Insilico Medicine Research Articles

In silico medicine and -omics strategies in nephrology: contributions and relevance to the diagnosis and prevention of chronic kidney disease.

Chronic kidney disease (CKD) has been increasing over the last years, with a rate between 0.49% to 0.87% new cases per year. Currently, the number of affected people is around 850 million worldwide. CKD is a slowly progressive disease that leads to irreversible loss of kidney function, end-stage kidney disease, and premature death. Therefore, CKD is considered a global health problem, and this sets the alarm for necessary efficient prediction, management, and disease prevention. At present, modern computer analysis, such as in silico medicine (ISM), denotes an emergent data science that offers interesting promise in the nephrology field. ISM offers reliable computer predictions to suggest optimal treatments in a case-specific manner. In addition, ISM offers the potential to gain a better understanding of the kidney physiology and/or pathophysiology of many complex diseases, together with a multiscale disease modeling. Similarly, -omics platforms (including genomics, transcriptomics, metabolomics, and proteomics), can generate biological data to obtain information on gene expression and regulation, protein turnover, and biological pathway connections in renal diseases. In this sense, the novel patient-centered approach in CKD research is built upon the combination of ISM analysis of human data, the use of in vitro models, and in vivo validation. Thus, one of the main objectives of CKD research is to manage the disease by the identification of new disease drivers, which could be prevented and monitored. This review explores the wide-ranging application of computational medicine and the application of -omics strategies in evaluating and managing kidney diseases.

PandaOmics: An AI-Driven Platform for Therapeutic Target and Biomarker Discovery.

PandaOmics is a cloud-based software platform that applies artificial intelligence and bioinformatics techniques to multimodal omics and biomedical text data for therapeutic target and biomarker discovery. PandaOmics generates novel and repurposed therapeutic target and biomarker hypotheses with the desired properties and is available through licensing or collaboration. Targets and biomarkers generated by the platform were previously validated in both in vitro and in vivo studies. PandaOmics is a core component of Insilico Medicine's Pharma.ai drug discovery suite, which also includes Chemistry42 for the de novo generation of novel small molecules, and inClinico─a data-driven multimodal platform that forecasts a clinical trial's probability of successful transition from phase 2 to phase 3. In this paper, we demonstrate how the PandaOmics platform can efficiently identify novel molecular targets and biomarkers for various diseases.

AACR 2024: Insilico Medicine to Present Preclinical Data for AI-Generated Cancer Drugs

AACR 2024: Insilico Medicine to Present Preclinical Data for AI-Generated Cancer Drugs

Chemistry42: An AI-Driven Platform for Molecular Design and Optimization.

Chemistry42 is a software platform for de novo small molecule design and optimization that integrates Artificial Intelligence (AI) techniques with computational and medicinal chemistry methodologies. Chemistry42 efficiently generates novel molecular structures with optimized properties validated in both in vitro and in vivo studies and is available through licensing or collaboration. Chemistry42 is the core component of Insilico Medicine's Pharma.ai drug discovery suite. Pharma.ai also includes PandaOmics for target discovery and multiomics data analysis, and inClinico─a data-driven multimodal forecast of a clinical trial's probability of success (PoS). In this paper, we demonstrate how the platform can be used to efficiently find novel molecular structures against DDR1 and CDK20.

A Phase 1a/1b First in Human Study of PHI-101, a Potent Small Molecule Inhibitor of FLT3 in Relapsed and Refractory Acute Myeloid Leukemia

Background: The FMS-like tyrosine kinase 3 (FLT3) is mutated in approximately 30% of acute myeloid leukemia (AML) patients either by internal tandem duplication (ITD) or by point mutations in the tyrosine kinase domain (TKD). PHI-101 is a Type I FLT3 inhibitor developed to overcome resistance in AML. PHI-101 showed potent cellular activity in vitro and in vivo for both FLT3 ITD mutations and a wide variety of FLT3 point mutations in TKD. In preclinical studies utilizing primary human FLT3-ITD leukemia cells, PHI-101 showed increased anti-leukemic activity compared to gilteritinib based on increased survival in mice transplanted with these samples.The Phase 1a and Phase 1b global clinical trial with PHI-101 (NCT04842370) is currently underway to assess the overall safety and efficacy of PHI-101 in refractory and relapsed AML.Study Design and Methods: The overall design of the AML clinical trial consists of two parts, phase 1a, the dose-escalation, and phase 1b, the dose-expansion trial utilizing oral PHI-101 tablets. Up to 5 dose levels are planned for phase 1a and a single subject enrolled per dose level until the occurrence of at least one subject with more than Grade 2 toxicity according to CTCAE5.0 criteria during the 28-day evaluation period. The dose escalation will then convert to a standard 3+3 scheme, with 3 to 6 subjects per dose level cohort over the 28-day cycle. If there are no dose-limiting toxicities (DLT), dose escalation to the next higher dose levels will proceed upon the recommendation of the Safety Monitoring Committee. Subjects with FLT3 mutations or FLT3 wild-type will be enrolled in the dose-escalation cohort, and blood samples are collected for the primary endpoint pharmacokinetics (PK) and for the exploratory pharmacodynamic (PD) endpoint evaluation, including plasma inhibitory assay and biomarker analysis.Results: The Phase 1a clinical trial of PHI-101 was initiated in June 2020 at a daily dose of 40mg for level 1. As of July 31, 2021, 8 patients with relapsed or refractory AML have been enrolled in the trial, all of whom had prior anti-leukemic treatments with intensive chemotherapy, hypomethylating agents and/or other FLT3 inhibitors. To date, 5 enrolled patients were available for safety assessment at three dose levels and have not experienced any dose-limiting toxicities (DLTs) from repeating doses of the 28-day DLT-assessment window. The leukemic blasts in marrow or peripheral blood were dramatically reduced by up to 98% with one cycle (28 days). The pharmacokinetic data of plasma PHI-101 varied in a dose-proportional manner and peak plasma concentrations (Cmax) were reached between 4 and 6 hours after once-daily dosing. The pharmacodynamic evaluation by plasma inhibitory assay showed that plasma from patients receiving at daily doses of PHI-101 inhibited more than 90% of the phosphorylation of the FLT3-ITD receptor. PHI-101 was well tolerated with these dose levels. This study is currently recruiting FLT3 mutation or wild-type AML patients at multiple sites in Korea and Australia.Conclusion: PHI-101 is a next-generation FLT3 inhibitor that showed a potent anti-leukemic activity and improved efficacy in primary AML samples harboring FLT3/ITD and FLT3/TKD mutations in preclinical studies. The current analysis of this trial indicated that PHI-101 is a very effective FLT3 inhibitor for both refractory and relapsed AML patients, including those that have relapsed on other FLT3 TKI. The assessments of safety, tolerability, and PK of PHI-101 to determine the recommended dose for expansion are ongoing in the Phase 1a clinical trial. DisclosuresLee: Astellas Pharma, Inc.: Consultancy, Honoraria, Other: Advisory board; AbbVie: Honoraria, Other: Advisory board; Korean Society of Hematology: Membership on an entity's Board of Directors or advisory committees. Im: Pharos iBio Co., Ltd.: Current Employment. Nam: Pharos iBio Co., Ltd.: Current holder of individual stocks in a privately-held company. Kim: Pharos iBio Co., Ltd.: Current holder of individual stocks in a privately-held company. Han: Pharos iBio Co., Ltd.: Current holder of individual stocks in a privately-held company. Yoon: Pharos iBio Co., Ltd.: Current holder of individual stocks in a privately-held company. Small: Pharos iBio Co., Ltd.: Consultancy, Other: Scientific Advisory Board. The arrangement has been reviewed and approved by Johns Hopkins University in accordance with its conflict-of-interest policies, Research Funding; InSilico Medicine: Other: Scientific Advisory Board.

Bone mineral density modeling via random field: Normality, stationarity, sex and age dependence

Background and ObjectiveCapturing the population variability of bone properties is of paramount importance to biomedical engineering. The aim of the present paper is to describe variability and correlations in bone mineral density with a spatial random field inferred from routine computed tomography data. MethodsRandom fields were simulated by transforming pairwise uncorrelated Gaussian random variables into correlated variables through the spectral decomposition of an age-detrended correlation matrix. The validity of the random field model was demonstrated in the spatiotemporal analysis of bone mineral density. The similarity between the computed tomography samples and those generated via random fields was analyzed with the energy distance metric. ResultsThe random field of bone mineral density was found to be approximately Gaussian/slightly left-skewed/strongly right-skewed at various locations. However, average bone density could be simulated well with the proposed Gaussian random field for which the energy distance, i.e., a measure that quantifies discrepancies between two distribution functions, is convergent with respect to the number of correlation eigenpairs. ConclusionsThe proposed random field model allows the enhancement of computational biomechanical models with variability in bone mineral density, which could increase the usability of the model and provides a step forward in in-silico medicine.

AI specialist Insilico hails drug candidate

Insilico Medicine, a specialist in artificial intelligence –enabled drug discovery, says it has identified a preclinical drug candidate to treat kidney fibrosis. Earlier this year, Insilico identified a novel drug target and novel compound to treat idiopathic pulmonary fibrosis using its AI drug discovery technology . Insilico CEO Alex Zhavoronkov told C&EN that the firm’s success in identifying preclinical candidates “shows you that AI can now really help medicinal chemists accelerate the drug discovery process.” The work also challenges the “myth” that AI will replace medicinal chemists, he says. The company, which has 16 other discovery projects in its pipeline, raised nearly $255 million in series C financing in June.

AI predicts a drug target and a small molecule to block it

Claiming it has achieved a first in drug discovery, Insilico Medicine says it will begin preclinical trials with a novel compound discovered by artificial intelligence; the drug candidate is aimed ...

Development of an FEA framework for analysis of subject-specific aortic compliance based on 4D flow MRI

This paper presents a subject-specific in-silico framework in which we uncover the relationship between the spatially varying constituents of the aorta and the non-linear compliance of the vessel during the cardiac cycle uncovered through our MRI investigations. A microstructurally motivated constitutive model is developed, and simulations reveal that internal vessel contractility, due to pre-stretched elastin and actively generated smooth muscle cell stress, must be incorporated, along with collagen strain stiffening, in order to accurately predict the non-linear pressure-area relationship observed in-vivo. Modelling of elastin and smooth muscle cell contractility allows for the identification of the reference vessel configuration at zero-lumen pressure, in addition to accurately predicting high- and low-compliance regimes under a physiological range of pressures. This modelling approach is also shown to capture the key features of elastin digestion and SMC activation experiments. The volume fractions of the constituent components of the aortic material model were computed so that the in-silico pressure-area curves accurately predict the corresponding MRI data at each location. Simulations reveal that collagen and smooth muscle volume fractions increase distally, while elastin volume fraction decreases distally, consistent with reported histological data. Furthermore, the strain at which collagen transitions from low to high stiffness is lower in the abdominal aorta, again supporting the histological finding that collagen waviness is lower distally. The analyses presented in this paper provide new insights into the heterogeneous structure-function relationship that underlies aortic biomechanics. Furthermore, this subject-specific MRI/FEA methodology provides a foundation for personalised in-silico clinical analysis and tailored aortic device development. Statement of significanceThis study provides a significant advance in in-silico medicine by capturing the structure/function relationship of the subject-specific human aorta presented in our previous MRI analyses. A physiologically based aortic constitutive model is developed, and simulations reveal that internal vessel contractility must be incorporated, along with collagen strain stiffening, to accurately predict the in-vivo non-linear pressure-area relationship. Furthermore, this is the first subject-specific model to predict spatial variation in the volume fractions of aortic wall constituents. Previous studies perform phenomenological hyperelastic curve fits to medical imaging data and ignore the prestress contribution of elastin, collagen, and SMCs and the associated zero-pressure reference state of the vessel. This novel MRI/FEA framework can be used as an in-silico diagnostic tool for the early stage detection of aortic pathologies.

Business Roundup

BUSINESS Business Roundup ShareShare onFacebookTwitterWechatLinked InRedditEmail C&EN, 2021, 99 (6), p 19February 22, 2021Cite this:C&EN 99, 6, 19AbstractElkem has acquired a new organofunctional silicones facility near Lyon, France, for more than $10 million. The metals and materials firm expects to start up the plant, bought from an undisclosed seller, by year’s end. Clariant has made “a significant contribution” to Swiss Federal Institute of Technology (ETH), Zurich, in exchange for 10 years of research into catalysis and sustainable chemistry. The goal is to better understand the effect of catalysts’ properties on performance. EnginZyme , a Swedish start-up developing catalysts based on enzymes fixed to packed-bed reactors, has increased its first round of funding to $13.2 million, from $7.0 million raised in April. Investors include Industrifonden, SEB Greentech VC, and Sofinnova Partners. Syngenta and Insilico Medicine, an artificial intelligence software company, are collaborating to accelerate the development of crop protection molecules. The project will work with AI and deep-learning techniques that Insilico has employed in pharmaceutical research. Snapdragon ChemistryView: PDF | Full Text HTML

Artificial intelligence finds drugs that could fight the new coronavirus

Two independent groups last week reported that they had used artificial intelligence in different ways to find possible treatments for the novel coronavirus, named 2019-nCoV. On Feb. 4, researchers from the AI drug discovery company BenevolentAI and Imperial College London reported that they had used AI software to find an already-approved drug that might limit the virus’s ability to infect people (Lancet 2020, DOI: 10.1016/S0140-6736(20)30304-4). On Feb. 6, Insilico Medicine announced that its AI algorithms had designed new molecules that could stop the virus from replicating in people’s bodies. The company says it submitted a paper to the bioRxiv server, but the preprint had not been posted there as of press time. BenevolentAI’s algorithms connect molecular structure data to biomedical information about relevant receptors and diseases to find potential drug targets. The group adapted its search to newly available information about 2019-nCoV and focused on the enzyme adaptor-associated protein kinase

AI for drug discovery came into its own

Artificial intelligence algorithms can make accurate predictions about unfamiliar systems using data they’ve been exposed to. In the vast and complex space of molecules, reactions, and biological interactions, that makes AI an attractive alternative to human drug hunters. But if AI will lead a revolution in drug discovery, it didn’t happen in 2019. “We have bits and pieces where people have made really good advances,” says Mike Tarselli, scientific director of the Society for Laboratory Automation and Screening, but he saw no transformative breakthrough. Among the advances: Researchers at the Massachusetts Institute of Technology showed off a system that uses AI and robotics to plan and execute syntheses of drug molecules, which could help chemists search for and evaluate potential therapies. The researchers behind Synthia, an AI-based retrosynthesis planner, introduced an algorithm that could help drug chemists avoid synthetic routes that are already patent protected. And the start-up Insilico Medicine

Latest advances in aging research and drug discovery.

An increasing aging population poses a significant challenge to societies worldwide. A better understanding of the molecular, cellular, organ, tissue, physiological, psychological, and even sociological changes that occur with aging is needed in order to treat age-associated diseases. The field of aging research is rapidly expanding with multiple advances transpiring in many previously disconnected areas. Several major pharmaceutical, biotechnology, and consumer companies made aging research a priority and are building internal expertise, integrating aging research into traditional business models and exploring new go-to-market strategies. Many of these efforts are spearheaded by the latest advances in artificial intelligence, namely deep learning, including generative and reinforcement learning. To facilitate these trends, the Center for Healthy Aging at the University of Copenhagen and Insilico Medicine are building a community of Key Opinion Leaders (KOLs) in these areas and launched the annual conference series titled “Aging Research and Drug Discovery (ARDD)” held in the capital of the pharmaceutical industry, Basel, Switzerland (www.agingpharma.org). This ARDD collection contains summaries from the 6th annual meeting that explored aging mechanisms and new interventions in age-associated diseases. The 7th annual ARDD exhibition will transpire 2nd-4th of September, 2020, in Basel.

Dissociation of the DNMT3A-HDAC1 Repressor Complex Induces PD-L1 Expression

Mutations of DNMT3A are one the most frequently observed alterations in AML patients. The DNMT3A R882 mutation appears to confer a dominant-negative loss-of function effect and changes the DNA binding preference according to recent studies. DNMT3A R882 mutations are also found at increasing frequency with age in healthy elderly populations and are one of the earliest pre-malignant alterations in the clonal evolution progression to leukemia. Studies have shown that DNMT3A mutations decrease overall DNA methylation and through less clear mechanisms, also impact the epigenetic landscape by effecting changes in histone acetylation. Here we investigated potential mechanisms by which loss of DNMT3A activity changes histone acetylation. DNMT3A binds with many proteins that regulate chromatin biology and gene transcription. Among those interacting proteins, we focused on the DNMT3A-HDAC interaction and their regulation of target gene suppression. To investigate whether there are differences in binding of wild-type vs. mutant DNMT3A to HDACs, we performed immunoprecipitation and Western blotting assays using Myc- tagged wild-type and R882 mutated DNMT3A. We found that DNMT3A R882 mutants showed reduced interaction with HDAC1 and 2. In addition, upon treatment with HDAC inhibitors (HDACi), DNMT3A mutant protein was more easily dissociated from HDAC1/2 than was wild-type DNMT3A. Intriguingly, covalent modification of DNMT3A R882 by SUMO1 protein was significantly enhanced relative to wild type DNMT3A. Together, we suggest that the weak complex formation between mutant DNMT3A and HDACs results from augmented SUMOylation of the R882 mutant. Because the DNMT3A R882 mutation reduces its methyl transferase activity, we investigated which genes would be upregulated from the DNMT3A repressor complex. To do this, we established isogenic TF-1 cell lines that harbor haploid DNMT3A knockout (DNMT3A+/-) using the Cripsr-Cas9 system. We also treated cells with HDACi and 5-azacytidine (5-aza) which inhibit HDAC and DNMT, respectively. Interestingly, we discovered that PD-L1 expression is induced by HDACi and 5-aza treatment. Chemical inhibition by 5-aza or genetic inhibition by knockout reduces DNMT3A activity and synergized with HDACi to increase PD-L1 expression. Flow cytometry analysis also demonstrated increased membrane PD-L1 expression in response to HDACi. We also found out that DNMT3A+/- resulted in higher Histone H3K27 acetylation, which is known as a gene activation mark. Higher H3K27 acetylation in DNMT3A+/- cell confirms the findings by other groups but the mechanisms by which this occurs are unknown. We suggest that haploinsufficiency of DNMT3A results in a reduced DNMT3A-HDAC interaction leading to higher H3K27 acetylation and increased PD-L1 expression. Our results also revealed that HDACi treatment induced cell cycle arrest, DNA damage and apoptosis at increasing levels in DNMT3A+/- cell. Even though the DNMT3A+/- TF-1 showed increased sensitivity to HDACi treatment, we observed a correlation of higher phosphor- ERK1/2 and PD-L1 levels in the surviving cells. The enhanced expression of PD-L1 and activation of ERK1/2 may explain in part how mutated DNMT3A contributes to drug resistance and immune checkpoint avoidance. Many oncology clinical trials are underway utilizing HDACi. However, the questions of which mutational backgrounds might be most sensitive to these agents and how to best combine them with other agents remain to be answered. To test whether reduced DNMT3A activity increases PD-L1 expression in vivo, we crossed floxed DNMT3A mice with Mx1-Cre mice. After 4 weeks of induction of Cre recombinase by injecting pIpC in the progeny carrying both genetically engineered changes, lineage depleted mouse BM cells were analyzed for PD-L1 expression using quantitative RT-PCR. BM cells derived from DNMT3A knockout mice showed increased expression of PD-L1 compared to wild-type mice. Treatment of these BM cells with an HDACi and/or 5-aza resulted in a synergistic induction of PD-L1 expression for the combination. Taken together, we suggest that mutant DNMT3A induces higher H3K27 acetylation along with PD-L1 expression due to a looser complex between HDAC1 and mutant DNMT3A. Therefore, we suggest that combined treatment with an HDACi and an immune checkpoint inhibitor targeting the PD-L1/PD-1 axis might be a promising strategy for treating DNMT3A mutant AML patients. Disclosures Small: InSilico Medicine: Membership on an entity's Board of Directors or advisory committees; Pharos I, B & T: Consultancy, Research Funding.

FLT3 Inhibitor Correlative Laboratory Assays Impact Outcomes in KMT2A-Rearranged Infant Acute Lymphoblastic Leukemia (ALL) Patients Treated with Lestaurtinib: AALL0631, a Children's Oncology Group Study

Background/Objectives Infants with ALL have poor outcomes, especially the 70-80% with KMT2A rearrangements (KMT2A-r). Intensification of chemotherapy has reached the limit of tolerability, and stem cell transplantation (SCT) has not improved outcomes. The FLT3 kinase is overexpressed and constitutively activated in KMT2A-r ALL. FLT3 inhibitors (FLT3i) selectively kill KMT2A-r ALL cells in vitro and in vivo, and synergize with chemotherapy. COG AALL0631 was a randomized trial that tested the hypothesis that the addition of the FLT3i lestaurtinib to post-induction chemotherapy would improve outcomes for KMT2A-r ALL.As previously reported, the addition of lestaurtinib did not improve outcomes [3-year disease-free/overall survival (DFS/OS) 39%/46% without lestaurtinib vs. 37%/45% with lestaurtinib, log-rank p=0.78/0.92]. We hypothesized that correlative laboratory measures of FLT3i pharmacodynamics and ex vivo FLT3i sensitivity would enhance our understanding of the clinical responses for KMT2A-r infants treated with lestaurtinib. Design/Methods FLT3i pharmacodynamics (PD) were determined using plasma inhibitory activity (PIA) measured at 5 trough time points during lestaurtinib treatment, using Western blotting and densitometry (normalized to pre-treatment levels). PIA ranged from 0% (no evidence of FLT3 inhibition) to 100% (full inhibition). Mean PIA of 5 troughs was calculated for each patient, and was categorized based on pre-defined protocol criteria as "adequate", or FLT3 inhibited (≥85%) vs "inadequate", or FLT3 uninhibited (<85%). FLT3i ex-vivo sensitivity (EVS) was determined using WST-1 cell proliferation assay [48 hour cytotoxicity of diagnostic blasts exposed to 50 nM lestaurtinib vs vehicle control; mean of triplicate wells; ranging from 0% (no cytotoxicity) to 100% (all cells killed)]. Median cytotoxicity was calculated for all assayed samples on the trial (n=151; median cytotoxicity 46%). Samples were categorized as "sensitive" (above median) or "resistant" (below median). Clinical outcomes examined included 3-year DFS and OS (measured from randomization) and minimal residual disease (MRD) measured in a central laboratory by flow cytometry at week 10 [after induction (6 weeks) followed by chemotherapy vs. chemotherapy/lestaurtinib intensification (4 weeks)]. MRD results were categorized as negative (<0.01%) or positive (≥0.01%). P-values for DFS and OS comparisons are based on log-rank test, and for MRD comparisons are based on Fisher's exact test. Results Of 210 eligible patients, 146 were KMT2A-r, and 78 were assigned to receive lestaurtinib. Of these, data were available for FLT3i PD (n=73), FLT3i EVS (n=62), MRD (n=55), and DFS/OS (n=78). FLT3i PD and FLT3i EVS significantly correlated with both survival and MRD. For FLT3i PD, 27 were "FLT3 inhibited" and 46 were "FLT3 uninhibited". The DFS/OS for inhibited was 59%/67% vs. 28%/37% for uninhibited (p=0.01/0.02). The week 10 MRD-negative rate was 90% for inhibited vs. 52% for uninhibited (p=0.003). For FLT3i EVS, of the 62 lestaurtinib-treated patients with data, 42 were "sensitive" and 20 were "resistant". The DFS/OS for sensitive was 52%/62% vs. 5%/11% for resistant (p<0.001 for both). The week 10 MRD-negative rate was 78% for sensitive vs. 42% for resistant (p=0.03). We then hypothesized that among patients treated with lestaurtinib, those with both "inhibited" FLT3i PD and "sensitive" FLT3i EVS would have particularly good clinical outcomes. Of 58 patients with data for both assays, 17 were both inhibited and sensitive. This cohort had DFS/OS of 88%/94% and week 10 MRD-negative rate of 100%. The remaining N=41 were uninhibited and/or resistant, and had DFS/OS of 20%/29% (p<0.001 for both, Figure 1 shows DFS) and week 10 MRD-negative rate of 58% (p=0.008). Table 1 summarizes all results. Conclusion While the addition of lestaurtinib to chemotherapy did not improve outcomes, patients achieving potent pharmacodynamic inhibition of FLT3 and those whose leukemia cells were sensitive to ex-vivo FLT3-inhibitor induced cytotoxicity did benefit from the addition of lestaurtinib. Selection of patients with FLT3-sensitive leukemia, real-time intra-patient dose escalation of patients with inadequate pharmacodynamics, and/or using a more potent FLT3 inhibitor may enhance the efficacy of FLT3 inhibition for KMT2A-r infant ALL. Disclosures Raetz: Pfizer: Research Funding. Loh:Medisix Therapeutics, Inc.: Membership on an entity's Board of Directors or advisory committees. Hunger:Bristol Myers Squibb: Consultancy; Amgen: Consultancy, Equity Ownership; Novartis: Consultancy; Jazz: Honoraria. Borowitz:Beckman Coulter: Honoraria. Small:Pharos I, B & T: Consultancy, Research Funding; InSilico Medicine: Membership on an entity's Board of Directors or advisory committees.

DOCK2 Expression Affects Leukemogenesis and Disease Progression in a Murine Model of FLT3/ITD Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is an aggressive hematologic neoplasm characterized by clonal expansion of myeloid blasts. The FMS-like tyrosine kinase-3 (FLT3) receptor gene is the most commonly mutated gene in AML, and an internal tandem duplication (ITD) mutation portends a poor prognosis. FLT3/ITD activates Rac1 GTPase, and both proteins play key roles in DNA damage response (DDR). Disease progression in FLT3/ITD AML is often heralded by copy-neutral loss of heterozygosity (CN-LOH) at the FLT3 locus, which is typically caused by duplication of the mutant allele through homologous recombination with loss of the wild-type (WT) allele. We previously identified dedicator of cytokinesis 2 (DOCK2) as a FLT3/ITD-interacting protein. The DOCK family of proteins acts as guanine nucleotide exchange factors for the Rho family of GTPases, including Rac1. Expression of DOCK2 is limited to hematopoietic cells, and it is expressed in leukemic blasts in FLT3/ITD AML. Knockdown of DOCK2 in FLT3/ITD cell lines results in reduced Rac1 activity, decreased expression and activity of FLT3 and DDR factors, and increased sensitivity to cytarabine. In this study, we used mouse models to further investigate the relationship between DOCK2 and FLT3/ITD. Mice with a Dock2 mutation (Dock2/Hsd) that results in a marked reduction in Dock2 expression were bred with mice that carry a Flt3/ITD mutation (Flt3+/ITD). Quantitative PCR of mouse bone marrow (BM) revealed that homozygous expression of Dock2/Hsd markedly decreased the expression of Flt3 and key DDR factors in both WT and Flt3+/ITD mice. Flow cytometric analysis of the BM demonstrated that Dock2/Hsd resulted in partial reversal of the myeloproliferative neoplasm that develops in Flt3+/ITD mice, with a significant reduction in multipotent progenitors (MPP) in Flt3+/ITD;Dock2Hsd/Hsd (0.026%) vs. Flt3+/ITD mice (0.075%; P=0.048) at 3 months. To assess the role of DOCK2 in leukemogenesis and disease progression, the Dock2/Hsd mutation was introduced into a murine model of FLT3/ITD acute leukemia that carries both Flt3/ITD and the Nup98-HoxD13 (NHD13) fusion gene. It was previously shown that Flt3+/ITD;NHD13 mice develop acute leukemia with 100% penetrance and short latency. We found that the median survival of Flt3+/ITD;NHD13 mice (117 days) was not significantly affected by heterozygous expression of Dock2/Hsd (123 days); however, the median survival of Flt3+/ITD; NHD13; Dock2Hsd/Hsd mice was extended to 198 days (P=0.001). In contrast, Dock2/Hsd provided no survival benefit to mice that carried either Flt3+/ITD or NHD13 alone. The prolonged survival of Flt3+/ITD;NHD13;Dock2Hsd/Hsd mice was accompanied by delayed onset of LOH at the Flt3/ITD locus. At 3 months, approximately half of Flt3+/ITD;NHD13;Dock2+/+ (47%; n=19) and Flt3+/ITD;NHD13;Dock2+/Hsd (57%; n=14) mice exhibited LOH in the peripheral blood, whereas only 11% of Flt3+/ITD; NHD13;Dock2Hsd/Hsd mice showed LOH (n=18; P=0.008 vs. Dock2+/Hsd and P=0.029 vs. Dock2+/+). Most of the moribund Flt3+/ITD;NHD13 mice (n=13) developed AML with minimal maturation (70%), and the remainder developed T-lymphoblastic leukemia with an early T-precursor phenotype (ETP-ALL). In contrast, moribund Flt3+/ITD;NHD13;Dock2Hsd/Hsd mice (n=14) exhibited a more diverse disease spectrum: 30% mixed phenotype acute leukemia, 21% AML with minimal maturation, 21% ETP-ALL, 21% AML with maturation, and 7% acute erythroid leukemia (AEL). All moribund Flt3+/ITD;NHD13 mice showed complete LOH in BM; however, only 68% of moribund Flt3+/ITD;NHD13;Dock2Hsd/Hsd mice exhibited complete LOH, with the rest showing partial or no LOH. Notably, all mice that exhibited partial or no LOH developed AEL or AML with maturation, diseases that were present in NHD13 mice but were not observed in Flt3+/ITD;NHD13 mice. This suggests that suppression of Dock2 mitigates the effects of Flt3/ITD, allowing the NHD13 mutation to drive the development of disease. These findings demonstrate that decreased DOCK2/Rac1 activity delays LOH at the FLT3 locus and disease progression in a murine model of FLT3/ITD leukemia. Thus, pharmacologic inhibition of the Rac1 signaling pathways via DOCK2 may provide a novel and promising therapeutic target for FLT3/ITD AML. Disclosures Small: Pharos I, B & T: Consultancy, Research Funding; InSilico Medicine: Membership on an entity's Board of Directors or advisory committees. Duffield:Boston Biomedical/Sumitomo Dainippon Pharma Co., Ltd.: Consultancy, Membership on an entity's Board of Directors or advisory committees; MedImmune: Consultancy.

Gilteritinib and Venetoclax Synergize to Eliminate FLT3/ITD+ Leukemia Cells through BIM

Gilteritinib and Venetoclax synergize to eliminate FLT3/ITD+ leukemia cells through BIM Abstract Acute myeloid leukemia (AML) is characterized by a clonal proliferation of immature myeloid cells in the bone marrow or other tissues. The most commonly mutated gene in AML is FMS-like tyrosine kinase (FLT3). FLT3 downstream signaling pathways include PI3K/AKT, STAT5 and MAPK, which affect apoptosis, differentiation and cell proliferation. An internal tandem duplication mutation in FLT3 (FLT3/ITD) is identified in approximately 25% of patients with AML, and this mutation is associated with a particularly poor prognosis. A subset of AML also has a point mutation in the tyrosine kinase domain of FLT3 (FLT3/TKD); however, this mutation does not have the same pronounced impact on prognosis as the FLT3/ITD mutation. Since tyrosine kinases are an attractive drug target, tyrosine kinase inhibitors (TKIs) that target FLT3 have been developed, including recent agents that show enhanced specificity, such as Gilteritinib. Despite these advances, TKI monotherapy continues to show limited success, indicating that combination therapy is likely necessary for the effective treatment of FLT3/ITD AML. As FLT3 signaling pathway activity is known to be anti-apoptotic, in this study we investigated the combinatorial effect of a FLT3-selective TKI and the BCL-2 inhibitor Venetoclax. The BCL-2 protein plays a key role in apoptosis, with anti-apoptotic/prosurvival effects. Venetoclax is a selective BCL-2 inhibitor, and is used clinically in the treatment of chronic lymphocytic leukemia and relapsed/refractory AML. We first investigated the combinatorial effect of treatment with Gilteritinb and Venetoclax in a FLT3/ITD+ leukemia cell line (Molm14). Combined treatment with Gilteritinib (20nM) and Venetoclax (80nM) reduced cell proliferation by 83.7%, as compared to Gilteritinib (62.2%, P<0.05) or Venetoclax (37.3%, P<0.05) alone. The drug combination demonstrated synergy (CI=0.42). Combined treatment shows that these agents act synergistically to enhance apoptosis (88.2%), as compared with Gilteritinib (52.1%, P<0.05) or Venetoclax (12.1%, P<0.05) alone. Moreover, the combined treatment also significantly reduced cell proliferation in patient samples with FLT3/ITD+ and FLT3/TKD mutations. Further experiments with cell line (Molm14 cells resistant to 60nM CEP-701) showed that Venetoclax can re-sensitize FLT3 TKI-resistant cell lines to TKI treatment. Western blot analysis indicates that this effect is mediated by inhibiting MAKP pathway-whose reactivation is an important reason for TKI resistance in FLT3/ITD+ patients. These data demonstrate that combined treatment with Gilteritinb and Venetoclax reduces cell proliferation and enhances apoptosis in a FLT3/ITD leukemia cell line. We next investigated the mechanism of action of this drug combination. BCL-2 and other antiapoptotic proteins in this subfamily (i.e. MCL-1, BCL-XL) exert their pro-survival effects by sequestering BIM. BIM is a proapoptotic protein, that, when released by BCL-2, can activate cell death mediators. These cell death mediators include BAX and BAK, which perforate the mitochondrial membrane, resulting in apoptosis. Venetoclax acts by binding to BCL-2 and displacing BIM, freeing BIM to associate with and activate cell death mediators, resulting in apoptosis. We found that in Molm14 cells, Venetoclax not only dissociates BIM from BCL2, but also decreases expression of BIM and enhances the binding of BIM and MCL-1. In contrast, we found that Gilteritinib increases expression of BIM, reduces the binding of BIM with MCL-1 via a reduction of MCL1 expression, and shows enhanced binding of BIM with BCL2. Cells treated with both Venetoclax and Gilteritinib show dissociation of BIM from both BCL2 and MCL-1, though the interaction between BIM and BCL-XL is not affected. Combination treatment also showed increased binding between BIM and the cell death mediator BAX, leading to increased apoptosis. These studies provide evidence that the addition of Venetoclax may enhance TKI therapy in the treatment of FLT3/ITD leukemia. Additionally, these findings suggest that enhanced cell death in FLT3/ITD AML cells treated with combination therapy occurs because Venetoclax mitigates unintentded pro-survival effects of the TKI, including an increase in BIM expression and increased association between BIM and BCL-2. Disclosures Duffield: MedImmune: Consultancy; Boston Biomedical/Sumitomo Dainippon Pharma Co., Ltd.: Consultancy, Membership on an entity's Board of Directors or advisory committees. Small:Pharos I, B & T: Consultancy, Research Funding; InSilico Medicine: Membership on an entity's Board of Directors or advisory committees.

AI identifies drug candidate in weeks

The artificial intelligence start-up Insilico Medicine has used machine learning to find credible drug candidates in a matter of weeks (Nat. Biotechnol. 2019, DOI: 10.1038/s41587-019-0224-x). Experts say it’s an important demonstration of what machine learning can do in drug discovery, but it isn’t a revolution. Insilico Medicine has been showing off its progress in teaching computers to find new drugs since its founding in 2014. The company’s latest effort involves generative reinforcement learning, a technique that uses rewards to guide an algorithm as it searches for molecules that satisfy its goals. In this case, the algorithm was hunting for small molecules that are inhibitors of discoidin domain receptor 1 (DDR1), a kinase linked to fibrosis. The researchers trained their algorithm using databases of known DDR1 inhibitors, kinase inhibitors, nonkinase inhibitors, and patent-protected molecules. Using measures of novelty and DDR1 inhibition, the algorithm proposed 30,000 potential drugs. This group was filtered

GSK amps up AI with Insilico pact

Aiming to bolster its abilities in the hot area of artificial intelligence research, GlaxoSmithKline will collaborate with Insilico Medicine. Under the deal, which follows a pilot pact, GSK will us...