Wang, Liguo

Abstract Motivation Illumina DNA methylation arrays have evolved rapidly, expanding genomic coverage while introducing backward incompatibilities by removing many CpG sites present in earlier versions. These changes result in systematic missing values when integrating data across array generations and substantially limiting the reuse of legacy datasets. Results We developed a two-stage framework for imputing missing DNA methylation values. The procedure first imputes randomly missing values using standard imputation techniques and then addresses systematic missingness using multi-output machine learning models, including support vector regression, nearest-neighbor methods, random forest models, and deep neural networks. When evaluated on real datasets with up to fifty percent induced missingness, the proposed framework consistently outperformed conventional imputation approaches. It also accurately imputes the missing CpG sites between methylation arrays and reduced representation bisulfite sequencing data, enabling robust cross-platform data integration. Analyses of large brain tumor methylation datasets demonstrate that the method restores array-specific methylation patterns while preserving biological complexity. Importantly, imputing missing methylation sites significantly improves the performance of epigenetic age prediction models. Availability and implementation This tool is implemented in the Python package “ultra-impute,” freely available at https://github.com/liguowang/ultra-impute. A code snippet demonstrating the usage of the ultra-impute package is provided in a Jupyter Notebook (https://github.com/liguowang/ultra-impute/blob/master/doc/Tutorial.ipynb).

Immunotherapy has been effective in many cancer types but has failed in multiple clinical trials in prostate cancers, with the underlying mechanisms remaining largely unclear. Here, we demonstrate that androgen receptor pathway inhibitor (ARPI) plus irradiation (IR) triggered robust anticancer immunity in prostate cancers in both patients and mice. We show that androgen-activated AR suppressed innate immune signaling by inducing inhibitor of nuclear factor kappa-B kinase subunit epsilon ( IKBKE ) gene repression through HDAC2 interaction with an IKBKE enhancer RNA ( IKBKE eRNA, or IKBKE-e ). ARPI treatment caused IKBKE derepression and enhanced an IR-induced innate immune response via action of RIG-I and MDA5 dsRNA sensors. IKBKE-e ablation largely enhanced innate immunity in prostate cancer cells in culture and anticancer immunity in mice. Our results revealed AR, HDAC2, and IKBKE eRNA as critical intrinsic immune suppressors in prostate cancer cells, suggesting that rejuvenating inhibitor of nuclear factor kappa-B kinase subunit epsilon (IKKε) signaling by targeting IKBKE-e is an actionable strategy to elicit synthetic anticancer immunity in immunologically “cold” cancers such as prostate cancer.

SUMMARY Artificial intelligence (AI)-based methods are increasingly critical in peptide drug discovery, but their applications are limited in a narrow scope such as antimicrobial peptides. For indications where peptide therapy inherently excels, such as metabolism, endocrinology, and tissue regeneration, AI-based pipeline for therapeutic peptide discovery is long awaited but still unmet. Here, we propose a deep learning-based pipeline, Deepeptide, capable of identifying therapeutic oligopeptides for various metabolism-related indications. Leveraging the intrinsical relationship of ‘disease indication — biological processes — molecular functions’, Deepeptide discovers oligopeptides with indication-ameliorating-related molecular functions as lead candidates for indication of interest. Deepeptide was applied in five representative indications of metabolism, endocrinology, and tissue regeneration: angiogenesis, lipid metabolism, osteogenesis, glucose metabolism, and anti-angiogenesis. Overall, 62% of the identified oligopeptide candidates demonstrated significant bioactivity in vitro, with most of them showing comparable potency to the first-line drugs. Notably, the heptapeptide AP7 exhibited angiogenic potency comparable to VEGF in excisional wound splinting mouse model by promoting cell migration rather than proliferation, and hexapeptide TP6 showed significant dual-efficacy against hyperlipidemia and obesity in high-fat diet mice by inhibiting lipid synthesis and regulating gut microbiota. These findings highlight the potential and generalizability of Deepeptide in therapeutic oligopeptide discovery for metabolic diseases.

Abstract Introduction: ASXL1-mutant (mt) Chronic Myelomonocytic Leukemia (CMML) is an aggressive myeloid neoplasm characterized by increased proliferation, resistance to epigenetic therapies, and poor survival. Epigenetic remodeling and up-regulation of leukemogenic driver genes are hallmarks of ASXL1mt CMML. Truncating mutations in ASXL1 activate enhancers that drive the expression of key leukemogenic driver genes such as HOXA9 and its co-factor MEIS1. Here, we investigate the chromatin accessibility of these leukemogenic enhancers by ASXL1 genotype at single-cell resolution (GoT-ChA-seq). Methods: After IRB approval, we performed bulk RNA-seq and ChIP-seq (H3K4me1, H3K4me3, H3K27ac, H3K27me3, H2AK119ub, H2BK20ac) on bone marrow (BM) mononuclear cells from 40 patients with CMML (19 ASXL1mt, 21 -wildtype/wt). After quality control, differential analyses of gene expression and histone modification occupancy between ASXL1mt and wt patients were conducted. A consensus peak calling framework was used to define genomic regions with ASXL1mt-specific histone modifications. Canonical enhancers were defined as genomic regions with a co-occupancy of H3K4me1 and H3K27ac. We then added regions marked by H2BK20ac to capture regulatory regions associated with p300 (Narita et al. Nat Genet 2023). We used an in silico prediction strategy to associate regulatory regions marked by strong histone modification occupancy with highly up-regulated leukemogenic driver genes (HOXA6-10, MEIS1). The ReMap database (Hammal et al. Nucleic Acids Res 2022) was queried for transcription factor binding in the genomic regions of interest. We then performed GoT-ChA-seq (Izzo et al. Nature 2024) to investigate chromatin accessibility for ASXL1mt (n=3) and wildtype BM single cells from CMML patients. In vitro colony forming assays were conducted on primary CMML BM cells (11- ASXL1mt and 9 wt CMML), assessing the efficacy of EP31670, a novel, oral, dual inhibitor of p300/BRD4. Results: This study included 40 patients with CMML, 19 ASXL1mt and 21 wt. The variant allele frequencies for ASXL1 were compatible with heterozygosity (median 0.41) and TET2 mutations were balanced between the two groups. We identified 149 candidate regulatory regions correlating with the expression of up-regulated target genes of interest (HOXA6-10, MEIS1) by bulk ChIP- and RNA-seq. Publicly available ChIP-seq data from myeloid cell lines demonstrated enrichment of p300 (in THP-1 cells, enrichment effect size +2.58, p=0.002) and BRD4 (in K-562 cells, enrichment effect size +1.58, p=0.045) in these candidate regulatory regions. Next, we interrogated the chromatin accessibility of ASXL1mt (n=13620) and ASXL1wt (n=24269) single cells from 3 CMML patients using GoT-ChA-seq. Differential accessibility analysis revealed 144 genomic regions, all of which were preferentially accessible in ASXL1mt cells compared to ASXL1wt cells (p<0.05 for all regions). Four of these preferentially accessible genomic regions (p<0.001 for all regions) overlapped with the previously identified candidate regulatory regions for MEIS1. The H3K27ac and H2BK20ac occupancy levels in these regions was strongly associated with the expression of MEIS1 (median Pearson r=0.73, p=1.16x10-7). These candidate regulatory regions were annotated in ENCODE as distal and proximal enhancers. Targeting this biology with the dual p300/BRD4 inhibitor EP31670 in progenitor colony forming assays led to a preferential therapeutic effect in ASXL1mt BM samples (median IC50 = 25nM) compared to in ASXL1wt samples (median IC50 = 74nM, p=0.03). Treatment with EP31670 was also associated with increased differentiation (increased blast forming units-erythroid; BFU-E) and organization of colonies in ASXL1mt vs wt CMML. Conclusions: ASXL1mt CMML is characterized by the up-regulation of several leukemogenic driver genes including HOXA9 and its co-factor MEIS1, secondary to the activation of genotype-specific enhancers. The preferential accessibility of these enhancers can be defined between ASXL1mt and ASXL1wt cells within patients. The enhancers of interest interact with p300 and BRD4 in myeloid cell lines and targeting them with the novel dual p300/BRD4 inhibitor, EP31670, induces therapeutic responses at low nanomolar drug concentrations in vitro, preferentially in ASXL1-mutant cells. These data solidify the biological rationale for targeting p300/BRD4 in ASXL1mt chronic myeloid neoplasms for therapeutic benefit (NCT05488548).

Abstract CDK12 is frequently deleted and/or mutated in human cancers, and CDK12 inactivation is associated with aggressive phenotypes, including the tendency to metastasis. In this study, we showed that CDK12 deficiency induced intronic polyadenylation (IPA) usage in BRD4 to upregulate expression of the bromodomain-containing protein 4 (BRD4) short isoform (BRD4S). BRD4S demonstrated an enhanced capacity to form phase separation despite lacking the intrinsically disordered region located in the C-terminal of full-length BRD4. Oligomerization of BRD4S mediated phase separation through a base-interacting structural domain. CUT&Tag and RNA sequencing analyses revealed that BRD4S preferentially bound to the promoters of TGFβ signaling genes, including TGFB2 and LTBP1, driving their expression. Consistently, BRD4S drove cancer cell dissemination in vitro and metastasis in vivo, which could be blocked by BET or TGFβ signaling pathway inhibitors. Overall, this study identifies IPA usage at the BRD4 gene locus induced by CDK12 deficiency as a mechanism responsible for BRD4S genesis, which represents an important mechanism driving cancer metastasis and a potential target for therapeutic intervention in CDK12-deficient cancers. Significance: The CDK12–BRD4S–TGFβ signaling axis represents a key regulatory module that promotes metastatic progression and can be targeted by BET or TGFβ pathway inhibitors for treating CDK12-deficient cancers.

CDK12 primarily functions as a transcription regulatory cyclin-dependent kinase (CDK) that controls mRNA elongation, splicing, and polyadenylation. The CDK12 gene is implicated in human cancers since it is frequently mutated and/or deleted in prostate and ovarian cancer but paradoxically amplified in breast cancer. Here, we demonstrate that CDK12 promotes serine-933 phosphorylation of DNA2, a nuclease/helicase critical for replication fork stress regulation, and the phosphorylation subsequently facilitates DNA2 polyubiquitination and degradation mediated by the APC/C CDC20 E3 ubiquitin ligase. CDK12 inactivation induces but amplification suppresses genome-wide expression of interferon response and antigen processing and presentation machinery genes in ovarian and breast cancer cells, respectively. Besides causing aberrant DNA2 stabilization, replication stress, genomic instability, and cytosolic double-stranded DNA (dsDNA) accumulation, CDK12 loss also triggers cGAS–STING activation and innate immune response, which can be reversed by forced expression of replication protein A (RPA) subunits or DNA2 depletion. Our findings identify DNA2 as a phosphorylation substrate of CDK12, connecting CDK12 to cell cycle regulation. These data also reveal DNA2 protein destruction as a critical mechanism that dictates genomic instability, cGAS–STING signaling activation, and innate immune response in CDK12-deregulated cancers.

Abstract Addressing the limited responsiveness to immune checkpoint blockade (ICB) in human cancers remains a critical challenge. Efforts over the past two decades have primarily focused on enhancing CD8 T cell function and preventing their exhaustion. However, less attention has been given to understanding the molecular mechanisms that contribute to the immunosuppressive tumor microenvironment (TME), which impedes CD8 T cell infiltration and fosters ICB resistance. The TME in ICB-resistant tumors is characterized by Transforming Growth Factor beta (TGF-β) cytokine signaling, leading to a fibrotic extracellular matrix (ECM) and the accumulation of immunosuppressive fibroblasts and myeloid cells. These features hinder CD8 T cell penetration, thereby promoting immune escape. The immune checkpoint molecule B7-H3, overexpressed in ICB-resistant cancers, has been linked to impaired CD8 T cell infiltration and poor prognosis. However, the molecular mechanisms governing B7-H3 function remain poorly understood. We investigated the relationship between B7-H3 gene expression and ICB response using transcriptomic data from two randomized phase 3 trials (IMvigor010 and IMvigor210). Bulk RNAseq data, tissue staining, and spectral flow cytometry were used to characterize the TME composition of B7-H3 high and low tumors. Multi-omic profiling of tumor cell lines with manipulated B7-H3 expression was performed to identify downstream targets of B7-H3. Syngeneic mouse tumor models were used to assess the impact of B7-H3 on tumor growth and antitumor immunity. High B7-H3 expression was associated with shorter overall survival in both IMvigor010 (HR=0.48, 95% CI: 0.33-0.70, p<0.0001) and IMvigor210 trials (HR=0.68, 95% CI: 0.51-0.92, p=0.01). In ICB-refractory cancers, B7-H3 expression correlated with high levels of fibroblasts, collagen deposits, and M2 macrophages. Gene set enrichment analysis revealed a positive association between B7-H3 and TGF-β signaling. Genetic deletion of B7-H3 in tumor cells reduced the surface expression of MMP14, a metalloprotease that activates TGF-β. Pan-cancer analysis identified MMP14 as the gene most correlated with B7-H3 (Pearson r=0.69, p<0.0001). TGF-β activation was reduced by 70% in B7-H3 knockout cells compared to wild-type. Loss of B7-H3 led to MMP14 accumulation in lysosomes, resulting in its degradation. We generated a B7-H3 inhibitor, 5D10, which promotes MMP14 degradation and inhibits TGF-β activation. In humanized B7-H3 mouse models, 5D10 significantly reduced TGF-β activation, enhancing CD8 T cell infiltration and impairing tumor growth. Our research identifies B7-H3 as a driver of immune exclusion and ICB resistance. We describe a novel, receptor-independent mechanism where B7-H3 promotes TGF-β activation via stabilization of MMP14. Targeting B7-H3 represents a strategy to modulate the TME, transforming immune-excluded tumors into immune-inflamed states that may respond better to ICB. Citation Format: Fabrice Lucien, Roxane Lavoie, Jack Korleski, Yohan Kim, Bharath Wootla, Liguo Wang, Ava Farrell, Kelly Harper, Martine Charbonneau, Claire Dubois, Igor Frank, Sounak Gupta, John Cheville, Jacob Orme, Stephen Boorjian, Parash Shah, Eugene Kwon, Sean Park, Haidong Dong. The immune checkpoint molecule B7-H3 is a driver of tumor immune exclusion by mediating TGF-β activation [abstract]. In: Proceedings of the AACR IO Conference: Discovery and Innovation in Cancer Immunology: Revolutionizing Treatment through Immunotherapy; 2025 Feb 23-26; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Immunol Res 2025;13(2 Suppl):Abstract nr B034.

Abstract Background: DNMT3A and TET2 are epigenetic regulator genes commonly mutated in age-related clonal hematopoiesis (CH). Despite having opposed epigenetic functions, these mutations are associated with increased all-cause mortality and a low risk for hematological neoplasms. In this study, we make use of a natural inflammatory response occurring during COVID-19 to define the impact of these mutations on inflammation and outcomes. Methods: A cohort of 243 community-based patients (Olmsted County, MN) with COVID-19 were identified. CH status was determined on PBMC using error corrected sequencing (lower detection limit 0.5% VAF). In a subset of patients, we use multi-omics interrogation of the genome, transcriptome, and epigenome at single cell (sc) resolution along with DNA methylation, in PBMC, from patients with COVID-19 and CH. Results: We detected 97 CH mutations in 72 (29.6%) of 243 patients with COVID-19. The most frequent being DNMT3A (n=30, 30%) and TET2 (n=26, 28%). The presence of CH negatively impacted overall survival when restricted to TET2 and DNMT3A mutations. However, only DNMT3Amt CH retained age and comorbidity independent prognostication (HR 2.84, p = 0.022). We then assessed for changes in DNA methylation using Illumina MethylationEPIC array on patients with COVID-19 and DNMT3A (n=3) or TET2mt (n=4) CH. Site specific DNA methylation changes revealed significant loss of methylation in DNMT3Amt CH in comparison to TET2mt patients, with 10,944 hypomethylated and 1,160 hypermethylated sites. We found that actively transcribed states (Tx, TxWk) were more commonly hypomethylated in DNMT3Amt CH while enhancers (Enh) and promoters (TssA, TssAFlnk) were more commonly hypomethylated in TET2mt CH. We then conducted scDNA-seq in DNMT3Amt (n=3) and TET2mt (n=1) CH in COVID-19 and observed that while in TET2mt CH, mutations were largely restricted to classical and intermediate monocytes, in DNMT3Amt CH, mutations were seen in myeloid and lymphoid lineage cells. We performed scRNA-seq on patients with COVID-19 and DNMT3A (n=3) or TET2mt (n=6) CH. Differential gene expression analysis identified 1,569 upregulated genes in DNMT3Amt patients and 205 downregulated. Pathway analysis demonstrated an upregulation of genes involved in T cell survival, function, and inflammation in DNMT3Amt patients compared to TET2mt patients with a significant overexpression of GIMAP1, GIMAP4, IRF2, IL32 in multiple cell types. We correlated transcriptomic changes with chromatin accessibility using the 10X Genomics Multiome platform to profile both gene expression and open chromatin from the same samples. Analysis of global distribution of cut sites and differentially accessible peaks showed increased chromatin accessibility in DNMT3Amt CH, especially CD4+ T lymphocytes and NK cells. Conclusion: We demonstrate an age, sex, and comorbidity-independent adverse impact on mortality associated with DNMT3Amt CH in the context of COVID-19 and highlight transcriptomic and epigenetic differences between DNMT3A and TET2mt CH, with single cell resolution. Citation Format: Jenna A Fernandez, Wazim Mohammed Ismail, Moritz Binder, Terra Lasho, Susan M. Geyer, Amelia Mazzone, Christy M. Finke, Abhishek Mangaonkar, Jeong-Heon Lee, Liguo Wang, Vernadette A. Simon, Fariborz Rakhshan Rohakthar, Amik Munankarmy, Jonathan J. Harrington, Melissa R. Snyder, Keith D. Robertson, Alexandre Gaspar-Maia, Mrinal Patnaik. Divergent impacts of DNMT3A and TET2 mutant clonal hematopoiesis on COVID-19 related outcomes [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: DNA Methylation, Clonal Hematopoiesis, and Cancer; 2025 Feb 1-4; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2025;85(3 Suppl):Abstract nr A012.

(1) Background: Prostate cancer treatment efficacy is significantly influenced by androgen receptor (AR) signaling pathways. SLC22A3, a membrane transporter, has been linked to SNP rs9364554 risk loci for drug efficacy in prostate cancer. (2) Methods: We examined the location of SNP rs9364554 in the genome and utilized TCGA and other publicly available datasets to analyze the association of this SNP with SLC22A3 transcription levels. We verified onco-mining findings in prostate cancer cell lines using quantitative PCR and Western blots. Additionally, we employed electrophoretic mobility shift assay (EMSA) to detect the binding affinity of transcription factors to this SNP. The ChIP-Seq was used to analyze the enrichment of H3K27ac on the SLC22A3 promoter. (3) Results: In this study, we revealed that SNP rs9364554 resides in the SLC22A3 gene and affects its transcription. The downregulation of SLC22A3 is associated with drug resistance. More importantly, we found that this SNP has different binding affinities with transcription factors, specifically FOXA1 and AR, which significantly affects their regulation of SLC22A3 transcription. (4) Conclusions: Our findings highlight the potential of using this SNP as a biomarker for predicting chemotherapeutic outcomes and uncover possible mechanisms underlying drug resistance in advanced prostate cancers. More importantly, it provides a clinical foundation for targeting FOXA1 to enhance drug efficacy in prostate cancer patients.