MYH knockdown in pancreatic cancer cells creates an exploitable DNA repair vulnerability

Abstract

Pancreatic ductal adenocarcinoma (PDAC) has a poor 5-year survival rate of just 13 %. Conventional therapies fail due to acquired chemoresistance. We previously identified MutY-Homolog (MYH), a protein that repairs oxidative DNA damage, as a therapeutic target that induces apoptosis in PDAC cells. However, we did not understand the mechanism driving these anti-PDAC effects, nor did we have a means to therapeutically inhibit MYH. In this study, we demonstrated that MYH inhibition induces DNA damage and checkpoint activation in PDAC cells. Using a clinically-relevant PDAC mouse model, we showed that therapeutic MYH-siRNA delivery using Star 3 nanoparticles increased intratumoural PDAC cell death, but did not inhibit tumour growth. Finally, we showed that MYH knockdown in PDAC cells sensitised them to the anti-proliferative and anti-clonogenic effects of oxaliplatin and olaparib. Our findings identify a potential novel therapeutic approach for PDAC that induces a therapeutically exploitable DNA repair vulnerability.

Keywords: Chemosensitisation; DNA repair; Oxidative stress; Pancreatic cancer.

Fig. 1

Base excision repair (BER) of 8-oxoguanine in DNA. The repair of 8-oxoguanine in DNA is primarily coordinated by the BER pathway glycosylases OGG1 and MYH. OGG1 detects and removes 8-oxoguanine in 8-oxoguanine:C mispairs (1). This forms a non-coding abasic site (2) that is cleaved by the enzyme apurinic/apyrimidinic endonuclease 1 (APE1). This cleavage generates a single-strand DNA break that is then protected from further damage by scaffolding proteins. A DNA polymerase is then recruited to restore the correct DNA base. BER is completed by a ligase that repairs the single-strand DNA break (3). However, if the 8-oxoguanine:C mispair is not repaired before DNA replication, DNA polymerases can incorrectly insert an adenine (A) base opposite 8-oxoguanine to form and 8-oxoguanine:A mispair (4). If left unrepaired, this can cause a permanent mutation. However, MYH recognises and excises this adenine (A) base from the DNA (5). MYH then directs downstream BER proteins to restore an 8-oxoguanine:C mispair (6). As described, 8-oxoguanine:C mispairs are acted upon by OGG1 (7)before BER proteins complete the DNA repair (3). Abbreviations: apurinic/apyrimidinic endonuclease 1 (APE1); base excision repair (BER); reactive oxygen species (ROS); MutY-Homolog (MYH); 8-oxoguanineuanine glycosylase (OGG1); 8-oxoguanineuanine (8-oxoguanine); adenine (A); thymine (T); guanine (G); cytosine (C).

Fig. 2

MYH Knockdown in PDAC cells induces DNA damage. A) Western blot comparing proteins involved in protection from oxidative DNA damage (MTH1: nucleotide pool sanitizer; MSH2: mismatch repair; OGG1, MYH: base excision repair glycosylase) in multiple PDAC cells. Alpha-tubulin was used as housekeeper for OGG1 and MTH1. GAPDH was housekeeper for MYH and MSH2. B-D) Western blots for phosphorylated Chk1 (P-Chk1) and Chk2 (P-Chk2) proteins in (B) MiaPaCa2 cells and (C-D) AsPC1 cells transfected with non silencing-siRNA (ns-siRNA) or MYH-siRNA. GAPDH was used as a housekeeper. Note that P-Chk1 was not detectable in MiaPaCa2 cells. Graphs show densitometry for P-Chk1 and P-Chk2 bands. E-F) Representative photos of immunofluorescence for gammaH2AX (DNA damage) and DAPI (nuclei) in (E) MiaPaCa2 cells and (F) AsPC1 cells transfected with ns-siRNA or MYH-siRNA (γH2AX foci per nucleus as a percent of ns-siRNA). Symbols in graphs indicate independent replicate experiments. Bars and lines = mean+s.e.m. Asterisks indicate significance (*p ≤ 0.05, student t-test).

Fig. 3

MYH is a potential therapeutic target in PDAC CAFs, but its expression in the tumour and stroma does not predict patient survival. A) Reference images for scoring of MYH expression (tumour and stroma scored on separate scales) in the Australian Pancreatic Cancer Genome Initiative International Cancer Genome Consortium PDAC cohort. B-D) Survival curves for patients expressing high (scores 2-3) versus low (scores 0-1) MYH expression in the (B) tumour compartment, (C) stroma compartment or (D) both. Numbers in brackets indicate the number of patients per group.

Fig. 4

Therapeutic MYH knockdown in PDAC tumours increased intratumoural cell death. All orthotopic tumours were co-injections of MiaPaCa-2 PDAC cells and human primary patient-derived CAFs. A) Model overview. B) Tumour luminescence in standardised groups at start of treatment. C) Tumour volume at therapeutic model endpoint, as assessed by calliper measurement (n = 8-9). D) Quantification of metastatic sites per mouse at model endpoint based on ex vivo luminescence (n = 8-9). E-G) Representative images and quantification of (E) MYH (inset = isotype control), (F) αSMA (CAFs, green; blue = nuclear stain; inset = isotype control; quantification of the average % aSMA positive area per region of interest [ROI], per mouse), (G) TUNEL (cell death, red). Lines in all graphs = mean ± s.e.m. Asterisks in all graphs indicate significance (ns: not significant, *p ≤ 0.05; student t-test). Panel A created with BioRender.com.

Fig. 5

MYH and MTH1 knockdown in PDAC cells increased their sensitivity to the anti-proliferative effects of tert‑butyl hydroperoxide (tBHP; oxidative stress). (A,C) Line graphs show PDAC cell proliferation 48 h-120 h post-transfection with non silencing-siRNA (ns-siRNA) or MYH-siRNA, as measured on an xCelligence platform (expressed as % of ns-siRNA control). Treatment with tBHP occurred at 48 h and 72 h post-transfection. The quantification and statistical comparison of cell proliferation at 96 h (latest time point at which knockdown was confirmed) is shown in (B,D). Bars and lines = mean+s.e.m (n = 4 independent replicates). Asterisks indicate significance relative to ns-siRNA at the same concentration of tBHP (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001; one-way ANOVA). Hashes indicate significance relative to the 0μM tBHP treatment of the same siRNA (^#p ≤ 0.05, ^##p ≤ 0.01; one-way ANOVA).

Fig. 6

MYH knockdown in PDAC cells increased sensitivity to oxaliplatin and olaparib. Clonogenic assays were performed for (A,E) MiaPaCa2 and (B,F) AsPC1 post-transfection with non-silencing siRNA (ns-siRNA) or MYH-siRNA, and co-treatment with oxaliplatin (platinum drug and oxidative stress inducer) or olaparib (PARP inhibitor). Graphs show colony count (as a % of ns-siRNA 0μM drug control). C-D, G-H) As per (A-B, E-F) except a trypan blue exclusion assay was performed at 96 h post-transfection and 48 h post-drug treatment. I) As per (A-B, E-F) except apoptosis analysis (AnnexinV/DPAI staining and flowcytometry) was performed at 96 h post-transfection and 48 h post-drug treatment. Symbols indicate independent replicate experiments. Bars and lines = mean+s.e.m (A-B, E-F), n = 3-5 independent replicates; C-D, G-In = 3 independent replicates). Asterisks indicate significance relative to ns-siRNA at the same concentration of drug or as indicated by lines (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001; one-way ANOVA). Hashes indicate significance relative to the 0μM drug treatment of the same siRNA (^#p ≤ 0.05, ^##p ≤ 0.01, ^###p ≤ 0.001, #### p ≤ 0.0001; one-way ANOVA).