Automated Generation of Novel Fragments Using Screening Data, a Dual SMILES Autoencoder, Transfer Learning and Syntax Correction.

Fragment-based hit identification (FBHI) allows proportionately greater coverage of chemical space using fewer molecules than traditional high-throughput screening approaches. However, effectively exploiting this advantage is highly dependent on the library design. Solubility, stability, chemical complexity, chemical/shape diversity, and synthetic tractability for fragment elaboration are all critical aspects, and molecule design remains a time-consuming task for computational and medicinal chemists. Artificial neural networks have attracted considerable attention in automated de novo design applications and could also prove useful for fragment library design. Chemical autoencoders are neural networks consisting of encoder and decoder parts, which respectively compress and decompress molecular representations. The decoder is applied to samples drawn from the space of compressed representations to generate novel molecules that can be scored for properties of interest. Here, we report an autoencoder model using a recurrent neural network architecture, which was trained using 486,565 fragments curated from commercial sources, to simultaneously reconstruct both SMILES and chemical fingerprints. To explore its utility in fragment design, we applied transfer learning to the fingerprint decoder layers to train a classifier using 66 frequent hitter fragments identified from our screening campaigns. Using a particle swarm optimization sampling approach, we compare the performance of this "dual" model to an architecture encoding SMILES only. The dual model produced valid SMILES with improved features, considering a range of properties including aromatic ring counts, heavy atom count, synthetic accessibility, and a new fragment complexity score we term Feature Complexity (FeCo). Additionally, we demonstrate that generative performance is further enhanced by use of a simple syntax-correction procedure during training, in which invalid and undesirable SMILES are spiked into the training set. Finally, we used the syntax-corrected model to generate a library of novel candidate privileged fragments.

A 'synthetic-sickness' screen for senescence re-engagement targets in mutant cancer backgrounds.

Senescence is a universal barrier to immortalisation and tumorigenesis. As such, interest in the use of senescence-induction in a therapeutic context has been gaining momentum in the past few years; however, senescence and immortalisation remain underserved areas for drug discovery owing to a lack of robust senescence inducing agents and an incomplete understanding of the signalling events underlying this complex process. In order to address this issue we undertook a large-scale morphological siRNA screen for inducers of senescence phenotypes in the human melanoma cell line A375P. Following rescreen and validation in a second cancer cell line, HCT116 colorectal carcinoma, a panel of 16 of the most robust hits were selected for further validation based on significance and the potential to be targeted by drug-like molecules. Using secondary assays for detection of senescence biomarkers p21, 53BP1 and senescence associated beta-galactosidase (SAßGal) in a panel of HCT116 cell lines carrying cancer-relevant mutations, we show that partial senescence phenotypes can be induced to varying degrees in a context dependent manner, even in the absence of p21 or p53 expression. However, proliferation arrest varied among genetic backgrounds with predominantly toxic effects in p21 null cells, while cells lacking PI3K mutation failed to arrest. Furthermore, we show that the oncogene ECT2 induces partial senescence phenotypes in all mutant backgrounds tested, demonstrating a dependence on activating KRASG13D for growth suppression and a complete senescence response. These results suggest a potential mechanism to target mutant KRAS signalling through ECT2 in cancers that are reliant on activating KRAS mutations and remain refractory to current treatments.

Cancer prevention and therapy through the modulation of the tumor microenvironment.

Cancer arises in the context of an in vivo tumor microenvironment. This microenvironment is both a cause and consequence of tumorigenesis. Tumor and host cells co-evolve dynamically through indirect and direct cellular interactions, eliciting multiscale effects on many biological programs, including cellular proliferation, growth, and metabolism, as well as angiogenesis and hypoxia and innate and adaptive immunity. Here we highlight specific biological processes that could be exploited as targets for the prevention and therapy of cancer. Specifically, we describe how inhibition of targets such as cholesterol synthesis and metabolites, reactive oxygen species and hypoxia, macrophage activation and conversion, indoleamine 2,3-dioxygenase regulation of dendritic cells, vascular endothelial growth factor regulation of angiogenesis, fibrosis inhibition, endoglin, and Janus kinase signaling emerge as examples of important potential nexuses in the regulation of tumorigenesis and the tumor microenvironment that can be targeted. We have also identified therapeutic agents as approaches, in particular natural products such as berberine, resveratrol, onionin A, epigallocatechin gallate, genistein, curcumin, naringenin, desoxyrhapontigenin, piperine, and zerumbone, that may warrant further investigation to target the tumor microenvironment for the treatment and/or prevention of cancer.

Mathematical model of a telomerase transcriptional regulatory network developed by cell-based screening: analysis of inhibitor effects and telomerase expression mechanisms.

Cancer cells depend on transcription of telomerase reverse transcriptase (TERT). Many transcription factors affect TERT, though regulation occurs in context of a broader network. Network effects on telomerase regulation have not been investigated, though deeper understanding of TERT transcription requires a systems view. However, control over individual interactions in complex networks is not easily achievable. Mathematical modelling provides an attractive approach for analysis of complex systems and some models may prove useful in systems pharmacology approaches to drug discovery. In this report, we used transfection screening to test interactions among 14 TERT regulatory transcription factors and their respective promoters in ovarian cancer cells. The results were used to generate a network model of TERT transcription and to implement a dynamic Boolean model whose steady states were analysed. Modelled effects of signal transduction inhibitors successfully predicted TERT repression by Src-family inhibitor SU6656 and lack of repression by ERK inhibitor FR180204, results confirmed by RT-QPCR analysis of endogenous TERT expression in treated cells. Modelled effects of GSK3 inhibitor 6-bromoindirubin-3'-oxime (BIO) predicted unstable TERT repression dependent on noise and expression of JUN, corresponding with observations from a previous study. MYC expression is critical in TERT activation in the model, consistent with its well known function in endogenous TERT regulation. Loss of MYC caused complete TERT suppression in our model, substantially rescued only by co-suppression of AR. Interestingly expression was easily rescued under modelled Ets-factor gain of function, as occurs in TERT promoter mutation. RNAi targeting AR, JUN, MXD1, SP3, or TP53, showed that AR suppression does rescue endogenous TERT expression following MYC knockdown in these cells and SP3 or TP53 siRNA also cause partial recovery. The model therefore successfully predicted several aspects of TERT regulation including previously unknown mechanisms. An extrapolation suggests that a dominant stimulatory system may programme TERT for transcriptional stability.

Plasma N-acetyl-glucosaminidase in advanced gastro-intestinal adenocarcinoma correlates with age, stage and outcome.

BACKGROUND: N-acetyl-glucosaminidase (NAG) is a potential marker of genotoxicity. We retrospectively analyzed plasma NAG and clinico-pathologic features in advanced gastrointestinal adenocarcinoma patients. METHODS: Plasma from 118 patients and 51 healthy volunteers was analyzed for associations between NAG levels and age, disease presence, stage, treatment responses and survival. RESULTS: Pretreatment NAG correlated with age but was independently increased in metastatic versus locally advanced disease, particularly in gastric/esophageal patients. NAG was also associated with reduced overall survival. In subgroup analysis, increased NAG activity between day 1 and 2 of chemotherapy cycle 1 correlated with treatment response. CONCLUSION: We demonstrated that NAG correlates with gastrointestinal cancer outcomes. Further studies are required to determine if plasma markers of genotoxicity can be useful for disease monitoring.

Implementation of an AI-assisted fragment-generator in an open-source platform.

We recently reported a deep learning model to facilitate fragment library design, which is critical for efficient hit identification. However, our model was implemented in Python. We have now created an implementation in the KNIME graphical pipelining environment which we hope will allow experimentation by users with limited programming knowledge.

Targeting the telomere and shelterin complex for cancer therapy: current views and future perspectives.

Aberrant telomere homeostasis is essential for cell immortality, enabling cells to evade telomere dependent senescence. Disruption of telomere structure and function in cancer cells is highly toxic as shown by detailed pre-clinical evaluation of telomerase inhibitors. Under telomerase inhibition, cells must divide sufficiently frequently to allow one or more telomeres to shorten to an unprotected length. Functioning telomeres are disguised from the DNA damage machinery by DNA remodelling and other activities of the telomere binding complex shelterin. Direct interference with shelterin has been shown to result in cell killing and small molecules directly targeting telomere DNA also have anti-tumour effects partially dependent on shelterin disruption. However, shelterin components have not generally been regarded as therapeutic targets in their own right. In this review, we explore the possibilities for therapeutic targeting of the shelterin complex.

Modulation of telomerase promoter tumor selectivity in the context of oncolytic adenoviruses.

The telomerase RNA (hTR) and reverse transcriptase (hTERT) promoters are active in most cancer cells, but not in normal cells, and are useful for transcriptional targeting in gene therapy models. Telomerase-specific conditionally replicating adenoviruses (CRAd) are attractive vectors because they should selectively lyse tumor cells. Here, we compare CRAds, in which either the hTR or hTERT promoter controls expression of the adenovirus E1A gene. In replication-defective reporter adenoviruses, the hTR promoter was up to 57-fold stronger in cancer cells than normal cells and up to 49-fold stronger than hTERT. In normal cells, hTERT promoter activity was essentially absent. Doses of telomerase-specific CRAds between 1.8 and 28 infectious units per cell efficiently killed cancer cells, but normal cells required higher doses. However, CRAd DNA replication and E1A expression were detected in both cancer and normal cells. Overall, tumor specificity of the CRAds was limited compared with nonreplicating vectors. Surprisingly, both CRAds expressed similar E1A levels and functional behavior, despite known differentials between hTR and hTERT promoter activities, suggesting that the promoters are deregulated. Rapid amplification of cDNA ends analysis of hTR-/hTERT-E1A transcripts ruled out cryptic transcription from the vector backbone. Blocking E1A translation partially restored the hTR-/hTERT-E1A mRNA differential, evidencing feedback regulation by E1A.

A Novel Pyrazolopyrimidine Ligand of Human PGK1 and Stress Sensor DJ1 Modulates the Shelterin Complex and Telomere Length Regulation.

Telomere signaling and metabolic dysfunction are hallmarks of cell aging. New agents targeting these processes might provide therapeutic opportunities, including chemoprevention strategies against cancer predisposition. We report identification and characterization of a pyrazolopyrimidine compound series identified from screens focused on cell immortality and whose targets are glycolytic kinase PGK1 and oxidative stress sensor DJ1. We performed structure-activity studies on the series to develop a photoaffinity probe to deconvolute the cellular targets. In vitro binding and structural analyses confirmed these targets, suggesting that PGK1/DJ1 interact, which we confirmed by immunoprecipitation. Glucose homeostasis and oxidative stress are linked to telomere signaling and exemplar compound CRT0063465 blocked hypoglycemic telomere shortening. Intriguingly, PGK1 and DJ1 bind to TRF2 and telomeric DNA. Compound treatment modulates these interactions and also affects Shelterin complex composition, while conferring cellular protection from cytotoxicity due to bleomycin and desferroxamine. These results demonstrate therapeutic potential of the compound series.

Transcriptional repression of telomerase RNA gene expression by c-Jun-NH2-kinase and Sp1/Sp3.

Telomerase is essential for immortalization of most human cancer cells. Expression of the core telomerase RNA (hTR) and reverse transcriptase (hTERT) subunits is mainly regulated by transcription. However, hTR transcriptional regulation remains poorly understood. We previously showed that the core hTR promoter is activated by Sp1 and is repressed by Sp3. Here, we show that the mitogen-activated protein kinase kinase kinase 1 (MEKK1)/c-Jun-NH(2)-kinase (JNK) pathway represses hTR expression by a mechanism that involves Sp1 and Sp3. Promoter activity was induced by the JNK inhibitor SP600125 and was repressed by activated MEKK1. Repression by MEKK1 was blocked by SP600125 or enhanced by coexpression of wild-type but not phosphoacceptor mutated JNK. SP600125 treatment also increased levels of endogenous hTR. Mutations in the hTR promoter Sp1/Sp3 binding sites attenuated SP600125-mediated promoter induction, whereas coexpression of MEKK1 with Sp3 enhanced hTR promoter repression. Chromatin immunoprecipitation showed that levels of immunoreactive Sp1 associated with the hTR promoter were low in comparison with Sp3 in control cells but increased after JNK inhibition with a reciprocal decrease in Sp3 levels. No corresponding changes in Sp1/Sp3 protein levels were detected. Thus, JNK represses hTR promoter activity and expression, apparently by enhancing repression through Sp3.

Virotherapy: cancer gene therapy at last?

For decades, effective cancer gene therapy has been a tantalising prospect; for a therapeutic modality potentially able to elicit highly effective and selective responses, definitive efficacy outcomes have often seemed out of reach. However, steady progress in vector development and accumulated experience from previous clinical studies has finally led the field to its first licensed therapy. Following a pivotal phase III trial, Imlygic (talimogene laherparepvec/T-Vec) received US approval as a treatment for cutaneous and subcutaneous melanoma in October 2015, followed several weeks later by its European authorisation. These represent the first approvals for an oncolytic virotherapy. Imlygic is an advanced-generation herpesvirus-based vector optimised for oncolytic and immunomodulatory activities. Many other oncolytic agents currently remain in development, providing hope that current success will be followed by other diverse vectors that may ultimately come to constitute a new class of clinical anti-cancer agents. In this review, we discuss some of the key oncolytic viral agents developed in the adenovirus and herpesvirus classes, and the prospects for further enhancing their efficacy by combining them with novel immunotherapeutic approaches.

Selective ablation of human cancer cells by telomerase-specific adenoviral suicide gene therapy vectors expressing bacterial nitroreductase.

Reactivation of telomerase maintains telomere function and is considered critical to immortalization in most human cancer cells. Elevation of telomerase expression in cancer cells is highly specific: transcription of both RNA (hTR) and protein (hTERT) components is strongly upregulated in cancer cells relative to normal cells. Therefore, telomerase promoters may be useful in cancer gene therapy by selectively expressing suicide genes in cancer cells and not normal cells. One example of suicide gene therapy is the bacterial nitroreductase (NTR) gene, which bioactivates the prodrug CB1954 into an active cytotoxic alkylating agent. We describe construction of adenovirus vectors harbouring the bacterial NTR gene under control of the hTR or hTERT promoters. Western blot analysis of NTR expression in normal and cancer cells infected with adenoviral vectors showed cancer cell-specific nitroreductase expression. Infection with adenoviral telomerase-NTR constructs in a panel of seven cancer cell lines resulted in up to 18-fold sensitization to the prodrug CB1954, an effect that was retained in two drug-resistant ovarian lines. Importantly, no sensitization was observed with either promoter in any of the four normal cell strains. Finally, an efficacious effect was observed in cervical and ovarian xenograft models following single intratumoural injection with low doses of vector, followed by injection with CB1954.

Properties of a telomerase-specific Cre/Lox switch for transcriptionally targeted cancer gene therapy.

Telomerase expression represents a good target for cancer gene therapy. The promoters of the core telomerase catalytic [human telomerase reverse transcriptase (hTERT)] and RNA [human telomerase RNA (hTR)] subunits show selective activity in cancer cells but not in normal cells. This property can be harnessed to express therapeutic transgenes in a wide range of cancer cells. Unfortunately, weak hTR and hTERT promoter activities in some cancer cells could limit the target cell range. Therefore, strategies to enhance telomerase-specific gene therapy are of interest. We constructed a Cre/Lox reporter switch coupling telomerase promoter specificity with Cytomegalovirus (CMV) promoter activity, which is generally considered to be constitutively high. In this approach, a telomerase-specific vector expressing Cre recombinase directs excisive recombination on a second vector, removing a transcriptional blockade to CMV-dependent luciferase expression. We tested switch activation in cell lines over a wide range of telomerase promoter activities. However, Cre/Lox-dependent luciferase expression was not enhanced relative to expression using hTR or hTERT promoters directly. Cell-specific differences between telomerase and CMV promoter activities and incomplete sigmoid switch activation were limiting factors. Notably, CMV activity was not always significantly stronger than telomerase promoter activity. Our conclusions provide a general basis for a more rational design of novel recombinase switches in gene therapy.

Identification of a Selective G1-Phase Benzimidazolone Inhibitor by a Senescence-Targeted Virtual Screen Using Artificial Neural Networks.

Cellular senescence is a barrier to tumorigenesis in normal cells, and tumor cells undergo senescence responses to genotoxic stimuli, which is a potential target phenotype for cancer therapy. However, in this setting, mixed-mode responses are common with apoptosis the dominant effect. Hence, more selective senescence inducers are required. Here we report a machine learning-based in silico screen to identify potential senescence agonists. We built profiles of differentially affected biological process networks from expression data obtained under induced telomere dysfunction conditions in colorectal cancer cells and matched these to a panel of 17 protein targets with confirmatory screening data in PubChem. We trained a neural network using 3517 compounds identified as active or inactive against these targets. The resulting classification model was used to screen a virtual library of ~2M lead-like compounds. One hundred and forty-seven virtual hits were acquired for validation in growth inhibition and senescence-associated ß-galactosidase assays. Among the found hits, a benzimidazolone compound, CB-20903630, had low micromolar IC50 for growth inhibition of HCT116 cells and selectively induced senescence-associated ß-galactosidase activity in the entire treated cell population without cytotoxicity or apoptosis induction. Growth suppression was mediated by G1 blockade involving increased p21 expression and suppressed cyclin B1, CDK1, and CDC25C. In addition, the compound inhibited growth of multicellular spheroids and caused severe retardation of population kinetics in long-term treatments. Preliminary structure-activity and structure clustering analyses are reported, and expression analysis of CB-20903630 against other cell cycle suppressor compounds suggested a PI3K/AKT-inhibitor-like profile in normal cells, with different pathways affected in cancer cells.

MicroRNA and senescence: the senectome, integration and distributed control.

Cellular senescence has attracted renewed interest in recent years in light of the realization that cancer cells are susceptible to senescence-like responses to stress including exposure to chemotherapeutic agents. Therefore, senescence is viewed as a potentially important target in cancer therapy. However, translation of senescence therapeutics will rely on identifying appropriate marker panels for assessing senescence responses in vivo. Although some core pathways governing senescence induction have been clearly identified, the full set of gene targets involved in its establishment and maintenance are not defined and their complex interactions are poorly understood. This view is reinforced by recently discovered roles of microRNA (miRNA) in senescence regulation and induction. Senescence involves coordination of diverse cellular processes ranging from telomere homeostasis and DNA damage to chromatin modifications, inflammatory signaling, and metabolic and cytoskeletal changes. Thus, from a target discovery perspective a "senectome" view is appropriate. miRNAs are well placed to act as distributed senectome control elements, fine-tuning regulation of multiple processes. Here, we examine the complex interactions among several of these subprocesses and subsequently consider senescence-associated miRNAs (SA-miRNAs) as distributed controllers of senectome regulation. Finally, we examine their potential as clinical biomarkers to accelerate development of senescence therapeutics.

Cancer cell senescence: a new frontier in drug development.

Senescence forms a universal block to tumorigenesis which impacts on all hallmarks of cancer, making it an attractive target for drug discovery. Therefore a strategy must be devised to focus this broad potential into a manageable drug discovery programme. Several issues remain to be addressed including the lack of robust senescence-inducing compounds and causally related biomarkers to measure cellular response. Here, we review the latest progress in translating senescence as a target for cancer therapy and some promising approaches to drug and biomarker discovery. Finally, we discuss the potential application of a senescence-induction therapy in a clinical setting.

Dynamic telomerase gene suppression via network effects of GSK3 inhibition.

BACKGROUND: Telomerase controls telomere homeostasis and cell immortality and is a promising anti-cancer target, but few small molecule telomerase inhibitors have been developed. Reactivated transcription of the catalytic subunit hTERT in cancer cells controls telomerase expression. Better understanding of upstream pathways is critical for effective anti-telomerase therapeutics and may reveal new targets to inhibit hTERT expression. METHODOLOGY/PRINCIPAL FINDINGS: In a focused promoter screen, several GSK3 inhibitors suppressed hTERT reporter activity. GSK3 inhibition using 6-bromoindirubin-3'-oxime suppressed hTERT expression, telomerase activity and telomere length in several cancer cell lines and growth and hTERT expression in ovarian cancer xenografts. Microarray analysis, network modelling and oligonucleotide binding assays suggested that multiple transcription factors were affected. Extensive remodelling involving Sp1, STAT3, c-Myc, NFkappaB, and p53 occurred at the endogenous hTERT promoter. RNAi screening of the hTERT promoter revealed multiple kinase genes which affect the hTERT promoter, potentially acting through these factors. Prolonged inhibitor treatments caused dynamic expression both of hTERT and of c-Jun, p53, STAT3, AR and c-Myc. CONCLUSIONS/SIGNIFICANCE: Our results indicate that GSK3 activates hTERT expression in cancer cells and contributes to telomere length homeostasis. GSK3 inhibition is a clinical strategy for several chronic diseases. These results imply that it may also be useful in cancer therapy. However, the complex network effects we show here have implications for either setting.