The proteomes of ovarian cancer ascitic cellular aggregates correlate with their ex vivo platinum sensitivities: a pilot study

The proteomes of ovarian cancer ascitic cellular aggregates correlate with their ex vivo platinum sensitivities: a pilot study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article The proteomes of ovarian cancer ascitic cellular aggregates correlate with their ex vivo platinum sensitivities: a pilot study Jack Scanlan, Parul Mittal, Noor A Lokman, Martin K Oehler, Peter Hoffmann, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6441929/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Dec, 2025 Read the published version in Journal of Proteome Research → Version 1 posted You are reading this latest preprint version Abstract The accumulation of malignant ascites in the peritoneal cavity is a hallmark of advanced epithelial ovarian cancer (EOC). This fluid contains three-dimensional multicellular aggregates known as spheroids, which contribute to chemoresistance and are an accessible source of tumour material. However, many studies use spheroids generated from primary cell suspensions to reduce heterogeneity. Here, we compare the proteomes and the response to chemotherapeutics of native spheroids directly collected from ascites to spheroids generated ex vivo . We demonstrate that the chemoresponse of native spheroids correlates with patients’ therapy responses in 4/5 cases. In contrast, all ex vivo- generated spheroids were resistant to carboplatin treatment and did not correlate with the clinical outcome. In addition, proteomics quantified over 6,300 proteins per sample, revealing that the global proteomes of native spheroids cluster according to their carboplatin response. A detailed analysis of the upregulated proteins highlights the potential role of extracellular matrix proteins in regulating chemoresponse. This pilot study suggests key proteins and biological pathways that may facilitate a global proteomics-based screening strategy for personalised EOC treatment. As such, native spheroids have the potential to be used to personalise the treatment of all diseases that cause malignant ascites. Health sciences/Medical research/Biomarkers/Predictive markers Health sciences/Oncology/Cancer/Gynaecological cancer/Ovarian cancer Ovarian cancer personalised medicine spheroids malignant ascites mass spectrometry proteomics Figures Figure 1 Figure 2 Figure 3 Figure 4 Background Epithelial ovarian cancer (EOC) is the most lethal gynaecological malignancy, with high-grade serous ovarian cancer (HGSOC) constituting three-quarters of all cases 1 . Patients typically receive standard first-line chemotherapy that consists of carboplatin (CBP) and paclitaxel (PTX) following cytoreductive surgery, with neo-adjuvant chemotherapy sometimes administered to pre-emptively reduce tumour burden 2 . CBP is one of several approved organoplatinum-based alkylating agents that inhibit DNA replication through cytotoxic inter-strand DNA cross-linkages 3 . PTX prevents cell division by stabilising microtubules, promoting G2/M cell cycle arrest, and inhibiting the repair of CBP-mediated DNA adducts 4 . Recent advancements in the treatment of stage III and IV HGSOC include the approval of bevacizumab—a vascular endothelial growth factor (VEGF) inhibitor that inhibits tumour growth by blocking nutrient delivery 5 — for administration alongside CBP and PTX. Pharmaceutical inhibition of poly (ADP-ribose) polymerase (PARP) was also recently approved as an additional first-line treatment in patients with BRCA1/2 mutations or homologous recombination deficiency (HRD) 6 . Although approximately 80 per cent of patients respond to first-line treatment, most develop resistance and require second-line chemotherapeutics. The absence of a clinical tool or marker to predict patients’ responses to first-line treatment promotes a trial-and-error-based approach. As this is an area of unmet clinical need, numerous genomics-based studies have been conducted to identify markers and mechanisms of chemoresponsiveness. These most often use immortalised 2D cell lines and have identified mutations in genes that encode DNA damage repair 7 , cell cycling 8 , drug efflux pathways 9 , and extracellular matrix (ECM) proteins 10 . However, the lack of validation of these markers in large patient cohorts means that none of these are used in clinical practice. Proteomic studies have also identified several proteins that correlate with chemoresponse. An early study of over 2000 proteins in the A2780 and 2008 cell lines and their cisplatin-resistant counterparts reported differential abundances of DNA repair, oxidative stress responses, apoptosis, translation, and cell cycling proteins 11 . A gel- and LC-MS/MS-based analysis of cisplatin-sensitive and resistant A2780 cell lines also identified ten differentially expressed proteins involved in ubiquitination, metabolism, signal transduction, and calcium and nucleotide binding 12 . Recently, our laboratory demonstrated a clear separation of the global proteomes of parental and cisplatin-resistant EOC cell lines using LC-MS/MS 13 . While protein abundances have also been linked with patient survival and treatment response 8 , 14 , investigations of molecular markers are often complicated by poor accessibility and availability of primary tumour samples. Malignant ascites (MA) describes the accumulation of peritoneal fluid that contains tumour cells and is drained during cytoreductive surgery or interval paracentesis to alleviate discomfort. It results from increased capillary permeability due to VEGF-mediated downregulation of the claudin 5 (CLDN5) tight junction protein and poor lymphatic drainage caused by tumour cell blockages 15 . Resident tumour cells, cancer-associated fibroblasts (CAFs), and tumour-associated macrophages (TAMs) can aggregate as heterogeneous spheroids 16 , herein referred to as native ascites-derived spheroids (ADS). Native ADS are proposed to aggregate in response to the upregulation of integrin subunit alpha 5 (ITGA5) and epidermal growth factor promoted by CAFs and TAMs, and contain quiescent and necrotic cores that are not effectively eliminated during treatment 16 . Their unique cell-cell interactions have been shown to promote treatment resistance and metastasis 17 , prompting the development of novel 3D pre-clinical EOC models. For example, spheroids can be generated from immortalised cell lines in hanging drop, ultra-low attachment (ULA), and nutator plates 18 . Artificial ADS have also been formed from tissues and demonstrate epithelial-to-mesenchymal transition and the expression of HGSOC markers such as paired box gene 8 (PAX8) 19 . Finally, retrospective functional drug assays on 3D models derived from dissociated primary tumours or ascites cell cultures have recapitulated patients’ clinical responses to platinum-based chemotherapy 20 – 23 , PTX 23 , 24 , PARP inhibitors 20 , monoclonal antibody treatments 22 , and kinase inhibitors 21 , which has successfully guided treatment selection in some instances 25 . However, artificial ADS often suffer poor generation rates, long lead times that reduce clinical actionability, and uncertainty regarding their representation of ovarian cancer biology. For instance, few differentially expressed genes were observed between HGSOC cells grown as spheroids using basement membrane extract (BME), collagen, and agarose supports from the same OVCAR8 cell lines 26 . Furthermore, no known studies have assessed protein-based markers of chemotherapy response in artificial ADS. Here, we establish a workflow to test the ex vivo drug responses of patient-derived cells and analyse their proteome in a timely manner. We first compare the treatment responses of artificial ADS formed in ULA plates from ascites cells grown in monolayer with chemotherapy-naïve patient-matched native ADS to demonstrate that the latter more accurately and rapidly recapitulate patients’ clinical treatment responses. We also show that the differential abundances of proteins and gene ontology (GO) pathways in native ADS give insights into potential molecular markers of treatment response. Only the global proteomes of native ADS cluster according to treatment response, which provides a path towards prospective treatment response prediction for all diseases that present with MA. Results Effects of chemotherapy dose-response on ovarian cancer spheroids Ex vivo CBP response testing was established by generating artificial spheroids from the parental TYK-nu and A2780 cell lines and their cisplatin-resistant TYK-nu.CPr and A2780-cis sublines. A2780 and A2780-cis cell lines formed significantly larger spheroids (p < 0.0001) than the TYKNU and TYK-nu.CPr cell lines (Fig. 1A), which aligned with their respective cell viabilities (Supplementary Figure S1). While TYK-nu.CPr spheroids were significantly larger (p < 0.0001) than TYK-nu spheroids, the opposite was observed for A2780 and A2780-cis cell lines. The EC 50 values of cisplatin-resistant TYK-nu.CPr and A2780-cis spheroids were 2 and 3.4 times higher than the parental cell lines, respectively (Fig. 1B-C). A proteomics analysis of all four cell lines grown as adherent and spheroidal cultures revealed a clear separation of TYK-nu/TYK-nu.CPr and A2780/A2780-cis cell lines along the first component of the PCA plot, while samples separated along the second component based on growth dimensionality (Supplementary Figure S2). Effects of chemotherapy dose response on artificial ADS We postulated that the generation of similarly-sized spheroids from primary ascites cells would provide a consistent measure of chemotherapy response. Although artificial ADS from the same patient were consistently sized, those from different patients varied significantly (p < 0.0001) without correlation to clinical response (Supplementary Figure S3). Artificial ADS from all patients had relatively high EC 50 values (greater than 140 µM), which did not correlate with clinical outcomes (Fig. 2). Artificial ADS generally demonstrated clear and contained borders, with the exception of those from patient 4, which appeared larger due to surrounding non-spheroidal matter. Native ADS originating from platinum-sensitive patients 1 and 2 were sensitive in ex vivo experiments, with respective EC 50 values of 14 and 71 µM. Patient 3 was clinically resistant but was most sensitive to CBP in ex vivo assessments with an EC 50 of 8 µM. The ex vivo EC 50 of 141 µM for the native ADS of patient 4 matched with their clinical resistance status. While patient 5 had an unknown clinical sensitivity, their native ADS were categorised as resistant due to an ex vivo EC 50 of 160 µM. Native and artificial ADS have distinct proteomes LC-MS/MS analysis of all native and artificial ADS samples identified 6,664 proteins overall (Fig. 3A), with 570 proteins being more abundant in artificial ADS and 928 being more abundant in native ADS (Fig. 3B). These differences were represented in the PCA plot, which clearly separated native and artificial ADS on the first component with negligible separation of artificial ADS (Fig. 3C). Native ADS separated along the second component based on ex vivo CBP sensitivity, with patients 1, 2, and 3 considered sensitive and 4 and 5 considered resistant. Figure 3D shows that the most significantly enriched GO terms in native ADS primarily pertained to DNA replication, such as nucleosomal DNA and RNA binding, and the organisation of chromatin, telomeres, and nucleosomes. The nucleus and cytosol hosted the most enriched CCs, except for the extracellular exosome, which was the most significantly enriched compartment in both native and artificial ADS. The most-enriched BPs in artificial ADS included the binding of collagens, proteases, integrins, and calcium ions, and ECM structural constituents. Relevant MFs included collagen fibril organisation and cell adhesion. Highly enriched CCs in artificial ADS included extracellular regions, such as the ECM, as well as the lumina of lysosomes and the endoplasmic reticulum. Native ADS proteomes cluster according to ex vivo treatment responses The greater separation of native ADS proteomes compared to artificial ADS proteomes (Fig. 3C) prompted an analysis of native ADS proteomes alone. Of the 6303 proteins identified (Fig. 4A), 700 were more abundant in resistant ADS and 496 were more abundant in sensitive ADS (Fig. 4B). The native ADS of patients 1–3 and 4 & 5 are separated into distinct clusters along the first component of the PCA plot (Fig. 4C), which reflects their categorisation of sensitivity and resistance in ex vivo CBP response assessments, respectively. Some of the most significantly enriched GO terms in resistant ADS were proteins at the interplay of cell migration, cytoskeletal organisation, and focal adhesions (Fig. 4D). At the individual protein level, we observed elevated abundances of several integrins in resistant native ADS, including ITGAX, ITGA2, ITGB4, ITGA6, ITGAM, ITGA, ITGB1, and ITGB2, which functionally link the cytoskeleton to the ECM. This was complemented by additional adhesion-related proteins such as annexin (ANX) A2, cadherins (CDH) 1 and 2, and the cell surface receptor CD44 being up in resistant native ADS, along with actin filament-associated protein 1-like 2 (AFAP1L2) and actins ACTB1 and ACTG1. The differential abundance of many ECM proteins prompted an analysis of those within the MatrixDB database. 29 Of all proteins in this database, 188 were differentially abundant between sensitive and resistant native ADS (Supplementary Table S1), including members of the serine protease inhibitor (SERPIN) superfamily, which were some of the most differentially abundant in this list. For example, SERPINs A1, A3, and H1 were significantly more abundant in sensitive samples and SERPINs B1, B2, B5, B6, B8, and B9 in resistant native ADS. Carcinoembryonic antigen-related cell adhesion molecules (CEACAM) 5–7 were also highly abundant in resistant native ADS. Relatedly, BPs relating to cytoskeletal organisation, such as actin filament binding, were most significantly enriched in resistant native ADS. Many intermediate filament-forming keratins had significantly higher abundances in resistant native ADS, with the gene expression of several being significantly associated with progression-free survival (PFS) and/or overall survival (OS) in stage III and IV HGSOC (Supplementary Table S2). GO terms that involve protein synthesis, including DNA unwinding, ribosomal biogenesis, and rRNA processing were enriched in sensitive native ADS. The four most significantly-enriched CCs were within mitochondria and mitochondrial translation was the most significantly-enriched BP. At the individual protein level, several large ribosomal subunit proteins, tRNA ligases, and ATP synthase subunits saw high fold changes. The enrichment of the ficolin-1-rich granule lumen cell compartment in resistant native ADS indicates an immune component, which was complemented by higher abundances of the branched-chain amino acid transaminase (BCAT) 1 and 2 immune-related proteins and proteoglycan 2 (PRG2)— the latter of which had the highest fold-increase compared to sensitive native ADS. In sensitive native ADS, immunoglobulin kappa variable cluster (IGKV) 3–20, IGKV3-7, IGKV2-28, and IGHV3OR16-9 constituted some of the most upregulated proteins, alongside the THY1 membrane glycoprotein, which had the highest fold change of all proteins. Effects of additional treatments on native ADS Given the availability of several first and second-line treatments, a future personalised medicine strategy that can predict patients’ responses to an expanded range of chemotherapy drugs would be valuable. The ex vivo responses of native ADS from a subset of four patients were assessed against PTX, olaparib, gemcitabine, doxorubicin, and topotecan (Table 1). Patient 2 in this cohort had a predictive BRCA2 mutation for olaparib senstivity, but had the second-highest ex vivo resistance. Sensitivity to PTX showed a distinct pattern, with patients 2 and 3 being far more sensitive than patients 1 and 4. Native ADS also showed differential chemoresponses to doxorubicin, topotecan, and gemcitabine. Table 1: Chemoresponses of native ADS to second-line chemotherapy agents. Green and red shading indicate predicted sensitivity and resistance to the PARP inhibitor olaparib based on genomic testing. Patient 2 had a predictive BRCA2 mutation. Discussion The poor survival outcomes for EOC patients necessitate a new model of care. Given the number of available chemotherapeutics and combinations thereof, an ex vivo disease model that can rapidly recapitulate patient sensitivities is sought to enable personalised medicine approaches that could guide clinicians in choosing the most effective treatment for each patient. Although 2D ex vivo models that use primary cells have been reported extensively 23 , the roles of ascitic multicellular spheroids in EOC treatment resistance and metastatic processes have prompted a shift to 3D models that can be generated by culturing cells in BME 21 , 24 , 31 – 33 , ULA plates 34 , or microfluidics devices 20 . Importantly, spheroids generated from primary tumour material can show differential responses to EOC treatments 24 , 31 . They also share genomic and phenotypic features with their tumours of origin, such as PAX8 and P53 expression 31 . Their responses to multiple first- and second-line chemotherapeutics correlate with clinical outcomes and have even been used to guide treatment in a limited number of prospective studies 22 , 25 . Importantly, studies have incorporated primary ascites cells, which are thought to be key drivers of treatment resistance and metastasis 15 . For example, ascitic spheroids show differential ex vivo responses to diverse chemotherapies and can correlate with patients’ clinical data 20 . While functional response assessments can assist in therapy decisions, the collection of proteomic data will expand our understanding of chemoresponse and may eventually permit chemoresponse prediction based solely on molecular data. In this pilot study, we established a workflow to assess the ex vivo dose responses of a novel 3D EOC model that does not require a spheroid generation step, which allows ex vivo chemoresponse and proteomics data to be obtained within 4 days of ascites collection. This combined approach bridges the gap between costly ex vivo testing and clinical application, which could facilitate the eventual creation of a proteomic knowledge base with which to guide treatment. A workflow to measure ex vivo chemoresponse was established with cisplatin-sensitive and -resistant ovarian cancer TYK-nu and A2780 cell lines. Although neither ATP concentrations nor the relationship between spheroid size and chemoresponse were consistent across the cell lines, this approach was sufficiently robust to monitor viability between treated and untreated cells. Together with the reproducibility of sample preparation for proteomic analyses, this gave confidence to use this approach for patient samples. Artificial ADS could be consistently generated from five of six patient ascites samples, which aligns with the success rates of previous studies. Interestingly, all artificial ADS required high CBP concentrations to induce cell death regardless of patient origin or spheroid size, indicating chemoresistance. The lack of separation between artificial ADS proteomes in the PCA plot further demonstrated their inability to distinguish patients based on chemoresponse. Given these factors, native ADS were used in all further experiments. Given that spheroid generation techniques can alter the transcriptome 26 , a different result may be possible using other generation methods. Native ADS were isolated from five of the six ascites samples. In four of the five analysed native ADS samples, the relative chemoresponses were consistent with the clinical data. The exception was native ADS isolated from patient 3, which had the lowest ex vivo EC 50 value against CBP of any sample despite being considered resistant to CBP + PTX treatment in the clinic. This discrepancy could arise from a variety of factors, including the selection of a subset of cells used in the assay or the particular clinical response indicator used in this study. For instance, de Witte et al. 24 demonstrated that the ex vivo responses of EOC organoids to standard doublet chemotherapy were significantly correlated with patient-matched histopathological, biochemical, and radiological data but not with the presently-used progression-free survival. The ability of proteomics to reflect dynamic cellular processes was observed in these results. For example, the observed pattern of N-cadherin upregulation and E-cadherin downregulation in artificial ADS compared to native ADS is known to trigger native ADS formation 16 . Despite this, the poor separation of artificial ADS proteomes prompted an investigation into the predictive value of native ADS. The clustering of native ADS proteomes according to ex vivo treatment response supports the utility of this approach in the context of crude clinical classifications and the notion that proteomics could be used to predict sensitivity in future prospective studies. At the individual protein level, pathways relating to the interplay between the ECM and actin cytoskeleton were of particular interest, given their known roles in cancer progression. Notably, members of all integrin classes, which act as an interface for the ECM and cytoskeleton, were elevated in resistant native ADS. This included leukocyte-specific (ITGAX, ITGAM, and ITGB2), collagen-binding (ITGA2), laminin-binding (ITGB4 and ITGA6), and RGD-binding (ITGA5) integrins, along with ITGB1. Integrin-mediated activation of focal adhesion kinase (FAK) has been reported to promote platinum resistance through the inhibition of apoptotic cell death by triggering Src kinase-mediated upregulation of MAPK/ERK and PI3K/Akt/mTOR signalling pathways 35 . For example, β1 integrin-mediated activation of the FAK/Akt pathway has been identified as a mechanism of resistance in spheroids generated from immortalised hepatocellular carcinoma cell lines 36 . The upregulation of FAK and steroid receptor coactivator (SRC) kinase family proteins in resistant ADS, such as SRC and cortactin, points to a potential role for this pathway in ascitic ADS. The stimulation of tyrosine kinases through ECM-mediated ITGB1 activation has also been shown to suppress apoptosis in small-cell lung cancer 37 . Finally, the overexpression of ITGA5 has been shown to mediate cisplatin resistance in spheroids derived from immortalised nasopharyngeal cancer-derived spheroids through the inactivation of caspase-3-mediated apoptosis 38 , further underscoring the likely importance of this receptor class in chemoresistance. Collagens are another constituent of the ECM that have been postulated to affect chemoresistance through cell adhesion-mediated drug resistance. For example, the upregulation of several collagens has been observed in cisplatin-resistant ovarian cancer models 10 , which elevates increases ECM stiffness to potentially promote drug efflux through altered multi-drug resistance 1 (MDR1) protein activity 39 . Another reported mechanism of chemoresistance in the context of collagen-mediated ECM stiffness is the prevented entry of chemotherapeutics and immune cells 40 . The elevated abundance of all identified collagens (COL1A1, 1A2, 4A2, 6A1, 6A2, and 6A3) in resistant native ADS indicates their potential role as a marker of chemoresponse in this model. The noted roles of ECM remodelling in chemorepsonse and the observed enrichment of extracellular GO terms in resistant native ADS prompted closer analysis of all ECM proteins using the MatrixDB database 29 . Several proteins from the SERPIN superfamily were some of the most differentially abundant proteins in native ADS. Those within the SERPIN ‘A’ clade are extracellular and primarily involved in pro-inflammatory processes and hormone transport. For example, SERPINA10 expression has been positively associated with platinum sensitivity in the TCGA ovarian cancer cohort, which was validated using immunohistochemistry on HGSOC tissues 41 . The observation of elevated abundances of SERPIN A1, A3, and H1 in sensitive native ADS supported this. ‘B’ SERPINs are instead intracellular and mainly inhibit apoptotic cell death through the inhibition of regulatory proteins. SERPINs B1, B2, B5, B6, B8, and B9 were all elevated in resistant native ADS, which supports previous associations with poor prognosis in ovarian cancers 42 . Notably, the paralogous SERPINB3 and B4 isoforms were not significantly different and not detected, respectively. Actin cytoskeleton remodelling is a hallmark of cancer and has also been observed to promote platinum chemoresponse through increased intracellular stiffness 43 , of which rho GTPases are key activators. Rho GTPase activators were significantly more abundant in resistant native ADS, including rho GTPase-activating proteins (ARHGAP) 12, 18, 27, and 45. The abundances of other actin cytoskeleton regulators also differed between sensitive and resistant native ADS. For example, gelsolin (GSN) is an actin filament-binding protein that mediates cytoskeletal remodelling and inhibits apoptosis, which was over 2.5-fold more abundant in resistant native ADS. This supports previous observations from our laboratory of differential GSN expression in cell lines and HGSOC tumour tissues based on platinum resistance 14 , and associations to poor OS and PFS in serous ovarian cancer patients 44 . Another critical cytoskeletal component is the keratin superfamily. Keratins form intermediate filaments within epithelial cells and have numerous structural and signalling functions, such as protection against chemotherapeutics and other cellular stressors. The abundance of KRT5 has been previously associated with poor outcomes at both the mRNA and protein levels 45 in stage III and IV serous ovarian cancers. Using the Kaplan-Meier (KM) plotter 46 , significantly higher expression of KRT6C and KRT85 were individually correlated with reduced PFS and OS, reflecting their increased abundances in resistant native ADS. Although immune proteins may have roles in chemoresponse due to the presence of immune cells in ascitic aggregates, their differential expression may result from the differential presence of cell-surface glycans. Further investigations of constituent immune cells of native ADS and cell surface glycans could clarify this matter. Finally, mitochondrial proteins have been implicated in the regulation of chemoresponse due to their roles as key regulators of critical intracellular processes such as metabolism, oxidative stress responses, and cell death. Mitochondrial involvement in platinum resistance remains relatively under-researched, despite frequent mitochondrial DNA (mtDNA) mutations and mtDNA depletion 47 that can promote the dysregulation of resistance pathways in cancer. The enrichment of translation, ATP synthesis, ribosomal biogenesis, and nucleic acid replication in sensitive ADS could represent a decreased ability for platinum-resistant ascitic spheroids to respond to pathogenic stimuli. This is evidenced by the decrease in several proteins that are protective against oxidative stress in resistant ADS, including glutathione peroxidases 3 and 7. Proteomics also identified the downregulation of several pro-apoptotic proteins in resistant ADS, including programmed cell death 2-like (PDCD) 2L, 5, 4, and apoptosis-inducing factor 1 (AIFM1). Interestingly, apoptotic peptidase activating factor 1 (APAF1) was upregulated in resistant ADS, despite a previous study using primary ovarian tumours which indicated that APAF-1 levels do not necessarily correlate with caspase-9 activation 48 . These findings underline the need for more research into the role of apoptotic proteins in treatment resistance. The use of native ADS as a 3D model to predict treatment response and collect molecular information more accurately and rapidly is limited by several factors. Firstly, Kim et al. debate the clonal relationship between the primary tumour and ADS 49 , suggesting early evolutionary divergence based on significant differences in copy number and single-nucleotide variants between ADS and patient-matched tumour tissues. However, this may be beneficial given the unique role of ascitic spheroids in chemoresistance and disease progression. Some EOC treatments would be incompatible with ex vivo response assessments, such as the VEGF inhibitor bevacizumab, which prevents micro-vascularisation that is not present in spheroids. 5 Nevertheless, expanding this approach to common second- and third-line treatments, as well as combinations of chemotherapy such as CBP + PTX, would be helpful for those without a predictive tool. Predicting PARP inhibitor response using a proteomic tool would also be welcome, given that the genomic tool based on BRCA1/2 mutation and HRD statuses often suffers from long turnaround times. This pilot study is the first to demonstrate that the global proteomes of native ADS cluster according to their ex vivo platinum responses. We show that these results are concordant with clinical data in most cases and that this model differentially responds to an expanded panel of EOC treatments. These findings suggest the potential utility of a spheroid-based personalised medicine approach for second- and third-line chemotherapeutics. The recruitment of further patients would allow for the development of a global proteomic strategy that may rely solely on PCA clustering to predict treatment responses and eventually bypass the need for individual molecular markers. This is especially useful given the use of a DIA proteomics approach, which could enable the generation of a public database against which clinicians could compare prospective samples. Nevertheless, the power of MS-based proteomics to identify changes in the abundance of individual proteins between sensitive and resistant native ADS identified ECM, cytoskeleton, mitochondria, and immune proteins and pathways. Given the prevalence of MA in other malignancies, including breast, liver, gastric, and colorectal cancers, the utility of this model could have far-reaching implications for personalised cancer treatments. Methods Ascites sample collection and processing Ascites samples were collected between 2019 and 2024 from six HGSOC patients during cytoreductive surgery at the Department of Gynaecological Oncology (Royal Adelaide Hospital, Adelaide, Australia) with written informed consent and approval by the hospital ethics committee (approval #R20181215) prior to storage in liquid nitrogen. Clinical platinum sensitivity was defined by the absence of progression six months after chemotherapy cessation, while relapse or no response within six months indicated clinical platinum resistance. Thawed ascites samples were cultured overnight in Advanced RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) (Bovogen, Vic, Australia), 1% penicillin-streptomycin, and 1% L-glutamine at 37⁰C with 5% CO 2 . Cells were detected by brightfield microscopy in five of six patient samples. Native ADS remained suspended in the medium and were isolated by gentle centrifugation for CBP response assessments or stored for proteomic analysis. Adherent cells were expanded for a maximum of 3 passages for artificial ADS generation. Cell lines Mycoplasma-free and authenticated A2780 (RRID:CVCL_0134), A2780-cis (RRID:CVCL_1942), TYK-nu (RRID:CVCL_1776), and TYK-nu.CPr (RRID:CVCL_3221) cell lines (Cell Bank Australia, Westmead, Australia) were cultured at 37⁰C with 5% CO 2 in Eagle’s Minimum Essential Medium (EMEM; Merck, NSW, Australia) that was supplemented with 10% FBS, 1% penicillin-streptomycin, and 1% L-glutamine. Cell lines were subcultured for a maximum of 10 passages before being trypsinised (Sigma Aldrich, Vic, Australia) for proteomic analyses or dilution and seeding in ULA plates spheroid generation. Artificial spheroids generation Artificial spheroids were generated over 72 hours from cell lines or adherent ascites cells by seeding 2 x 10 3 cells in 90 µL of appropriate medium within each well of 96-well Nunclon Sphera ULA plates (Thermo Fisher Scientific, Vic, Australia). Spheroids were imaged using an IN Cell Analyzer 2200 (Cytiva, MA, United States of America) with a 10x brightfield objective and sized using the semi-automated MATLAB-based SpheroidSizer 27 program. Artificial spheroids were collected for proteomic analyses using a manual pipette and gentle centrifugation. Artificial ADS generated from the ascites cells of patient 4 were only used in dose-response testing and were excluded from sizing due to the abundance of interfering non-spheroidal matter. Ex vivo chemoresponse assessments Artificial and native ADS collected directly from MA were treated with CBP (Hospira, NSW, Australia) at concentrations of 300, 250, 200, 150, 100, 75, 50, 25, 12.5, 6.3, 1.6, and 0.8 µM and compared to a PBS control. Native ADS were treated with olaparib at the same concentrations using a 1:1 PBS:DMSO control. Native ADS were also treated with gemcitabine, topotecan, doxorubicin, and PTX ( at 10, 7.5, 3.16, 1, 0.316, 0.01, 0.0316, 0.01, 0.00316, and 0.001 µM, in separate experiments with a 1:1 PBS:DMSO control. All assessments were conducted in triplicate in total volume of 100 µL for 72 hours at 37 o C with 5% CO 2 . ATP concentration was used to indicate viability by incubating samples with an equal volume of CellTitre-Glo 3D (Promega, WI, United States of America) at room temperature for 5 mins at 300 rpm followed by 25 mins without rotation. During incubation, spheroids were transferred to white-coated 96-well plates for luminescence assessment using the VICTOR Nivo plate reader (Revvity, MA, United States of America). Half maximal dose-response values (EC 50 ) were calculated using GraphPad Prism (v10.1.2) using a four-parameter logistical regression. Patients were designated CBP-sensitive or -resistant based on EC 50 values against CBP. Values over the maximum dose ( x ) were reported as ‘> x µM’. Sample preparation for proteomic analyses Sample preparation was performed using the S-Trap™ Micro protocol (PROTIFI, NY, United States of America) with minor modifications. Briefly, cell pellets were sonicated for 15 mins in 5% SDS and 50 mM Tris at pH 8.5. Proteins were reduced with 10 mM dithiothreitol in the dark for 1 hour, and alkylated with 15 mM chloroacetamide for 30 mins in the dark. Protein concentration was determined using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Vic, Australia). Approximately 50–100 µg of protein was loaded onto the S-Trap™ Micro column and digested overnight at 37 o C with 5 µg of a Trypsin/LysC mix (Promega, WI, United States of America) in 50 mM ammonium bicarbonate. The resulting peptides were eluted, dried under vacuum, resuspended in 0.1% formic acid (FA), and quantified using the Nanodrop One/One c (Thermo Fisher Scientific, Vic, Australia). The peptide solutions were transferred to high-performance liquid chromatography vials for further analysis. Mass spectrometry data acquisition Approximately 200 ng of peptides from each sample were injected in duplicate into the UltiMate 3000 nanoLC (Thermo Fisher Scientific, Vic, Australia) that was coupled online to a timsTOF fleX (Bruker Daltonics, MA, United States of America) with a nano-electrospray Captive Spray ion source and 20 µm Classic Emitter with the capillary voltage set to 1500 v. Instruments were controlled with Bruker Compass 4.1 (v6.2, build 1.2) and Compass HyStar 6.3 (v6.3.1.8). A 150 µm ID × 150 mm (1.5 µm, 100 Å Reprosil Saphir C18) reversed-phase column (PremierLCMS, CA, United States of America) was used for chromatographic separation at 1 µL/min at 40 o C. Mobile phases A and B were 0.1% FA in water and 0.1% FA in 80% acetonitrile, respectively, with a linear gradient of phase B from 3.8–25% over 45 mins, then 87.3% in 3 mins, which was held for 5 mins for a total run time of 70 mins. Data were acquired in the diaPASEF mode with a mass scan range of 100–1700 m/z , an ion mobility window of 1/K0 0.6–1.6, ramp time of 100 ms, and 100% duty cycle. Accumulation time was 2 ms with 3 L/min dry gas flow at 180 o C. Additional settings included deflection 1 Δ of 70.0 V, funnel 1 RF of 300.0 Vpp, funnel 2 RF of 200.0 Vpp, and CID energy of 0.0 eV. Ion mobility parameters were set to ∆t6 of 55.0 V, funnel 1 RF of 450.0 Vpp, and a collision cell voltage of 300 V. Analysis of mass spectrometry data Proteins were inferred from DIA (data-independent acquisition) mass spectrometry data on Spectronaut (Biognosis, Zurich, Switzerland; version v18.7.240506.55695) using the library-free directDIA mode and the canonical Homo sapiens UniProt knowledgebase containing 20,596 entries (downloaded 08/11/2023). Two missed cleavage sites, two variable modifications (N-terminal acetylation and methionine oxidation), and 7 to 52 amino acids were permitted per peptide. Protein identification relied on the presence of at least two peptides per protein and the cysteine carbamidomethylation as a fixed modification. The false discovery rate was set to 1%. Peptides between 200 and 3000 m/z were quantified using the three most abundant peptides with a minimum intensity threshold of 1%. The five most significant biological process (BP), cellular compartment (CC), and molecular function (MF) GO terms, according to their Benjamini-Hochberg-corrected p-values, were identified by inputting lists of differentially expressed proteins into DAVID (National Institutes of Health, United States of America) 28 . ECM proteins were identified with MatrixDB 29 and principal component analysis (PCA) plots were generated with ClustVis 30 . Statistical analysis Statistical analyses were performed using GraphPad Prism (version 10.1.2, GraphPad Software, Inc., CA, USA) with an alpha level of 0.05. The means of two approximately normally distributed data sets were compared using unpaired t-tests. In contrast, comparisons of three or more means were analysed using ordinary one-way ANOVA with Bonferroni’s correction. Differences in protein abundances measured by mass spectrometry were considered significant when there was a two-fold or greater change, and the q-value was < 0.05. P-values sourced from the KM plotting tool 32 used the log-rank hypothesis test to compare the PFS and OS of stage III and IV high-grade (grade III) serous ovarian cancer patients with high or low gene expression, with an alpha level of 0.05 considered as significant. Declarations Additional information The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Human Research Ethics Committee of the Central Adelaide Local Health Network, South Australia, Australia, under approval number R20181215. Informed consent was obtained from all participants prior to their inclusion in the study. All participants provided informed consent for the publication of their data in this study. The authors declare no competing interests. Author Contribution Author contributions included JS (conceptualisation, data collection, writing the original draft), PM (conceptualisation, data collection, supervision, reviewing, and editing), NAL (methodology, data analysis, reviewing, and editing), MKO (conceptualisation, supervision, data analysis, reviewing, and editing), PH (conceptualisation, supervision, data analysis, reviewing, and editing), and MKH (conceptualisation, supervision, data analysis, reviewing, editing, and project administration). All authors read and approved the final manuscript. Acknowledgement This study was supported by the Letitia Linke Research Foundation, Tour de Cure, and an Australian Government Research Training Program (RTP) Scholarship. The authors acknowledge all patients treated at the Royal Adelaide Hospital for their donation of biological material for this project. Financial support was provided by Bioplatforms Australia, the Government of South Australia and the University of South Australia towards the National Collaborative Research Infrastructure Strategy (NCRIS) node for Tissue Imaging Mass Spectrometry. Data Availability The datasets generated during and analysed during the current study are available from the corresponding author on reasonable request. References Torre, L. A. et al. Ovarian Cancer Statistics, C.A. Cancer. J. Clin. 68, 284–96 (2018). (2018). Neesham, D., Richards, A. & McGauran, M. Advances in epithelial ovarian cancer. Aust J. Gen. Pract. 49 , 665–669 (2020). Kemp, Z. & Ledermann, J. Update on first-line treatment of advanced ovarian carcinoma. Int. J. Womens Health . 5 , 45–51 (2013). Jiang, S. et al. Paclitaxel Enhances Carboplatin-DNA Adduct Formation and Cytotoxicity. Chem. Res. Toxicol. 28 , 2250–2252 (2015). Shih, T. & Lindley, C. Bevacizumab: an angiogenesis inhibitor for the treatment of solid malignancies. Clin. Ther. 28 , 1779–1802 (2006). PBAC Meeting Outcomes July. Pharmaceutical Benefits Advisory Committee https:// (2023). (2023). Pennington, K. P. et al. Germline and Somatic Mutations in Homologous Recombination Genes Predict Platinum Response and Survival in Ovarian, Fallopian Tube, and Peritoneal Carcinomas. Clin. Cancer Res. 20 , 764–775 (2014). Chowdhury, S. et al. Proteogenomic analysis of chemo-refractory high-grade serous ovarian cancer. Cell 186 , 3476–3498 (2023). Ween, M. P., Armstrong, M. A., Oehler, M. K. & Ricciardelli, C. The role of ABC transporters in ovarian cancer progression and chemoresistance. Crit. Rev. Oncol. Hematol. 96 , 220–256 (2015). Januchowski, R. et al. Increased Expression of Several Collagen Genes is Associated with Drug Resistance in Ovarian Cancer Cell Lines. J. Cancer . 7 , 1295–1310 (2016). Fitzpatrick, D. P. G. et al. Searching for potential biomarkers of cisplatin resistance in human ovarian cancer using a label-free LC/MS-based protein quantification method. Proteom. – Clin. Appl. 1 , 246–263 (2007). Gong, F. et al. Proteomic analysis of cisplatin resistance in human ovarian cancer using 2-DE method. Mol. Cell. Biochem. 348 , 141–147 (2011). Acland, M. et al. Chemoresistant Cancer Cell Lines Are Characterized by Migratory, Amino Acid Metabolism, Protein Catabolism and IFN1 Signalling Perturbations. Cancers 14 , 2763 (2022). Arentz, G. et al. Label-Free Quantification Mass Spectrometry Identifies Protein Markers of Chemotherapy Response in High-Grade Serous Ovarian Cancer. Cancers 15 , 2172 (2023). Ford, C. E., Werner, B., Hacker, N. F. & Warton, K. The untapped potential of ascites in ovarian cancer research and treatment. Br. J. Cancer . 123 , 9–16 (2020). Rakina, M., Kazakova, A., Villert, A., Kolomiets, L. & Larionova, I. Spheroid Formation and Peritoneal Metastasis in Ovarian Cancer: The Role of Stromal and Immune Components. Int. J. Mol. Sci. 23 , 6215 (2022). Acland, M. et al. Mass Spectrometry Analyses of Multicellular Tumor Spheroids. Proteom. – Clin. Appl. 12 , 1700124 (2018). Raghavan, S. et al. Comparative analysis of tumor spheroid generation techniques for differential in vitro drug toxicity. Oncotarget . 7 , 16948–16961 (2016). Maenhoudt, N. et al. Developing Organoids from Ovarian Cancer as Experimental and Preclinical Models. Stem Cell. Rep. 14 , 717–729 (2020). Gerton, T. J. et al. Development of a Patient-Derived 3D Immuno-Oncology Platform to Potentiate Immunotherapy Responses in Ascites-Derived Circulating Tumor Cells. Cancers 15 , 4128 (2023). Phan, N. et al. A simple high-throughput approach identifies actionable drug sensitivities in patient-derived tumor organoids. Commun. Biol. 2 (2019). Bi, J. et al. Successful Patient-Derived Organoid Culture of Gynecologic Cancers for Disease Modeling and Drug Sensitivity Testing. Cancers 13 , 2901 (2021). den Ouden, J. E. et al. Chemotherapy sensitivity testing on ovarian cancer cells isolated from malignant ascites. Oncotarget 11 , 4570–4581 (2020). de Witte, C. J. et al. Patient-Derived Ovarian Cancer Organoids Mimic Clinical Response and Exhibit Heterogeneous Inter- and Intrapatient Drug Responses. Cell. Rep. 31 (2020). Chen, W. et al. Effective Treatment for Recurrent Ovarian Cancer Guided by Drug Sensitivity from Ascites-Derived Organoid: A Case Report. Int. J. Womens Health . 15 , 1047–1057 (2023). Kerslake, R. et al. Transcriptional Landscape of 3D vs. 2D Ovarian Cancer Cell Models. Cancers 15 , 3350 (2023). Chen, W. et al. High-throughput Image Analysis of Tumor Spheroids: A User-friendly Software Application to Measure the Size of Spheroids Automatically and Accurately. J. Vis. Exp. 89 , 51639 (2014). Sherman, B. T. et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res. 50 , 216–221 (2022). Chautard, E., Fatoux-Ardore, M., Ballut, L., Thierry-Mieg, N. & Ricard-Blum, S. MatrixDB, the extracellular matrix interaction database. Nucleic Acids Res. 39 , 235–240 (2011). Metsalu, T. & Vilo, J. ClustVis: a web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap. Nucleic Acids Res. 43 , 566–570 (2015). Senkowski, W. et al. A platform for efficient establishment and drug-response profiling of high-grade serous ovarian cancer organoids. Dev. Cell. 58 , 1106–1121 (2023). Chen, H. Short-term organoid culture for drug sensitivity testing of high-grade serous carcinoma. Gynecol. Oncol. 157 , 783–792 (2020). Chen, L-Y. et al. In vitro drug testing using patient-derived ovarian cancer organoids. J. Ovarian Res. 17 , 194 (2024). Åkerlund, E. et al. The drug efficacy testing in 3D cultures platform identifies effective drugs for ovarian cancer patients. N P J. Precis Oncol. 7 (2023). Nasimi Shad, A. & Moghbeli, M. Integrins as the pivotal regulators of cisplatin response in tumor cells. Cell. Commun. Signal. 22 , 265 (2024). Tian, T. et al. β1 integrin-mediated multicellular resistance in hepatocellular carcinoma through activation of the FAK/Akt pathway. J. Int. Med. Res. 46 , 1311–1325 (2018). Sethi, T. et al. Extracellular matrix proteins protect small cell lung cancer cells against apoptosis: A mechanism for small cell lung cancer growth and drug resistance in vivo. Nat. Med. 5 , 662–668 (1999). Ngaokrajang, U., Janvilisri, T., Sae-Ueng, U., Prungsak, A. & Kiatwuthinon, P. Integrin α5 mediates intrinsic cisplatin resistance in three-dimensional nasopharyngeal carcinoma spheroids via the inhibition of phosphorylated ERK /caspase-3 induced apoptosis. Exp. Cell. Res. 406 , 112765 (2021). Kuermanbayi, S. et al. In situ monitoring of functional activity of extracellular matrix stiffness-dependent multidrug resistance protein 1 using scanning electrochemical microscopy. Chem. Sci. 13 , 10349–10360 (2022). Prakash, J. & Shaked, Y. The Interplay between Extracellular Matrix Remodeling and Cancer Therapeutics. Cancer Discov . 14 , 1375–1388 (2024). Guo, W. et al. High Serpin Family A Member 10 Expression Confers Platinum Sensitivity and Is Associated With Survival Benefit in High-Grade Serous Ovarian Cancer: Based on Quantitative Proteomic Analysis. Front. Oncol. 11 , 761960 (2021). Park, S. J. et al. SERPINB11 Expression Is Associated With Prognosis of High-grade Serous and Clear Cell Carcinoma of the Ovary. Vivo 35 , 2647–2653 (2021). Sharma, S., Santiskulvong, C., Rao, J., Gimzewski, J. K. & Dorigo, O. The role of Rho GTPase in cell stiffness and cisplatin resistance in ovarian cancer cells. Integr. Biol. 6 , 611–617 (2014). Abedini, M. R. et al. Cell fate regulation by gelsolin in human gynecologic cancers. Proc. Natl. Acad. Sci. U.S.A. 111, 14442–14447 (2014). Ricciardelli, C. et al. Keratin 5 overexpression is associated with serous ovarian cancer recurrence and chemotherapy resistance. Oncotarget 8 , 17819–17832 (2017). Győrffy, B. Discovery and ranking of the most robust prognostic biomarkers in serous ovarian cancer. GeroScience 45 , 1889–1898 (2023). Guerra, F., Arbini, A. A. & Moro, L. Mitochondria and cancer chemoresistance. Biochim. Biophys. Acta . 1858 , 686–699 (2017). Liu, J. R. et al. Dysfunctional Apoptosome Activation in Ovarian Cancer: Implications for Chemoresistance. Cancer Res. 62 , 924–931 (2002). Kim, S. et al. Evaluating Tumor Evolution via Genomic Profiling of Individual Tumor Spheroids in a Malignant Ascites. Sci. Rep. 8 , 12724 (2018). Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterials.docx Cite Share Download PDF Status: Published Journal Publication published 22 Dec, 2025 Read the published version in Journal of Proteome Research → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6441929","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":452826351,"identity":"3c0d60cb-25df-4e30-97d2-659a58da9c7f","order_by":0,"name":"Jack Scanlan","email":"","orcid":"","institution":"University of South Australia","correspondingAuthor":false,"prefix":"","firstName":"Jack","middleName":"","lastName":"Scanlan","suffix":""},{"id":452826352,"identity":"147ab48e-5555-4979-b15c-e753e4f75d5a","order_by":1,"name":"Parul Mittal","email":"","orcid":"","institution":"University of South Australia","correspondingAuthor":false,"prefix":"","firstName":"Parul","middleName":"","lastName":"Mittal","suffix":""},{"id":452826353,"identity":"9f2f8e65-a39d-4c9a-8415-6c39ca8c280a","order_by":2,"name":"Noor A Lokman","email":"","orcid":"","institution":"University of Adelaide","correspondingAuthor":false,"prefix":"","firstName":"Noor","middleName":"A","lastName":"Lokman","suffix":""},{"id":452826354,"identity":"e7c6505f-b026-49be-94e7-8fe11973ca24","order_by":3,"name":"Martin K Oehler","email":"","orcid":"","institution":"Royal Adelaide Hospital","correspondingAuthor":false,"prefix":"","firstName":"Martin","middleName":"K","lastName":"Oehler","suffix":""},{"id":452826355,"identity":"bc5f0335-50c6-415d-b37f-88f3b55de716","order_by":4,"name":"Peter Hoffmann","email":"","orcid":"","institution":"University of South Australia","correspondingAuthor":false,"prefix":"","firstName":"Peter","middleName":"","lastName":"Hoffmann","suffix":""},{"id":452826356,"identity":"c1b3596e-4439-41e7-9f8d-b72934f2c048","order_by":5,"name":"Manuela Klingler-Hoffmann","email":"data:image/png;base64,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","orcid":"","institution":"University of South Australia","correspondingAuthor":true,"prefix":"","firstName":"Manuela","middleName":"","lastName":"Klingler-Hoffmann","suffix":""}],"badges":[],"createdAt":"2025-04-14 02:53:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6441929/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6441929/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1021/acs.jproteome.5c00771","type":"published","date":"2025-12-23T00:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":82352505,"identity":"fd6e98f6-0d72-4073-a94a-9f4999a8fce5","added_by":"auto","created_at":"2025-05-09 11:01:46","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":332356,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDose responses of spheroids generated from ovarian cancer cell lines and cisplatin-resistant sublines against CBP. \u003c/strong\u003e(A) A2780 (n=84) and A2780-cis (n=118) spheroids were significantly larger than TYK-nu (n=83) and TYK-NU.CPr (n=137) spheroids 72 hours after seeding. Spheroids generated from (B) TYK-nu and (C) A2780 parental cell lines were 2- and 3.4-fold more sensitive to CBP than their cisplatin-resistant sublines (n=3). Two independent experiments were conducted\u003cem\u003e \u003c/em\u003efor each cell line. Values are mean ± SD; ****p\u0026lt;0.0001.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6441929/v1/b3d9e3be5bf3dfce2e26d697.png"},{"id":82352506,"identity":"a51f8598-37f5-409f-b728-9e15019e0ff3","added_by":"auto","created_at":"2025-05-09 11:01:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":888052,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEx vivo dose responses of patient-matched artificial and native ADS against CBP after treatment for 72 hours. \u003c/strong\u003e(A)\u003cstrong\u003e \u003c/strong\u003eRepresentative images of spheroids treated with 0 µM and 300 µM CBP. Images of native ADS from patient 4 were unavailable due to instrument downtime. Scale bars = 200 µm. Green and red shading indicates clinical chemosensitivity and chemoresistance, respectively, while grey signifies unknown clinical chemoresponse. (B) Dose-response curves from artificial and native ADS. Values are mean ± SD (n=3).\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6441929/v1/a446020d08a5274905f0d1f9.png"},{"id":82356311,"identity":"87174382-bab9-4c41-b933-3914829e3dac","added_by":"auto","created_at":"2025-05-09 11:17:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":376529,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferentially abundant proteins and representative pathways in artificial and native ADS. \u003c/strong\u003eProteins with significantly different abundances are represented in (A) a volcano plot and (B) a Venn diagram. (C) PCA separated matched native and artificial ADS samples along the first component and sensitive and resistant native ADS samples on the second component. A=artificial ADS (blue); N=native ADS (green). (D) The five most significant GO terms from lists of differentially abundant proteins between artificial and native ADS using the DAVID bioinformatics tool.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6441929/v1/5039ca1e750620ba8622b7e4.png"},{"id":82354268,"identity":"119876fe-dffa-4604-b79f-0de30c34fb5c","added_by":"auto","created_at":"2025-05-09 11:09:46","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":452328,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferentially abundant proteins and representative pathways in sensitive and resistant native ADS. \u003c/strong\u003eThe differentially abundant proteins are illustrated in (A) a volcano plot and (B) a Venn diagram. (C) PCA analysis distinguishes between CBP-sensitive and -resistant native ADS samples along the first component. The 'N' suffix indicates native ADS. (D) The five most significant GO terms from lists of differentially abundant proteins between sensitive and resistant native ADS using the DAVID bioinformatics tool.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6441929/v1/932822d4ffc9aa23fdd9efe6.png"},{"id":99551293,"identity":"d687a65f-7ddb-4dc6-beb3-1e95e9d585f5","added_by":"auto","created_at":"2026-01-05 17:13:28","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3081235,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6441929/v1/f67d8340-dddd-4870-b0d1-a9c7aa23c3f5.pdf"},{"id":82352509,"identity":"b6cf8be5-b76e-42f3-a8f1-9247c3908cd5","added_by":"auto","created_at":"2025-05-09 11:01:46","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":194452,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-6441929/v1/73c753116e020ec9a5113c73.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The proteomes of ovarian cancer ascitic cellular aggregates correlate with their ex vivo platinum sensitivities: a pilot study","fulltext":[{"header":"Background","content":"\u003cp\u003eEpithelial ovarian cancer (EOC) is the most lethal gynaecological malignancy, with high-grade serous ovarian cancer (HGSOC) constituting three-quarters of all cases\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Patients typically receive standard first-line chemotherapy that consists of carboplatin (CBP) and paclitaxel (PTX) following cytoreductive surgery, with neo-adjuvant chemotherapy sometimes administered to pre-emptively reduce tumour burden\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. CBP is one of several approved organoplatinum-based alkylating agents that inhibit DNA replication through cytotoxic inter-strand DNA cross-linkages\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. PTX prevents cell division by stabilising microtubules, promoting G2/M cell cycle arrest, and inhibiting the repair of CBP-mediated DNA adducts\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Recent advancements in the treatment of stage III and IV HGSOC include the approval of bevacizumab\u0026mdash;a vascular endothelial growth factor (VEGF) inhibitor that inhibits tumour growth by blocking nutrient delivery\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e\u0026mdash; for administration alongside CBP and PTX. Pharmaceutical inhibition of poly (ADP-ribose) polymerase (PARP) was also recently approved as an additional first-line treatment in patients with \u003cem\u003eBRCA1/2\u003c/em\u003e mutations or homologous recombination deficiency (HRD)\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Although approximately 80 per cent of patients respond to first-line treatment, most develop resistance and require second-line chemotherapeutics.\u003c/p\u003e \u003cp\u003eThe absence of a clinical tool or marker to predict patients\u0026rsquo; responses to first-line treatment promotes a trial-and-error-based approach. As this is an area of unmet clinical need, numerous genomics-based studies have been conducted to identify markers and mechanisms of chemoresponsiveness. These most often use immortalised 2D cell lines and have identified mutations in genes that encode DNA damage repair\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, cell cycling\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, drug efflux pathways\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, and extracellular matrix (ECM) proteins\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. However, the lack of validation of these markers in large patient cohorts means that none of these are used in clinical practice. Proteomic studies have also identified several proteins that correlate with chemoresponse. An early study of over 2000 proteins in the A2780 and 2008 cell lines and their cisplatin-resistant counterparts reported differential abundances of DNA repair, oxidative stress responses, apoptosis, translation, and cell cycling proteins\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. A gel- and LC-MS/MS-based analysis of cisplatin-sensitive and resistant A2780 cell lines also identified ten differentially expressed proteins involved in ubiquitination, metabolism, signal transduction, and calcium and nucleotide binding\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Recently, our laboratory demonstrated a clear separation of the global proteomes of parental and cisplatin-resistant EOC cell lines using LC-MS/MS\u003csup\u003e13\u003c/sup\u003e. While protein abundances have also been linked with patient survival and treatment response\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, investigations of molecular markers are often complicated by poor accessibility and availability of primary tumour samples.\u003c/p\u003e \u003cp\u003eMalignant ascites (MA) describes the accumulation of peritoneal fluid that contains tumour cells and is drained during cytoreductive surgery or interval paracentesis to alleviate discomfort. It results from increased capillary permeability due to VEGF-mediated downregulation of the claudin 5 (CLDN5) tight junction protein and poor lymphatic drainage caused by tumour cell blockages\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Resident tumour cells, cancer-associated fibroblasts (CAFs), and tumour-associated macrophages (TAMs) can aggregate as heterogeneous spheroids\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, herein referred to as native ascites-derived spheroids (ADS). Native ADS are proposed to aggregate in response to the upregulation of integrin subunit alpha 5 (ITGA5) and epidermal growth factor promoted by CAFs and TAMs, and contain quiescent and necrotic cores that are not effectively eliminated during treatment\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Their unique cell-cell interactions have been shown to promote treatment resistance and metastasis\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, prompting the development of novel 3D pre-clinical EOC models. For example, spheroids can be generated from immortalised cell lines in hanging drop, ultra-low attachment (ULA), and nutator plates\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Artificial ADS have also been formed from tissues and demonstrate epithelial-to-mesenchymal transition and the expression of HGSOC markers such as paired box gene 8 (PAX8)\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Finally, retrospective functional drug assays on 3D models derived from dissociated primary tumours or ascites cell cultures have recapitulated patients\u0026rsquo; clinical responses to platinum-based chemotherapy\u003csup\u003e\u003cspan additionalcitationids=\"CR21 CR22\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, PTX\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, PARP inhibitors\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, monoclonal antibody treatments\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, and kinase inhibitors\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, which has successfully guided treatment selection in some instances\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. However, artificial ADS often suffer poor generation rates, long lead times that reduce clinical actionability, and uncertainty regarding their representation of ovarian cancer biology. For instance, few differentially expressed genes were observed between HGSOC cells grown as spheroids using basement membrane extract (BME), collagen, and agarose supports from the same OVCAR8 cell lines\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Furthermore, no known studies have assessed protein-based markers of chemotherapy response in artificial ADS.\u003c/p\u003e \u003cp\u003eHere, we establish a workflow to test the \u003cem\u003eex vivo\u003c/em\u003e drug responses of patient-derived cells and analyse their proteome in a timely manner. We first compare the treatment responses of artificial ADS formed in ULA plates from ascites cells grown in monolayer with chemotherapy-na\u0026iuml;ve patient-matched native ADS to demonstrate that the latter more accurately and rapidly recapitulate patients\u0026rsquo; clinical treatment responses. We also show that the differential abundances of proteins and gene ontology (GO) pathways in native ADS give insights into potential molecular markers of treatment response. Only the global proteomes of native ADS cluster according to treatment response, which provides a path towards prospective treatment response prediction for all diseases that present with MA.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\"\u003e\n \u003ch2\u003eEffects of chemotherapy dose-response on ovarian cancer spheroids\u003c/h2\u003e\n \u003cp\u003e\u003cem\u003eEx vivo\u003c/em\u003e CBP response testing was established by generating artificial spheroids from the parental TYK-nu and A2780 cell lines and their cisplatin-resistant TYK-nu.CPr and A2780-cis sublines. A2780 and A2780-cis cell lines formed significantly larger spheroids (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) than the TYKNU and TYK-nu.CPr cell lines (Fig.\u0026nbsp;1A), which aligned with their respective cell viabilities (Supplementary Figure S1). While TYK-nu.CPr spheroids were significantly larger (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) than TYK-nu spheroids, the opposite was observed for A2780 and A2780-cis cell lines. The EC\u003csub\u003e50\u003c/sub\u003e values of cisplatin-resistant TYK-nu.CPr and A2780-cis spheroids were 2 and 3.4 times higher than the parental cell lines, respectively (Fig.\u0026nbsp;1B-C).\u003c/p\u003e\n \u003cp\u003eA proteomics analysis of all four cell lines grown as adherent and spheroidal cultures revealed a clear separation of TYK-nu/TYK-nu.CPr and A2780/A2780-cis cell lines along the first component of the PCA plot, while samples separated along the second component based on growth dimensionality (Supplementary Figure S2).\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eEffects of chemotherapy dose response on artificial ADS\u003c/h3\u003e\n\u003cp\u003eWe postulated that the generation of similarly-sized spheroids from primary ascites cells would provide a consistent measure of chemotherapy response. Although artificial ADS from the same patient were consistently sized, those from different patients varied significantly (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) without correlation to clinical response (Supplementary Figure S3). Artificial ADS from all patients had relatively high EC\u003csub\u003e50\u003c/sub\u003e values (greater than 140 \u0026micro;M), which did not correlate with clinical outcomes (Fig.\u0026nbsp;2). Artificial ADS generally demonstrated clear and contained borders, with the exception of those from patient 4, which appeared larger due to surrounding non-spheroidal matter. Native ADS originating from platinum-sensitive patients 1 and 2 were sensitive in \u003cem\u003eex vivo\u003c/em\u003e experiments, with respective EC\u003csub\u003e50\u003c/sub\u003e values of 14 and 71 \u0026micro;M. Patient 3 was clinically resistant but was most sensitive to CBP in \u003cem\u003eex vivo\u003c/em\u003e assessments with an EC\u003csub\u003e50\u003c/sub\u003e of 8 \u0026micro;M. The \u003cem\u003eex vivo\u003c/em\u003e EC\u003csub\u003e50\u003c/sub\u003e of 141 \u0026micro;M for the native ADS of patient 4 matched with their clinical resistance status. While patient 5 had an unknown clinical sensitivity, their native ADS were categorised as resistant due to an \u003cem\u003eex vivo\u003c/em\u003e EC\u003csub\u003e50\u003c/sub\u003e of 160 \u0026micro;M.\u003c/p\u003e\n\u003ch3\u003eNative and artificial ADS have distinct proteomes\u003c/h3\u003e\n\u003cp\u003eLC-MS/MS analysis of all native and artificial ADS samples identified 6,664 proteins overall (Fig.\u0026nbsp;3A), with 570 proteins being more abundant in artificial ADS and 928 being more abundant in native ADS (Fig.\u0026nbsp;3B). These differences were represented in the PCA plot, which clearly separated native and artificial ADS on the first component with negligible separation of artificial ADS (Fig.\u0026nbsp;3C). Native ADS separated along the second component based on \u003cem\u003eex vivo\u003c/em\u003e CBP sensitivity, with patients 1, 2, and 3 considered sensitive and 4 and 5 considered resistant. Figure\u0026nbsp;3D shows that the most significantly enriched GO terms in native ADS primarily pertained to DNA replication, such as nucleosomal DNA and RNA binding, and the organisation of chromatin, telomeres, and nucleosomes. The nucleus and cytosol hosted the most enriched CCs, except for the extracellular exosome, which was the most significantly enriched compartment in both native and artificial ADS. The most-enriched BPs in artificial ADS included the binding of collagens, proteases, integrins, and calcium ions, and ECM structural constituents. Relevant MFs included collagen fibril organisation and cell adhesion. Highly enriched CCs in artificial ADS included extracellular regions, such as the ECM, as well as the lumina of lysosomes and the endoplasmic reticulum.\u003c/p\u003e\n\u003ch3\u003eNative ADS proteomes cluster according to ex vivo treatment responses\u003c/h3\u003e\n\u003cp\u003eThe greater separation of native ADS proteomes compared to artificial ADS proteomes (Fig.\u0026nbsp;3C) prompted an analysis of native ADS proteomes alone. Of the 6303 proteins identified (Fig.\u0026nbsp;4A), 700 were more abundant in resistant ADS and 496 were more abundant in sensitive ADS (Fig.\u0026nbsp;4B). The native ADS of patients 1\u0026ndash;3 and 4 \u0026amp; 5 are separated into distinct clusters along the first component of the PCA plot (Fig.\u0026nbsp;4C), which reflects their categorisation of sensitivity and resistance in \u003cem\u003eex vivo\u003c/em\u003e CBP response assessments, respectively.\u003c/p\u003e\n\u003cp\u003eSome of the most significantly enriched GO terms in resistant ADS were proteins at the interplay of cell migration, cytoskeletal organisation, and focal adhesions (Fig.\u0026nbsp;4D). At the individual protein level, we observed elevated abundances of several integrins in resistant native ADS, including ITGAX, ITGA2, ITGB4, ITGA6, ITGAM, ITGA, ITGB1, and ITGB2, which functionally link the cytoskeleton to the ECM. This was complemented by additional adhesion-related proteins such as annexin (ANX) A2, cadherins (CDH) 1 and 2, and the cell surface receptor CD44 being up in resistant native ADS, along with actin filament-associated protein 1-like 2 (AFAP1L2) and actins ACTB1 and ACTG1.\u003c/p\u003e\n\u003cp\u003eThe differential abundance of many ECM proteins prompted an analysis of those within the MatrixDB database.\u003csup\u003e29\u003c/sup\u003e Of all proteins in this database, 188 were differentially abundant between sensitive and resistant native ADS (Supplementary Table S1), including members of the serine protease inhibitor (SERPIN) superfamily, which were some of the most differentially abundant in this list. For example, SERPINs A1, A3, and H1 were significantly more abundant in sensitive samples and SERPINs B1, B2, B5, B6, B8, and B9 in resistant native ADS. Carcinoembryonic antigen-related cell adhesion molecules (CEACAM) 5\u0026ndash;7 were also highly abundant in resistant native ADS. Relatedly, BPs relating to cytoskeletal organisation, such as actin filament binding, were most significantly enriched in resistant native ADS. Many intermediate filament-forming keratins had significantly higher abundances in resistant native ADS, with the gene expression of several being significantly associated with progression-free survival (PFS) and/or overall survival (OS) in stage III and IV HGSOC (Supplementary Table S2).\u003c/p\u003e\n\u003cp\u003eGO terms that involve protein synthesis, including DNA unwinding, ribosomal biogenesis, and rRNA processing were enriched in sensitive native ADS. The four most significantly-enriched CCs were within mitochondria and mitochondrial translation was the most significantly-enriched BP. At the individual protein level, several large ribosomal subunit proteins, tRNA ligases, and ATP synthase subunits saw high fold changes. The enrichment of the ficolin-1-rich granule lumen cell compartment in resistant native ADS indicates an immune component, which was complemented by higher abundances of the branched-chain amino acid transaminase (BCAT) 1 and 2 immune-related proteins and proteoglycan 2 (PRG2)\u0026mdash; the latter of which had the highest fold-increase compared to sensitive native ADS. In sensitive native ADS, immunoglobulin kappa variable cluster (IGKV) 3\u0026ndash;20, IGKV3-7, IGKV2-28, and IGHV3OR16-9 constituted some of the most upregulated proteins, alongside the THY1 membrane glycoprotein, which had the highest fold change of all proteins.\u003c/p\u003e\n\u003ch3\u003eEffects of additional treatments on native ADS\u003c/h3\u003e\n\u003cp\u003eGiven the availability of several first and second-line treatments, a future personalised medicine strategy that can predict patients\u0026rsquo; responses to an expanded range of chemotherapy drugs would be valuable. The \u003cem\u003eex vivo\u003c/em\u003e responses of native ADS from a subset of four patients were assessed against PTX, olaparib, gemcitabine, doxorubicin, and topotecan (Table 1). Patient 2 in this cohort had a predictive \u003cem\u003eBRCA2\u003c/em\u003e mutation for olaparib senstivity, but had the second-highest \u003cem\u003eex vivo\u003c/em\u003e resistance. Sensitivity to PTX showed a distinct pattern, with patients 2 and 3 being far more sensitive than patients 1 and 4. Native ADS also showed differential chemoresponses to doxorubicin, topotecan, and gemcitabine.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1: Chemoresponses of native ADS to second-line chemotherapy agents.\u0026nbsp;\u003c/strong\u003eGreen and red shading indicate predicted sensitivity and resistance to the PARP inhibitor olaparib based on genomic testing. Patient 2 had a predictive \u003cem\u003eBRCA2\u003c/em\u003e mutation.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img1746727766.png\"\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe poor survival outcomes for EOC patients necessitate a new model of care. Given the number of available chemotherapeutics and combinations thereof, an \u003cem\u003eex vivo\u003c/em\u003e disease model that can rapidly recapitulate patient sensitivities is sought to enable personalised medicine approaches that could guide clinicians in choosing the most effective treatment for each patient.\u003c/p\u003e \u003cp\u003eAlthough 2D \u003cem\u003eex vivo\u003c/em\u003e models that use primary cells have been reported extensively\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, the roles of ascitic multicellular spheroids in EOC treatment resistance and metastatic processes have prompted a shift to 3D models that can be generated by culturing cells in BME\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan additionalcitationids=\"CR32\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, ULA plates\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, or microfluidics devices\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Importantly, spheroids generated from primary tumour material can show differential responses to EOC treatments\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. They also share genomic and phenotypic features with their tumours of origin, such as PAX8 and P53 expression\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Their responses to multiple first- and second-line chemotherapeutics correlate with clinical outcomes and have even been used to guide treatment in a limited number of prospective studies\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eImportantly, studies have incorporated primary ascites cells, which are thought to be key drivers of treatment resistance and metastasis\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. For example, ascitic spheroids show differential \u003cem\u003eex vivo\u003c/em\u003e responses to diverse chemotherapies and can correlate with patients\u0026rsquo; clinical data\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. While functional response assessments can assist in therapy decisions, the collection of proteomic data will expand our understanding of chemoresponse and may eventually permit chemoresponse prediction based solely on molecular data. In this pilot study, we established a workflow to assess the \u003cem\u003eex vivo\u003c/em\u003e dose responses of a novel 3D EOC model that does not require a spheroid generation step, which allows \u003cem\u003eex vivo\u003c/em\u003e chemoresponse and proteomics data to be obtained within 4 days of ascites collection. This combined approach bridges the gap between costly \u003cem\u003eex vivo\u003c/em\u003e testing and clinical application, which could facilitate the eventual creation of a proteomic knowledge base with which to guide treatment.\u003c/p\u003e \u003cp\u003eA workflow to measure \u003cem\u003eex vivo\u003c/em\u003e chemoresponse was established with cisplatin-sensitive and -resistant ovarian cancer TYK-nu and A2780 cell lines. Although neither ATP concentrations nor the relationship between spheroid size and chemoresponse were consistent across the cell lines, this approach was sufficiently robust to monitor viability between treated and untreated cells. Together with the reproducibility of sample preparation for proteomic analyses, this gave confidence to use this approach for patient samples.\u003c/p\u003e \u003cp\u003eArtificial ADS could be consistently generated from five of six patient ascites samples, which aligns with the success rates of previous studies. Interestingly, all artificial ADS required high CBP concentrations to induce cell death regardless of patient origin or spheroid size, indicating chemoresistance. The lack of separation between artificial ADS proteomes in the PCA plot further demonstrated their inability to distinguish patients based on chemoresponse. Given these factors, native ADS were used in all further experiments. Given that spheroid generation techniques can alter the transcriptome\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, a different result may be possible using other generation methods.\u003c/p\u003e \u003cp\u003eNative ADS were isolated from five of the six ascites samples. In four of the five analysed native ADS samples, the relative chemoresponses were consistent with the clinical data. The exception was native ADS isolated from patient 3, which had the lowest \u003cem\u003eex vivo\u003c/em\u003e EC\u003csub\u003e50\u003c/sub\u003e value against CBP of any sample despite being considered resistant to CBP\u0026thinsp;+\u0026thinsp;PTX treatment in the clinic. This discrepancy could arise from a variety of factors, including the selection of a subset of cells used in the assay or the particular clinical response indicator used in this study. For instance, de Witte et al.\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e demonstrated that the \u003cem\u003eex vivo\u003c/em\u003e responses of EOC organoids to standard doublet chemotherapy were significantly correlated with patient-matched histopathological, biochemical, and radiological data but not with the presently-used progression-free survival.\u003c/p\u003e \u003cp\u003eThe ability of proteomics to reflect dynamic cellular processes was observed in these results. For example, the observed pattern of N-cadherin upregulation and E-cadherin downregulation in artificial ADS compared to native ADS is known to trigger native ADS formation\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Despite this, the poor separation of artificial ADS proteomes prompted an investigation into the predictive value of native ADS. The clustering of native ADS proteomes according to \u003cem\u003eex vivo\u003c/em\u003e treatment response supports the utility of this approach in the context of crude clinical classifications and the notion that proteomics could be used to predict sensitivity in future prospective studies.\u003c/p\u003e \u003cp\u003eAt the individual protein level, pathways relating to the interplay between the ECM and actin cytoskeleton were of particular interest, given their known roles in cancer progression. Notably, members of all integrin classes, which act as an interface for the ECM and cytoskeleton, were elevated in resistant native ADS. This included leukocyte-specific (ITGAX, ITGAM, and ITGB2), collagen-binding (ITGA2), laminin-binding (ITGB4 and ITGA6), and RGD-binding (ITGA5) integrins, along with ITGB1. Integrin-mediated activation of focal adhesion kinase (FAK) has been reported to promote platinum resistance through the inhibition of apoptotic cell death by triggering Src kinase-mediated upregulation of MAPK/ERK and PI3K/Akt/mTOR signalling pathways\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. For example, β1 integrin-mediated activation of the FAK/Akt pathway has been identified as a mechanism of resistance in spheroids generated from immortalised hepatocellular carcinoma cell lines\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. The upregulation of FAK and steroid receptor coactivator (SRC) kinase family proteins in resistant ADS, such as SRC and cortactin, points to a potential role for this pathway in ascitic ADS. The stimulation of tyrosine kinases through ECM-mediated ITGB1 activation has also been shown to suppress apoptosis in small-cell lung cancer\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Finally, the overexpression of ITGA5 has been shown to mediate cisplatin resistance in spheroids derived from immortalised nasopharyngeal cancer-derived spheroids through the inactivation of caspase-3-mediated apoptosis\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, further underscoring the likely importance of this receptor class in chemoresistance.\u003c/p\u003e \u003cp\u003eCollagens are another constituent of the ECM that have been postulated to affect chemoresistance through cell adhesion-mediated drug resistance. For example, the upregulation of several collagens has been observed in cisplatin-resistant ovarian cancer models\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, which elevates increases ECM stiffness to potentially promote drug efflux through altered multi-drug resistance 1 (MDR1) protein activity\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Another reported mechanism of chemoresistance in the context of collagen-mediated ECM stiffness is the prevented entry of chemotherapeutics and immune cells\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. The elevated abundance of all identified collagens (COL1A1, 1A2, 4A2, 6A1, 6A2, and 6A3) in resistant native ADS indicates their potential role as a marker of chemoresponse in this model.\u003c/p\u003e \u003cp\u003eThe noted roles of ECM remodelling in chemorepsonse and the observed enrichment of extracellular GO terms in resistant native ADS prompted closer analysis of all ECM proteins using the MatrixDB database\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. Several proteins from the SERPIN superfamily were some of the most differentially abundant proteins in native ADS. Those within the SERPIN \u0026lsquo;A\u0026rsquo; clade are extracellular and primarily involved in pro-inflammatory processes and hormone transport. For example, SERPINA10 expression has been positively associated with platinum sensitivity in the TCGA ovarian cancer cohort, which was validated using immunohistochemistry on HGSOC tissues\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. The observation of elevated abundances of SERPIN A1, A3, and H1 in sensitive native ADS supported this. \u0026lsquo;B\u0026rsquo; SERPINs are instead intracellular and mainly inhibit apoptotic cell death through the inhibition of regulatory proteins. SERPINs B1, B2, B5, B6, B8, and B9 were all elevated in resistant native ADS, which supports previous associations with poor prognosis in ovarian cancers\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Notably, the paralogous SERPINB3 and B4 isoforms were not significantly different and not detected, respectively.\u003c/p\u003e \u003cp\u003eActin cytoskeleton remodelling is a hallmark of cancer and has also been observed to promote platinum chemoresponse through increased intracellular stiffness\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, of which rho GTPases are key activators. Rho GTPase activators were significantly more abundant in resistant native ADS, including rho GTPase-activating proteins (ARHGAP) 12, 18, 27, and 45. The abundances of other actin cytoskeleton regulators also differed between sensitive and resistant native ADS. For example, gelsolin (GSN) is an actin filament-binding protein that mediates cytoskeletal remodelling and inhibits apoptosis, which was over 2.5-fold more abundant in resistant native ADS. This supports previous observations from our laboratory of differential GSN expression in cell lines and HGSOC tumour tissues based on platinum resistance\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, and associations to poor OS and PFS in serous ovarian cancer patients\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Another critical cytoskeletal component is the keratin superfamily. Keratins form intermediate filaments within epithelial cells and have numerous structural and signalling functions, such as protection against chemotherapeutics and other cellular stressors. The abundance of KRT5 has been previously associated with poor outcomes at both the mRNA and protein levels\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e in stage III and IV serous ovarian cancers. Using the Kaplan-Meier (KM) plotter\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, significantly higher expression of KRT6C and KRT85 were individually correlated with reduced PFS and OS, reflecting their increased abundances in resistant native ADS.\u003c/p\u003e \u003cp\u003eAlthough immune proteins may have roles in chemoresponse due to the presence of immune cells in ascitic aggregates, their differential expression may result from the differential presence of cell-surface glycans. Further investigations of constituent immune cells of native ADS and cell surface glycans could clarify this matter.\u003c/p\u003e \u003cp\u003eFinally, mitochondrial proteins have been implicated in the regulation of chemoresponse due to their roles as key regulators of critical intracellular processes such as metabolism, oxidative stress responses, and cell death. Mitochondrial involvement in platinum resistance remains relatively under-researched, despite frequent mitochondrial DNA (mtDNA) mutations and mtDNA depletion\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e that can promote the dysregulation of resistance pathways in cancer. The enrichment of translation, ATP synthesis, ribosomal biogenesis, and nucleic acid replication in sensitive ADS could represent a decreased ability for platinum-resistant ascitic spheroids to respond to pathogenic stimuli. This is evidenced by the decrease in several proteins that are protective against oxidative stress in resistant ADS, including glutathione peroxidases 3 and 7. Proteomics also identified the downregulation of several pro-apoptotic proteins in resistant ADS, including programmed cell death 2-like (PDCD) 2L, 5, 4, and apoptosis-inducing factor 1 (AIFM1). Interestingly, apoptotic peptidase activating factor 1 (APAF1) was upregulated in resistant ADS, despite a previous study using primary ovarian tumours which indicated that APAF-1 levels do not necessarily correlate with caspase-9 activation\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. These findings underline the need for more research into the role of apoptotic proteins in treatment resistance.\u003c/p\u003e \u003cp\u003eThe use of native ADS as a 3D model to predict treatment response and collect molecular information more accurately and rapidly is limited by several factors. Firstly, Kim et al. debate the clonal relationship between the primary tumour and ADS\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, suggesting early evolutionary divergence based on significant differences in copy number and single-nucleotide variants between ADS and patient-matched tumour tissues. However, this may be beneficial given the unique role of ascitic spheroids in chemoresistance and disease progression. Some EOC treatments would be incompatible with \u003cem\u003eex vivo\u003c/em\u003e response assessments, such as the VEGF inhibitor bevacizumab, which prevents micro-vascularisation that is not present in spheroids.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e Nevertheless, expanding this approach to common second- and third-line treatments, as well as combinations of chemotherapy such as CBP\u0026thinsp;+\u0026thinsp;PTX, would be helpful for those without a predictive tool. Predicting PARP inhibitor response using a proteomic tool would also be welcome, given that the genomic tool based on BRCA1/2 mutation and HRD statuses often suffers from long turnaround times.\u003c/p\u003e \u003cp\u003eThis pilot study is the first to demonstrate that the global proteomes of native ADS cluster according to their \u003cem\u003eex vivo\u003c/em\u003e platinum responses. We show that these results are concordant with clinical data in most cases and that this model differentially responds to an expanded panel of EOC treatments. These findings suggest the potential utility of a spheroid-based personalised medicine approach for second- and third-line chemotherapeutics. The recruitment of further patients would allow for the development of a global proteomic strategy that may rely solely on PCA clustering to predict treatment responses and eventually bypass the need for individual molecular markers. This is especially useful given the use of a DIA proteomics approach, which could enable the generation of a public database against which clinicians could compare prospective samples. Nevertheless, the power of MS-based proteomics to identify changes in the abundance of individual proteins between sensitive and resistant native ADS identified ECM, cytoskeleton, mitochondria, and immune proteins and pathways. Given the prevalence of MA in other malignancies, including breast, liver, gastric, and colorectal cancers, the utility of this model could have far-reaching implications for personalised cancer treatments.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eAscites sample collection and processing\u003c/h2\u003e \u003cp\u003eAscites samples were collected between 2019 and 2024 from six HGSOC patients during cytoreductive surgery at the Department of Gynaecological Oncology (Royal Adelaide Hospital, Adelaide, Australia) with written informed consent and approval by the hospital ethics committee (approval #R20181215) prior to storage in liquid nitrogen. Clinical platinum sensitivity was defined by the absence of progression six months after chemotherapy cessation, while relapse or no response within six months indicated clinical platinum resistance. Thawed ascites samples were cultured overnight in Advanced RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) (Bovogen, Vic, Australia), 1% penicillin-streptomycin, and 1% L-glutamine at 37⁰C with 5% CO\u003csub\u003e2\u003c/sub\u003e. Cells were detected by brightfield microscopy in five of six patient samples. Native ADS remained suspended in the medium and were isolated by gentle centrifugation for CBP response assessments or stored for proteomic analysis. Adherent cells were expanded for a maximum of 3 passages for artificial ADS generation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eCell lines\u003c/h2\u003e \u003cp\u003eMycoplasma-free and authenticated A2780 (RRID:CVCL_0134), A2780-cis (RRID:CVCL_1942), TYK-nu (RRID:CVCL_1776), and TYK-nu.CPr (RRID:CVCL_3221) cell lines (Cell Bank Australia, Westmead, Australia) were cultured at 37⁰C with 5% CO\u003csub\u003e2\u003c/sub\u003e in Eagle\u0026rsquo;s Minimum Essential Medium (EMEM; Merck, NSW, Australia) that was supplemented with 10% FBS, 1% penicillin-streptomycin, and 1% L-glutamine. Cell lines were subcultured for a maximum of 10 passages before being trypsinised (Sigma Aldrich, Vic, Australia) for proteomic analyses or dilution and seeding in ULA plates spheroid generation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eArtificial spheroids generation\u003c/h2\u003e \u003cp\u003eArtificial spheroids were generated over 72 hours from cell lines or adherent ascites cells by seeding 2 x 10\u003csup\u003e3\u003c/sup\u003e cells in 90 \u0026micro;L of appropriate medium within each well of 96-well Nunclon Sphera ULA plates (Thermo Fisher Scientific, Vic, Australia). Spheroids were imaged using an IN Cell Analyzer 2200 (Cytiva, MA, United States of America) with a 10x brightfield objective and sized using the semi-automated MATLAB-based SpheroidSizer\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e program. Artificial spheroids were collected for proteomic analyses using a manual pipette and gentle centrifugation. Artificial ADS generated from the ascites cells of patient 4 were only used in dose-response testing and were excluded from sizing due to the abundance of interfering non-spheroidal matter.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eEx vivo chemoresponse assessments\u003c/h2\u003e \u003cp\u003eArtificial and native ADS collected directly from MA were treated with CBP (Hospira, NSW, Australia) at concentrations of 300, 250, 200, 150, 100, 75, 50, 25, 12.5, 6.3, 1.6, and 0.8 \u0026micro;M and compared to a PBS control. Native ADS were treated with olaparib at the same concentrations using a 1:1 PBS:DMSO control. Native ADS were also treated with gemcitabine, topotecan, doxorubicin, and PTX ( at 10, 7.5, 3.16, 1, 0.316, 0.01, 0.0316, 0.01, 0.00316, and 0.001 \u0026micro;M, in separate experiments with a 1:1 PBS:DMSO control. All assessments were conducted in triplicate in total volume of 100 \u0026micro;L for 72 hours at 37\u003csup\u003eo\u003c/sup\u003e C with 5% CO\u003csub\u003e2\u003c/sub\u003e. ATP concentration was used to indicate viability by incubating samples with an equal volume of CellTitre-Glo 3D (Promega, WI, United States of America) at room temperature for 5 mins at 300 rpm followed by 25 mins without rotation. During incubation, spheroids were transferred to white-coated 96-well plates for luminescence assessment using the VICTOR Nivo plate reader (Revvity, MA, United States of America). Half maximal dose-response values (EC\u003csub\u003e50\u003c/sub\u003e) were calculated using GraphPad Prism (v10.1.2) using a four-parameter logistical regression. Patients were designated CBP-sensitive or -resistant based on EC\u003csub\u003e50\u003c/sub\u003e values against CBP. Values over the maximum dose (\u003cem\u003ex\u003c/em\u003e) were reported as \u0026lsquo;\u0026gt; \u003cem\u003ex\u003c/em\u003e \u0026micro;M\u0026rsquo;.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSample preparation for proteomic analyses\u003c/h2\u003e \u003cp\u003eSample preparation was performed using the S-Trap\u0026trade; Micro protocol (PROTIFI, NY, United States of America) with minor modifications. Briefly, cell pellets were sonicated for 15 mins in 5% SDS and 50 mM Tris at pH 8.5. Proteins were reduced with 10 mM dithiothreitol in the dark for 1 hour, and alkylated with 15 mM chloroacetamide for 30 mins in the dark. Protein concentration was determined using the Pierce\u0026trade; BCA Protein Assay Kit (Thermo Fisher Scientific, Vic, Australia). Approximately 50\u0026ndash;100 \u0026micro;g of protein was loaded onto the S-Trap\u0026trade; Micro column and digested overnight at 37\u003csup\u003eo\u003c/sup\u003e C with 5 \u0026micro;g of a Trypsin/LysC mix (Promega, WI, United States of America) in 50 mM ammonium bicarbonate. The resulting peptides were eluted, dried under vacuum, resuspended in 0.1% formic acid (FA), and quantified using the Nanodrop One/One\u003csup\u003ec\u003c/sup\u003e (Thermo Fisher Scientific, Vic, Australia). The peptide solutions were transferred to high-performance liquid chromatography vials for further analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eMass spectrometry data acquisition\u003c/h2\u003e \u003cp\u003eApproximately 200 ng of peptides from each sample were injected in duplicate into the UltiMate 3000 nanoLC (Thermo Fisher Scientific, Vic, Australia) that was coupled online to a timsTOF fleX (Bruker Daltonics, MA, United States of America) with a nano-electrospray Captive Spray ion source and 20 \u0026micro;m Classic Emitter with the capillary voltage set to 1500 v. Instruments were controlled with Bruker Compass 4.1 (v6.2, build 1.2) and Compass HyStar 6.3 (v6.3.1.8). A 150 \u0026micro;m ID \u0026times; 150 mm (1.5 \u0026micro;m, 100 \u0026Aring; Reprosil Saphir C18) reversed-phase column (PremierLCMS, CA, United States of America) was used for chromatographic separation at 1 \u0026micro;L/min at 40\u003csup\u003eo\u003c/sup\u003e C. Mobile phases A and B were 0.1% FA in water and 0.1% FA in 80% acetonitrile, respectively, with a linear gradient of phase B from 3.8\u0026ndash;25% over 45 mins, then 87.3% in 3 mins, which was held for 5 mins for a total run time of 70 mins. Data were acquired in the diaPASEF mode with a mass scan range of 100\u0026ndash;1700 \u003cem\u003em/z\u003c/em\u003e, an ion mobility window of 1/K0 0.6\u0026ndash;1.6, ramp time of 100 ms, and 100% duty cycle. Accumulation time was 2 ms with 3 L/min dry gas flow at 180\u003csup\u003eo\u003c/sup\u003e C. Additional settings included deflection 1 Δ of 70.0 V, funnel 1 RF of 300.0 Vpp, funnel 2 RF of 200.0 Vpp, and CID energy of 0.0 eV. Ion mobility parameters were set to ∆t6 of 55.0 V, funnel 1 RF of 450.0 Vpp, and a collision cell voltage of 300 V.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eAnalysis of mass spectrometry data\u003c/h2\u003e \u003cp\u003eProteins were inferred from DIA (data-independent acquisition) mass spectrometry data on Spectronaut (Biognosis, Zurich, Switzerland; version v18.7.240506.55695) using the library-free directDIA mode and the canonical \u003cem\u003eHomo sapiens\u003c/em\u003e UniProt knowledgebase containing 20,596 entries (downloaded 08/11/2023). Two missed cleavage sites, two variable modifications (N-terminal acetylation and methionine oxidation), and 7 to 52 amino acids were permitted per peptide. Protein identification relied on the presence of at least two peptides per protein and the cysteine carbamidomethylation as a fixed modification. The false discovery rate was set to 1%. Peptides between 200 and 3000 \u003cem\u003em/z\u003c/em\u003e were quantified using the three most abundant peptides with a minimum intensity threshold of 1%. The five most significant biological process (BP), cellular compartment (CC), and molecular function (MF) GO terms, according to their Benjamini-Hochberg-corrected p-values, were identified by inputting lists of differentially expressed proteins into DAVID (National Institutes of Health, United States of America)\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. ECM proteins were identified with MatrixDB\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e and principal component analysis (PCA) plots were generated with ClustVis\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using GraphPad Prism (version 10.1.2, GraphPad Software, Inc., CA, USA) with an alpha level of 0.05. The means of two approximately normally distributed data sets were compared using unpaired t-tests. In contrast, comparisons of three or more means were analysed using ordinary one-way ANOVA with Bonferroni\u0026rsquo;s correction. Differences in protein abundances measured by mass spectrometry were considered significant when there was a two-fold or greater change, and the q-value was \u0026lt;\u0026thinsp;0.05. P-values sourced from the KM plotting tool\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e used the log-rank hypothesis test to compare the PFS and OS of stage III and IV high-grade (grade III) serous ovarian cancer patients with high or low gene expression, with an alpha level of 0.05 considered as significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAdditional information\u003c/h2\u003e\n\u003cp\u003eThe study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Human Research Ethics Committee of the Central Adelaide Local Health Network, South Australia, Australia, under approval number R20181215. Informed consent was obtained from all participants prior to their inclusion in the study. All participants provided informed consent for the publication of their data in this study. The authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eAuthor contributions included JS (conceptualisation, data collection, writing the original draft), PM (conceptualisation, data collection, supervision, reviewing, and editing), NAL (methodology, data analysis, reviewing, and editing), MKO (conceptualisation, supervision, data analysis, reviewing, and editing), PH (conceptualisation, supervision, data analysis, reviewing, and editing), and MKH (conceptualisation, supervision, data analysis, reviewing, editing, and project administration). All authors read and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eThis study was supported by the Letitia Linke Research Foundation, Tour de Cure, and an Australian Government Research Training Program (RTP) Scholarship. The authors acknowledge all patients treated at the Royal Adelaide Hospital for their donation of biological material for this project. Financial support was provided by Bioplatforms Australia, the Government of South Australia and the University of South Australia towards the National Collaborative Research Infrastructure Strategy (NCRIS) node for Tissue Imaging Mass Spectrometry.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eThe datasets generated during and analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eTorre, L. A. et al. Ovarian Cancer Statistics, \u003cem\u003eC.A. Cancer. J. Clin.\u003c/em\u003e 68, 284\u0026ndash;96 (2018). (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeesham, D., Richards, A. \u0026amp; McGauran, M. Advances in epithelial ovarian cancer. \u003cem\u003eAust J. Gen. Pract.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, 665\u0026ndash;669 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKemp, Z. \u0026amp; Ledermann, J. Update on first-line treatment of advanced ovarian carcinoma. \u003cem\u003eInt. J. Womens Health\u003c/em\u003e. \u003cb\u003e5\u003c/b\u003e, 45\u0026ndash;51 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, S. et al. Paclitaxel Enhances Carboplatin-DNA Adduct Formation and Cytotoxicity. \u003cem\u003eChem. Res. Toxicol.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 2250\u0026ndash;2252 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShih, T. \u0026amp; Lindley, C. Bevacizumab: an angiogenesis inhibitor for the treatment of solid malignancies. \u003cem\u003eClin. Ther.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 1779\u0026ndash;1802 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePBAC Meeting Outcomes July. \u003cem\u003ePharmaceutical Benefits Advisory Committee\u003c/em\u003e https:// (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003c/span\u003e\u003cspan address=\"http://www.pbs.gov.au/industry/listing/elements/pbac-meetings/pbac-outcomes/2023-07/pbac-web-outcomes-07-2023-v2.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePennington, K. P. et al. Germline and Somatic Mutations in Homologous Recombination Genes Predict Platinum Response and Survival in Ovarian, Fallopian Tube, and Peritoneal Carcinomas. \u003cem\u003eClin. Cancer Res.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 764\u0026ndash;775 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChowdhury, S. et al. Proteogenomic analysis of chemo-refractory high-grade serous ovarian cancer. \u003cem\u003eCell\u003c/em\u003e \u003cb\u003e186\u003c/b\u003e, 3476\u0026ndash;3498 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeen, M. P., Armstrong, M. A., Oehler, M. K. \u0026amp; Ricciardelli, C. The role of ABC transporters in ovarian cancer progression and chemoresistance. \u003cem\u003eCrit. Rev. Oncol. Hematol.\u003c/em\u003e \u003cb\u003e96\u003c/b\u003e, 220\u0026ndash;256 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanuchowski, R. et al. Increased Expression of Several Collagen Genes is Associated with Drug Resistance in Ovarian Cancer Cell Lines. \u003cem\u003eJ. Cancer\u003c/em\u003e. \u003cb\u003e7\u003c/b\u003e, 1295\u0026ndash;1310 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFitzpatrick, D. P. G. et al. Searching for potential biomarkers of cisplatin resistance in human ovarian cancer using a label-free LC/MS-based protein quantification method. \u003cem\u003eProteom. \u0026ndash; Clin. Appl.\u003c/em\u003e \u003cb\u003e1\u003c/b\u003e, 246\u0026ndash;263 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGong, F. et al. Proteomic analysis of cisplatin resistance in human ovarian cancer using 2-DE method. \u003cem\u003eMol. Cell. Biochem.\u003c/em\u003e \u003cb\u003e348\u003c/b\u003e, 141\u0026ndash;147 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAcland, M. et al. Chemoresistant Cancer Cell Lines Are Characterized by Migratory, Amino Acid Metabolism, Protein Catabolism and IFN1 Signalling Perturbations. \u003cem\u003eCancers\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 2763 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArentz, G. et al. Label-Free Quantification Mass Spectrometry Identifies Protein Markers of Chemotherapy Response in High-Grade Serous Ovarian Cancer. \u003cem\u003eCancers\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 2172 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFord, C. E., Werner, B., Hacker, N. F. \u0026amp; Warton, K. The untapped potential of ascites in ovarian cancer research and treatment. \u003cem\u003eBr. J. Cancer\u003c/em\u003e. \u003cb\u003e123\u003c/b\u003e, 9\u0026ndash;16 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRakina, M., Kazakova, A., Villert, A., Kolomiets, L. \u0026amp; Larionova, I. Spheroid Formation and Peritoneal Metastasis in Ovarian Cancer: The Role of Stromal and Immune Components. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 6215 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAcland, M. et al. Mass Spectrometry Analyses of Multicellular Tumor Spheroids. \u003cem\u003eProteom. \u0026ndash; Clin. Appl.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 1700124 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaghavan, S. et al. Comparative analysis of tumor spheroid generation techniques for differential in vitro drug toxicity. \u003cem\u003eOncotarget\u003c/em\u003e. \u003cb\u003e7\u003c/b\u003e, 16948\u0026ndash;16961 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaenhoudt, N. et al. Developing Organoids from Ovarian Cancer as Experimental and Preclinical Models. \u003cem\u003eStem Cell. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 717\u0026ndash;729 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerton, T. J. et al. Development of a Patient-Derived 3D Immuno-Oncology Platform to Potentiate Immunotherapy Responses in Ascites-Derived Circulating Tumor Cells. \u003cem\u003eCancers\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 4128 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePhan, N. et al. A simple high-throughput approach identifies actionable drug sensitivities in patient-derived tumor organoids. \u003cem\u003eCommun. Biol.\u003c/em\u003e 2 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBi, J. et al. Successful Patient-Derived Organoid Culture of Gynecologic Cancers for Disease Modeling and Drug Sensitivity Testing. \u003cem\u003eCancers\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 2901 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eden Ouden, J. E. et al. Chemotherapy sensitivity testing on ovarian cancer cells isolated from malignant ascites. \u003cem\u003eOncotarget\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 4570\u0026ndash;4581 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Witte, C. J. et al. Patient-Derived Ovarian Cancer Organoids Mimic Clinical Response and Exhibit Heterogeneous Inter- and Intrapatient Drug Responses. \u003cem\u003eCell. Rep.\u003c/em\u003e 31 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, W. et al. Effective Treatment for Recurrent Ovarian Cancer Guided by Drug Sensitivity from Ascites-Derived Organoid: A Case Report. \u003cem\u003eInt. J. Womens Health\u003c/em\u003e. \u003cb\u003e15\u003c/b\u003e, 1047\u0026ndash;1057 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerslake, R. et al. Transcriptional Landscape of 3D vs. 2D Ovarian Cancer Cell Models. \u003cem\u003eCancers\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 3350 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, W. et al. High-throughput Image Analysis of Tumor Spheroids: A User-friendly Software Application to Measure the Size of Spheroids Automatically and Accurately. \u003cem\u003eJ. Vis. Exp.\u003c/em\u003e \u003cb\u003e89\u003c/b\u003e, 51639 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSherman, B. T. et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cb\u003e50\u003c/b\u003e, 216\u0026ndash;221 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChautard, E., Fatoux-Ardore, M., Ballut, L., Thierry-Mieg, N. \u0026amp; Ricard-Blum, S. MatrixDB, the extracellular matrix interaction database. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 235\u0026ndash;240 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetsalu, T. \u0026amp; Vilo, J. ClustVis: a web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e, 566\u0026ndash;570 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSenkowski, W. et al. A platform for efficient establishment and drug-response profiling of high-grade serous ovarian cancer organoids. \u003cem\u003eDev. Cell.\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e, 1106\u0026ndash;1121 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, H. Short-term organoid culture for drug sensitivity testing of high-grade serous carcinoma. \u003cem\u003eGynecol. Oncol.\u003c/em\u003e \u003cb\u003e157\u003c/b\u003e, 783\u0026ndash;792 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, L-Y. et al. In vitro drug testing using patient-derived ovarian cancer organoids. \u003cem\u003eJ. Ovarian Res.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e, 194 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u0026Aring;kerlund, E. et al. The drug efficacy testing in 3D cultures platform identifies effective drugs for ovarian cancer patients. \u003cem\u003eN P J. Precis Oncol.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNasimi Shad, A. \u0026amp; Moghbeli, M. Integrins as the pivotal regulators of cisplatin response in tumor cells. \u003cem\u003eCell. Commun. Signal.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 265 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian, T. et al. β1 integrin-mediated multicellular resistance in hepatocellular carcinoma through activation of the FAK/Akt pathway. \u003cem\u003eJ. Int. Med. Res.\u003c/em\u003e \u003cb\u003e46\u003c/b\u003e, 1311\u0026ndash;1325 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSethi, T. et al. Extracellular matrix proteins protect small cell lung cancer cells against apoptosis: A mechanism for small cell lung cancer growth and drug resistance in vivo. \u003cem\u003eNat. Med.\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, 662\u0026ndash;668 (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNgaokrajang, U., Janvilisri, T., Sae-Ueng, U., Prungsak, A. \u0026amp; Kiatwuthinon, P. Integrin α5 mediates intrinsic cisplatin resistance in three-dimensional nasopharyngeal carcinoma spheroids via the inhibition of phosphorylated ERK /caspase-3 induced apoptosis. \u003cem\u003eExp. Cell. Res.\u003c/em\u003e \u003cb\u003e406\u003c/b\u003e, 112765 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuermanbayi, S. et al. In situ monitoring of functional activity of extracellular matrix stiffness-dependent multidrug resistance protein 1 using scanning electrochemical microscopy. \u003cem\u003eChem. Sci.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 10349\u0026ndash;10360 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrakash, J. \u0026amp; Shaked, Y. The Interplay between Extracellular Matrix Remodeling and Cancer Therapeutics. \u003cem\u003eCancer Discov\u003c/em\u003e. \u003cb\u003e14\u003c/b\u003e, 1375\u0026ndash;1388 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo, W. et al. High Serpin Family A Member 10 Expression Confers Platinum Sensitivity and Is Associated With Survival Benefit in High-Grade Serous Ovarian Cancer: Based on Quantitative Proteomic Analysis. \u003cem\u003eFront. Oncol.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 761960 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark, S. J. et al. SERPINB11 Expression Is Associated With Prognosis of High-grade Serous and Clear Cell Carcinoma of the Ovary. \u003cem\u003eVivo\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, 2647\u0026ndash;2653 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma, S., Santiskulvong, C., Rao, J., Gimzewski, J. K. \u0026amp; Dorigo, O. The role of Rho GTPase in cell stiffness and cisplatin resistance in ovarian cancer cells. \u003cem\u003eIntegr. Biol.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 611\u0026ndash;617 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbedini, M. R. et al. Cell fate regulation by gelsolin in human gynecologic cancers. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 111, 14442\u0026ndash;14447 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRicciardelli, C. et al. Keratin 5 overexpression is associated with serous ovarian cancer recurrence and chemotherapy resistance. \u003cem\u003eOncotarget\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 17819\u0026ndash;17832 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGyőrffy, B. Discovery and ranking of the most robust prognostic biomarkers in serous ovarian cancer. \u003cem\u003eGeroScience\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e, 1889\u0026ndash;1898 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuerra, F., Arbini, A. A. \u0026amp; Moro, L. Mitochondria and cancer chemoresistance. \u003cem\u003eBiochim. Biophys. Acta\u003c/em\u003e. \u003cb\u003e1858\u003c/b\u003e, 686\u0026ndash;699 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, J. R. et al. Dysfunctional Apoptosome Activation in Ovarian Cancer: Implications for Chemoresistance. \u003cem\u003eCancer Res.\u003c/em\u003e \u003cb\u003e62\u003c/b\u003e, 924\u0026ndash;931 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, S. et al. Evaluating Tumor Evolution via Genomic Profiling of Individual Tumor Spheroids in a Malignant Ascites. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 12724 (2018).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Ovarian cancer, personalised medicine, spheroids, malignant ascites, mass spectrometry, proteomics","lastPublishedDoi":"10.21203/rs.3.rs-6441929/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6441929/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe accumulation of malignant ascites in the peritoneal cavity is a hallmark of advanced epithelial ovarian cancer (EOC). This fluid contains three-dimensional multicellular aggregates known as spheroids, which contribute to chemoresistance and are an accessible source of tumour material. However, many studies use spheroids generated from primary cell suspensions to reduce heterogeneity. Here, we compare the proteomes and the response to chemotherapeutics of native spheroids directly collected from ascites to spheroids generated \u003cem\u003eex vivo\u003c/em\u003e. We demonstrate that the chemoresponse of native spheroids correlates with patients\u0026rsquo; therapy responses in 4/5 cases. In contrast, all \u003cem\u003eex vivo-\u003c/em\u003egenerated spheroids were resistant to carboplatin treatment and did not correlate with the clinical outcome. In addition, proteomics quantified over 6,300 proteins per sample, revealing that the global proteomes of native spheroids cluster according to their carboplatin response. A detailed analysis of the upregulated proteins highlights the potential role of extracellular matrix proteins in regulating chemoresponse. This pilot study suggests key proteins and biological pathways that may facilitate a global proteomics-based screening strategy for personalised EOC treatment. As such, native spheroids have the potential to be used to personalise the treatment of all diseases that cause malignant ascites.\u003c/p\u003e","manuscriptTitle":"The proteomes of ovarian cancer ascitic cellular aggregates correlate with their ex vivo platinum sensitivities: a pilot study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-09 11:01:42","doi":"10.21203/rs.3.rs-6441929/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8fa9b3f0-8c94-478b-8533-136dbdaab9b3","owner":[],"postedDate":"May 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":48158137,"name":"Health sciences/Medical research/Biomarkers/Predictive markers"},{"id":48158138,"name":"Health sciences/Oncology/Cancer/Gynaecological cancer/Ovarian cancer"}],"tags":[],"updatedAt":"2026-01-05T17:13:22+00:00","versionOfRecord":{"articleIdentity":"rs-6441929","link":"https://doi.org/10.1021/acs.jproteome.5c00771","journal":{"identity":"journal-of-proteome-research","isVorOnly":true,"title":"Journal of Proteome Research"},"publishedOn":"2025-12-23 00:00:00","publishedOnDateReadable":"December 23rd, 2025"},"versionCreatedAt":"2025-05-09 11:01:42","video":"","vorDoi":"10.1021/acs.jproteome.5c00771","vorDoiUrl":"https://doi.org/10.1021/acs.jproteome.5c00771","workflowStages":[]},"version":"v1","identity":"rs-6441929","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6441929","identity":"rs-6441929","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}