Patient-derived Explants Research Articles

Patient-Derived Tumor Explants As a "Live" Preclinical Platform for Predicting Drug Resistance in Patients.

An understanding of drug resistance and the development of novel strategies to sensitize highly resistant cancers rely on the availability of suitable preclinical models that can accurately predict patient responses. One of the disadvantages of existing preclinical models is the inability to contextually preserve the human tumor microenvironment (TME) and accurately represent intratumoral heterogeneity, thus limiting the clinical translation of data. By contrast, by representing the culture of live fragments of human tumors, the patient-derived explant (PDE) platform allows drug responses to be examined in a three-dimensional (3D) context that mirrors the pathological and architectural features of the original tumors as closely as possible. Previous reports with PDEs have documented the ability of the platform to distinguish chemosensitive from chemoresistant tumors, and it has been shown that this segregation is predictive of patient responses to the same chemotherapies. Simultaneously, PDEs allow the opportunity to interrogate molecular, genetic, and histological features of tumors that predict drug responses, thereby identifying biomarkers for patient stratification as well as novel interventional approaches to sensitize resistant tumors. This paper reports PDE methodology in detail, from collection of patient samples through to endpoint analysis. It provides a detailed description of explant derivation and culture methods, highlighting bespoke conditions for particular tumors, where appropriate. For endpoint analysis, there is a focus on multiplexed immunofluorescence and multispectral imaging for the spatial profiling of key biomarkers within both tumoral and stromal regions. By combining these methods, it is possible to generate quantitative and qualitative drug response data that can be related to various clinicopathological parameters and thus potentially be used for biomarker identification.

ELOVL5 Is a Critical and Targetable Fatty Acid Elongase in Prostate Cancer.

The androgen receptor (AR) is the key oncogenic driver of prostate cancer, and despite implementation of novel AR targeting therapies, outcomes for metastatic disease remain dismal. There is an urgent need to better understand androgen-regulated cellular processes to more effectively target the AR dependence of prostate cancer cells through new therapeutic vulnerabilities. Transcriptomic studies have consistently identified lipid metabolism as a hallmark of enhanced AR signaling in prostate cancer, yet the relationship between AR and the lipidome remains undefined. Using mass spectrometry-based lipidomics, this study reveals increased fatty acyl chain length in phospholipids from prostate cancer cells and patient-derived explants as one of the most striking androgen-regulated changes to lipid metabolism. Potent and direct AR-mediated induction of ELOVL fatty acid elongase 5 (ELOVL5), an enzyme that catalyzes fatty acid elongation, was demonstrated in prostate cancer cells, xenografts, and clinical tumors. Assessment of mRNA and protein in large-scale data sets revealed ELOVL5 as the predominant ELOVL expressed and upregulated in prostate cancer compared with nonmalignant prostate. ELOVL5 depletion markedly altered mitochondrial morphology and function, leading to excess generation of reactive oxygen species and resulting in suppression of prostate cancer cell proliferation, 3D growth, and in vivo tumor growth and metastasis. Supplementation with the monounsaturated fatty acid cis-vaccenic acid, a direct product of ELOVL5 elongation, reversed the oxidative stress and associated cell proliferation and migration effects of ELOVL5 knockdown. Collectively, these results identify lipid elongation as a protumorigenic metabolic pathway in prostate cancer that is androgen-regulated, critical for metastasis, and targetable via ELOVL5. SIGNIFICANCE: This study identifies phospholipid elongation as a new metabolic target of androgen action that is critical for prostate tumor metastasis.

Development of a patient-derived explant model for prediction of drug responses in endometrial cancer

ObjectiveTo undertake a pilot study to develop a novel Patient-Derived-Explant (PDE) model system for use in endometrial cancer (EC) that is capable of monitoring differential drug responses in a pre-clinical setting. MethodsFresh tumour was obtained post-hysterectomy from 27 patients with EC. Tumours were cut into 1–3 mm3 explants that were cultured at the air-liquid interface for 16–24 h in culture media. Explants were cultured in different media conditions to optimise viability. Explants were also treated with carboplatin/paclitaxel or pembrolizumab for 24 h and processed into histology slides. Multiplexed immunofluorescence for Ki67 (proliferation marker), cPARP (apoptosis marker) and CAM 5.2 (tumour mask) was performed followed by image analysis and quantitation of biomarker expression. ResultsEC samples are amenable to PDE culture with preserved histological architecture and PDE viability for up to 48 h, with the addition of autologous serum in culture media facilitating EC-PDE viability. Our PDE platform provides evidence of differential drug-response to conventional chemotherapeutics and immune checkpoint inhibition, and these responses can be assessed in the context of a preserved tumour microenvironment. ConclusionsOur PDE platform represents a rapid, low-cost pre-clinical model which can be easily integrated into drug development pipelines. PDE culture preserves original tumour architecture and enables evaluation of spatial relationships in the tumour microenvironment. PDE culture has the potential for personalised drug-testing in a pre-clinical setting which is increasingly important in an era of personalised medicine in the treatment of EC.

Patient-Derived Prostate Cancer Explants: A Clinically Relevant Model to Assess siRNA-Based Nanomedicines.

Over the last thirty years, research in nanomedicine has widely been focused on applications in cancer therapeutics. However, despite the plethora of reported nanoscale drug delivery systems that can successfully eradicate solid tumor xenografts in vivo, many of these formulations have not yet achieved clinical translation. This issue particularly pertains to the delivery of small interfering RNA (siRNA), a highly attractive tool for selective gene targeting. One of the likely reasons behind the lack of translation is that current in vivo models fail to recapitulate critical elements of clinical solid tumors that may influence drug response, such as cellular heterogeneity in the tumor microenvironment. This study incorporates a more clinically relevant model for assessing siRNA delivery systems; ex vivo culture of prostate cancer harvested from patients who have undergone radical prostatectomy, denoted patient-derived explants (PDE). The model retains native human tissue architecture, microenvironment, and cell signaling pathways. Porous silicon nanoparticles (pSiNPs) behavior in this model is investigated and compared with commonly used 3D cancer cell spheroids for their efficacy in the delivery of siRNA directed against the androgen receptor (AR), a key driver of prostate cancer.

Patient-derived ovarian cancer explants: preserved viability and histopathological features in long-term agitation-based cultures

Ovarian carcinoma (OvC) remains a major therapeutic challenge due to its propensity to develop resistance after an initial response to chemotherapy. Interactions of tumour cells with the surrounding microenvironment play a role in tumour survival, invasion capacity and drug resistance. Cancer models that retain tissue architecture and tumour microenvironment components are therefore essential to understand drug response and resistance mechanisms. Herein, our goal was to develop a long-term OvC patient-derived explant (OvC-PDE) culture strategy in which architecture and cell type heterogeneity of the original tumour would be retained. Samples from 25 patients with distinct OvC types and one with a benign tumour, were cultured for 30 days in agitation-based culture systems with 100% success rate. OvC-PDE cultures retained the original tumour architecture and main cellular components: epithelial cells, fibroblasts and immune cells. Epithelial cells kept their original levels of proliferation and apoptosis. Moreover, the major extracellular components, such as collagen-I and -IV, were retained in explants. OvC-PDE cultures were exposed to standard-of-care chemotherapeutics agents for 2 weeks, attesting the ability of the platform for drug assays employing cyclic drug exposure regimens. We established an OvC-PDE dynamic culture in which tumour architecture and cell type heterogeneity were preserved for the different OvC types, replicating features of the original tumour and compatible with long-term drug exposure for drug efficacy and resistance studies.

An Aptamer for Broad Cancer Targeting and Therapy.

Simple SummaryRecent efforts to improve chemotherapy’s antitumor effects have increasingly focused on targeted therapies, where the drug is modified with an agent able to specifically deliver it to the tumor while limiting its accumulation in normal tissue. Aptamers, comprised of short pieces of RNA or DNA, are ideal for this type of drug targeting due in part to their ease of chemical synthesis. The E3 aptamer was previously conjugated to highly toxic chemotherapeutics and shown to target and treat prostate tumors. Here, we show that E3 is not limited to prostate cancer targeting but appears to broadly target cancer cells. E3 highly toxic drug conjugates also efficiently kill a broad range of cancer types, and E3 targets tumors that closely model patient tumors. Thus, the E3 aptamer appears to be a general agent for specific delivery of chemotherapy to tumors and should improve antitumor treatment while reducing unwanted toxicities in other tissues. Recent advances in chemotherapy treatments are increasingly targeted therapies, with the drug conjugated to an antibody able to deliver it directly to the tumor. As high-affinity chemical ligands that are much smaller in size, aptamers are ideal for this type of drug targeting. Aptamer-highly toxic drug conjugates (ApTDCs) based on the E3 aptamer, selected on prostate cancer cells, target and inhibit prostate tumor growth in vivo. Here, we observe that E3 also broadly targets numerous other cancer types, apparently representing a universal aptamer for cancer targeting. Accordingly, ApTDCs formed by conjugation of E3 to the drugs monomethyl auristatin E (MMAE) or monomethyl auristatin F (MMAF) efficiently target and kill a range of different cancer cells. Notably, this targeting extends to both patient-derived explant (PDX) cancer cell lines and tumors, with the E3 MMAE and MMAF conjugates inhibiting PDX cell growth in vitro and with the E3 aptamer targeting PDX colorectal tumors in vivo.

Dynamic differences between DNA damage repair responses in primary tumors and cell lines

The study of DNA damage repair response (DDR) in prostate cancer is restricted by the limited number of prostate cancer cell lines and lack of surrogates for heterogeneity in clinical samples. Here, we sought to leverage our experience with patient derived explants (PDEs) cultured ex vivo to study dynamics of DDR in primary tumors following application of clinically relevant doses of ionizing radiation (IR) to tumor cells in their native 3-dimensional microenvironment. We compared DDR dynamics between prostate cancer cell lines, PDEs and xenograft derived explants (XDEs) following treatment with IR (2Gy) either alone or in combination with pharmacological modulators of DDR. We have shown that following treatment with 2Gy, DDR can be consistently detected in PDEs from multiple solid tumors, including prostate, kidney, testes, lung and breast, as evidenced by γ-H2AX, 53BP1, phospho-ATM and phospho-DNA-PKcs foci. By examining kinetics of resolution of IR-induced foci, we have shown that DDR in prostate PDEs (complete resolution in 8 h) is much faster than in prostate cancer cell lines (<50% resolution in 8 h). The transcriptional profile of DDR genes following 2Gy IR appears to be distinct between PDEs and cell lines. Pre-treatment with drugs targeting DDR pathways differentially alter the kinetics of DDR in the PDEs and cell lines, as evidenced by altered kinetics of foci resolution. This study highlights the utility of PDEs as a robust model system for short-term evaluation of DDR in primary solid tumors in clinically relevant microenvironment.

Fatty Acid Oxidation Is an Adaptive Survival Pathway Induced in Prostate Tumors by HSP90 Inhibition

HSP90 is a molecular chaperone required for stabilization and activation of hundreds of client proteins, including many known oncoproteins. AUY922 (luminespib), a new-generation HSP90 inhibitor, exhibits potent preclinical efficacy against several cancer types including prostate cancer. However, clinical use of HSP90 inhibitors for prostate cancer has been limited by toxicity and treatment resistance. Here, we aimed to design an effective combinatorial therapeutic regimen that utilizes subtoxic doses of AUY922, by identifying potential survival pathways induced by AUY922 in clinical prostate tumors. We conducted a proteomic analysis of 30 patient-derived explants (PDE) cultured in the absence and presence of AUY922, using quantitative mass spectrometry. AUY922 significantly increased the abundance of proteins involved in oxidative phosphorylation and fatty acid metabolism in the PDEs. Consistent with these findings, AUY922-treated prostate cancer cell lines exhibited increased mitochondrial mass and activated fatty acid metabolism processes. We hypothesized that activation of fatty acid oxidation is a potential adaptive response to AUY922 treatment and that cotargeting this process will sensitize prostate cancer cells to HSP90 inhibition. Combination treatment of AUY922 with a clinical inhibitor of fatty acid oxidation, perhexiline, synergistically decreased viability of several prostate cancer cell lines, and had significant efficacy in PDEs. The novel drug combination treatment induced cell-cycle arrest and apoptosis, and attenuated the heat shock response, a known mediator of HSP90 treatment resistance. This combination warrants further preclinical and clinical investigation as a novel strategy to overcome resistance to HSP90 inhibition. IMPLICATIONS: Metabolic pathways induced in tumor cells by therapeutic agents may be critical, but targetable, mediators of treatment resistance.

The use of patient-derived breast tissue explants to study macrophage polarization and the effects of environmental chemical exposure.

Ex vivo mammary explant systems are an excellent model to study interactions between epithelium and stromal cell types because they contain physiologically relevant heterotypic interactions in the background of genetically diverse patients. The intact human mammary tissue, termed patient‐derived explant (PDE), can be used to investigate cellular responses to a wide variety of external stimuli in situ. For this study, we examined the impact of cytokines or environmental chemicals on macrophage phenotypes. We demonstrate that we can polarize macrophages within human breast tissue PDEs toward M1 or M2 through the addition of interferon‐γ (IFNγ) + lipopolysaccharide (LPS) or interleukin (IL)‐4 + IL‐13, respectively. Elevated expression levels of M(IFNγ + LPS) markers (HLADRA and CXCL10) or M(IL‐4 + IL‐13) markers (CD209 and CCL18) were observed in cytokine‐treated tissues. We also examined the impact of the endocrine‐disrupting chemical, benzophenone‐3, on PDEs and measured significant, yet varying effects on macrophage polarization. Furthermore, a subset of the PDEs respond to IL‐4 + IL‐13 through downregulation of E‐cadherin and upregulation of vimentin which is reminiscent of epithelial‐to‐mesenchymal transition (EMT) changes. Finally, we were able to show immortalized nonmalignant breast epithelial cells can exhibit EMT characteristics when exposed to growth factors secreted by M(IL‐4 + IL‐13) macrophages. Taken together, the PDE model system is an outstanding preclinical model to study early tissue‐resident immune responses and effects on epithelial and stromal responses to stimuli found both endogenously in the breast and exogenously as a result of exposures.

1385P Analysis of drug-induced RNA expression changes in NSCLC patient-derived explants as a potential tool for personalized therapy choice

1385P Analysis of drug-induced RNA expression changes in NSCLC patient-derived explants as a potential tool for personalized therapy choice

Abstract 6403: Characterization of small molecule inhibitors of PELP1 for treating advanced breast cancer

Abstract Background: Breast cancer (BC) is the most commonly diagnosed cancer and the second leading cause of cancer death in American women. BC is composed of distinct molecular subtypes, such as ER positive BC (ER+) and triple negative BC (TNBC). Development of novel effective therapies for patients with therapy resistant BC (TR-BC) and TNBC remains the highest unmet need in patient treatment and survivorship. Proline, glutamic acid-, and leucine-rich protein 1 (PELP1) plays a critical role in multiple nuclear receptor functions leading to TR-BC and TNBC progression. PELP1 expression is dysregulated in BC, is a prognostic indicator of poorer BC survival, and its deregulation contributes to TR-BC. The objective of this study is characterization of a small molecule inhibitor of PELP1 (SMIP34) as novel therapeutic for treating advanced BC. Methods: Yeast two-hybrid screening was used to identify PELP1 Inhibitory Peptide 1 (PIP1). Direct binding of PIP1 to PELP1 was confirmed using biotin pull-down assays and inhibition of BC proliferation confirmed using MTT assays. Hit-Ligand interaction site with PIP1 hot spot residues based on 3D alignment and morphology was used to generate a library of small chemical molecules that function as peptidomimetics of PIP1. In vitro activity was assessed using CellTiter Glo, MTT, colony formation and matrigel invasion assays in multiple BC models. Mechanistic studies were conducted using Western blot, and reporter gene assays. Xenograft and patient derived explant (PDEX) assays were used for preclinical evaluation. Results: Screening utilizing MTT assays lead to the selection of SMIP34 as an inhibitor of PELP1. SMIP34 treatment reduced cell viability at an IC50 of 3-10µM in a panel of BC cells. Additionally, SMIP34 showed no activity in human mammary epithelial cells. Knockdown of PELP1 in BC cells significantly reduced the SMIP34 activity, confirming target specificity. Furthermore, SMIP34 treatment significantly reduced the invasiveness and colony formation of TR-BC and TNBC cells. Mechanistic studies using Western blot analysis confirmed that SMIP34 binding to PELP1 contributes to its degradation. In combination studies, SMIP34 displayed synergy and enhanced the efficacy of current chemotherapeutics Cisplatin and Paclitaxel. In PDEX assays, SMIP34 significantly decreased tumor proliferation as measured by Ki67 staining. In xenograft models, SMIP34 (10mg/kg/s.c.) treatment resulted in significant reduction in tumor volume compared to vehicle control. Furthermore, overall mouse body weight in both control and SMIP34 treated groups were similar, suggesting no overt signs of toxicity. Conclusion: We have developed a first-in-class small molecule inhibitor of PELP1 (SMIP34) displaying therapeutic efficacy against TR-BC and TNBC in vitro, ex vivo, and in vivo. Supported by CPRIT Predoctoral Fellowship CPRIT RTA; RP170345 (K.A. Altwegg) and VA grant I01BX004545 (R.K.V) Citation Format: Kristin Ann Altwegg, Suryavathi Viswanadhapalli, Monica Mann, Uday P. Pratap, Mengxing Li, Junhao Liu, Yiliao Luo, Gangadhara R. Sareddy, Hariprasad Vankayalapati, Ratna K. Vadlamudi. Characterization of small molecule inhibitors of PELP1 for treating advanced breast cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6403.

Abstract 1764: KDM1A inhibition activates estrogen receptor beta pathway to suppress ovarian cancer progression

Abstract Background: Ovarian cancer (OCa) is the deadliest gynecologic cancer. Recent studies suggest that OCa cells express estrogen receptor beta (ESR2), which functions as a tumor suppressor. However, ESR2 expression decreases during tumor progression and under the selection pressure of chemotherapy; this decrease occurs via epigenetic mechanisms. The lysine-specific histone demethylase 1A (KDM1A), an epigenetic regulator, is overexpressed in OCa. We reason that agents that restore the expression/functions of ESR2 may provide a novel therapeutic opportunity for suppression of OCa. In this study we tested the hypothesis that KDM1A suppresses ESR2 expression/functions and that inhibition of KDM1A potentiates ESR2 mediated tumor suppression. Methods: We examined the utility of combination therapy of KDM1A inhibitor and ESR2 agonist on established and patient derived OCa cells using cell viability, survival, apoptosis and invasion assays. KDM1A knockout (KDM1A-KO) cells were generated using the CRISPR/Cas9 system. Mechanistic studies were conducted using RNA-seq, ERE-Luc reporter assays, RT-qPCR, co-IP, ChIP and Western blot analysis. The efficacy of combination therapy on primary ovarian tumors was examined using patient derived explant (PDEx) models. In vivo efficacy of KDM1A inhibitor and ESR2 agonist was studied using orthotopic OCa and patient derived xenograft models. Results: Analysis of the gene expression profiles of OCa patients showed that KDM1A negatively correlates with ESR2. Combination of KDM1A inhibitor NCD-38 and ESR2 agonist LY500307 synergistically reduced cell viability, survival and invasion and increased apoptosis of established and patient derived primary OCa cells. Inhibition or knockout of KDM1A sensitized OCa cells to ESR2 agonist treatment and significantly increased the expression of ESR2 and its target genes. Further, we found that KDM1A is recruited to the ESR2 0N promoter and interacts with ESR2. In addition, inhibition of KDM1A increased the active histone methylation mark H3K4-me2 at the ESR2 0N promoter. RNA-seq analysis demonstrated that ESR2 agonist treatment resulted in modulation of pathways related to cell cycle, apoptosis and DNA damage. Importantly, combination treatment significantly reduced the tumor growth in orthotopic and patient derived xenograft models and reduced the proliferation of OCa cells in PDEx models. Conclusions: Our results demonstrate that KDM1A inhibition enhances ESR2 expression/functions and combination therapy of KDM1A inhibitor and ESR2 agonist is an attractive therapy for treating OCa. Citation Format: Prabhakar Pitta Venkata, Bridgitte E. Palacios, Yihong Chen, Salvador Alejo, Yi He, Ilanna Loeffel, Uday Pratap, Kristin Altwegg, Suryavathi Viswanadhapalli, Takayoshi Suzuki, Rajeshwar Tekmal, Edward Kost, Gangadhara Sareddy. KDM1A inhibition activates estrogen receptor beta pathway to suppress ovarian cancer progression [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1764.

Abstract 562: Novel combination therapy for treating TNBC using LIFR and HDAC Inhibitors

Abstract Background: Triple-negative breast cancer (TNBC) lacks targeted therapies and represents a disproportional share of the breast cancer (BC) mortality rate. TNBC exhibits autocrine stimulation of the LIF/LIFR axis and overexpression of LIF is associated with poorer relapse-free survival in BC patients. Histone deacetylase inhibitors (HDACIs) are emerging as promising multifunctional agents in TNBC to elicit cytotoxic actions. Recent studies have shown that cancer cells elicit feedback activation of leukemia inhibitory factor receptor (LIFR) which in turn curtails response to HDACIs. We developed a first-in-class inhibitor of LIFR, EC359 that directly interacts with LIFR and effectively blocks LIFR downstream signaling. The objective of this study is to examine the therapeutic efficacy of combination therapy using preclinical and patient-derived xenograft (PDX) models. Methods: We tested utility of combination therapy using multiple HDACIs that are currently in clinical trails along with EC359. The effect of combination therapy was evaluated using MTT, invasion, colony formation, and Caspase3/7 assays. Mechanistic studies were performed using Western blotting, qRT-PCR, and STAT3 reporter assays. The efficacy of combination therapy in vivo was examined using xenograft, PDX, and patient-derived explant (PDEx) models. Results: Immunohistochemical analyses of breast tumors using tissue microarrays revealed significant expression of LIFR in TNBC tissues. Treatment of TNBC model cells with four different HDACIs increased the expression of LIFR. LIFR inhibitor EC359 at nM concentration is additive to HDACIs in reducing cell viability. Knockdown of LIFR or treatment with EC359 significantly enhanced the efficacy of HDACIs in reducing the cell viability, colony formation ability, and invasiveness as well as promoted apoptosis compared to monotherapy in TNBC model cells. On the contrary, treatment with STAT3 inhibitor requires µM concentrations to reduce the cell viability of TNBC cells and is not additive to HDACIs. Mechanistic studies utilizing STAT3 reporter gene assays and biochemical studies using multiple TNBC model cells exhibited activation of the LIFR signaling pathway upon HDACIs treatment but was attenuated by EC359 therapy. Treatment of human TNBC utilizing PDEx assays showed that EC359 enhanced the ability of HDACIs to decrease proliferation (Ki-67 positivity) compared to monotherapy. Using TNBC xenografts and PDX models, we demonstrated that EC359 treatment enhanced the ability of HDACIs to reduce in vivo tumor growth compared to monotherapy. Conclusions: Our results suggest that the combination therapy of HDACIs and EC359 provides therapeutic utility in overcoming the limitation of feedback activation of LIFR observed in the treatment of HDACIs in treating TNBC. Supported by DOD BCRP grant W81XWH-18-1-0016 (R.K. Vadlamudi; K.J. Nickisch) Citation Format: Suryavathi Viswanadhapalli, Mengxing Li, Bindu Santhamma, Uday P. Pratap, Yiliao Luo, Junhao Liu, Kristin A. Altwegg, Xiaonan Li, Ahmed Gulzar, Hui Yan, Zhenming Xu, Andrew Brenner, Gangadhara R. Sareddy, Manjeet K. Rao, Rajeshwar R. Tekmal, Hareesh B. Nair, Klaus J. Nickisch, Ratna K. Vadlamudi. Novel combination therapy for treating TNBC using LIFR and HDAC Inhibitors [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 562.

Abstract 6168: Characterization of ATM inhibitor (AZD1390) distribution and pharmacodynamic changes in brain and glioblastoma in preclinical models to inform on mechanism of action

Abstract Ataxia telangiectasia mutated (ATM) kinase is a key factor in DNA double strand break response and is constitutively activated in glioblastoma, leading to marked radioresistance [Bartkova, Carruthers] and therefore an attractive radiosensitisation target in this disease. AZD1390, a potent ATM inhibitor optimized to cross the blood brain barrier, is being evaluated with radiotherapy in Phase 1 (NCT03423628). Positron emission tomography (PET) studies demonstrated AZD1390 brain exposure in healthy human subjects. This study characterises ATM inhibitor penetration and distribution in brain and GBM, as well as pharmacodynamic biomarker changes in human orthotopic GBM patient-derived explant (PDX) models. Radiation +/- AZD1390 treatment was evaluated for compound concentration in plasma and relative abundance in brain/ GBM regions by mass spectrometry imaging (MSI). Parallel samples were analysed by immunohistochemistry for biomarkers of DNA damage response and ATM signalling. We show AZD1390 penetrates brain tissue with preferential compound accumulation into tumour. Higher ratios of AZD1390 in GBM versus normal brain correlate with improved efficacy and greater levels of DNA damage. Moreover, AZD1390 in the GBM region was heterogeneously distributed with MSI characterized specific metabolic clusters having varied AZD1390 intensities. As expected, radiation treatment activated ATM signalling (indicated by increase in pRAD50 and γH2AX positive cells) which was effectively inhibited by AZD1390 co-treatment. Our work provides novel insights into ATM inhibitor pharmacokinetics in brain tumours and demonstrates the benefit of using MSI in conjunction with pharmacodynamic markers to provide a more complete picture of AZD1390 mechanism of action for future translational studies on clinical GBM samples. Citation Format: Lenka Oplustil O'Connor, Gregory Hamm, Maria Udriste, Stephanie Ling, Joanne Wilson, Steve Durant, Gareth Hughes, Sabina Cosulich, Wenlin Shao, Martin Pass, Gemma Jones, Andrew Pierce, Richard Goodwin, Mark O'Connor, Carl Barrett, Elizabeth Harrington. Characterization of ATM inhibitor (AZD1390) distribution and pharmacodynamic changes in brain and glioblastoma in preclinical models to inform on mechanism of action [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6168.

Abstract 2076: Phospholipid profiling of clinical prostate tissues reveals targetable alterations in membrane lipid composition accompanying tumorigenesis

Abstract Membranes are unique phospholipid (PL) interfaces that play a central role in cancer cell biology. However, PL composition of clinical tumors, and its dynamic regulation, remains a critical gap in the molecular profiling of this disease. Here, we used mass spectrometry-based spatial imaging and lipidomics to generate PL profiles of clinical prostate tissues upon cancer development and in patient-derived tumor explants (PDEs; n=43) cultured with the current clinical agent enzalutamide. Analysis of more than 100 PL species in tumors and matched normal tissues from 21 patients revealed a complexity of cancer-related changes in the tissue PL landscape. Phosphatidylcholine (PC)-based PLs were the most abundant PL class, and their relative abundance was increased in tumors compared to matched benign tissues. A greater proportion of monounsaturated fatty acid chains was characteristic of PLs in tumors, most evident in the PC and phosphatidylethanolamine (PE) PLs, as was striking elongation of fatty acid chains in the phosphatidylinositol (PI) and phosphatidylserine (PS) classes. The abundance of specific PL species and saturation groups was significantly associated with malignancy in both this and an independent cohort of unmatched patient tissues (n=47). Notably, the PL profile was significantly associated with the ERG molecular subtype and basal proliferative index (Ki67) in tumors. Furthermore, we identified individual PL species that were associated with the antiproliferative response to enzalutamide, in cultured PDEs. In light of these PL features, we proposed acetyl coA carboxylase (ACC) as a novel therapeutic target in prostate cancer. Indeed, blocking lipid biosynthesis and elongation in prostate tumors with an ACC inhibitor (ACCi, PF-05175157) significantly reduced tumor cell proliferation in PDE tissues (n=13), concomitant with increased pACC1 staining and a decreased proportion of longer chain PLs. Our findings suggest that the clinical PCa lipidome is not only a marker of malignant transformation and aggressiveness, but also therapeutic response to enzalutamide. Moreover, defining this unique biology identified further clinically actionable targets that may improve prostate cancer outcomes. Citation Format: Lisa M. Butler, Chui Yan Mah, Jonas Dehairs, Andrew Vincent, Shadrack Mutuku, Xander Spotbeen, Rajdeep Das, Zeyad Nassar, Luke Selth, Paul Trim, Marten Snel, David Lynn, Lisa Horvath, Wayne Tilley, Margaret Centenera, Johannes Swinnen. Phospholipid profiling of clinical prostate tissues reveals targetable alterations in membrane lipid composition accompanying tumorigenesis [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2076.

Abstract 237: DECR1: The rate limiting enzyme of polyunsaturated fatty acid metabolism and a novel therapeutic target in prostate cancer

Abstract Introduction: Fatty acid oxidation (FAO) is the dominant bioenergetic pathway in prostate cancer (PCa). Previous studies have demonstrated efficacy in targeting key FAO enzymes using in vitro and in vivo models of PCa. However, clinical evidence supporting FAO targeting in PCa is lacking. A recent study reported that FAO is upregulated in PCa cells and is essential for cell viability. Here, we demonstrated therapeutic efficacy of targeting FAO in clinical prostate tumors and identified DECR1 as a potential FAO-related survival factor. Methods and Results: To examine the therapeutic efficacy of targeting FAO in a more clinically-relevant setting, patient-derived prostate explants (PDE) were treated with etomoxir (FAO inhibitor). Results showed significant inhibition of cell proliferation (evaluated using Ki67 immunohistochemistry staining) in drug-treated PDE by a mean of 48% (n=13, p&amp;lt;0.05). To identify key targets of FAO in PCa, a meta-analysis was performed using four clinical RNA-sequencing datasets comprising of malignant and matched normal PCa patient tissues. DECR1 – the rate limiting enzyme of polyunsaturated fatty acid (PUFA) oxidation was identified as a robustly overexpressed FAO gene in PCa tissues; its expression is significantly associated with shorter biochemical recurrence-free survival and overall PCa patient survival. Consistent with the vital role of DECR1 in PUFA oxidation, knockdown of DECR1 impaired the cell's capacity to metabolize PUFAs; furthermore, FAO inhibition via DECR1 knockdown is specific to PUFA oxidation. Notably, DECR1 knockdown significantly inhibited PCa cell proliferation, migration, and invasion, and significantly reduced cellular proliferation in tumors. Mechanistically, DECR1 knockdown results in the induction of lipid peroxidation and cell death via ferroptosis in PCa cells. In contrast, overexpression of DECR1 significantly decreased mitochondrial oxidative stress and reduced levels of malondialdehyde. Treatment of PCa cells with ferrostatin successfully rescued PCa cells from death by DECR1 inhibition. Taken together, these data suggest that overexpression of DECR1 may offer PCa cells a survival advantage by protecting the cells from ferroptosis. Conclusion: This study demonstrated for the first time clinically-relevant evidence that targeting FAO is efficacious and provides evidence for the importance of FAO, and specifically PUFA oxidation, in PCa progression. Importantly, this study identified DECR1 as a promising and novel therapeutic target for PCa. Citation Format: Chui Yan Mah, Zeyad D. Nassar, Ingrid J. Burvenich, Swati Irani, Margaret M. Centenera, Max Moldovan, David J. Lynn, Andrew J. Hoy, Johannes V. Swinnen, Lisa M. Butler. DECR1: The rate limiting enzyme of polyunsaturated fatty acid metabolism and a novel therapeutic target in prostate cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 237.

Abstract 1115: Patient-derived explant cultures (PDECs) as a model system for immuno-oncology studies

Abstract Despite the great promise of immunotherapy, only relatively few patients show long-term complete remission. In addition, immunotherapy is not effective against every cancer type. Cancer types with low mutational burden such as breast cancer (BCa), for example, are thought to not elicit an immune reaction and are therefore considered non-immunogenic. A major hurdle in the advancement of immunotherapy is the lack of preclinical models that faithfully preserve the tumor microenvironment and recapitulate human disease. We have developed a method to grow 3D cultures of intact fragments of patient-derived tissue (Patient-Derived Explant Cultures; PDECs), which has been used to assess the efficacy of novel treatment regimens on real human tumors (Haikala et al., Nature Comm. 2019; Tervonen et al., Oncogene 2016). PDECs offer many advantages over other cancer model systems; PDECs preserve the human tumor microenvironment, reflect inter- and intra-patient tumor heterogeneity and results are obtained within one week. Up to 16 experimental samples can be obtained from one clinical sample, allowing the same ‘patient' to be experimentally treated with different drug regimens and/or measurement of more than one endpoint.To assess whether PDECs are also a suitable model for IO studies we performed immunoprofiling experiments using FACS, RT-PCR, cytokine profiling and immunofluorescence, showing that the immune-contexture, i.e. the number and activation status of different immune cell types, is similar in PDECs than in the primary tumor they were derived from. Subsequently, we addressed whether the resident immune cells in PDECs were functional. Preliminary data showed that direct T cell activation in PDECs led to apoptosis of cancer cells and invoked a complex immune response, as assessed using cytokine profiling and FACS. Treatment with the immune checkpoint inhibitor anti-PD-1 had more moderate effects. These data reflect the clinical reality where very few BCa patients respond to immune checkpoint monotherapy, but also show that T cells in BCa can be activated to kill the cancer cells, suggesting that with the right treatment strategies, BCa can be targeted with immunotherapy. All in all, our data show that the tumor's immunocontexture is conserved in PDECs and that resident immune cells in PDECs can be activated to attack cancer cells, and thus support the use of PDECs as an ex vivo model system to study immuno-oncology in real human cancers for the development of the immunotherapies of tomorrow. Citation Format: Rita Turpin, Pauliina Munne, Ilida Suleymanova, Satu Mustjoki, Jeroen Pouwels, Juha Klefström. Patient-derived explant cultures (PDECs) as a model system for immuno-oncology studies [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1115.

Heparanase Promotes Syndecan-1 Expression to Mediate Fibrillar Collagen and Mammographic Density in Human Breast Tissue Cultured ex vivo.

Mammographic density (MD) is a strong and independent factor for breast cancer (BC) risk and is increasingly associated with BC progression. We have previously shown in mice that high MD, which is characterized by the preponderance of a fibrous stroma, facilitates BC xenograft growth and metastasis. This stroma is rich in extracellular matrix (ECM) factors, including heparan sulfate proteoglycans (HSPGs), such as the BC-associated syndecan-1 (SDC1). These proteoglycans tether growth factors, which are released by heparanase (HPSE). MD is positively associated with estrogen exposure and, in cell models, estrogen has been implicated in the upregulation of HPSE, the activity of which promotes SDC expression. Herein we describe a novel measurement approach (single-sided NMR) using a patient-derived explant (PDE) model of normal human (female) mammary tissue cultured ex vivo to investigate the role(s) of HPSE and SDC1 on MD. Relative HSPG gene and protein analyses determined in patient-paired high vs. low MD tissues identified SDC1 and SDC4 as potential mediators of MD. Using the PDE model we demonstrate that HPSE promotes SDC1 rather than SDC4 expression and cleavage, leading to increased MD. In this model system, synstatin (SSTN), an SDC1 inhibitory peptide designed to decouple SDC1-ITGαvβ3 parallel collagen alignment, reduced the abundance of fibrillar collagen as assessed by picrosirius red viewed under polarized light, and reduced MD. Our results reveal a potential role for HPSE in maintaining MD via its direct regulation of SDC1, which in turn physically tethers collagen into aligned fibers characteristic of MD. We propose that inhibitors of HPSE and/or SDC1 may afford an opportunity to reduce MD in high BC risk individuals and reduce MD-associated BC progression in conjunction with established BC therapies.

Nuclear PDCD4 Expression Defines a Subset of Luminal B-Like Breast Cancers with Good Prognosis.

The hormone receptor-positive (estrogen and/or progesterone receptor (PR)-positive) and HER2-negative breast cancer (BC) subtype is a biologically heterogeneous entity that includes luminal A-like (LumA-like) and luminal B-like (LumB-like) subtypes. Decreased PR levels is a distinctive biological feature of LumB-like tumors. These tumors also show reduced sensitivity to endocrine therapies and poorer prognosis than LumA-like tumors. Identification of biomarkers to accurately predict disease relapse in these subtypes is crucial in order to select effective therapies. We identified the tumor suppressor PDCD4 (programmed cell death 4), located in the nucleus (NPDCD4), as an independent prognostic factor of good clinical outcome in LumA-like and LumB-like subtypes. NPDCD4-positive LumB-like tumors presented overall and disease-free survival rates comparable to those of NPDCD4-positive LumA-like tumors, indicating that NPDCD4 improves the outcome of LumB-like patients. In contrast, NPDCD4 loss increased the risk of disease recurrence and death in LumB-like compared with LumA-like tumors. This, along with our results showing that LumB-like tumors present lower NPDCD4 positivity than LumA-like tumors, suggests that NPDCD4 loss contributes to endocrine therapy resistance in LumB-like BCs. We also revealed that PR induces PDCD4 transcription in LumB-like BC, providing a mechanistic explanation to the low PDCD4 levels in LumB-like BCs lacking PR. Finally, PDCD4 silencing enhanced BC cell survival in a patient-derived explant model of LumB-like disease. Our discoveries highlight NPDCD4 as a novel biomarker in LumA- and LumB-like subtypes, which could be included in the panel of immunohistochemical markers used in the clinic to accurately predict the prognosis of LumB-like tumors.

SAT-119 Targeting Glutamate Metabolism and Signaling in ER+, Endocrine Therapy-Resistant Breast Cancer

Estrogen receptor-positive (ER+) breast cancer is the most commonly diagnosed form of this malignancy. Aromatase inhibitors and selective estrogen receptor modulators or degraders (SERMS, SERDs) can be highly effective in treating ER+ breast cancer, but de novo and acquired resistance to these interventions is a persistent clinical problem. Endocrine therapy resistant breast cancer cells rewire their metabolism to support cellular demands associated with rapid proliferation and/or increased invasion and metastasis. An important feature of this metabolic flexibility is conversion of glutamine to glutamate, an amino acid integral to protection of cells from oxidative stress. Consistent with this, we show multiple cellular models of ER+, endocrine resistant breast cancer cells markedly increase glutamate release and upregulate expression of essential glutamine/glutamate metabolic enzymes and transporters, including the glutamate/cystine antiporter xCT, glutamate dehydrogenase (GLUD1/2), and/or the glutamine importer SLC1A5. Riluzole (RIL) is FDA-approved for the treatment of amyotrophic lateral sclerosis (ALS), and has several proposed mechanisms of action, including suppression of glutamate release and increased glutamate uptake. We show ER+, endocrine responsive and resistant breast cancer cells are growth-inhibited by RIL. This is due to an increase in cell death, particularly in endocrine resistant breast cancer cells, and cell cycle arrest. Interestingly, histologic subtype confers a different cell cycle arrest profile, with invasive ductal cancer (IDC) models arresting in G1 but invasive lobular cancer (ILC) models arresting in G2/M. Isobologram analysis of RIL plus SERMs or SERDs shows additive-to-synergistic activity in a subset of ER+ cell line models, and preliminary studies show combination activity in patient-derived explants (PDEs). Mechanistically, we tested whether signaling through metabotropic glutamate receptors (mGluRs, GRMs) and/or cystine import contribute to RIL’s growth-inhibitory phenotype. Antagonists of mGluRs/GRMs don’t phenocopy the effects of RIL, suggesting extracellular glutamate signaling through these receptors is not a key mechanism. Rescue experiments with β-mercaptoethanol to promote cystine uptake through transporters other than xCT show partial reversal of RIL-mediated cell cycle arrest in some cells, suggesting xCT may contribute to RIL-induced growth inhibition. In summary, we show RIL may be a viable addition to endocrine therapy in ER+ breast cancer. Ongoing studies will test additional mechanism(s) by which RIL may attenuate the growth of ER+ breast cancer models in vitro, including inhibition of protein kinase C and casein kinase 1 delta. We are further testing RIL efficacy alone and in combination with a SERD in primary tumors and lung metastases in a ER+ patient-derived xenograft (PDX) model.