are currently no clinical endonuclease inhibitors of APE1. We have used a newly developed high-throughput protein X-ray crystallography-based fragment screen to obtain starting points for the design of molecules to block APE1 function. Starting with a proprietary fragment library, we obtained high quality fragment-bound crystal structures showing diversity of chemical matter and hit location, representing the first experimental 3D structures of APE1 bound to drug-like molecules, thereby resolving a primary bottleneck in the path to inhibitor development. The implementation of this unique lead discovery campaign has facilitated three independent strategies toward the development of APE1 inhibitors, including (i) fragment growing and elaboration of hits bound at the endonuclease site; (ii) linking of fragments bound to distinct but proximally located sites, and (iii) use of fragments for the design of hooks to use in targeted protein degradation (TPD) strategies. We are using a combination of computational and medicinal chemistry, structural biology, and biochemical and biophysical studies and will discuss our progress towards these goals.

The DNA damage is crucial for the emergence of cancer cells. If the DNA damage response is defective, the DNA damage is converted to fixed mutations. Some of these mutations drive tumorigenesis and are called driver mutations. However, the extent of consequential DNA damage per tumor, i.e. the number of various kinds of driver mutations, is not known. We have utilized the largest database of human cancer mutations – TCGA PanCanAtlas, multiple popular algorithms for cancer driver prediction and several custom scripts to estimate the number of various kinds of driver mutations in primary tumors. We have found that there are on average 19.6 driver mutations per patient’s tumor, of which 2.4 are hyperactivating SNA mutations in oncogenes, 9.2 are CNA amplifications of oncogenes, 0.6 have both in the same oncogene, 0.2 are homozygous inactivating SNA mutations in tumor suppressors, 1.1 have inactivating SNA mutation in one allele and CNA deletion in the other allele of a tumor suppressor, 1.5 are driver chromosome losses, 2 are driver chromosome gains, 1 is driver chromosome arm loss, and 1.6 are driver chromosome arm gains. The number of driver mutations per tumor increased with age, from 12.5 for <25 y.o. to 23.6 for >85 y.o. There was no big difference between genders (19.9 in males vs 19.2 in females). The number of driver mutations per tumor varied strongly between cancer types, from 1.5 in thyroid carcinoma to 43 in lung squamous cell carcinoma. Overall, our results provide valuable insights into the extent of functional DNA damage in tumors.

Abstract

Esophageal cancer occurs with extraordinarily high incidence and dismal survival in high-risk regions of Northern China, highlighting the unmet need for early detection to improve patient survival. Combining the enrichment of genetic component strategy and unbiased whole-exome sequencing (WES) approaches, we performed next-generation sequencing (NGS) analysis for 186 familial esophageal squamous cell carcinoma (ESCC) cases involving two generations with ≥2 family members diagnosed with ESCC, including the proband. We identified multiple candidate cancer predisposition genes (CPGs) for familial ESCC, including BRCA2, POLQ, and MSH2 involved in DNA repair. Further data mining of the WES data in the discovery phase by applying loss-of-function (LOF) filtering strategy, RAD50 was prioritized as the top CPG for validation in a larger cohort of 3103 individuals. The combined study consisted of 3289 Henan individuals (2118 ESCC cases and 1171 controls). We did not observe a significant difference in the frequency of LOF variants considering the entire RAD50 gene in familial ESCC patients compared to sporadic ESCC patients and controls. However, two pathogenic RAD50 LOF variants, p.Q672X and the other recurrent p.K722fs variant (4/2088, 0.19%) at the zinc hook domain, were associated with increased risk of familial ESCC compared to sporadic ESCC and controls (0/4490, 0%) (p = 0.01). An increased risk of familial ESCC was also observed, when compared to East Asians from the gnomAD database (4/19954, 0.06%) (OR 9.57, p = 4.1x10^-3) and all populations from gnomAD (5/251308, 0.028%) (OR 96, p = 5.6x10^-7). Further functional characterization suggested that the Q672X variant contributed a dominant negative effect in the DNA repair. Our study suggested RAD50 LOF variants in the zinc hook domain associate with higher risk of familial ESCC in Chinese. Screening of the two pathogenic LOF RAD50 variants may be utilized to improve cancer detection and prognosis among familial ESCC patients.

Many chemotherapeutic drugs induce oxidative stress by accelerating the accumulation of reactive oxygen species (ROS), which triggers the death of cancer cells and then causes severe DNA damage in cancer cells. Here, we proposed using a preclinical microfluidic model to evaluate the combination of doxorubicin and aspirin (DA) for anti-inflammatory therapy using patient-derived circulating tumor cell (CTC) clusters. The preclinical model could perform high-throughput screening of drug combinations and used valves to regulate media inflow for CTC cluster formation. We demonstrated that low-dose aspirin (445–500 mg/ml) and a suboptimal dose of doxorubicin (0.5 D) for seven days could produce higher killing efficacy and significantly reduced the proportion of cancer stem cells and colony-forming ability. Compared with the treatment with doxorubicin alone, the intracellular oxidative activity in the sample under combinatorial DA treatment was reduced, as demonstrated by the intensity of Calcein AM. We demonstrated that the treatment outcomes were mediated by the reduction of COX-2, which was associated with inflammation triggered by ROS. Overall, the preclinical model could be used as a proof of concept to demonstrate the efficacy of anti-inflammatory combinatorial therapies by influencing oxidative stress. Similar research could provide a basis for more DNA-related cancer treatment research in the future.

Background:

Neuroblastoma (NB) is a rare childhood cancer derived but accounts for 15% of paediatric cancer deaths. MYCN amplification and/or ATM loss through 11q deletion may cause increase in replication stress (RS) in a substantial proportion (~80%) of high-risk NB. RS creates a dependency on ATR-mediated S and G2 checkpoint control. This study aimed to determine if MYCN amplification or ATM loss identifies cell which are sensitive to ATR inhibition (ATRi). As PARP inhibition causes RS through unrepaired single strand DNA breaks progressing to replication, we also examined the effect of ATRi on PARP inhibitor (PARPi) cytotoxicity and PARPi-induced RS, cell cycle arrest and homologous recombination repair (HRR) activity.

Methods:

Cell proliferation in response to 72 hours treatment with the ATR inhibitor, VE-821, was assessed by XTT (Roche) and clonogenic survival assays in a panel of 11 NB cell lines and the effect of VE-821 on growth inhibition caused by the PARPi, olaparib, in 4 cell lines. CHK1^S345 (marker of ATR activity) and RPA2^S8 and H2AX^S129 phosphorylation (RS markers) were assessed by Western blotting and immunofluorescent microscopy. HRR was examined by the formation of RAD51 foci by immunofluorescent microscopy. Cell cycle analysis was carried out by flow cytometry.

Results:

VE-821-induced growth inhibition and cell death was significantly increased in MYCN amplified cell lines and cell lines with low ATM protein expression (p<0.05 Mann-Whitney U test). Olaparib (5 µM) treatment increased CHK1^S345 and H2AX^S129 phosphorylation after 24 hours treatment in all cell lines. ATR inhibition prevented CHK1^S345 phosphorylation and reduced olaparib-induced RAD51 foci formation. VE-821 abrogated olaparib-induced S and G2 checkpoint arrest.

Conclusion:

MYCN amplification and low ATM protein expression are determinants of ATRi sensitivity in NB cell lines. ATRi sensitises NB cells to PARPi by abrogating S/G2 checkpoint arrest and impairing HRR.