Search

Abstract Objectives There is an urgent need to be able to identify individuals with asymptomatic Leishmania donovani infection, so their risk of progressing to VL and transmitting parasites can be managed. This study examined transcriptional markers expressed by CD4+ T cells that could distinguish asymptomatic individuals from endemic controls and visceral leishmaniasis (VL) patients. Methods CD4+ T cells were isolated from individuals with asymptomatic L. donovani infection, endemic controls and VL patients. RNA was extracted and RNAseq employed to identify differentially expressed genes. The expression of one gene and its protein product during asymptomatic infection were evaluated. Results Amphiregulin (AREG) was identified as a distinguishing gene product in CD4+ T cells from individuals with asymptomatic L. donovani infection, compared to VL patients and healthy endemic control individuals. AREG levels in plasma and antigen‐stimulated whole‐blood assay cell culture supernatants were significantly elevated in asymptomatic individuals, compared to endemic controls and VL patients. Regulatory T (Treg) cells were identified as an important source of AREG amongst CD4+ T‐cell subsets in asymptomatic individuals. Conclusion Increased Treg cell AREG expression was identified in individuals with asymptomatic L. donovani infection, suggesting the presence of an ongoing inflammatory response in these individuals required for controlling infection and that AREG may play an important role in preventing inflammation‐induced tissue damage and subsequent disease in asymptomatic individuals.

T-cell-engaging bispecific antibody (T-BsAb, also known as BiTE) therapy has emerged as a powerful therapeutic modality against multiple myeloma. Given that T-BsAb therapy redirects endogenous T cells to eliminate tumor cells, reinvigorating dysfunctional T cells may be a potential approach to improve the efficacy of T-BsAb. While various immunostimulatory cytokines can potentiate effector T-cell functions, the optimal cytokine treatment for T-BsAb therapy is yet to be established, partly due to a concern of cytokine release syndrome driven by aberrant interferon (IFN)-γ production. Here, we functionally screen immunostimulatory cytokines to determine an ideal combination partner for T-BsAb therapy. This approach reveals IL-21 as a potential immunostimulatory cytokine with the ability to augment T-BsAb-mediated release of granzyme B and perforin, without increasing IFN-γ release. Transcriptome profiling and functional characterization strongly support that IL-21 selectively targets the cytotoxic granule exocytosis pathway, but not pro-inflammatory responses. Notably, IL-21 modulates multiple steps of cytotoxic effector functions including upregulation of co-activating CD226 receptor, increasing cytotoxic granules, and promoting cytotoxic granule delivery at the immunological synapse. Indeed, T-BsAb-mediated myeloma-killing is cytotoxic granule-dependent, and IL-21 priming significantly augments cytotoxic activities. Furthermore, in vivo IL-21 treatment induces cytotoxic effector reprogramming in bone marrow T cells, showing synergistic anti-myeloma effects in combination with T-BsAb therapy. Together, harnessing the cytotoxic granule exocytosis pathway by IL-21 may be a potential approach to achieve better responses by T-BsAb therapy.

Lakis et al. report novel findings related to the epigenetic landscape of pancreatic neuroendocrine cancer. This relatively large cohort provides functional insights into the epigenetic wiring of pancreatic neuroendocrine tumor sub-types with the potential ability to stratify tumours of different prognosis.

Abstract For complex machine learning (ML) algorithms to gain widespread acceptance in decision making, we must be able to identify the features driving the predictions. Explainability models allow transparency of ML algorithms, however their reliability within high-dimensional data is unclear. To test the reliability of the explainability model SHapley Additive exPlanations (SHAP), we developed a convolutional neural network to predict tissue classification from Genotype-Tissue Expression (GTEx) RNA-seq data representing 16,651 samples from 47 tissues. Our classifier achieved an average F1 score of 96.1% on held-out GTEx samples. Using SHAP values, we identified the 2423 most discriminatory genes, of which 98.6% were also identified by differential expression analysis across all tissues. The SHAP genes reflected expected biological processes involved in tissue differentiation and function. Moreover, SHAP genes clustered tissue types with superior performance when compared to all genes, genes detected by differential expression analysis, or random genes. We demonstrate the utility and reliability of SHAP to explain a deep learning model and highlight the strengths of applying ML to transcriptome data.

Abstract Cells within the tumour microenvironment (TME) can impact tumour development and influence treatment response. Computational approaches have been developed to deconvolve the TME from bulk RNA-seq. Using scRNA-seq profiling from breast tumours we simulate thousands of bulk mixtures, representing tumour purities and cell lineages, to compare the performance of nine TME deconvolution methods (BayesPrism, Scaden, CIBERSORTx, MuSiC, DWLS, hspe, CPM, Bisque, and EPIC). Some methods are more robust in deconvolving mixtures with high tumour purity levels. Most methods tend to mis-predict normal epithelial for cancer epithelial as tumour purity increases, a finding that is validated in two independent datasets. The breast cancer molecular subtype influences this mis-prediction. BayesPrism and DWLS have the lowest combined numbers of false positives and false negatives, and have the best performance when deconvolving granular immune lineages. Our findings highlight the need for more single-cell characterisation of rarer cell types, and suggest that tumour cell compositions should be considered when deconvolving the TME.

Abstract Oesophageal adenocarcinoma is a poor prognosis cancer and the molecular features underpinning response to treatment remain unclear. We investigate whole genome, transcriptomic and methylation data from 115 oesophageal adenocarcinoma patients mostly from the DOCTOR phase II clinical trial (Australian New Zealand Clinical Trials Registry-ACTRN12609000665235), with exploratory analysis pre-specified in the study protocol of the trial. We report genomic features associated with poorer overall survival, such as the APOBEC mutational and RS3-like rearrangement signatures. We also show that positron emission tomography non-responders have more sub-clonal genomic copy number alterations. Transcriptomic analysis categorises patients into four immune clusters correlated with survival. The immune suppressed cluster is associated with worse survival, enriched with myeloid-derived cells, and an epithelial-mesenchymal transition signature. The immune hot cluster is associated with better survival, enriched with lymphocytes, myeloid-derived cells, and an immune signature including CCL5, CD8A, and NKG7. The immune clusters highlight patients who may respond to immunotherapy and thus may guide future clinical trials.

Abstract Uncertainty estimation is crucial for understanding the reliability of deep learning (DL) predictions, and critical for deploying DL in the clinic. Differences between training and production datasets can lead to incorrect predictions with underestimated uncertainty. To investigate this pitfall, we benchmarked one pointwise and three approximate Bayesian DL models for predicting cancer of unknown primary, using three RNA-seq datasets with 10,968 samples across 57 cancer types. Our results highlight that simple and scalable Bayesian DL significantly improves the generalisation of uncertainty estimation. Moreover, we designed a prototypical metric—the area between development and production curve (ADP), which evaluates the accuracy loss when deploying models from development to production. Using ADP, we demonstrate that Bayesian DL improves accuracy under data distributional shifts when utilising ‘uncertainty thresholding’. In summary, Bayesian DL is a promising approach for generalising uncertainty, improving performance, transparency, and safety of DL models for deployment in the real world.

Abstract Background High-grade serous ovarian carcinomas (HGSCs) are a heterogeneous subtype of epithelial ovarian cancers and include serous cancers arising in the fallopian tube and peritoneum. These cancers are now subdivided into homologous recombination repair (HR)-deficient and proficient subgroups as this classification impacts on management and prognosis. PARP inhibitors (PARPi) have shown significant clinical efficacy, particularly as maintenance therapy following response to platinum-based chemotherapy in BRCA-mutant or homologous recombination (HR)-deficient HGSCs in both the 1st and 2nd line settings. However, PARPi have limited clinical benefit in HR-proficient HGSCs which make up almost 50% of HGSC and improving outcomes in these patients is now a high priority due to the poor prognosis with ineffectiveness of the current standard of care. There are a number of potential lines of investigation including efforts in sensitizing HR-proficient tumors to PARPi. Herein, we aimed to develop a novel combination therapy by targeting SSRP1 using a small molecule inhibitor CBL0137 with PARPi in HR-proficient HGSCs. Experimental design We tested anti-cancer activity of CBL0137 monotherapy using a panel of HGSC cell lines and patient-derived tumor cells in vitro. RNA sequencing was used to map global transcriptomic changes in CBL0137-treated patient-derived HR-proficient HGSC cells. We tested efficacy of CBL0137 in combination with PARPi using HGSC cell lines and patient-derived tumor cells in vitro and in vivo. Results We show that SSRP1 inhibition using a small molecule, CBL0137, that traps SSRP1 onto chromatin, exerts a significant anti-growth activity in vitro against HGSC cell lines and patient-derived tumor cells, and also reduces tumor burden in vivo. CBL0137 induced DNA repair deficiency via inhibition of the HR repair pathway and sensitized SSRP1-high HR-proficient HGSC cell lines and patient-derived tumor cells/xenografts to the PARPi, Olaparib in vitro and in vivo. CBL0137 also enhanced the efficacy of DNA damaging platinum-based chemotherapy in HGSC patient-derived xenografts. Conclusion Our findings strongly suggest that combination of CBL0137 and PARP inhibition represents a novel therapeutic strategy for HR-proficient HGSCs that express high levels of SSRP1 and should be investigated in the clinic.

Comparing single cell transcriptome data to clinical characteristics and in vitro growth profiles uncovers a potential role for divergent mesenchymal-derived stromal fibroblast maturation in endometriosis susceptibility.

Control of intracellular parasites responsible for malaria requires host IFN-γ+T-bet+CD4+ T cells (Th1 cells) with IL-10 produced by Th1 cells to mitigate the pathology induced by this inflammatory response. However, these IL-10–producing Th1 (induced type I regulatory [Tr1]) cells can also promote parasite persistence or impair immunity to reinfection or vaccination. Here, we identified molecular and phenotypic signatures that distinguished IL-10–Th1 cells from IL-10+Tr1 cells in Plasmodium falciparum–infected people who participated in controlled human malaria infection studies, as well as C57BL/6 mice with experimental malaria caused by P. berghei ANKA. We also identified a conserved Tr1 cell molecular signature shared between patients with malaria, dengue, and graft-versus-host disease. Genetic manipulation of primary human CD4+ T cells showed that the transcription factor cMAF played an important role in the induction of IL-10, while BLIMP-1 promoted the development of human CD4+ T cells expressing multiple coinhibitory receptors. We also describe heterogeneity of Tr1 cell coinhibitory receptor expression that has implications for targeting these molecules for clinical advantage during infection. Overall, this work provides insights into CD4+ T cell development during malaria that offer opportunities for creation of strategies to modulate CD4+ T cell functions and improve antiparasitic immunity.

Abstract Treatment options for COVID-19 remain limited, especially during the early or asymptomatic phase. Here, we report a novel SARS-CoV-2 viral replication mechanism mediated by interactions between ACE2 and the epigenetic eraser enzyme LSD1, and its interplay with the nuclear shuttling importin pathway. Recent studies have shown a critical role for the importin pathway in SARS-CoV-2 infection, and many RNA viruses hijack this axis to re-direct host cell transcription. LSD1 colocalized with ACE2 at the cell surface to maintain demethylated SARS-CoV-2 spike receptor-binding domain lysine 31 to promote virus–ACE2 interactions. Two newly developed peptide inhibitors competitively inhibited virus–ACE2 interactions, and demethylase access to significantly inhibit viral replication. Similar to some other predominantly plasma membrane proteins, ACE2 had a novel nuclear function: its cytoplasmic domain harbors a nuclear shuttling domain, which when demethylated by LSD1 promoted importin-α-dependent nuclear ACE2 entry following infection to regulate active transcription. A novel, cell permeable ACE2 peptide inhibitor prevented ACE2 nuclear entry, significantly inhibiting viral replication in SARS-CoV-2-infected cell lines, outperforming other LSD1 inhibitors. These data raise the prospect of post-exposure prophylaxis for SARS-CoV-2, either through repurposed LSD1 inhibitors or new, nuclear-specific ACE2 inhibitors.