Biology subjects

KIM, C. H.

Publications and source records attributed to KIM, C. H..

5 recordsLinked to original sources

bioRxiv2026

simCRISPR: Modeling experimental complexity in pooled CRISPR screens

Pooled CRISPR screens are widely used to investigate gene function and uncover genetic interactions. However, benchmarking computational methods for detecting gene-by-environment (GxE) interactions remains difficult because ground truth is rarely available and existing simulation tools are not designed for GxE screening contexts. To address this, we developed simCRISPR, a flexible simulation framework for generating pooled CRISPR screen data under complex experimental designs. Using simulated datasets informed by empirical CRISPR screen designs, we evaluated commonly used analysis methods, comparing normalization strategies based on safe-harbor versus non-targeting sgRNAs and assessing empirical log2FC thresholds as an additional effect-size criterion. We found that safe-harbor-based normalization improved interaction detection when DNA damage-related effects were present, particularly when combined with empirical log2FC thresholding for DESeq2. Application of this workflow to a doxorubicin GxE screen further showed that safe-harbor-based normalization reduced bias in log2FC distributions and identified additional biologically relevant candidates. simCRISPR is available at https://github.com/bachergroup/simCRISPR.

bioinformatics↗

bioRxiv2025

Genome-wide CRISPR screens identified C18orf32 as a novel regulator of lipid metabolism that mediates PFOA-induced toxicity

Perfluorooctanoic acid (PFOA) remains a public health concern due to its persistence in the environment, despite global production bans. Epidemiological, animal, and in vitro studies have consistently linked PFOA exposure to hepatotoxicity characterized by lipid dysregulation; however, the molecular mechanisms underlying these adverse effects in humans remain poorly understood. To address this gap, we employed an integrated functional toxicogenomics framework combining CRISPR screening (for gene target identification), target validation, and assessment of target-derived cellular and molecular phenotypes. From genome-wide CRISPR screens in HepG2/C3A human liver cells, we identified 319 candidate genes (140 sensitive and 179 resistant) that modulate PFOA toxicity. Among these, C18orf32--encoding a lipid droplet (LD)-associated protein and the top resistant candidate--was selected for further mechanistic investigation based on its potential functional involvement in lipid metabolism. Targeted knockout (KO) of C18orf32 conferred marked cellular resistance to PFOA, accompanied by reduced LD and Triglyceride accumulation in cells under the exposure condition, suggesting a functional role of C18orf32p in lipid dysregulation. To elucidate molecular mechanisms underlying the cellular phenotypes, we conducted systematic transcriptomics profiling of wild-type (WT) and C18orf32 KO cells in the absence and presence of PFOA exposure. C18orf32 KO caused extensive gene expression reprogramming, with over one-third of coding genes differentially expressed. Notably, a range of lipid metabolism pathways including cholesterol metabolism, PPAR signaling, fatty acid metabolism, and peroxisome {beta}-oxidation--known mediators of PFOA-induced hepatotoxicity--were significantly downregulated in C18orf32 KO cells, contrasting to upregulation of these pathways in PFOA exposed WT cells. Collectively, our results identify C18orf32p as a previously unrecognized regulator of hepatic lipid metabolism and a key genetic determinant of PFOA-induced lipid dysregulation and hepatotoxicity in humans.

pharmacology and toxicology↗

bioRxiv2025

Identification of Functional Genetic Components Modulating Toxicity Response to PFOS using Genome-wide CRISPR Screens in HepG2/C3A cells

Perfluorooctane sulfonate (PFOS) poses significant health and environmental risks due to its persistence and widespread use and has been linked to various adverse outcomes, such as liver toxicity. Although the molecular responses and toxicity effects of PFOS exposure have been extensively studied, considerable uncertainty remains regarding the causal mechanisms leading to PFOS-associated adverse effects. To help bridge this gap, we conducted CRISPR screens in HepG2/C3A human liver cells exposed to IC25 (170 {micro}M) of PFOS to identify genes and pathways influencing PFOS-induced cytotoxicity. Using a genome-wide CRISPR knockout library targeting 18,819 genes, we identified 340 candidate genes that modulate PFOS-induced cytotoxicity when genetically disrupted (189 gene disruptions increased sensitivity and 151 gene disruptions increased resistance). From these candidate genes, we individually disrupted two candidate genes, SLC6A9 which encodes the glycine transporter GlyT1, and CPSF2, and confirmed increased resistance to PFOS exposure. Further, molecular docking analysis predicts that PFOS directly binds to GlyT1 and functional inhibition of GlyT1 also increases resistance to PFOS exposure. Gene-Disease outcome association analysis using the Comparative Toxicogenomics Database (CTD) indicated an enrichment of candidate genes associated with cancer-related and liver disease phenotypes. KEGG and STRING enrichment analyses found over representation of several biological pathways including DNA damage response and cell cycle. Lastly, cross-species conservation analysis using the top two validated gene targets found that their pathways were highly conserved in several environmentally relevant species. These findings provide new mechanistic and functional insights into PFOS-induced cytotoxicity, highlight potential molecular targets for toxicity mitigation, and establish a foundation for cross-species toxicogenomic modeling of PFOS health effects.

pharmacology and toxicology↗

bioRxiv2025

Characterizing common loss-of-function genes and their potential utility in assessing population variability and chemical susceptibility

Inter-individual and population variability in susceptibility to chemical exposures confounds determination of threshold exposure levels to protect the most vulnerable. Current risk assessment frameworks, in the absence of empiric chemical-specific data, generally recommend default or probabilistic adjustment factors to account for such variability. We present an experimental approach to incorporate common genetic variants potentially impacting population-level differences in toxicant susceptibility into human cell-based models for any cellular apical endpoint of interest. We focus on the genes with the most common aggregate loss-of-function (LoF) alleles in the gnomAD v3.0 data which we designated as the PopVarLoF set. Unexpectedly, enrichment analysis of these genes found significant overrepresentation of gene products playing important functional roles in toxicology. Interrogation of GWAS and PheWAS databases found that these genes are associated with diverse metabolic phenotypes consistent with the relevance of the PopVarLoF set in studying variability of toxicant response in human populations. We further characterized the PopVarLoF set by developing custom lentiviral CRISPR knockout libraries targeting the PopVarLoF genes to assess their functional essentiality in the HepG2/C3A cell line. Functional disruption of 14 of the PopVarLoF genes ([~]1 %) without toxicant exposure resulted in significant growth defects in this cell line, consistent with the majority of PopVarLoF gene products having non-essential roles. The development of human cell-based toxicity assays or other NAMs which include the empiric assessment of common genetic sources of population variability in susceptibility to chemical exposure could contribute to more robust risk assessment which protects vulnerable populations while reducing uncertainty. Impact statementWe characterize common loss of function genetic variants which could impact toxicant susceptibility and describe an approach to incorporate them into NAMs to enable empiric estimates of the contribution of genetic variability to diverse toxicity endpoints.

pharmacology and toxicology↗

bioRxiv2025

Physiologically Relevant 3D CRISPR Screening Enhances Mechanistic Insight into Chemical Toxicity Compared to 2D Screening

Many omics-based approaches in toxicology research primarily rely on correlative data, often lacking functional relationships or causal links between genotypes and phenotypes. CRISPR-based approaches can overcome this limitation by establishing direct causal connections between genes and toxicological phenotypes. Moreover, CRISPR screens enable scalable and systematic interrogation of gene function and associated mechanisms following chemical exposure, predominantly using in vitro models. In line with the paradigm of new approach methodologies (NAMs) in toxicology research, CRISPR screens hold promise to provide an in vitro cell-based functional toxicogenomics approach. One of the main limitations of conventional in vitro assays is their compromised physiological relevance to humans due to their inability to fully recapitulate in vivo physiology. To improve the functional and physiological relevance of the toxicogenomics approach, we developed a 3D CRISPR screening system using HepG2/C3A spheroids generated and cultivated in a continuously rotating bioreactor (ClinoStar). We first performed time-course 3D CRISPR screens to identify genes that confer growth disadvantage or advantage, influencing spheroid development compared to 2D cultures. We then applied this approach to a chemical toxicity study using doxorubicin, comparing the performance of the 3D and 2D systems in identifying chemical-specific mechanisms. The results showed that the 3D system captured more candidate genetic determinants and biological pathways related to DNA damage processes--a known toxicity mechanism of doxorubicin--demonstrating improved performance in identifying chemical-specific pathways over the 2D counterpart. In our screens, we employed custom CRISPR sgRNA libraries representing common human loss-of-function genetic variants (mean allele frequency > 0.1% in all individuals catalogued in the genome aggregation database), which potentially affect toxicity responses. By comparing our CRISPR screen results with previously reported genetic associations for doxorubicin response, we found that the 3D system identified more known associated genes than the 2D system. Together, the 3D CRISPR screening system demonstrated its feasibility and utility for physiologically relevant functional toxicogenomics. This platform enables in vitro NAMs, by providing a scalable and effective approach to identify causal genetic determinants and biological pathways that modulate chemical-induced toxicity.

pharmacology and toxicology↗