# paper.pdf ## Metadata - PDFFormatVersion=1.7 - Language=en-US - IsLinearized=true - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - Author=Iuliia Pavlyk, George Field, Matthew Young, Josephine Carpentier, Emilia A. Szlosarek, Michaela R. O'Keeffe-Brown, Timothy Crook, Nelofer Syed, John S. Bomalaski, Pui Ying Chan, Peter W. Szlosarek, Iuliia Pavlyk, George Field, Matthew Young, Josephine Carpentier, Emilia A. Szlosarek, Michaela R. O'Keeffe-Brown, Timothy Crook, Nelofer Syed, John S. Bomalaski, Pui Ying Chan, Peter W. Szlosarek - CreationDate=D:20260619184456+05'30' - Creator=Adobe InDesign 19.5 (Windows) - ModDate=D:20260619184523+05'30' - Producer=Adobe PDF Library 17.0 - Title=Pegargiminase Suppresses the Fanconi Anemia Pathway and Promotes Melphalan‐Induced DNA Double‐Strand Breaks in Uveal Melanoma - Trapped=False - Custom.WPS-ARTICLEDOI=10.1111/pcmr.70104 - Custom.WPS-JOURNALDOI=10.1111/(ISSN)1755-148X - Custom.WPS-PROCLEVEL=3 - jav:journal_article_version=VoR - prism:doi=10.1111/pcmr.70104 - pdfx:wps-articledoi=10.1111/pcmr.70104 - pdfx:wps-journaldoi=10.1111/(ISSN)1755-148X - pdfx:wps-proclevel=3 - dc:title=Pegargiminase Suppresses the Fanconi Anemia Pathway and Promotes Melphalan‐Induced DNA Double‐Strand Breaks in Uveal Melanoma ## Contents ### Page 1 ### Page 2 Pigment Cell & Melanoma Research ORIGINAL ARTICLE[ ![Open Access logo]() ](https://doi.org/10.1111/pcmr.70104) # Pegargiminase Suppresses the Fanconi Anemia Pathwayand Promotes Melphalan-Induced DNA Double-StrandBreaks in Uveal Melanoma Iuliia Pavlyk1 | George Field1 | Matthew Young1[ ![ORCID logo]() ](https://orcid.org/0000-0002-9775-1956) | Josephine Carpentier1 | Emilia A. Szlosarek1,2 |Michaela R. O'Keeffe-Brown1 | Timothy Crook3 | Nelofer Syed4 | John S. Bomalaski5 | Pui Ying Chan6[ ![ORCID logo]() ](https://orcid.org/0000-0001-7804-4482) |Peter W. Szlosarek1,3,7[ ![ORCID logo]() ](https://orcid.org/0000-0002-5039-711X) 1Center for Cancer Biomarkers and Biotherapeutics, Barts Cancer Institute (BCI)—A Cancer Research UK Center of Excellence, John Vane Science Center,Queen Mary University of London, London, UK |2University of East Anglia, Norwich, UK |3The London Clinic, London, UK |4Department of BrainSciences, Hammersmith Hospital, Imperial College London, London, UK |5Polaris Pharmaceuticals Inc., San Diego, California, USA |6Wellcome SangerInstitute, Hinxton, UK |7Department of Medical Oncology, Barts Health NHS Trust, St Bartholomew's Hospital, London, UK Correspondence: Peter W. Szlosarek () Received: 6 April 2025 | Revised: 9 May 2026 | Accepted: 26 May 2026 Keywords: ADI-PEG20 | arginine dependency | DSBs | Fanconi anemia | melphalan | pegargiminase | uveal melanoma ABSTRACT Uveal melanoma is a hard-to-treat arginine-dependent cancer secondary to argininosuccinate synthetase 1 (ASS1) loss with halfof patients succumbing to liver-dominant metastases. Arginine deprivation with pegargiminase is a novel antimetabolite strategyfor patients with uveal melanoma. We investigated the preclinical rationale for combining pegargiminase with melphalan, analkylating agent approved recently for the treatment of hepatic-centric disease. Drug sensitivity of ASS1-deficient uveal mela-noma cell lines was performed in 2D culture using proliferation and cytotoxicity assays, with analysis of cell death, cell cycle,DNA double-strand breaks, and interrogation of the molecular mechanism of action by RNA-seq. ADI-PEG20 and melphalansuppressed uveal melanoma cell line proliferation and triggered cytotoxicity, effects which were enhanced with the drug combi-nation. ADI-PEG20 downregulated multiple genes of the Fanconi anemia pathway and synergized with melphalan to increaseDNA double-strand breaks. Melphalan and pegargiminase is a rational new drug combination that warrants clinical testing inuveal melanoma. ## 1 | Introduction Uveal melanoma (UM) is the commonest ocular malignancyin adulthood known for poor survival outcomes secondary tomultifocal liver metastases and liver failure developing in ap-proximately 50% of patients (Jager et al.). Treatments arelimited and include immune checkpoint blockade (ICB) andtebentafusp, with clinical benefit observed in a subset of pa-tients (Hassel et al.; Pelster et al.). The median over-all survival is 10–22 months, with 5-year survivals remaininglow at around 10%–20% compared with over 50% in cutane-ous metastatic melanoma with immunotherapy over the lastdecade (Dimitriou et al.; Khoja et al.; Lane et al.;Wolchok et al.). Local therapies, including liver metastasectomy and radio-frequency ablation provide benefits to selected patients witholigometastatic disease (Hand et al.). More recently, che-mosaturation with percutaneous hepatic perfusion of mel-phalan has been approved by the FDA for metastatic UM, whichis also being studied in the context of immunotherapy (OlofssonBagge et al.). Melphalan, as a classic alkylating agent, in-duces DNA damage, thereby potentially enhancing antigenicityfor ICB therapy (Patel and Minn). ![WILEY logo]() Pigment Cell & Melanoma Research, 2026; 39:e70104 This is an open access article under the terms of the License, which permits use, distribution and reproduction in any medium,provided the original work is properly cited. © 2026 The Author(s). Pigment Cell & Melanoma Research published by John Wiley & Sons Ltd. ### Page 3 Previously, we and others reported an encouraging 60%–70%disease control rate in patients with UM using the arginine-depleting enzyme, pegylated arginine deiminase (ADI-PEG20,pegargiminase) (Chan et al.; Ott et al.). ADI-PEG20exploits the known arginine auxotrophy of argininosucci-nate synthetase 1 (ASS1) cancers of which UM is an exemplar(Delage et al.). We have shown that ADI-PEG20 potenti-ates the effect of antifolate chemotherapy and increases over-all survival in patients with ASS1- deficient mesothelioma,but not in UM (Allen et al.; Chan et al.; Szlosareket al.). Thus, alternative combinations of ADI-PEG20 inUM are needed. Here, we explored the preclinical rationale forcombining ADI-PEG20 with melphalan using a human UMcell line panel. Our data highlight a novel mechanism of ac-tion involving double- strand breaks (DSBs) in DNA and sup-port further clinical development of ADI-PEG20 in the contextof DNA- damaging agents. ## 2 | Results ### 2.1 | Increased Antiproliferative Activityof ADI-PEG20 Plus Melphalan in UM Cell Lines First, we screened a panel of four UM cell lines identifyingASS1 deficiency in 92.1 and MP41, while MP38 and MP46were ASS1-proficient (Figure). We tested the proliferationof the ASS1- deficient cell lines in 2D culture by analyzing cellconfluency using an IncuCyte imaging platform. As expected,ADI-PEG20 significantly inhibited the growth of the 92.1 andMP41, while melphalan also displayed growth- suppressive ac-tivity. Notably, the combination of ADI-PEG20 and melphalanwas at least additive in both the 92.1 and MP41 (Figure)cell lines. ### 2.2 | Increased Cytotoxicity and Apoptosis of 92.1UM Cells With ADI-PEG20 and Melphalan Next, we determined the IC50 of ADI-PEG20 and melphalanin the ASS1-deficient UM cell lines (Figure) and analyzedcell viability confirming increased cytotoxicity of the combi-nation by 48 h in the 92.1 cell line but not the MP41 cell line(Figure; Figure). Cell death was caspase-dependent andpoly(ADP-Ribose) polymerase (PARP)-dependent in the 92.1cell line (Figure) and was validated by Annexin V stainingas a measure of increased apoptosis in the ADI-PEG20 and mel-phalan combination compared to either drug alone (Figure).There was no increased cleaved PARP or caspase-3 with ADI-PEG20 plus melphalan in MP41 cells indicating differential drugeffects despite common ASS1 deficiency in the two cell lines(Figure). ### 2.3 | ADI-PEG20 Promotes DSBs by MelphalanWith S and G2/M Cell Cycle Arrest in 92.1 Cells Since melphalan is a well-known DNA- damaging alkylatingagent, we addressed the further impact of pegargiminase byanalyzing the extent of DSBs in DNA. Our analysis revealedthat indeed melphalan induced a significant increase in DSBs,whereas this was minimal with pegargiminase, but that therewas a fivefold synergistic increase in DSBs compared to controlin UM cells treated with the combination of ADI-PEG20 andmelphalan (Figure). Moreover, there was a correspond-ing increase of DSBs detected in the cytoplasm of pegargimi-nase and melphalan treated UM cells. Cell cycle analysis wasbased on nuclear surface and volume and DAPI intensity, andindicated cell cycle arrest at the G2/S phase with melphalanalone and the drug combination, whereas ADI-PEG20 trig-gered arrest at G0/G1 (Figure). Moreover, cells in G2/Mand S phases were characterized by higher dSTRIDE countswhen compared to G0/G1 phases (Figure). Similarly, therewas a pronounced increase in the dSTRIDE counts in the cy-toplasm of UM cells in G2/M and S phases treated with ADI-PEG20 and melphalan compared with control, melphalan, orADI-PEG20 (Figure). ### 2.4 | Major Downregulation of the FanconiAnemia Pathway by ADI-PEG20 in 92.1 Cells To understand the mechanism underlying the increased DSBswith ADI-PEG20 and melphalan, we employed RNA sequenc-ing (RNA- seq). Bioinformatic analysis by WIKI revealedmodulation of the DNA damage response pathway and byReactome modulation of DNA repair, transcriptional regula-tion by p53 and cell cycle checkpoint pathways. Several KEGGdatabase pathways were also altered in the drug combinationincluding p53 signaling and the cell cycle; however, the mostrepresented and statistically significant was widespread sup-pression of the Fanconi Anemia (FA) pathway (Figure).Furthermore, ADI-PEG20 and not melphalan was the keydrug downregulating the FA genes: FANCD2, UBE2T, EME1,BRIP, FANCA, FANCE, FANCI, and FANCB (Figure).A protein–protein interaction of the FA pathway was derivedin response to the drug combination compared to the controltreatment (Figure). We validated the FA pathway as the top pathway regulatedby ADI-PEG20 in the 92.1 cell line by qPCR (Figure) andwestern blotting (Figure). Notably, we identified thatADI-PEG20 was critical in reducing the expression of multi-ple FA proteins—with a modest inhibitory effect of melphalanon FANCD2 and BRIP1 levels—and that the combination ofADI-PEG20 and melphalan had a further suppressive andcytotoxic effect in 92.1 but not in MP41 cells. Additionally, Significance • Uveal melanoma is an arginine-dependent cancer sec-ondary to ASS1 loss and is sensitive to arginine depri-vation with pegargiminase (ADI-PEG20). • Here, we show that pegargiminase suppresses mul-tiple Fanconi anemia pathway genes and synergizesto increase DNA double-strand breaks in uveal mela-noma cells, with enhanced cytotoxicity and PARP andcaspase-dependent apoptosis. • Clinical trials of pegargiminase plus melphalan arewarranted in uveal melanoma. ### Page 4 ASS1 overexpression in 92.1 cells abrogated the ADI-PEG20-mediated suppression of the FA pathway, confirming that thedownregulation of the FA pathway in response to argininedeprivation is ASS1- dependent. ## 3 | Discussion Chemosaturation with melphalan is an FDA approved therapyfor hepatic-dominant UM, a disease that remains an area of highunmet need with a minority of patients surviving 5 years. In thisstudy, we have shown that arginine deprivation with pegargimi-nase downregulates multiple components of the FA pathwayand potentiates the DNA-damaging effects of melphalan asevidenced by a marked increase in DNA DSBs. UM cells char-acterized by ASS1 loss were highly sensitive to the combinationof ADI-PEG20 and melphalan triggering arrest in G2/M and Sphase and apoptosis. Previous studies in hematological cancers have shown thatthe FA pathway is a key driver of melphalan resistance (Yardeet al.). In solid cancers, such as pancreatic cancer, defi-ciency of several FANC genes increased tumor cell susceptibil-ity specifically to interstrand crosslinkers, including melphalan,cisplatin, and mitomycin- C (van der Heijden et al.). Incontrast, cancer cells with FA-deficient genes were insensitiveto a range of non-alkylating chemotherapeutic agents, includ-ing gemcitabine, 5-fluorouracil, vinorelbine, and paclitaxel. The ![image]() FIGURE 1 | ADI-PEG20 and melphalan inhibit proliferation of ASS1- deficient UM cell lines. (a) Mutation status. (b) ASS1 western blotting. (c)Cell proliferation in 92.1 cell line and MP41 cell lines with melphalan (0.77 μM) and ADI-PEG20 (500 ng/mL) alone or in combination and comparedto untreated cells (representative of three experiments performed in triplicate). (d) Typical cell confluency images of 92.1 and MP41 respectively att = 0, 36, and 66 h for assessment of proliferation. Scale bar 600 μm. ### Page 5 FANCI gene downregulated by ADI-PEG20 in our study is crit-ical for DNA repair via FANCD2 heterodimerization, and theirinhibition leads to DNA damage and chemosensitization (Liet al.; Smogorzewska et al.). More broadly, the FA genes uncovered using RNA-Seq arepart of the homologous recombination (HR) repair pathway,with mutations directly impacting HR repair mechanismsthat lead to genomic instability, cancer predisposition and in-creased chemotherapy-associated toxicities (Niraj et al.).Additionally, germline and somatic FA gene mutations in can-cer have been linked to improved chemotherapy outcomesin patients with several solid cancers exposed to alkylatingagents, including breast, pancreatic, bladder, and liver cancer(Lai et al.; O'Reilly et al.; Plimack et al.; Rodleret al.). A recent study of germline variants in patients withmelanoma reported on uncommon alterations in BRIP1 (1/83)and FANCE (1/83) with a pathogenic role in UM tumorigenesis(Johansson et al.). Defects in the FA pathway genes mayexplain sporadic responses reported to alkylating drugs in UMakin to responses described for gemcitabine in MBD4-mutatedUM (Chabot et al.; Schinzari et al.). Studies of arginine deprivation over the last two decades have re-vealed a cohort of cancers with substantial ASS1 deficiency andsensitivity to the arginine-catabolizing enzymes ADI-PEG20 andarginase (Field et al.). Previously, we showed that UM displaysnear universal loss of ASS1 with evidence of promoter methylation ![image]() FIGURE 2 | ADI-PEG20 and melphalan increase 92.1 cell line cytotoxicity and apoptosis. (a) IC50 for ADI-PEG20 (48 h; n = 3). (b) IC50 for mel-phalan (48 h; n = 3). (c) Cell viability for control, ADI-PEG20 (500 ng/mL), melphalan (5 μM), and the drug combination (n = 9) with **p < 0.001,***p = 0.003, and ****p < 0.0001. (d) Immunoblotting for PARP and caspase 3 following treatment with control, ADI-PEG20 (500 ng/mL), melphalan(5 μM), and the drug combination. (e) Analysis of Annexin V positive cells by flow cytometry upon treatment with control, ADI-PEG20 (500 ng/mL),melphalan (5 μM), and the drug combination (n = 6) with **p < 0.001 and ****p < 0.0001. ### Page 6 ![image]() FIGURE 3 | ADI-PEG20 enhances melphalan-induced DNA DSBs in 92.1 cells. (a) DNA DSBs revealed as yellow fluorescent foci by confocalmicroscopy (scale bar 20 μm). (b) DSBs in the nucleus. (c) DSBs in the cytoplasm. (d) Percentage of cells in a given cell cycle phase. *p < 0.05 and****p < 0.0001. ### Page 7 ![image]() FIGURE 4 | ADI-PEG20 suppresses genes of the FA pathway in 92.1 cells. Volcano plots of the differentially expressed genes of 92.1 cell line treat-ed with (a) the combination of ADI-PEG20 and melphalan compared to control, and (b) the combination compared to melphalan only; vertical linesillustrate differentially expressed genes with a log2 fold change of ±1; and horizontal lines describe the genes that where significantly differentiallyexpressed at a false discovery rate of q < 0.05 by Benjamini–Hochberg method. (c) Pathway overrepresentation analyses of the differentially expressedgenes identified in the comparison of combination treatment against melphalan treatment (volcano plot b); WebGestalt used an FDR of q < 0.05within the Benjamini–Hochberg method to determine significantly overrepresented pathways in the KEGG, REACTOME, and WIKI databases.(d) Heatmap highlighting the differential expression of genes associated with the KEGG pathway annotation FA pathway in each of the treatmentgroups relative to untreated. (e) String database protein–protein interactions of the FA pathway hits from the combination against untreated differen-tial expression analysis in (a). (f) qPCR analysis of FA pathway genes in 92.1 and MP41 cells treated with ADI-PEG20 (500 ng/mL over 48 h). Relativeexpression levels normalized to housekeeping gene, PPIA. Bars represent the mean ± SEM (n = 3); *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001,one-way ANOVA with Dunnett's Multiple comparison test. (g) Representative western blot analysis of 92.1 and MP41 cells showing protein levels ofFANCI, FANCD2, BRIP1, EME1, and UBE2T treated with ADI-PEG20 (500 ng/mL), Melphalan (5 μM), and the combination (n = 3); HSP70 is shownas a loading control. (h) Representative western blot analysis of ASS1- overexpressing 92.1 cells showing protein levels of FANCI, FANCD2, BRIP1,EME1, and UBE2T treated with ADI-PEG20 (500 ng/mL), Melphalan (5 μM), and the combination (n = 3); HSP70 is shown as a loading control. ASS1protein expression after lentiviral transduction in 92.1 is shown on the right-hand side; HSP70 is shown as a loading control. ### Page 8 ![Figure 4]() ### Page 9 in metastatic disease (Chan et al.). Here, the ASS1-negativeMP41 and 92.1 cell lines displayed reduced cell proliferation andviability to ADI-PEG20; however, only the latter exhibited signif-icant apoptosis upon drug combination therapy that was caspaseand PARP-dependent. Cell type is a key determinant of ADI-PEG20sensitivity, with variable effects noted across and between cell linestudies (Lin et al.; Zheng et al.). Moreover, several ADI-PEG20 resistance mechanisms have been identified, ranging fromupregulation of ASS1—reported in patients with UM progressingon pegargiminase—and the antiapoptotic protein Bcl-xL to activa-tion of autophagy and the supply of the salvage precursor argini-nosuccinate via tumor-associated macrophages (Chan et al.;Kim et al.; Panda et al.; Phillips et al.). ADI-PEG20 plus melphalan adds to the number of rationalcombinations that exploit synergies as well as nonredundantanticancer drug mechanisms incorporating arginine depriva-tion. To date, we have succeeded in translating pegargiminasewith pemetrexed plus platinum for the treatment of chemo-refractory mesothelioma. This exploits the inhibitory effectof ADI-PEG20 on key enzymes involved in thymidine syn-thesis thereby enhancing pemetrexed cytotoxicity in meso-thelioma—the first arginine-lowering chemotherapy to seekFDA approval (Szlosarek et al.). A global phase 3 studyis underway of another antimetabolite-based regimen, namelyADI-PEG20 with gemcitabine and docetaxel in the therapy ofASS1- deficient soft tissue sarcoma. This chemotherapy reg-imen capitalizes on the taxane and ADI-PEG20 stabilizinghENT1 surface expression resulting in enhanced gemcitabineuptake and tumor cytotoxicity (Prudner et al.). The al-kylating agent temozolomide also enhances the effects ofADI-PEG20 with the combination now in frontline clinicaltrials for patients receiving surgery and radiation for glioblas-toma multiforme (Przystal et al.). Therefore, the appli-cation of chemosaturation therapy with melphalan, whichdelivers a 10–100-fold higher peak perfusate drug concentra-tion than systemic dosing and has the single highest drug ac-tivity in uveal melanoma of up to 60%, merits further studyin the context of arginine depletion using pegargiminase andalso immunotherapies (Artzner et al.; Chang et al.;Kraehenbuehl et al.; Minor et al.). Since our drug combination data were confined to only one oftwo ASS1- deficient UM cell lines, namely 92.1, we expandedour ASS1 negative UM cell line panel to include Mel270 andconfirmed decreased cell proliferation and viability, andincreased DNA DSBs with ADI-PEG20 and melphalan co-treatment compared with monotherapy (Figure). However,unlike 92.1 cells which demonstrated caspase 3 and PARPcleavage with the drug combination, this was not observed inMel270 cells indicating likely caspase-independent cell death.The latter mechanism of cell death has been reported witharginine- degrading enzymes in various solid cancer cell lines(Field et al.). Further functional validation of our datawith ASS1 knockdown/knockout in the MP38 and MP46 celllines was not possible due to extremely poor growth in vitroconsistent with the known tumor suppressive role of ASS1.Additionally, further study of the ADI-PEG20- depleted FApathway is needed to address which gene combination is crit-ical to the increased melphalan sensitivity. Lastly, althoughthere was no enhanced cytotoxicity nor apoptosis from addingADI-PEG20 to melphalan in the MP41 UM cell line, alterna-tive druggable pathways with arginine deprivation deservescrutiny. In particular, the known activating mutations ofGNAQ or GNA11 sensitize UM to the PKC inhibitor darova-sertib currently in clinical trials (Visser et al.). Bypassmechanisms of resistance for PKC in UM include PI3K whichis also a driver of ASS1 induction in melanoma via c-Myc, thusproviding a rationale for studying ADI-PEG20 with kinase in-hibition, which is an expanding area in UM therapeutics (Parket al.; Tsai et al.; Baqai et al.). Collectively, our data provide a novel and readily translatabledrug combination therapy for liver-dominant metastases fromuveal melanoma that exploits inhibition of DNA repair by ADI-PEG20 thereby potentiating melphalan cytotoxicity. A clinicaltrial of pegargiminase plus hepatic chemosaturation with mel-phalan is planned (ATOMIC-UM). ## 4 | Materials and Methods ### 4.1 | Cell Culture and Drug Treatments UM cell lines were purchased from ATCC and maintained inendotoxin-free RPMI 1640 medium (Gibco, #61870010) for 92.1and MP4, and DMEM (Gibco, #10566016) for MP46 and MP38,supplemented with 10% (v/v) heat-inactivated fetal bovineserum (Gibco, #16140071) and 1% (v/v) penicillin–streptomycin.Cells were cultured at 37°C in a humidified incubator with 5%CO2. All cell lines were checked regularly for Mycoplasma con-tamination. ADI-PEG20 was from Polaris Pharmaceuticals Inc.(San Diego, USA). Melphalan was purchased from CambridgeBioscience (#16665). All drug treatments were performed 24 hfollowing cell plating, when medium was replaced with freshmedium alone or with ADI-PEG20 (500 ng/mL), melphalan(5 μM unless specified otherwise), or the drug combination for48 h. Relevant controls and validation of ADI-PEG20 have beenperformed previously within our group (Phillips et al.;Szlosarek et al.). ### 4.2 | 2D IncuCyte Live Cell Proliferation Assay Cells were trypsinized and seeded at 2 × 103 cells/well in a 96-well plate and grown overnight. Twenty-four hours later, wellswere replaced with fresh medium alone or with ADI-PEG20(500 ng/mL), melphalan (0.77 μM), or the drug combination. Theplate was then placed into the IncuCyte S3TM live cell imagingsystem (Essen Bioscience, MI, USA), and images of well con-fluency were recorded every 2 h, up to a duration of 96 h. Wellconfluency was calculated automatically using the IncuCyteS3TM and then normalized against t = 0 h to obtain the prolif-eration rate. ### 4.3 | Cell Viability Assay (CTG) Cells were trypsinized and seeded at 2 × 103 cells/well in a 96-well plate and grown overnight prior to drug treatment. Cell via-bility was measured using the CellTiter- Glo Cell Viability Assay(Promega, #G7571) at 48 h, according to the manufacturer's ### Page 10 instructions. Percent cell viability was normalized to controluntreated cells. ### 4.4 | ASS1 Overexpression ASS1 overexpressing 92.1 cells were generated using lentiviraltransduction. Briefly, lentiviral particles were produced by tran-sient co- transfection of HeLa cells with ASS1 transcript variant1 cloned into the pLenti-P2A- Puro Lentiviral Gene ExpressionVector driven by the EF1α promoter (OriGene, #PS100142),together with the Lenti-vpak packaging system (OriGene,#TR30037). Transfection was performed using TurboFectin8.0 (OriGene) in a 6-well plate format according to the man-ufacturer's protocol. Viral supernatants were harvested 48 hpost-transfection and passed through a 0.45 μm syringe filter toeliminate cellular debris. For lentiviral transduction, 92.1 cellswere seeded in 10 cm dishes and incubated with the filteredlentiviral supernatant supplemented with 8 μg/mL polybrene(Sigma, #TR-1003) for 18 h. Following transduction, the mediumwas replaced with fresh complete medium, and cells were sub-sequently expanded in 10 cm dishes and subjected to treatmentwith ADI-PEG20, melphalan, or their combination as indicated.ASS1 overexpression was verified by western blot analysis usingan anti-ASS1 antibody (Cell Signaling Technology, #70720; dilu-tion 1:1000). ### 4.5 | Reverse Transcription Quantitative PCR(qPCR) 3 × 105 cells were plated in 6-well plates in 2 mL RPMI (10%FBS, 1% penicillin–streptomycin). Cells were treated withADI-PEG20 (500 ng/mL) or 1× PBS for 48 h. The QiagenRNAeasy mini kit was used to extract RNA from 6-wellplates according to manufacturer protocol. After the RNAextraction, total RNA concentration and purity were deter-mined using Nanodrop (ThermoFisher). Purity was deter-mined by measuring the absorbance at 230, 260, and 280 nm.Only RNA samples with a 260/230 ratio and a 260/280 ratioabove 1.5 were used. To generate cDNA, the High- CapacitycDNA Reverse Transcription Kit (ThermoFisher, #4368814)was used as per manufacturer instructions. qPCR was per-formed using TaqMan probes using the QuantStudio Real-Time PCR Systems (Applied Biosystems). All qPCR reactionswere carried out in duplicate using MicroAmp Optical 96-Well Reaction Plates (Applied Biosystems, Life Technologies)sealed with MicroAmp Optical Adhesive Films (AppliedBiosystems, Life Technologies). Each well of the 96-well re-action plate contained 0.5 μL of appropriate cDNA, 0.5 μL ofappropriate TaqMan Probe, 4 μL of ddH2O, 5 μL TaqMan 2×Master Mix. List of RT- qPCR probes are as follows: BRIP1(ThermoFisher, #Hs00908143_m1), FANCA (ThermoFisher,#Hs01116668_m1), FANCB (ThermoFisher, #Hs00537483_m1), FANCD2 (ThermoFisher, #Hs00276992_m1), FANCE(ThermoFisher, #Hs01562763_m1), FANCI (ThermoFisher,#Hs00383049_m1), EME1 (ThermoFisher, #Hs01103357_g1), UBE2S (ThermoFisher, #Hs02797093_g1), UBE2T(ThermoFisher, #Hs00928040_m1), PPIA (ThermoFisher,#Hs04194521_s1). The ∆∆CT method was used for the analy-sis within GraphPad Prism. ### 4.6 | Western Blot Analysis Plates were seeded at a density of 2.2 × 106 cells. Cells were lysedat 48 h following drug treatment in RIPA lysis buffer (50 mM TrisHCL [pH 8], 150 mM NaCl, 1% NP40, 1× protease inhibitor, de-ionized water) for 15 min. Lysates were centrifuged for 15 min at12,000 rpm, and the supernatant was collected. Protein concen-tration was determined using the BSA protein assay (Bio Rad,#500-0116). Samples were reduced with 10× reducing agent(ThermoFisher, #NP0009) and 4× LDS buffer (ThermoFisher,#NP0007). Protein samples were denatured on a heating blockfor 5 min at 95°C, separated by SDS-PAGE, and transferred to anitrocellulose membrane. Primary antibodies were used as fol-lows: ASS1 (1:1000), apoptosis western blot cocktail (pro/p17-caspase-3, cleaved PARP1, muscle actin) (Abcam #136812),FANCD2 (1:3000, ThermoFisher, #MA1-16570), FANCI (1:3000,ThermoFisher, #20789–1-AP), EME1 (1:1000, ThermoFisher,#PA5-79202), UBE2T (1:1000, ThermoFisher, #10105–2-AP),BRIP1 (1:1000, ThermoFisher, #PA5-27096) in 5% milk for 1 h atroom temperature or overnight. After washing (3 times for 10 minin TBS-T), the membranes were incubated in either anti-rabbitor anti-mouse Horseradish Peroxidase (HRP) conjugated sec-ondary antibody (Dako anti-rabbit #P0448 or anti-mouse, Dako,#P0447) diluted at a concentration of 1:5000 in 5% milk for 1 hat room temperature. After incubation with the secondary anti-body, the membranes were washed again three times for 10 min inTBS-T. Immunoreactivity was detected using chemiluminescencesubstrate (SuperSignal West Pico Chemiluminescent substrate,Pierce, #32209) on Amersham Imager 600 Chemi-doc. Expressionlevels of β-Actin or HSP70 were used as protein loading controls. ### 4.7 | Annexin Staining Cells were trypsinized and seeded at 2.2 × 106 cells/well in 10 cmplates and grown overnight prior to drug treatment. Forty-eighthours later, cells were trypsinized, washed with PBS, and re-suspended in 100 μL Annexin V binding buffer (Invitrogen,#V13241). Cells were stained with 5 μL of Alexa Fluor 488Annexin V and 1 μL 100 μg/mL PI (Invitrogen, #V13241).Staining was analyzed by flow cytometry on a LSR Fortessa1flow cytometer (BD Biosciences). Data were analyzed by FlowJosoftware (BD Biosciences). ### 4.8 | dSTRIDE Analysis Cells were seeded on cover glasses in 12-well plates at a den-sity 1 × 105 cells/well and incubated overnight followed bydrug treatment for 48 h. Next, all samples were fixed withice- cold 70% EtOH (Fisher Scientific, BP2818100) and storedat −20°C in preparation for dSTRIDE analysis by intoDNA(Krakow, Poland) (Kordon et al.). The dSTRIDE pro-cedure was performed followed by DAPI (ThermoFisherScientific, #62248) staining. The coverslips with fixed cellswere then mounted with Vectashield (Vector Laboratories,#H-1000-10) and stored at 4°C until imaging. Fields of view forimaging (n = 10 per sample) were chosen randomly over thecoverslip surface. The images were then collected as three-dimensional (3D) confocal stacks using Cell Discoverer 7 LSM900 microscope, Zeiss. ### Page 11 ### 4.9 | RNA Sequencing (RNA-Seq) SampleProcessing Cells were trypsinized and seeded at 0.3 × 106 cells/well using6-well plates and grown overnight prior to drug treatment.Following 48 h of drug incubation, RNA was isolated usingRNeasy Kit for RNA purification (Qiagen, #74104) and DNAasetreated (Qiagen, #79254). Samples were sent to Imperial Collegefor transcriptome library preparation and RNA-seq. ### 4.10 | Bioinformatics Salmon alignment tool (v1.9.0) and GENCODE release 45 wereused to quantify the transcript abundance estimates usingthe computing systems at Queen Mary University of London.Tximport was used to convert the abundance estimates intogene expression data with R in R-Studio. edgeR (v4.0.16) wasused for differential expression analysis. All significantly dif-ferentially expressed genes (FDR q < 0.05) that had a log2 foldchange ≤ −1 or ≥ 1 were used of overrepresentation analysesusing WEB-based GEne SeT AnaLysis Toolkit (WEBGESTALT)for the Gene Ontology terms; Molecular Function, BiologicalProcess, and Cellular Component. The pathway databasesKEGG, PANTHER, and REACTOME were also investigatedusing this same method. ### 4.11 | Statistical Analysis All statistical analysis (mean ± SD) was undertaken in GraphPadPrism (version 10). Three independent experiments were per-formed unless indicated. Experiments with greater than twogroups were analyzed by two-way ANOVA unless specifiedotherwise. dSTRIDE data was analyzed using nonparametricKruskal–Wallis and Dunn's Test. A p value of < 0.05 was consid-ered statistically significant. ## Author Contributions Iuliia Pavlyk: investigation, methodology, validation, writing – reviewand editing, formal analysis, supervision, data curation. George Field:investigation, writing – original draft, methodology, validation, writing –review and editing, software, formal analysis, data curation. MatthewYoung: investigation, validation, writing – review and editing, formalanalysis, data curation. Josephine Carpentier: investigation, method-ology, validation, writing – review and editing, formal analysis, data cura-tion, supervision. Emilia A. Szlosarek: investigation, validation, writing– review and editing. Michaela R. O'Keeffe-Brown: investigation, vali-dation, writing – review and editing. Timothy Crook: conceptualization,writing – review and editing. Nelofer Syed: resources, writing – reviewand editing. John S. Bomalaski: funding acquisition, writing – reviewand editing. Pui Ying Chan: resources, writing – review and editing.Peter W. Szlosarek: conceptualization, writing – original draft, writing– review and editing, validation, formal analysis, supervision, resources,data curation, project administration, visualization. ## Acknowledgments The authors are grateful to Dr. Roberto Bellelli of the Barts CancerInstitute for helpful manuscript discussions. This work was supportedby a grant from Polaris Pharmaceuticals Inc. (MIMR1A3S). The authorsthank the Barts Cancer Institute Flow Cytometry Core Facility and theImperial BRC Genomics facility (London, UK) for integrated supportfor translational next-generation sequencing projects. The authorsalso thank Dr. Anna Uherek and Dr. Olga Wójcikowska at intoDNA(Krakow, Poland) for the characterization of DNA DSBs and cell cycleusing the STRIDE platform. ## Funding This work was supported by a grant from Polaris Pharmaceuticals Inc.(MIMR1A3S). ## Conflicts of Interest Peter W. Szlosarek is a recipient of research funding from PolarisPharmaceuticals Inc., and John S. Bomalaski is an employee of PolarisPharmaceuticals Inc. The other authors declare no conflicts of interest. ## Data Availability Statement The data that support the findings of this study are available on requestfrom the corresponding author. The data are not publicly available dueto privacy or ethical restrictions. ## References Allen, M. D., P. Luong, C. Hudson, et al. 2014. “Prognostic andTherapeutic Impact of Argininosuccinate Synthetase 1 Control inBladder Cancer as Monitored Longitudinally by PET Imaging.”Cancer Research 74, no. 3: 896–907.. Artzner, C., O. Mossakowski, G. Hefferman, et al. 2019.“Chemosaturation With Percutaneous Hepatic Perfusion of Melphalanfor Liver-Dominant Metastatic Uveal Melanoma: A Single CenterExperience.” Cancer Imaging 19, no. 1: 31.. Baqai, U., A. M. Kurimchak, I. V. Trachtenberg, et al. 2023. “KinomeProfiling Identifies MARK3 and STK10 as Potential TherapeuticTargets in Uveal Melanoma.” Journal of Biological Chemistry 299, no.12: 105418.. Chabot, T., F. Nemati, A. Herbette, et al. 2022. “Cytidine Analogs AreSynthetic Lethal With Base Excision Repair Default due to MBD4Deficiency.” npj Precision Oncology 6, no. 1: 81.. Chan, P. Y., M. M. Phillips, S. Ellis, et al. 2022. “A Phase 1 Study of ADI-PEG20 (Pegargiminase) Combined With Cisplatin and Pemetrexed inASS1-Negative Metastatic Uveal Melanoma.” Pigment Cell & MelanomaResearch 35, no. 4: 461–470. . Chang, K. Y., N. J. Chiang, S. Y. Wu, et al. 2021. “Phase 1b Study ofPegylated Arginine Deiminase (ADI-PEG 20) Plus Pembrolizumab inAdvanced Solid Cancers.” Oncoimmunology 10, no. 1: 1943253.. Delage, B., D. A. Fennell, L. Nicholson, et al. 2010. “Arginine Deprivationand Argininosuccinate Synthetase Expression in the Treatment ofCancer.” International Journal of Cancer 126, no. 12: 2762–2772. . Dimitriou, F., M. M. Orloff, E. C. Koch Hein, et al. 2025. “TreatmentSequence With Tebentafusp and Immune Checkpoint Inhibitors inPatients With Metastatic Uveal Melanoma and Metastatic GNA11/GNAQ Mutant Melanocytic Tumors.” European Journal of Cancer 214:115161.. Field, G. C., I. Pavlyk, and P. W. Szlosarek. 2023. “Bench-to-BedsideStudies of Arginine Deprivation in Cancer.” Molecules 28, no. 5: 2150.. ### Page 12 Hand, F., S. Doherty, G. Gullo, et al. 2020. “Metastatic Uveal Melanoma:A Valid Indication for Liver Resection.” Journal of BUON 25, no. 2:1161–1165. Hassel, J. C., S. Piperno-Neumann, P. Rutkowski, et al. 2023.“Three-Year Overall Survival With Tebentafusp in Metastatic UvealMelanoma.” New England Journal of Medicine 389, no. 24: 2256–2266.. Jager, M. J., C. L. Shields, C. M. Cebulla, et al. 2020. “Uveal Melanoma.”Nature Reviews Disease Primers 6, no. 1: 24.. Johansson, P. A., J. M. Palmer, L. McGrath, et al. 2025. “GermlineVariants in Patients Affected by Both Uveal and Cutaneous Melanoma.”Pigment Cell & Melanoma Research 38, no. 1: e13199.. Khoja, L., E. G. Atenafu, S. Suciu, et al. 2019. “Meta-Analysis inMetastatic Uveal Melanoma to Determine Progression Free and OverallSurvival Benchmarks: An International Rare Cancers Initiative (IRCI)Ocular Melanoma Study.” Annals of Oncology 30, no. 8: 1370–1380.. Kim, R. H., J. M. Coates, T. L. Bowles, et al. 2009. “Arginine Deiminaseas a Novel Therapy for Prostate Cancer Induces Autophagy andCaspase-Independent Apoptosis.” Cancer Research 69, no. 2: 700–708.. Kordon, M. M., M. Zarebski, K. Solarczyk, H. Ma, T. Pederson, and J. W.Dobrucki. 2020. “STRIDE—A Fluorescence Method for Direct, SpecificIn Situ Detection of Individual Single- Or Double- Strand DNA Breaksin Fixed Cells.” Nucleic Acids Research 48, no. 3: e14.. Kraehenbuehl, L., A. Holland, E. Armstrong, et al. 2022. “Pilot Trial ofArginine Deprivation Plus Nivolumab and Ipilimumab in Patients WithMetastatic Uveal Melanoma.” Cancers 14, no. 11: 2638.. Lai, Y. H., K. C. Tung, and S. C. Chen. 2022. “Durable Response toOlaparib Combined Low-Dose Cisplatin in Advanced HepatocellularCarcinoma With FANCA Mutation: A Case Report.” Medicine 101, no.39: e30719.. Lane, A. M., I. K. Kim, and E. S. Gragoudas. 2018. “Survival Rates inPatients After Treatment for Metastasis From Uveal Melanoma.” JAMAOphthalmology 136, no. 9: 981–986.. Li, Y., Y. Zhang, Y. Q. Yang, et al. 2022. “Silencing of FANCI PromotesDNA Damage and Sensitizes Ovarian Cancer Cells to Carboplatin.”Current Cancer Drug Targets 22, no. 7: 591–602.. Lin, R., Y. Mo, H. Zha, et al. 2017. “CLOCK Acetylates ASS1 to DriveCircadian Rhythm of Ureagenesis.” Molecular Cell 68, no. 1: 198–209.e6.. Minor, D. R., R. E. Allen, D. Alberts, Y. M. Peng, G. Tardelli, and J.Hutchinson. 1985. “A Clinical and Pharmacokinetic Study of IsolatedLimb Perfusion With Heat and Melphalan for Melanoma.” Cancer 55,no. 11: 2638–2644.. Niraj, J., A. Farkkila, and A. D. D'Andrea. 2019. “The Fanconi AnemiaPathway in Cancer.” Annual Review of Cancer Biology 3: 457–478.. Olofsson Bagge, R., A. Nelson, A. Shafazand, et al. 2024. “A PhaseIb Randomized Multicenter Trial of Isolated Hepatic Perfusion inCombination With Ipilimumab and Nivolumab for Uveal MelanomaMetastases (SCANDIUM II Trial).” ESMO Open 9, no. 7: 103623.. O'Reilly, E. M., J. W. Lee, M. Zalupski, et al. 2020. “Randomized,Multicenter, Phase II Trial of Gemcitabine and Cisplatin With orWithout Veliparib in Patients With Pancreas Adenocarcinoma and aGermline BRCA/PALB2 Mutation.” Journal of Clinical Oncology 38, no.13: 1378–1388.. Ott, P. A., R. D. Carvajal, N. Pandit-Taskar, et al. 2013. “Phase I/IIStudy of Pegylated Arginine Deiminase (ADI-PEG 20) in Patients WithAdvanced Melanoma.” Investigational New Drugs 31, no. 2: 425–434.. Panda, P. K., A. C. Paschoalini Mafra, A. C. S. Bastos, et al. 2025. “Bcl-XlProtects ASS1-Deficient Cancers From Arginine Starvation InducedApoptosis.” Clinical Cancer Research 31: 1333–1345.. Park, J. J., S. A. Hamad, A. Stewart, M. S. Carlino, S. Y. Lim, and H.Rizos. 2024. “PKC-Independent PI3K Signalling Diminishes PKCInhibitor Sensitivity in Uveal Melanoma.” Oncogene 13, no. 1: 9.. Patel, S. A., and A. J. Minn. 2018. “Combination Cancer Therapy WithImmune Checkpoint Blockade: Mechanisms and Strategies.” Immunity48, no. 3: 417–433.. Pelster, M. S., S. K. Gruschkus, R. Bassett, et al. 2021. “Nivolumab andIpilimumab in Metastatic Uveal Melanoma: Results From a Single-ArmPhase II Study.” Journal of Clinical Oncology 39, no. 6: 599–607.. Phillips, M. M., I. Pavlyk, M. Allen, et al. 2023. “A Role for MacrophagesUnder Cytokine Control in Mediating Resistance to ADI-PEG20(Pegargiminase) in ASS1-Deficient Mesothelioma.” PharmacologicalReports 75, no. 3: 570–584.. Plimack, E. R., C. Tangen, M. Plets, et al. 2024. “Correlative Analysisof ATM, RB1, ERCC2, and FANCC Mutations and Pathologic CompleteResponse After Neoadjuvant Chemotherapy in Patients With Muscle-Invasive Bladder Cancer: Results From the SWOG S1314 Trial.”European Urology 86, no. 4: 297–300.. Prudner, B. C., R. Rathore, A. M. Robinson, et al. 2019. “ArginineStarvation and Docetaxel Induce c-Myc-Driven hENT1 SurfaceExpression to Overcome Gemcitabine Resistance in ASS1-NegativeTumors.” Clinical Cancer Research 25, no. 16: 5122–5134.. Przystal, J. M., N. Hajji, C. Khozoie, et al. 2018. “Efficacy of ArginineDepletion by ADI-PEG20 in an Intracranial Model of GBM.” Cell Death& Disease 9, no. 12: 1192.. Rodler, E., P. Sharma, W. E. Barlow, et al. 2023. “Cisplatin WithVeliparib or Placebo in Metastatic Triple-Negative Breast Cancer andBRCA Mutation-Associated Breast Cancer (S1416): A Randomised,Double-Blind, Placebo- Controlled, Phase 2 Trial.” Lancet Oncology 24,no. 2: 162–174.. Schinzari, G., E. Rossi, A. Cassano, et al. 2017. “Cisplatin, Dacarbazineand Vinblastine as First Line Chemotherapy for Liver Metastatic UvealMelanoma in the Era of Immunotherapy: A Single Institution Phase IIStudy.” Melanoma Research 27, no. 6: 591–595.. Smogorzewska, A., S. Matsuoka, P. Vinciguerra, et al. 2007.“Identification of the FANCI Protein, a Monoubiquitinated FANCD2Paralog Required for DNA Repair.” Cell 129, no. 2: 289–301.. Szlosarek, P. W., B. C. Creelan, T. Sarkodie, et al. 2024. “PegargiminasePlus First-Line Chemotherapy in Patients With Nonepithelioid PleuralMesothelioma: The ATOMIC-Meso Randomized Clinical Trial.”JAMA Oncology 10, no. 4: 475–483.. ### Page 13 Tsai, W. B., I. Aiba, Y. Long, et al. 2012. “Activation of Ras/PI3K/ERKPathway Induces c-Myc Stabilization to Upregulate ArgininosuccinateSynthetase, Leading to Arginine Deiminase Resistance in MelanomaCells.” Cancer Research 72, no. 10: 2622–2633.. van der Heijden, M. S., J. R. Brody, D. A. Dezentje, et al. 2005. “In VivoTherapeutic Responses Contingent on Fanconi Anemia/BRCA2 Statusof the Tumor.” Clinical Cancer Research 11, no. 20: 7508–7515.. Visser, M., J. P. N. Papillon, M. Luzzio, et al. 2024. “Discovery ofDarovasertib (NVP-LXS196), a Pan-PKC Inhibitor for the Treatment ofMetastatic Uveal Melanoma.” Journal of Medicinal Chemistry 67, no. 2:1447–1459.. Wolchok, J. D., V. Chiarion- Sileni, R. Gonzalez, et al. 2022. “Long-Term Outcomes With Nivolumab Plus Ipilimumab or Nivolumab AloneVersus Ipilimumab in Patients With Advanced Melanoma.” Journal ofClinical Oncology 40, no. 2: 127–137.. Yarde, D. N., V. Oliveira, L. Mathews, et al. 2009. “Targeting the FanconiAnemia/BRCA Pathway Circumvents Drug Resistance in MultipleMyeloma.” Cancer Research 69, no. 24: 9367–9375.. Zheng, L., E. D. MacKenzie, S. A. Karim, et al. 2013. “ReversedArgininosuccinate Lyase Activity in Fumarate Hydratase-DeficientCancer Cells.” Cancer & Metabolism 1, no. 1: 12.. Supporting Information Additional supporting information can be found online in theSupporting Information section. Figure S1: Effect of ADI-PEG20 andmelphalan on MP41 cell line toxicity. Figure S2: Analysis of PARP andcaspase-3 in MP41 cells. Figure S3: Nuclear dSTRIDE foci count (meanvalues) according to 92.1 UM cell cycle phase. Figure S4: CytoplasmicdSTRIDE foci count (mean values) according to 92.1 UM cell cyclephase. Figure S5: Additional validation in the ASS1 negative Mel270cell line.