RNA helicase DDX5 modulates sorafenib sensitivity in hepatocellular carcinoma via the Wnt/β-catenin-ferroptosis axis

Reduced expression of the RNA helicase DDX5 associated with increased hepatocellular carcinoma (HCC) tumor grade and poor patient survival following treatment with sorafenib. While immunotherapy is the first-line treatment for HCC, sorafenib and other multi-tyrosine kinase inhibitors (mTKIs) are widely used when immunotherapy is contra-indicated or fails. Herein, we elucidate the role of DDX5 in sensitizing HCC to sorafenib, offering new therapeutic strategies. Treatment of various human HCC cell lines with sorafenib/mTKIs downregulated DDX5 in vitro and in preclinical HCC models. Conversely, DDX5 overexpression reduced the viability of sorafenib-treated cells via ferroptosis, suggesting a role for DDX5 in sorafenib sensitivity. RNAseq of wild-type vs. DDX5-knockdown cells treated with or without sorafenib identified a set of common genes repressed by DDX5 and upregulated by sorafenib. This set significantly overlaps with Wnt signaling genes, including Disheveled-1 (DVL1), an indispensable Wnt activator and prognostic indicator of poor survival for sorafenib-treated patients. DDX5-knockout (DDX5KO) HCC cells exhibited DVL1 induction, Wnt/β-catenin pathway activation, and ferroptosis upon inhibition of canonical Wnt signaling. Consistently, xenograft HCC tumors exhibited reduced growth by inhibition of Wnt/β-catenin signaling via induction of ferroptosis. Significantly, overexpression of DDX5 in HCC xenografts repressed DVL1 expression and increased ferroptosis, resulting in reduced tumor growth by sorafenib. We conclude that DDX5 downregulation by sorafenib mediates adaptive resistance by activating Wnt/β-catenin signaling, leading to ferroptosis escape. Conversely, overexpression of DDX5 in vivo enhances the anti-tumor efficacy of sorafenib by suppressing Wnt/β-catenin activation and induction of ferroptosis. Thus, DDX5 overexpression in combination with mTKIs is a promising therapeutic strategy for HCC.

Over-expression of miR-183-5p or miR-492 triggers invasion and proliferation and loss of polarity in non-neoplastic breast epithelium

microRNAs (miRNAs) serve as novel noninvasive cancer biomarkers. In an HMT-3522 S1 (S1) breast epithelial risk-progression three-dimensional (3D) culture model, non-neoplastic S1 cells form a fully polarized epithelium. When silenced for the gap junction and tumor suppressor Cx43, Cx43-KO-S1 cells recapitulate pre-neoplastic phenotypes observed in tissues at risk for breast cancer in vivo. To delineate the role of miRNAs in breast tumorigenesis and identify key miRNA players in breast epithelial polarity, the miRNA profile specific to Cx43 loss in Cx43-KO-S1 compared to S1 cells was sequenced, revealing 65 differentially expressed miRNAs. A comparative analysis was conducted between these miRNAs and tumor-associated miRNAs from a young Lebanese patient validation cohort. miR-183-5p, downstream of Cx43 loss, was commonly upregulated in the patient cohort and the 3D culture model. miR-492, not attributed to Cx43 loss, was only specifically up-regulated in the young Lebanese patients. Ectopic expression of either miR-183-5p or miR-492 in S1 cells, through pLenti-III-miR-GPF vectors, resulted in the formation of larger multi-layered acini devoid of lumen, with disrupted epithelial polarity, as shown by an altered localization of Cx43, ß-catenin and Scrib, and decreased nuclear circularity in 3D cultures. Enhanced proliferation and invasion capacity were also observed. Over-expression of miR-183-5p or miR-492, therefore, induces pre-neoplastic phenotypes similar to those reported upon Cx43 loss, and may act as oncomiRs and possible biomarkers of increased breast cancer risk.

Engineered natural killer cells impede the immunometabolic CD73-adenosine axis in solid tumors

Immunometabolic reprogramming due to adenosine produced by CD73 (encoded by the 5’-ectonucleotidase gene NT5E) is a recognized immunosuppressive mechanism contributing to immune evasion in solid tumors. Adenosine is not only known to contribute to tumor progression, but it has specific roles in driving dysfunction of immune cells, including natural killer (NK) cells. Here, we engineered human NK cells to directly target the CD73-adenosine axis by blocking the enzymatic activity of CD73. In doing so, the engineered NK cells not only impaired adenosinergic metabolism driven by the hypoxic uptake of ATP by cancer cells in a model of non-small-cell lung cancer, but also mediated killing of tumor cells due to the specific recognition of overexpressed CD73. This resulted in a ‘single agent’ immunotherapy that combines antibody specificity, blockade of purinergic signaling, and killing of targets mediated by NK cells. We also showed that CD73-targeted NK cells are potent in vivo and result in tumor arrest, while promoting NK cell infiltration into CD73⁺ tumors and enhancing intratumoral activation.

Epigenetic aberrations of gene expression in a rat model of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is mostly triggered by environmental and life-style factors and may involve epigenetic aberrations. However, a comprehensive documentation of the link between the dysregulated epigenome, transcriptome, and liver carcinogenesis is lacking. In the present study, Fischer-344 rats were fed a choline-deficient (CDAA, cancer group) or choline-sufficient (CSAA, healthy group) L-amino acid-defined diet. At the end of 52 weeks, transcriptomic alterations in livers of rats with HCC tumours and healthy livers were investigated by RNA sequencing. DNA methylation and gene expression were assessed by pyrosequencing and quantitative reverse-transcription PCR (qRT-PCR), respectively. We discovered 1,848 genes that were significantly differentially expressed in livers of rats with HCC tumours (CDAA) as compared with healthy livers (CSAA). Upregulated genes in the CDAA group were associated with cancer-related functions, whereas macronutrient metabolic processes were enriched by downregulated genes. Changes of highest magnitude were detected in numerous upregulated genes that govern key oncogenic signalling pathways, including Notch, Wnt, Hedgehog, and extracellular matrix degradation. We further detected perturbations in DNA methylating and demethylating enzymes, which was reflected in decreased global DNA methylation and increased global DNA hydroxymethylation. Four selected upregulated candidates, Mmp12, Jag1, Wnt4, and Smo, demonstrated promoter hypomethylation with the most profound decrease in Mmp12. MMP12 was also strongly overexpressed and hypomethylated in human HCC HepG2 cells as compared with primary hepatocytes, which coincided with binding of Ten-eleven translocation 1 (TET1). Our findings provide comprehensive evidence for gene expression changes and dysregulated epigenome in HCC pathogenesis, potentially revealing novel targets for HCC prevention/treatment.

Multispecific targeting of glioblastoma with tumor microenvironment-responsive multifunctional engineered NK cells

Tumor antigen heterogeneity, a severely immunosuppressive tumor microenvironment (TME) and lymphopenia resulting in inadequate immune intratumoral trafficking, have rendered glioblastoma (GBM) highly resistant to therapy. To address these obstacles, here we describe a unique, sophisticated combinatorial platform for GBM: a cooperative multifunctional immunotherapy based on genetically engineered human natural killer (NK) cells bearing multiple antitumor functions including local tumor responsiveness that addresses key drivers of GBM resistance to therapy: antigen escape, immunometabolic reprogramming of immune responses, and poor immune cell homing. We engineered dual-specific chimeric antigen receptor (CAR) NK cells to bear a third functional moiety that is activated in the GBM TME and addresses immunometabolic suppression of NK cell function: a tumor-specific, locally released antibody fragment which can inhibit the activity of CD73 independently of CAR signaling and decrease the local concentration of adenosine. The multifunctional human NK cells targeted patient-derived GBM xenografts, demonstrated local tumor site-specific activity in the tissue, and potently suppressed adenosine production. We also unveil a complex reorganization of the immunological profile of GBM induced by inhibiting autophagy. Pharmacologic impairment of the autophagic process not only sensitized GBM to antigenic targeting by NK cells but promoted a chemotactic profile favorable to NK infiltration. Taken together, our study demonstrates a promising NK cell-based combinatorial strategy that can target multiple clinically recognized mechanisms of GBM progression simultaneously.

TIM-3 Expression Is Downregulated on Human NK Cells in Response to Cancer Targets in Synergy with Activation

Among natural killer (NK) cell receptors, the T-cell immunoglobulin and mucin-containing domain (TIM-3) has been associated with both inhibitory and activating functions, depending on context and activation pathway. Ex vivo and in vitro, expression of TIM-3 is inducible and depends on activation stimulus. Here, we report that TIM-3 expression can be downregulated on NK cells under specific conditions. When NK cells are exposed to cancer targets, they synergize with stimulation conditions to induce a substantial decrease in TIM-3 expression on their surface. We found that such downregulation occurs following prior NK activation. Downregulated TIM-3 expression correlated to lower cytotoxicity and lower interferon gamma (IFN-γ) expression, fueling the notion that TIM-3 might function as a benchmark for human NK cell dysfunction.

Abstract 278: Deciphering circular RNA axes in early-onset breast cancer using a breast epithelial risk-progression 3D culture model

One of the known circular RNAs (circRNAs) functions is to sponge microRNAs (miRNAs). Studies reported dysregulated mRNA-circRNA-miRNAs axes as biomarker signatures in breast cancer (BC). Signatures in early-onset BC have not been characterized, while its incidence increased drastically worldwide. To address this gap, we performed circRNAs microarrays and miRNA-Sequencing on 3D breast epithelial risk-progression HMT-3522 S1 (S1) culture model to identify dysregulated post-transcriptional axes. S1 (non-neoplastic) cells form fully-polarized epithelium and closely mimic normal in vivo mammary epithelial morphology, while (pre-tumorigenic) Cx43KO-S1 (S1 silenced with shRNA for gap junction Cx43) lose epithelial polarity, multi-layer and mimic tumor-initiated in vivo mammary epithelial morphology. We detected through circRNA microarray (Arraystar Human circRNA Array V2) 75 significantly upregulated and 46 significantly downregulated circRNAs (Fold Change > 2, p-value < 0.05) in 6 samples of Cx43KO-S1 versus S1 cells in 3D. We then selected 10 upregulated and 8 downregulated circRNAs for validation. Of these, hsa_circ_0007961 and hsa_circ_0081481 were most significantly upregulated while hsa_circ_0060055, hsa_circ_0005185, hsa_circ_0083442 and hsa_circ_0077755 were most significantly downregulated by RT-qPCR. We focus on “possible” sponging activity of validated circRNAs to their predicted (in silico) target miRNAs, and cross-compared their dysregulation to our validated miRNA datasets (from young patients and 3D cell line). When circRNAs are upregulated, they downregulate the function of their sponged miRNAs, hence preventing them from deregulating their target mRNAs. For upregulated hsa_circ_0007961 (originating from SPRED2 mRNA), its predicted miR-99a-3p was indeed downregulated in young Lebanese and matched U.S. BC patient tissues (and in Cx43KO-S1 versus S1 microarrays) and predicted miR-653 possessed treatment potential in BC cells. For the second upregulated hsa_circ_0081481 (originating from FBXO24 mRNA), its predicted miR-3960 was exclusively dysregulated in young Lebanese BC patients. For downregulated hsa_circ_0077755 (originating from Cx43mRNA), its predicted miR-182 was commonly upregulated in young Lebanese and matched US patients and predicted miR-203a-3p was reported as promising prognostic biomarker in triple negative BC. Additionally, all predicted miRNAs targeted by all detected circRNAs were predominantly enriched in pathways in cancer. Thus, we suggest that our 3D risk-progression culture model might help delineate post-transcriptional axes to better understand early-onset BC initiation. We reveal three validated axes: i) SPRED2/hsa_circ_0007961/miR-99a-3p or miR-653-5p, ii) FBXO24/hsa_circ_0081481/miR-3960 and iii) Cx43/hsa_circ_0077755/miR-182 or miR-203, which could serve as potential biomarker signatures in early-onset BC.

Citation Format: Nataly Naser Al Deen, Nadia Atallah Lanman, Shirisha Chittiboyina, Sophie Lelièvre, Heinrich Zu Dohna, Mounir AbouHaidar, Rabih Talhouk. Deciphering circular RNA axes in early-onset breast cancer using a breast epithelial risk-progression 3D culture model [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 278.

scTree: An R package to generate antibody-compatible classifiers from single-cell sequencing data

Single-cell RNA sequencing (scRNA-seq) is now a commonly used technique to measure the transcriptome of populations of cells. Clustering heterogeneous cells based on these transcriptomes enables identification of cell populations (Butler, Hoffman, Smibert, Papalexi, & Satija, 2018; Trapnell et al., 2014). There are multiple methods available to identify "marker" genes that differ between these populations (Butler et al., 2018; Love, Huber, & Anders, 2014; Robinson, McCarthy, & Smyth, 2009). However, there are usually too many genes in these lists to directly suggest an experimental follow-up strategy for selecting them from a bulk population (e.g. via FACS (Tung et al., 2007)). Here we present scTree, a tool that aims to provide biologists using the R programming language and scRNA-seq analysis programs a minimal set of genes that can be used in downstream experiments. The package is free, open source and available though GitHub at github.com/jspaezp/sctree.

Ductal Carcinoma In Situ Progression in Dog Model of Breast Cancer

The mechanisms that drive ductal carcinoma in situ (DCIS) progression to invasive cancer are not clear. Studying DCIS progression in humans is challenging and not ethical, thus necessitating the characterization of an animal model that faithfully resembles human disease. We have characterized a canine model of spontaneous mammary DCIS and invasive cancer that shares histologic, molecular, and diagnostic imaging characteristics with DCIS and invasive cancer in women. The purpose of the study was to identify markers and altered signaling pathways that lead to invasive cancer and shed light on early molecular events in breast cancer progression and development. Transcriptomic studies along the continuum of cancer progression in the mammary gland from healthy, through atypical ductal hyperplasia (ADH), DCIS, and invasive carcinoma were performed using the canine model. Gene expression profiles of preinvasive DCIS lesions closely resemble those of invasive carcinoma. However, certain genes, such as SFRP2, FZD2, STK31, and LALBA, were over-expressed in DCIS compared to invasive cancer. The over-representation of myoepithelial markers, epithelial-mesenchymal transition (EMT), canonical Wnt signaling components, and other pathways induced by Wnt family members distinguishes DCIS from invasive. The information gained may help in stratifying DCIS as well as identify actionable targets for primary and tertiary prevention or targeted therapy.