Histone deacetylase inhibition is synthetically lethal with arginine deprivation in pancreatic cancers with low argininosuccinate synthetase 1 expression

Arginine (Arg) deprivation is a promising therapeutic approach for tumors with low argininosuccinate synthetase 1 (ASS1) expression. However, its efficacy as a single agent therapy needs to be improved as resistance is frequently observed. Methods: A tissue microarray was performed to assess ASS1 expression in surgical specimens of pancreatic ductal adenocarcinoma (PDAC) and its correlation with disease prognosis. An RNA-Seq analysis examined the role of ASS1 in regulating the global gene transcriptome. A high throughput screen of FDA-approved oncology drugs identified synthetic lethality between histone deacetylase (HDAC) inhibitors and Arg deprivation in PDAC cells with low ASS1 expression. We examined HDAC inhibitor panobinostat (PAN) and Arg deprivation in a panel of human PDAC cell lines, in ASS1-high and -knockdown/knockout isogenic models, in both anchorage-dependent and -independent cultures, and in multicellular complex cultures that model the PDAC tumor microenvironment. We examined the effects of combined Arg deprivation and PAN on DNA damage and the protein levels of key DNA repair enzymes. We also evaluated the efficacy of PAN and ADI-PEG20 (an Arg-degrading agent currently in Phase 2 clinical trials) in xenograft models with ASS1-low and -high PDAC tumors. Results: Low ASS1 protein level is a negative prognostic indicator in PDAC. Arg deprivation in ASS1-deficient PDAC cells upregulated asparagine synthetase (ASNS) which redirected aspartate (Asp) from being used for de novo nucleotide biosynthesis, thus causing nucleotide insufficiency and impairing cell cycle S-phase progression. Comprehensively validated, HDAC inhibitors and Arg deprivation showed synthetic lethality in ASS1-low PDAC cells. Mechanistically, combined Arg deprivation and HDAC inhibition triggered degradation of a key DNA repair enzyme C-terminal-binding protein interacting protein (CtIP), resulting in DNA damage and apoptosis. In addition, S-phase-retained ASS1-low PDAC cells (due to Arg deprivation) were also sensitized to DNA damage, thus yielding effective cell death. Compared to single agents, the combination of PAN and ADI-PEG20 showed better efficacy in suppressing ASS1-low PDAC tumor growth in mouse xenograft models. Conclusion: The combination of PAN and ADI-PEG20 is a rational translational therapeutic strategy for treating ASS1-low PDAC tumors through synergistic induction of DNA damage.

HEMA) (Sigma). Briefly, Poly-HEMA was dissolved at 20 mg/mL concentration in 95% Ethanol and added to the plate to fully cover growth area and dried overnight.

Colony formation assay.
Cells were seeded in 12-well plates at 50,000 cells/well in duplicate. After 24 hour treatment, media was replaced and cells were incubated for 7 days. Cells were washed with PBS and fixed in 4% PFA. Cells were then stained with 0.1% Crystal Violet (Sigma) for 30 minutes and washed with water.
In vitro tumor cell-fibroblast 3D co-culture. Reader. Images were taken using a CX41 Inverted Microscope with a DP26 Digital Camera (Olympus).

Generation of ASS1 knockout isogenic cells.
Four sgRNA sites were designed for the exon4 of the ASS1 gene. The recombinant plasmids of Lentiviral vector2-ASS1-sgRNA were constructed. The constructed vector were transfected into SU8686 cells. then 24 hr later cells were treated with puromycin for 3 days. The two gRNAs (#1: TATGTGTCCCACGGCGCCAC and #3: ATACTTGGCCCCCTCCCGCT) with best knockout effect of ASS1 gene in SU8686 cells were selected. Single cell cloning was performed with limited dilution of cells into 96-well plates. At least 2 single clones of each gRNA were selected for knockout confirmation with Western blot.

Western blotting.
Cells were lysed in cold RIPA lysis buffer with protease and phosphatase inhibitors (Thermo Scientific). Protein extracts were resolved on SDS-PAGE and then electrotransferred to Immun-Blot Nitrocellulose membrane (Bio-Rad, Hercules, CA, USA). After blocking in 5% milk, membranes were incubated in primary antibody solution at 4 C overnight, and then with horseradish peroxidase (HRP)-conjugated secondary antibody solution at room temperature for 1h. Blots were developed using Pierce ECL Substrate (Thermo Scientific, Rockford, IL) and imaged on the LI-COR Odyssey imaging system.

RNA-Seq.
Samples were prepared in triplicate. Cells were trypsinized and collected on ice after 24 h treatment for RNA extraction and analysis. Libraries for RNA-Seq were prepared with KAPA Stranded mRNA Kit. The workflow consisted of mRNA capture, cDNA generation, and end repair to generate blunt ends, A-tailing, adaptor ligation and PCR amplification. The data was sequenced on Illumina HiSeq 3000 for a single-read 50 bp run. Data quality check was performed on Illumina SAV. The reads were mapped to the latest UCSC transcript set using Bowtie2 version 2.1.0 (1) and the gene expression level was estimated using RSEM v1.2.15 (2).
Normalized gene expression data were fed into Ingenuity Pathway Analysis (3).

Sample preparation for LC-MS/MS-MRM analysis of nucleotide pools and incorporation
of labeled nucleotides into newly replicated DNA and RNA.

Quantification.
The areas for nucleotide measurements were obtained from extracted ion chromatograms of MRM ion transitions. These measurements were normalized from the spiked internal standards ([ 15 N3]dCMP and [ 15 N3]dCTP). Nucleotide data were normalized to cell number and displayed as relative amount per cell compared to untreated. For DNA, the areas for the hydrolyzed labeled nucleosides were obtained from extracted ion chromatograms of MRM ion transitions, and normalized to total ion current at that retention time.