Genetic Redundancy, Functional Compensation, and Cancer Vulnerability

An interpretable artificial intelligence framework for designing synthetic lethality-based anti-cancer combination therapies

INTRODUCTION

Synthetic lethality (SL) provides an opportunity to leverage different genetic interactions when designing synergistic combination therapies. To further explore SL-based combination therapies for cancer treatment, it is important to identify and mechanistically characterize more SL interactions. Artificial intelligence (AI) methods have recently been proposed for SL prediction, but the results of these models are often not interpretable such that deriving the underlying mechanism can be challenging.

OBJECTIVES

This study aims to develop an interpretable AI framework for SL prediction and subsequently utilize it to design SL-based synergistic combination therapies.

METHODS

We propose a knowledge and data dual-driven AI framework for SL prediction (KDDSL). Specifically, we use gene knowledge related to the SL mechanism to guide the construction of the model and develop a method to identify the most relevant gene knowledge for the predicted results.

RESULTS

Experimental and literature-based validation confirmed a good balance between predictive and interpretable ability when using KDDSL. Moreover, we demonstrated that KDDSL could help to discover promising drug combinations and clarify associated biological processes, such as the combination of MDM2 and CDK9 inhibitors, which exhibited significant anti-cancer effects in vitro and in vivo.

CONCLUSION

These data underscore the potential of KDDSL to guide SL-based combination therapy design. There is a need for biomedicine-focused AI strategies to combine rational biological knowledge with developed models.

Tissue-Predisposition to Cancer Driver Mutations

Driver mutations are considered the cornerstone of cancer initiation. They are defined as mutations that convey a competitive fitness advantage, and hence, their mutation frequency in premalignant tissue is expected to exceed the basal mutation rate. In old terms, that translates to "the survival of the fittest" and implies that a selective process underlies the frequency of cancer driver mutations. In that sense, each tissue is its own niche that creates a molecular selective pressure that may favor the propagation of a mutation or not. At the heart of this stands one of the biggest riddles in cancer biology: the tissue-predisposition to cancer driver mutations. The frequency of cancer driver mutations among tissues is non-uniform: for instance, mutations in APC are particularly frequent in colorectal cancer, and 99% of chronic myeloid leukemia patients harbor the driver BCR-ABL1 fusion mutation, which is rarely found in solid tumors. Here, we provide a mechanistic framework that aims to explain how tissue-specific features, ranging from epigenetic underpinnings to the expression of viral transposable elements, establish a molecular basis for selecting cancer driver mutations in a tissue-specific manner.

Cancer genomes tolerate deleterious coding mutations through somatic copy number amplifications of wild-type regions

Cancers evolve under the accumulation of thousands of somatic mutations and chromosomal aberrations. While most coding mutations are deleterious, almost all protein-coding genes lack detectable signals of negative selection. This raises the question of how tumors tolerate such large amounts of deleterious mutations. Using 8,690 tumor samples from The Cancer Genome Atlas, we demonstrate that copy number amplifications frequently cover haploinsufficient genes in mutation-prone regions. This could increase tolerance towards the deleterious impact of mutations by creating safe copies of wild-type regions and, hence, protecting the genes therein. Our findings demonstrate that these potential buffering events are highly influenced by gene functions, essentiality, and mutation impact and that they occur early during tumor evolution. We show how cancer type-specific mutation landscapes drive copy number alteration patterns across cancer types. Ultimately, our work paves the way for the detection of novel cancer vulnerabilities by revealing genes that fall within amplifications likely selected during evolution to mitigate the effect of mutations.

Semantic clustering analysis of E3-ubiquitin ligases in gastrointestinal tract defines genes ontology clusters with tissue expression patterns

Background

Ubiquitin ligases (Ub-ligases) are essential intracellular enzymes responsible for the regulation of proteome homeostasis, signaling pathway crosstalk, cell differentiation and stress responses. Individual Ub-ligases exhibit their unique functions based on the nature of their substrates. They create a complex regulatory network with alternative and feedback pathways to maintain cell homeostasis, being thus important players in many physiological and pathological conditions. However, the functional classification of Ub-ligases needs to be revised and extended.

Methods

In the current study, we used a novel semantic biclustering technique for expression profiling of Ub-ligases and ubiquitination-related genes in the murine gastrointestinal tract (GIT). We accommodated a general framework of the algorithm for finding tissue-specific gene expression clusters in GIT. In order to test identified clusters in a biological system, we used a model of epithelial regeneration. For this purpose, a dextran sulfate sodium (DSS) mouse model, following with in situ hybridization, was used to expose genes with possible compensatory features. To determine cell-type specific distribution of Ub-ligases and ubiquitination-related genes, principal component analysis (PCA) and Uniform Manifold Approximation and Projection technique (UMAP) were used to analyze the Tabula Muris scRNA-seq data of murine colon followed by comparison with our clustering results.

Results

Our established clustering protocol, that incorporates the semantic biclustering algorithm, demonstrated the potential to reveal interesting expression patterns. In this manner, we statistically defined gene clusters consisting of the same genes involved in distinct regulatory pathways vs distinct genes playing roles in functionally similar signaling pathways. This allowed us to uncover the potentially redundant features of GIT-specific Ub-ligases and ubiquitination-related genes. Testing the statistically obtained results on the mouse model showed that genes clustered to the same ontology group simultaneously alter their expression pattern after induced epithelial damage, illustrating their complementary role during tissue regeneration.

Conclusions

An optimized semantic clustering protocol demonstrates the potential to reveal a readable and unique pattern in the expression profiling of GIT-specific Ub-ligases, exposing ontologically relevant gene clusters with potentially redundant features. This extends our knowledge of ontological relationships among Ub-ligases and ubiquitination-related genes, providing an alternative and more functional gene classification. In a similar way, semantic cluster analysis could be used for studding of other enzyme families, tissues and systems.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-022-02265-2.

Artificial Intelligence in Bulk and Single-Cell RNA-Sequencing Data to Foster Precision Oncology

Artificial intelligence, or the discipline of developing computational algorithms able to perform tasks that requires human intelligence, offers the opportunity to improve our idea and delivery of precision medicine. Here, we provide an overview of artificial intelligence approaches for the analysis of large-scale RNA-sequencing datasets in cancer. We present the major solutions to disentangle inter- and intra-tumor heterogeneity of transcriptome profiles for an effective improvement of patient management. We outline the contributions of learning algorithms to the needs of cancer genomics, from identifying rare cancer subtypes to personalizing therapeutic treatments.

Transcriptional Silencing of ALDH2 Confers a Dependency on Fanconi Anemia Proteins in Acute Myeloid Leukemia

Hundreds of genes become aberrantly silenced in acute myeloid leukemia (AML), with most of these epigenetic changes being of unknown functional consequence. Here, we demonstrate how gene silencing can lead to an acquired dependency on the DNA repair machinery in AML. We make this observation by profiling the essentiality of the ubiquitination machinery in cancer cell lines using domain-focused CRISPR screening, which revealed Fanconi anemia (FA) proteins UBE2T and FANCL as unique dependencies in AML. We demonstrate that these dependencies are due to a synthetic lethal interaction between FA proteins and aldehyde dehydrogenase 2 (ALDH2), which function in parallel pathways to counteract the genotoxicity of endogenous aldehydes. We show DNA hypermethylation and silencing of ALDH2 occur in a recurrent manner in human AML, which is sufficient to confer FA pathway dependency. Our study suggests that targeting of the ubiquitination reaction catalyzed by FA proteins can eliminate ALDH2-deficient AML. SIGNIFICANCE: Aberrant gene silencing is an epigenetic hallmark of human cancer, but the functional consequences of this process are largely unknown. In this study, we show how an epigenetic alteration leads to an actionable dependency on a DNA repair pathway through the disabling of genetic redundancy.This article is highlighted in the In This Issue feature, p. 2113.

Prodigiosin/PU-H71 as a novel potential combined therapy for triple negative breast cancer (TNBC): preclinical insights

Prodigiosin, a secondary metabolite red pigment produced by Serratia marcescens, has an interesting apoptotic efficacy against cancer cell lines with low or no toxicity on normal cells. HSP90α is known as a crucial and multimodal target in the treatment of TNBC. Our research attempts to assess the therapeutic potential of prodigiosin/PU-H71 combination on MDA-MB-231 cell line. The transcription and protein expression levels of different signalling pathways were assessed. Treatment of TNBC cells with both drugs resulted in a decrease of the number of adherent cells with apoptotic effects. Prodigiosin/PU-H71 combination increased the levels of caspases 3,8 and 9 and decreased the levels of mTOR expression. Additionally, there was a remarkable decrease of HSP90α transcription and expression levels upon treatment with combined therapy. Also, EGFR and VEGF expression levels decreased. This is the first study to show that prodigiosin/PU-H71 combination had potent cytotoxicity on MDA-MB-231 cells; proving to play a paramount role in interfering with key signalling pathways in TNBC. Interestingly, prodigiosin might be a potential anticancer agent to increase the sensitivity of TNBC cells to apoptosis. This study provides a new basis for upcoming studies to overcome drug resistance in TNBC cells.

A Division of Labor between YAP and TAZ in Non-Small Cell Lung Cancer

Lung cancer is the leading cause of cancer-related deaths worldwide. The paralogous transcriptional cofactors Yes-associated protein (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ, also called WWTR1), the main downstream effectors of the Hippo signal transduction pathway, are emerging as pivotal determinants of malignancy in lung cancer. Traditionally, studies have tended to consider YAP and TAZ as functionally redundant transcriptional cofactors with similar biological impact. However, there is growing evidence that each of them also possesses distinct attributes. Here we sought to systematically characterize the division of labor between YAP and TAZ in non-small cell lung cancer (NSCLC), the most common histological subtype of lung cancer. Representative NSCLC cell lines as well as patient-derived data showed that the two paralogs orchestrated nonoverlapping transcriptional programs in this cancer type. YAP preferentially regulated gene sets associated with cell division and cell-cycle progression, whereas TAZ preferentially regulated genes associated with extracellular matrix organization. Depletion of YAP resulted in growth arrest, whereas its overexpression promoted cell proliferation. Likewise, depletion of TAZ compromised cell migration, whereas its overexpression enhanced migration. The differential effects of YAP and TAZ on key cellular processes were also associated with differential response to anticancer therapies. Uncovering the different activities and downstream effects of YAP and TAZ may thus facilitate better stratification of patients with lung cancer for anticancer therapies. SIGNIFICANCE: Thease findings show that oncogenic paralogs YAP and TAZ have distinct roles in NSCLC and are associated with differential response to anticancer drugs, knowledge that may assist lung cancer therapy decisions.

KIF11 and KIF15 mitotic kinesins are potential therapeutic vulnerabilities for malignant peripheral nerve sheath tumors

Background

Malignant peripheral nerve sheath tumor (MPNST) constitutes the leading cause of neurofibromatosis type 1–related mortality. MPNSTs contain highly rearranged hyperploid genomes and exhibit a high division rate and aggressiveness. We have studied in vitro whether the mitotic kinesins KIF11, KIF15, and KIF23 have a functional role in maintaining MPNST cell survival and can represent potential therapeutic vulnerabilities.

Methods

We studied the expression of kinesin mRNAs and proteins in tumors and cell lines and used several in vitro functional assays to analyze the impact of kinesin genetic suppression (KIF15, KIF23) and drug inhibition (KIF11) in MPNST cells. We also performed in vitro combined treatments targeting KIF11 together with other described MPNST targets.

Results

The studied kinesins were overexpressed in MPNST samples. KIF15 and KIF23 were required for the survival of MPNST cell lines, which were also more sensitive than benign control fibroblasts to the KIF11 inhibitors ispinesib and ARRY-520. Co-targeting KIF11 and BRD4 with ARRY-520 and JQ1 reduced MPNST cell viability, synergistically killing a much higher proportion of MPNST cells than control fibroblasts. In addition, genetic suppression of KIF15 conferred an increased sensitivity to KIF11 inhibitors alone or in combination with JQ1.

Conclusions

The mitotic spindle kinesins KIF11 and KIF15 and the cytokinetic kinesin KIF23 play a clear role in maintaining MPNST cell survival and may represent potential therapeutic vulnerabilities. Although further in vivo evidences are still mandatory, we propose a simultaneous suppression of KIF11, KIF15, and BRD4 as a potential therapy for MPNSTs.

Next Generation Networks: Featuring the Potential Role of Emerging Applications in Translational Oncology

Nowadays, networks are pervasively used as examples of models suitable to mathematically represent and visualize the complexity of systems associated with many diseases, including cancer. In the cancer context, the concept of network entropy has guided many studies focused on comparing equilibrium to disequilibrium (i.e., perturbed) conditions. Since these conditions reflect both structural and dynamic properties of network interaction maps, the derived topological characterizations offer precious support to conduct cancer inference. Recent innovative directions have emerged in network medicine addressing especially experimental omics approaches integrated with a variety of other data, from molecular to clinical and also electronic records, bioimaging etc. This work considers a few theoretically relevant concepts likely to impact the future of applications in personalized/precision/translational oncology. The focus goes to specific properties of networks that are still not commonly utilized or studied in the oncological domain, and they are: controllability, synchronization and symmetry. The examples here provided take inspiration from the consideration of metastatic processes, especially their progression through stages and their hallmark characteristics. Casting these processes into computational frameworks and identifying network states with specific modular configurations may be extremely useful to interpret or even understand dysregulation patterns underlying cancer, and associated events (onset, progression) and disease phenotypes.

Neglected Functions of TFCP2/TFCP2L1/UBP1 Transcription Factors May Offer Valuable Insights into Their Mechanisms of Action

In recent years, the TFCP2 (transcription factor cellular promoter 2)/TFCP2L1 (TFCP2-like 1)/UBP1 (upstream binding protein 1) subfamily of transcription factors has been attracting increasing attention in the scientific community. These factors are very important in cancer, Alzheimer’s disease, and other human conditions, and they can be attractive targets for drug development. However, the interpretation of experimental results is complicated, as in principle, any of these factors could substitute for the lack of another. Thus, studying their hitherto little known functions should enhance our understanding of mechanisms of their functioning, and analogous mechanisms might govern their functioning in medically relevant contexts. For example, there are numerous parallels between placental development and cancer growth; therefore, investigating the roles of TFCP2, TFCP2L1, and UBP1 in the placenta may help us better understand their functioning in cancer, as is evidenced by the studies of various other proteins and pathways. Our review article aims to call the attention of the scientific community to these neglected functions, and encourage further research in this field. Here, we present a systematic review of current knowledge of the TFCP2/TFCP2L1/UBP1 subfamily in reproduction, embryonic development, renal function, blood-pressure regulation, brain function, and other processes, where their involvement has not been studied much until now.

TFCP2/TFCP2L1/UBP1 transcription factors in cancer

The TFCP2/Grainyhead family of transcription factors is divided into two distinct subfamilies, one of which includes the Grainyhead-like 1-3 (GRHL1-3) proteins and the other consists of TFCP2 (synonyms: CP2, LSF, LBP-1c), TFCP2L1 (synonyms: CRTR-1, LBP-9) and UBP1 (synonyms: LBP-1a, NF2d9). Transcription factors from the TFCP2/TFCP2L1/UBP1 subfamily are involved in various aspects of cancer development. TFCP2 is a pro-oncogenic factor in hepatocellular carcinoma, pancreatic cancer and breast cancer, may be important in cervical carcinogenesis and in colorectal cancer. TFCP2 can also act as a tumor suppressor, for example, it inhibits melanoma growth. Furthermore, TFCP2 is involved in epithelial-mesenchymal transition and enhances angiogenesis. TFCP2L1 maintains pluripotency and self-renewal of embryonic stem cells and was implicated in a wide variety of cancers, including clear cell renal cell carcinoma, breast cancer and thyroid cancer. Here we present a systematic review of current knowledge of this protein subfamily in the context of cancer. We also discuss potential challenges in investigating this family of transcription factors. These challenges include redundancies between these factors as well as their interactions with each other and their ability to modulate each other's activity.

Synthetic lethal interaction between the tumour suppressor STAG2 and its paralog STAG1

Cohesin is a multi-protein complex that tethers sister chromatids during mitosis and mediates DNA repair, genome compartmentalisation and regulation of gene expression. Cohesin subunits frequently acquire cancer loss-of-function alterations and act as tumour suppressors in several tumour types. This has led to increased interest in cohesin as potential target in anti-cancer therapy. Here we show that the loss-of-function of STAG2, a core component of cohesin and an emerging tumour suppressor, leads to synthetic dependency of mutated cancer cells on its paralog STAG1. STAG1 and STAG2 share high sequence identity, encode mutually exclusive cohesin subunits and retain partially overlapping functions. We inhibited STAG1 and STAG2 in several cancer cell lines where the two genes have variable mutation and copy number status. In all cases, we observed that the simultaneous blocking of STAG1 and STAG2 significantly reduces cell proliferation. We further confirmed the synthetic lethal interaction developing a vector-free CRISPR system to induce STAG1/STAG2 double gene knockout. We provide strong evidence that STAG1 is a promising therapeutic target in cancers with inactivating alterations of STAG2.

Fitness effects of mutation: testing genetic redundancy in Arabidopsis thaliana

Screens of organisms with disruptive mutations in a single gene often fail to detect phenotypic consequences for the majority of mutants. One explanation for this phenomenon is that the presence of paralogous loci provides genetic redundancy. However, it is also possible that the assayed traits are affected by few loci, that effects could be subtle or that phenotypic effects are restricted to certain environments. We assayed a set of T-DNA insertion mutant lines of Arabidopsis thaliana to determine the frequency with which mutation affected fitness-related phenotypes. We found that between 8% and 42% of the assayed lines had altered fitness from the wild type. Furthermore, many of these lines exhibited fitness greater than the wild type. In a second experiment, we grew a subset of the lines in multiple environments and found whether a T-DNA insert increased or decreased fitness traits depended on the assay environment. Overall, our evidence contradicts the hypothesis that genetic redundancy is a common phenomenon in A. thaliana for fitness traits. Evidence for redundancy from prior screens of knockout mutants may often be an artefact of the design of the phenotypic assays which have focused on less complex phenotypes than fitness and have used single environments. Finally, our study adds to evidence that beneficial mutations may represent a significant component of the mutational spectrum of A. thaliana.