Integrating Reinforcement Learning and Monte Carlo Tree Search for enhanced neoantigen vaccine design.

Recent advances in cancer immunotherapy have highlighted the potential of neoantigen-based vaccines. However, the design of such vaccines is hindered by the possibility of weak binding affinity between the peptides and the patient's specific human leukocyte antigen (HLA) alleles, which may not elicit a robust adaptive immune response. Triggering cross-immunity by utilizing peptide mutations that have enhanced binding affinity to target HLA molecules, while preserving their homology with the original one, can be a promising avenue for neoantigen vaccine design. In this study, we introduced UltraMutate, a novel algorithm that combines Reinforcement Learning and Monte Carlo Tree Search, which identifies peptide mutations that not only exhibit enhanced binding affinities to target HLA molecules but also retains a high degree of homology with the original neoantigen. UltraMutate outperformed existing state-of-the-art methods in identifying affinity-enhancing mutations in an independent test set consisting of 3660 peptide-HLA pairs. UltraMutate further showed its applicability in the design of peptide vaccines for Human Papillomavirus and Human Cytomegalovirus, demonstrating its potential as a promising tool in the advancement of personalized immunotherapy.

HLA-A+ tertiary lymphoid structures with reactivated tumor infiltrating lymphocytes are associated with a positive immunotherapy response in esophageal squamous cell carcinoma.

BACKGROUND: Immune checkpoint blockade (ICB) therapy provides remarkable clinical benefits for multiple cancer types. However, the overall response rate to ICB therapy remains low in esophageal squamous cell carcinoma (ESCC). This study aimed to identify biomarkers of ICB therapy for ESCC and interrogate its potential clinical relevance. METHODS: We investigated gene expression in 42 treatment-naïve ESCC tumor tissues and identified differentially expressed genes, tumor-infiltrating lymphocytes and immune-related genes signatures associated with differential immunotherapy responses. We systematically assessed the tumor microenvironment using the NanoString GeoMx digital spatial profiler, single-cell RNA-seq and multiplex immunohistochemistry in ESCC. Finally, we evaluated the associations between HLA-A-positive tertiary lymphoid structures (TLSs) and patients' responses to ICB in 60 ESCC patients. RESULTS: Tumor infiltrating B lymphocytes and several immune-related gene signatures, such as the antigen presenting machinery (APM) signature, are significantly elevated in ICB treatment responders. Multiplex immunohistochemistry identified the presence of HLA-A+ TLSs and showed that TLS-resident cells increasingly express HLA-A as TLSs mature. Most TLS-resident HLA-A+ cells are tumor-infiltrating T (TIL-T) or tumor-infiltrating B (TIL-B) lymphocytes. Digital spatial profiling of spatially distinct TIL-T lymphocytes and single-cell RNA-seq data from 60 ESCC tumor tissues revealed that CXCL13-expressing exhausted TIL-Ts inside TLSs are reactivated with elevated expression of the APM signature as TLSs mature. Finally, we demonstrated that HLA-A+ TLSs and their major cellular components, TIL-Ts and TIL-Bs, are associated with a clinical benefit from ICB treatment for ESCC. CONCLUSIONS: HLA-A+ TLSs are present in ESCC tumor tissues. TLS-resident TIL-Ts with elevated expression of the APM signature may be reactivated. HLA-A+ TLSs and their major cellular components, TIL-Ts and TIL-Bs, may serve as biomarkers for ICB-treated ESCC patients.

Distinct epigenetic modulation of differentially expressed genes in the adult mouse brain following prenatal exposure to low-dose bisphenol A.

Bisphenol A (BPA) is a common component in the manufacture of daily plastic consumer goods. Recent studies have suggested that prenatal exposure to BPA can increase the susceptibility of offspring to mental illness, although the underlying mechanisms remain unclear. In this study, we performed transcriptomic and epigenomic profiling in the adult mouse brain following prenatal exposure to low-dose BPA. We observed a sex-specific transcriptional dysregulation in the cortex, with more significant differentially expressed genes was observed in adult cortex from male offspring. Moreover, the upregulated genes primarily influenced neuronal functions, while the downregulated genes were significantly associated with energy metabolism pathways. More evidence supporting impaired mitochondrial function included a decreased ATP level and a reduced number of mitochondria in the cortical neuron of the BPA group. We further investigated the higher-order chromatin regulatory patterns of DEGs by incorporating published Hi-C data. Interestingly, we found that upregulated genes exhibited more distal interactions with multiple enhancers, while downregulated genes displayed relatively short-range interactions among adjacent genes. Our data further revealed decreased H3K9me3 signal on the distal enhancers of upregulated genes, whereas increased DNA methylation and H3K27me3 signals on the promoters of downregulated genes. In summary, our study provides compelling evidence for the potential health risks associated with prenatal exposure to BPA, and uncovers sex-specific transcriptional changes with a complex interplay of multiple epigenetic mechanisms.

Genotype-dependent epigenetic regulation of DLGAP2 in alcohol use and dependence.

Alcohol misuse is a major public health problem originating from genetic and environmental risk factors. Alterations in the brain epigenome may orchestrate changes in gene expression that lead to alcohol misuse and dependence. Through epigenome-wide association analysis of DNA methylation from human brain tissues, we identified a differentially methylated region, DMR-DLGAP2, associated with alcohol dependence. Methylation within DMR-DLGAP2 was found to be genotype-dependent, allele-specific and associated with reward processing in brain. Methylation at the DMR-DLGAP2 regulated expression of DLGAP2 in vitro, and Dlgap2-deficient mice showed reduced alcohol consumption compared with wild-type controls. These results suggest that DLGAP2 may be an interface for genetic and epigenetic factors controlling alcohol use and dependence.

Artificial neural network model to predict post-hepatectomy early recurrence of hepatocellular carcinoma without macroscopic vascular invasion.

BACKGROUND: The accurate prediction of post-hepatectomy early recurrence (PHER) of hepatocellular carcinoma (HCC) is vital in determining postoperative adjuvant treatment and monitoring. This study aimed to develop and validate an artificial neural network (ANN) model to predict PHER in HCC patients without macroscopic vascular invasion. METHODS: Nine hundred and three patients who underwent curative liver resection for HCC participated in this study. They were randomly divided into derivation (n = 679) and validation (n = 224) cohorts. The ANN model was developed in the derivation cohort and subsequently verified in the validation cohort. RESULTS: PHER morbidity in the derivation and validation cohorts was 34.8 and 39.2%, respectively. A multivariable analysis revealed that hepatitis B virus deoxyribonucleic acid load, γ-glutamyl transpeptidase level, α-fetoprotein level, tumor size, tumor differentiation, microvascular invasion, satellite nodules, and blood loss were significantly associated with PHER. These factors were incorporated into an ANN model, which displayed greater discriminatory abilities than a Cox's proportional hazards model, preexisting recurrence models, and commonly used staging systems for predicting PHER. The recurrence-free survival curves were significantly different between patients that had been stratified into two risk groups. CONCLUSION: When compared to other models and staging systems, the ANN model has a significant advantage in predicting PHER for HCC patients without macroscopic vascular invasion.

SETDB1 regulates short interspersed nuclear elements and chromatin loop organization in mouse neural precursor cells.

BACKGROUND: Transposable elements play a critical role in maintaining genome architecture during neurodevelopment. Short Interspersed Nuclear Elements (SINEs), a major subtype of transposable elements, are known to harbor binding sites for the CCCTC-binding factor (CTCF) and pivotal in orchestrating chromatin organization. However, the regulatory mechanisms controlling the activity of SINEs in the developing brain remains elusive. RESULTS: In our study, we conduct a comprehensive genome-wide epigenetic analysis in mouse neural precursor cells using ATAC-seq, ChIP-seq, whole genome bisulfite sequencing, in situ Hi-C, and RNA-seq. Our findings reveal that the SET domain bifurcated histone lysine methyltransferase 1 (SETDB1)-mediated H3K9me3, in conjunction with DNA methylation, restricts chromatin accessibility on a selective subset of SINEs in neural precursor cells. Mechanistically, loss of Setdb1 increases CTCF access to these SINE elements and contributes to chromatin loop reorganization. Moreover, de novo loop formation contributes to differential gene expression, including the dysregulation of genes enriched in mitotic pathways. This leads to the disruptions of cell proliferation in the embryonic brain after genetic ablation of Setdb1 both in vitro and in vivo. CONCLUSIONS: In summary, our study sheds light on the epigenetic regulation of SINEs in mouse neural precursor cells, suggesting their role in maintaining chromatin organization and cell proliferation during neurodevelopment.

Deep learning on tertiary lymphoid structures in hematoxylin-eosin predicts cancer prognosis and immunotherapy response.

Tertiary lymphoid structures (TLSs) have been associated with favorable immunotherapy responses and prognosis in various cancers. Despite their significance, their quantification using multiplex immunohistochemistry (mIHC) staining of T and B lymphocytes remains labor-intensive, limiting its clinical utility. To address this challenge, we curated a dataset from matched mIHC and H&E whole-slide images (WSIs) and developed a deep learning model for automated segmentation of TLSs. The model achieved Dice coefficients of 0.91 on the internal test set and 0.866 on the external validation set, along with intersection over union (IoU) scores of 0.819 and 0.787, respectively. The TLS ratio, defined as the segmented TLS area over the total tissue area, correlated with B lymphocyte levels and the expression of CXCL13, a chemokine associated with TLS formation, in 6140 patients spanning 16 tumor types from The Cancer Genome Atlas (TCGA). The prognostic models for overall survival indicated that the inclusion of the TLS ratio with TNM staging significantly enhanced the models' discriminative ability, outperforming the traditional models that solely incorporated TNM staging, in 10 out of 15 TCGA tumor types. Furthermore, when applied to biopsied treatment-naïve tumor samples, higher TLS ratios predicted a positive immunotherapy response across multiple cohorts, including specific therapies for esophageal squamous cell carcinoma, non-small cell lung cancer, and stomach adenocarcinoma. In conclusion, our deep learning-based approach offers an automated and reproducible method for TLS segmentation and quantification, highlighting its potential in predicting immunotherapy response and informing cancer prognosis.

CRISPR-based targeted haplotype-resolved assembly of a megabase region.

Constructing high-quality haplotype-resolved genome assemblies has substantially improved the ability to detect and characterize genetic variants. A targeted approach providing readily access to the rich information from haplotype-resolved genome assemblies will be appealing to groups of basic researchers and medical scientists focused on specific genomic regions. Here, using the 4.5 megabase, notoriously difficult-to-assemble major histocompatibility complex (MHC) region as an example, we demonstrated an approach to construct haplotype-resolved assembly of the targeted genomic region with the CRISPR-based enrichment. Compared to the results from haplotype-resolved genome assembly, our targeted approach achieved comparable completeness and accuracy with reduced computing complexity, sequencing cost, as well as the amount of starting materials. Moreover, using the targeted assembled personal MHC haplotypes as the reference both improves the quantification accuracy for sequencing data and enables allele-specific functional genomics analyses of the MHC region. Given its highly efficient use of resources, our approach can greatly facilitate population genetic studies of targeted regions, and may pave a new way to elucidate the molecular mechanisms in disease etiology.

DNA hypomethylation of a transcription factor binding site within the promoter of a gout risk gene NRBP1 upregulates its expression by inhibition of TFAP2A binding.

BACKGROUND: Genome-wide association studies (GWASs) have identified dozens of loci associated with gout, but for most cases, the risk genes and the underlying molecular mechanisms contributing to these associations are unknown. This study sought to understand the molecular mechanism of a common genetic variant, rs780093, in the development of gout, both in vitro and in vivo. RESULTS: Nuclear receptor binding protein 1 (NRBP1), as a gout risk gene, and its regulatory region, 72 bp upstream of the transcription start site, designated as B1, were identified through integrative analyses of genome-wide genotype and DNA methylation data. We observed elevated NRBP1 expression in human peripheral blood mononuclear cells (PBMCs) from gout patients. In vitro luciferase reporter and protein pulldown assay results showed that DNA methylation could increase the binding of the transcription factor TFAP2A to B1, leading to suppressed gene expression. There results were further confirmed by in vivo bisulfite pyrosequencing showing that hypomethylation on B1 is associated with increased NRBP1 expression in gout patients. CONCLUSIONS: Hypomethylation at the promoter region of NRBP1 reduces the binding of TFAP2A and thus leads to elevated NRBP1 expression, which might contribute to the development of gout.

DNA methylation mediates genotype and smoking interaction in the development of anti-citrullinated peptide antibody-positive rheumatoid arthritis.

BACKGROUND: Multiple factors, including interactions between genetic and environmental risks, are important in susceptibility to rheumatoid arthritis (RA). However, the underlying mechanism is not fully understood. This study was undertaken to evaluate whether DNA methylation can mediate the interaction between genotype and smoking in the development of anti-citrullinated peptide antibody (ACPA)-positive RA. METHODS: We investigated the gene-smoking interactions in DNA methylation using 393 individuals from the Epidemiological Investigation of Rheumatoid Arthritis (EIRA). The interaction between rs6933349 and smoking in the risk of developing ACPA-positive RA was further evaluated in a larger portion of the EIRA (1119 controls and 944 ACPA-positive patients with RA), and in the Malaysian Epidemiological Investigation of Rheumatoid Arthritis (MyEIRA) (1556 controls and 792 ACPA-positive patients with RA). Finally, mediation analysis was performed to investigate whether DNA methylation of cg21325723 mediates this gene-environment interaction on the risk of developing of ACPA-positive RA. RESULTS: We identified and replicated one significant gene-environment interaction between rs6933349 and smoking in DNA methylation of cg21325723. This gene-smoking interaction is a novel interaction in the risk of developing ACPA-positive in both Caucasian (multiplicative P value = 0.056; additive P value = 0.016) and Asian populations (multiplicative P value = 0.035; additive P value = 0.00027), and it is mediated through DNA methylation of cg21325723. CONCLUSIONS: We showed that DNA methylation of cg21325723 can mediate the gene-environment interaction between rs6933349 and smoking, impacting the risk of developing ACPA-positive RA, thus being a potential regulator that integrates both internal genetic and external environmental risk factors.