Comprehensive Analysis of Variant Effects, Mutation Spectrum, and Functional Enrichment in Familial Hypercholesterolemia- Associated Mutations

ABSTRACT

This genomics project investigated genetic variants associated with familial hypercholesterolemia (FH), a hereditary disorder characterized by elevated LDL cholesterol and increased cardiovascular risk. Using publicly available sequence data on genes commonly implicated in FH, such as LDLR, APOB, and PCSK9, pathogenic variants were identified and classified through bioinformatics tools including Variant Effect Predictor (VEP), ClinVar, and gene ontology (GO) enrichment software. Three distinct analyses were conducted: (1) a distribution of predicted variant consequences, (2) a GO enrichment analysis to examine the biological processes, molecular functions, and cellular components impacted by the variants, and (3) a mutation spectrum analysis assessing single-nucleotide variants (SNVs), transition/transversion ratios, and insertion and deletion (indel) patterns. These analyses revealed functional trends and mutation dynamics that offer insights into the genetic architecture of FH. These findings contribute to a deeper understanding of high-impact variants and may inform the development of targeted gene therapies for individuals with FH.

INTRODUCTION.

Familial hypercholesterolemia (FH) is a common autosomal co-dominant disorder that is associated with lifelong high levels of low-density lipoprotein cholesterol (LDL-C) and a remarkably elevated risk for premature atherosclerotic cardiovascular disease (ASCVD). Present in about 1 in 250 individuals with heterozygous FH (HeFH) and 1 in 160,000–300,000 with homozygous FH (HoFH), the condition mostly results from pathogenic variants in LDLR (>80%), APOB (~5–10%), and PCSK9 (~1–2%) [1]. Increased involvement of LDLRAP1 and rare APOE variants emphasizes genetic heterogeneity influencing FH phenotypes. Among more than 2,000 LDLR variants described, five functional classes (I–V) of receptor synthesis, trafficking, binding, internalization, or recycling explain the variability of LDL‑C levels in FH patients [2].

Under normal physiological conditions, circulating LDL particles are cleared from the bloodstream primarily through the LDL receptor (LDLR) pathway in the liver. LDL particles contain apolipoprotein B (ApoB), which serves as the primary ligand allowing LDL to bind to LDL receptors on hepatocyte surfaces. Once bound, the LDL–LDLR complex is internalized through endocytosis, after which LDL particles are degraded in lysosomes and the receptor is recycled back to the cell surface to continue clearing cholesterol from circulation. Proprotein convertase subtilisin/kexin type 9 (PCSK9) regulates this pathway by binding to LDL receptors and targeting them for lysosomal degradation rather than recycling, thereby reducing the number of functional receptors available to remove LDL from the bloodstream. Disruption of this tightly regulated pathway, whether through impaired receptor function, defective ApoB binding, or excessive PCSK9-mediated receptor degradation, leads to reduced LDL clearance and the accumulation of circulating LDL-C that characterizes FH.

While traditional therapies, statins, ezetimibe, PCSK9 inhibitors, bile acid sequestrants, and LDL apheresis, are effective at reducing LDL-C levels to a degree, especially in HeFH, they are often less effective in many HoFH patients, particularly those with severely reduced or absent LDL receptor function [3]. Because these treatments depend largely on the presence of functional LDL receptors to enhance LDL clearance, their therapeutic impact can be limited in individuals with receptor-defective or receptor-negative variants. These limitations highlight the need for better and more efficient treatment options.

Gene-based therapies provide a paradigm-shifting solution for plugging the gap by directly repairing or supplementing genetic defects. Strategies include viral and non-viral delivery of endogenous LDLR cDNA, RNA-targeting antisense or RNA interference, and precision genome editing with CRISPR-Cas9, base editing, and prime editing [4]. Encouraging animal and early human data, e.g., AAV-mediated LDL replacement, PCSK9 and ANGPTL3 inhibition by lipid-nanoparticle–delivered base editors (e.g., VERVE-101, VERVE-201), and companion siRNA/ASO therapies like inclisiran, induce substantial LDL-C decreases and provide a foundation for long-lasting cures [3].

However, the efficacy of gene therapies is strongly associated with causative FH genetic variants. Different types of variants, including missense, nonsense, splicing, insertion/deletion, and copy number variants, can alter gene function in distinct ways that range from partial reduction of receptor activity to complete loss of function. These differences create unique mechanistic challenges for treatment development because therapeutic strategies and genome-editing approaches must be tailored to the specific mutation type and its functional consequences [5]. Standard variant classification, established by ClinGen and ACMG-AMP specifications, becomes imperative to stratify patients, guide variant-directed therapy, and ensure therapeutic safety and efficacy [6].

This study aims to systematically delineate a comprehensive panel of FH-associated variants in LDLR, APOB, PCSK9, and related genes, and assess their predicted functional impact using computational analysis of previously published datasets. Coupled with variant-specific functional annotations and emerging gene therapy paradigms, we aim to enable the rationale-based design of precision therapies for each genetic class, toward the goal of curative therapy for FH.

MATERIALS AND METHODS.

Data Compilation and Sources.

To characterize the genetic landscape of FH, we assembled a comprehensive dataset of 12,334 unique variants which were compiled from a combination of publicly available genomic repositories. ClinVar, Ensembl, and dbSNP, as well as four peer-reviewed studies focused on FH genetics and LDL-C levels. These included: (1) Genetic spectrum of FH and correlations with clinical expression by Di Taranto et al., which contributed detailed variant-level data from 22 unrelated FH patients; (2) Family-specific aggregation of lipid GWAS variants by Nikkola et al., providing 65 variants with rsIDs and associated risk alleles from a large Austrian family; (3) A multi-ancestry genome-wide association study of LDL-C levels by Umlai et al., offering 8,043 variants with detailed statistical metadata; and (4) Large-scale functional characterization of LDLR variants by Islam et al., which contributed ~500 experimentally validated LDLR variants. These sources ensured broad representation across ancestries, variant types, and experimental contexts, enriching the reliability of downstream analyses.

Variant Annotation and Filtering.

Variants were reformatted into a tab-delimited VCF-compatible input file, including chromosome, position, REF, and ALT alleles, and then processed through the Ensembl Variant Effect Predictor (VEP) to determine molecular consequences. Variants with incomplete annotations (e.g., missing gene information) were excluded unless manually validated [7].

Pathogenicity Classification and Molecular Consequence Analysis.

ClinVar annotations were used to categorize each variant as “pathogenic” or “likely pathogenic.” This works by pulling down variant information that has been previously categorized in human test subjects. Molecular consequence categories output by VEP (e.g., missense, nonsense, frameshift, splice site) were aggregated and analyzed to determine the prevalence of high-impact changes. These effects are determined by comparing reference gene sequences to alternate mutant sequences and then identifying where in the coding sequence a mutation has changed and how that would impact the subsequent protein sequence. Variants contributing at least 1% of the overall consequence frequency were retained for primary visualizations. Frequency distribution plots were generated using R (v4.5.2) [8] with ggplot2 [9].

Gene Ontology Enrichment Analysis.

Gene Ontology (GO) enrichment analysis was conducted to identify overrepresented biological processes, molecular functions, and cellular components among genes harboring pathogenic FH variants. The packages clusterProfiler [10] and org.Hs.eg.db [11] were necessary to perform gene ontology enrichment analysis, with p-values adjusted using the Benjamini–Hochberg method. Only terms with adjusted p-values < 0.05 and q-values < 0.20 were considered significant. The top five GO terms per category (Biological Process, Molecular Function, and Cellular Component) were visualized via horizontal bar plots.

Mutation Spectrum and Structural Variation Characterization.

To explore underlying mutational mechanisms, we performed a detailed mutation spectrum analysis. Custom R scripts categorized single-nucleotide variants (SNVs) into six base substitution classes (e.g., A>G, C>T) and calculated the transition/transversion (Ti/Tv) ratio. Additionally, we examined substitution trends at A/T versus G/C nucleotide sites. Insertion and deletions (indels) were quantified separately, with metrics including insertion-to-deletion ratios and size distributions. The final visualization comprised SNV spectrum bar charts, Ti/Tv ratio plots, G/C vs A/T bar plots, and an indel size histogram.

Statistical and Visualization Tools.

All statistical analysis and data visualization were conducted using RStudio (v4.5.2). Plots were generated using ggplot2, and composite figures were arranged using gridExtra [12]. Enrichment significance was determined through adjusted p-values, with multiple hypothesis correction applied where applicable. Summary tables and figures were exported for integration into supplementary materials.

RESULTS.

Predicted Molecular Consequences of FH-Associated Variants.

The curated dataset of 12,334 unique variants revealed a wide distribution of predicted functional consequences, as classified by Ensembl’s Variant Effect Predictor (VEP) (Figure 1). The majority of variants (n = 6,339, 51.4%) were missense mutations, altering the amino acid sequence without necessarily truncating the protein, and are thus critical for fine-tuned protein function and interaction. Intronic variants (n = 2,117) and synonymous substitutions (n = 1,984) followed in frequency, with the former potentially affecting splicing regulation and the latter often acting as silent mutations, though some may still impact mRNA stability or translational efficiency.

Other notable categories included splice region variants (n = 1,247) and regulatory region variants (n = 849), both of which may influence gene expression levels or exon usage. High-impact mutations such as frameshifts (n = 849) and stop-gained variants (n = 617) were also present in moderate numbers, resulting in premature truncation and loss-of-function effects. Mutations in untranslated regions (3′ and 5′ UTRs) and upstream/downstream modifiers collectively accounted for a smaller but non-negligible portion of the mutational landscape, further highlighting the genetic complexity of FH.

GO Enrichment Highlights Lipid Metabolism, Transport, and Cellular Localization.

Gene Ontology (GO) enrichment analysis provided critical insights into the functional roles of genes affected by pathogenic and likely pathogenic variants. This analysis revealed consistent and statistically significant enrichment across the Cellular Component, Molecular Function, and Biological Process  categories (Figure 2A–C), illustrating how structural components of lipid transport systems, the molecular interactions they perform, and the downstream metabolic pathways they influence collectively contribute to the pathophysiology of FH.

Within the Cellular Component category (Figure 2C), significant enrichment was observed in plasma lipoprotein particles, lipoprotein complexes, and high-density lipoprotein particles, all directly related to lipid transport in circulation. These structures represent the physical context in which LDL particles circulate and interact with cellular receptors in the bloodstream and liver. Additionally, components involved in vesicular trafficking, such as clathrin-coated vesicles and lysosomal membranes, were overrepresented, aligning with the cellular mechanism by which LDL receptors mediate cholesterol internalization and degradation through the LDL uptake pathway described in the introduction.

The Molecular Function enrichment (Figure 2B) emphasized the functional impact of the observed variants on lipid-interacting domains within these cellular structures. Top terms included cholesterol transfer activity, lipid binding, and glucuronosyltransferase activity. The enrichment of transmembrane transporter activity and receptor binding functions further supports the hypothesis that many variants may affect LDL receptor function or ligand affinity, including interactions between ApoB-containing LDL particles and the LDL receptor, leading to impaired endocytosis and reduced clearance of circulating LDL.

Finally, in the Biological Process domain (Figure 2A), several key lipid-associated pathways were enriched, including lipid homeostasis, cholesterol homeostasis, and regulation of lipid metabolic processes. These pathways represent downstream physiological outcomes of disruptions in cellular localization and molecular interactions involved in LDL handling. Additional processes, such as steroid metabolic process and cellular response to lipids, reflect broader metabolic consequences and adaptive responses that arise when cholesterol clearance through the LDL pathway is impaired. Together, these enriched processes reinforce how mutations affecting LDL-associated cellular components and molecular functions ultimately propagate through larger lipid metabolic networks that are central to FH pathology.

Single-Nucleotide Variant and Indel Spectra Reveal Mutational Biases

Further stratification of the variant dataset provided insight into nucleotide-level mutational dynamics (Figure 3A–D). The mutation spectrum (Figure 3A) was dominated by C>T (n = 4,707) and A>G transitions (n = 3,054), a finding consistent with deamination of methylated cytosines at CpG dinucleotides and typical for germline SNVs [13]. Less frequent were transversions such as C>A (n = 1,778), C>G (n = 1,279), and A>T (n = 648), which are typically associated with oxidative stress or environmental mutagens. This helps to better understand that the failure of germline machinery and not an exogenous stressor is likely the cause of these variants. Analysis of the transition-to-transversion (Ti/Tv) ratio (Figure 3B) revealed a value of 2.0, in line with expectations for human germline datasets [13]. A breakdown of mutations by nucleotide context (Figure 3C) revealed that GC sites harbored a disproportionately higher number of mutations (n = 7,794) compared to AT sites (n = 4,469). This imbalance aligns with known patterns of hypermutability at GC-rich regions, potentially due to methylation status and higher rates of spontaneous transitions. These findings suggest that genomic regions rich in CpG content may serve as mutational hotspots in FH-related genes. Alternatively, genes are known to be GC dense and as such variants found within genes may show a resulting GC bias. In short, the spectrum of SNVs found in this dataset likely indicate that spontaneous germline mutations are primarily driving the emergence of FH in the population.

Indel Landscape Suggests Balanced Insertion and Deletion Events.

Analysis of indels revealed a relatively balanced landscape (Figure 3D), with 820 insertions and 797 deletions, producing an insertion-to-deletion ratio of approximately 1.03. Most indels were small in size, with the majority under 10 base pairs, and the distribution was heavily skewed toward shorter events. The median indel size across all events was 3 bp, indicating that even small insertions and deletions can have large effects on pathogenicity. Notably, frameshift-inducing indels, though fewer in number, likely contribute disproportionately to functional disruption, especially when occurring in ligand-binding or transmembrane domains of LDLR or APOB. This is evidence to support the idea that mutations which affect the coding open reading frame seem to be more likely to be related to FH.

DISCUSSION.

The results of this study uncover key implications for understanding the genetic underpinnings and potential therapeutic directions for FH. One of the most striking observations was the high prevalence of noncoding variants, specifically intronic and intergenic mutations (Figure 1). While coding mutations in genes such as LDLR, APOB, and PCSK9 have been traditionally emphasized in FH pathogenesis [14], the dominant presence of noncoding variants in this dataset underscores the likely importance of regulatory and splicing-related disruptions. Introns, although constituting roughly 24% of the genome[15], appear to harbor an outsized number of potentially functional variants, suggesting nonrandom distribution and functional relevance. Intergenic variants, though often considered non-functional, may influence transcription factor binding, chromatin accessibility, or long-range enhancer-promoter interactions [16]. This high proportion supports previous findings indicating the central role of missense variants in FH, particularly within LDLR, APOB, and PCSK9 genes.

Gene Ontology (GO) enrichment analysis revealed several biological processes associated with lipid metabolism, including fatty acid biosynthesis and cholesterol homeostasis, which are hallmarks of FH [17] (Figure 2a). In addition to these expected pathways, notable enrichment was observed in processes related to xenobiotic detoxification and response to toxic substances (Figure 2a). This raises the possibility that environmental exposure, including pharmaceuticals and pollutants, may modify FH phenotype expression which is an idea supported by growing literature on gene-environment interactions in metabolic diseases [18].

The molecular function and cellular component analyses (Figure 2b and 2c) reinforced the involvement of lipid-binding proteins, receptor activities, and plasma membrane-localized components which are consistent with the pathophysiology of impaired LDL receptor-mediated cholesterol uptake. Such findings highlight precise targets for future functional assays and molecular modeling.

Mutation spectrum analysis revealed a predominance of C>T transitions (Figure 3a), consistent with spontaneous deamination of methylated cytosines at CpG sites which are the most common spontaneous mutations in the human germline [13][19]. This aligns with the notion that not all FH mutations are inherited; some may arise spontaneously, particularly in patients without family history. The calculated transition/transversion (Ti/Tv) ratio (Figure 3b) and base composition analysis (Figure 3c) fall within expected genomic parameters [20], suggesting that patterns of mutation traditionally associated with spontaneous mutation are likely the cause of variants associated with FH. .

The observed indel distributions (Figure 3d) suggests that both insertions and deletions contribute meaningfully to the FH mutational landscape, with a range of sizes indicating varied mutational origins and consequences. These findings carry implications for gene-editing technologies, which must account for these structural variant types when designing therapeutic interventions.

These results have several translational implications. First, they support expanding variant screening beyond exonic regions in clinical FH diagnostics. Second, they suggest new targets for pharmacogenomics research, especially concerning lipid-lowering therapies and environmental modulators. Finally, the findings reinforce the relevance of noncoding and regulatory variants as both diagnostic biomarkers and therapeutic entry points.

This comprehensive mutational overview highlights the multifaceted nature of FH genetics, ranging from subtle missense changes to disruptive indels and splicing defects. Collectively, the results emphasize the importance of evaluating variant function not only at the sequence level, but also through broader functional and structural lenses that incorporate gene ontology, pathway interaction, and cellular localization.

CONCLUSION.

In summary, this work provides a broad genomic and bioinformatic view of FH, characterizing the diversity of its variants and their biological implications. The results highlight how different mutation types which range from coding to regulatory, collectively influence lipid regulation pathways and receptor function.

These findings not only reinforce known mechanisms behind FH but also open new avenues for exploration, particularly in understanding how noncoding and population-specific variants contribute to disease risk. Moving forward, expanding analyses to include transcriptomic data, variant impact modeling, and patient-specific genotype-phenotype correlations could further clarify FH’s molecular complexity.

Future directions include integrating this variant database into precision medicine efforts to support targeted therapeutic design and individualized treatment strategies. By bridging genetic data with clinical application, this study lays groundwork for more effective, variant-informed approaches to managing and potentially curing FH.

ACKNOWLEDGMENTS.

I would like to thank my teachers and parents for their support in my academic journey. Additionally I would like to thank Rob Melde and the Polygence mentor team for support during this project.

REFERENCES.

1. Q. Fu, L. Hu, T. Shen, R. Yang, L. Jiang, Recent Advances in Gene Therapy for Familial Hypercholesterolemia: An Update Review. Journal of Clinical Medicine. 11, 6773 (2022).
2. N. Parsamanesh et al., Gene and cell therapy approaches for familial hypercholesterolemia: An update. Drug Discovery Today. 28, 103470 (2022).
3. M. W. Huff, J. M. Assini, R. A. Hegele, Gene Therapy for Hypercholesterolemia. Circulation Research. 115, 542–545 (2014).
4. M. Hurwitz, O. Ugoala, A. Kulkarni, J. Patel, Emerging Gene Therapies For Familial Hypercholesterolemia. American College of Cardiology. (2025).
5. R. Chen, S. Lin, X. Chen, The promising novel therapies for familial hypercholesterolemia. Journal of Clinical Laboratory Analysis. 36, e24552 (2022).
6. J. R. Chora et al., The Clinical Genome Resource (ClinGen) Familial Hypercholesterolemia Variant Curation Expert Panel consensus guidelines for LDLR variant classification. Genetics in Medicine. 24, 293–306 (2021).
7. McLaren, W., et al. The Ensembl Variant Effect Predictor. Genome Biology, 17, 122. (2016).
8. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, (2024).
9. H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, (2016).
10. S Xu, et al. Using clusterProfiler to characterize multiomics data. Nature Protocols. 19, 3292-3320. (2024).
11. Carlson M, org.Hs.eg.db: Genome wide annotation for Human. R package version 3.20.0. (2024).
12. Auguie B, gridExtra: Miscellaneous Functions for “Grid” Graphics. R package version 2.3. (2017).
13. R. Rahbari et al., Timing, rates and spectra of human germline mutation. Nature Genetics. 48, 126–133 (2015).
14. M. A. Austin, Genetic causes of monogenic heterozygous familial hypercholesterolemia: A HuGE prevalence review. American Journal of Epidemiology. 160, 407–420 (2004).
15. M. Sakharkar, V. T. Chow, P. Kangueane, Distributions of exons and introns in the human genome. In Silico Biology, 4, 387–393. (2004).
16. L. D. Ward, M. Kellis, Interpreting noncoding genetic variation in complex traits and human disease. Nature Biotechnology. 30, 1095–1106 (2012).
17. M. S. Brown, J. L. Goldstein, A receptor-mediated pathway for cholesterol homeostasis. Science. 232, 34–47 (1986).
18. M. A. Austin, C. M. Hutter, R. L. Zimmern, S. E. Humphries, Genetic Causes of Monogenic Heterozygous Familial Hypercholesterolemia: A HuGE Prevalence Review. American Journal of Epidemiology. 160, 407–420 (2004).
19. K. Harris, R. Nielsen, Error-prone polymerase activity causes multinucleotide mutations in humans. Genome Research. 24, 1445–1454 (2014).
20. G. Lunter, C. P. Ponting, J. Hein, Genome-wide identification of human functional DNA using a neutral indel model. PLoS Computational Biology. 2, e5 (2006).

 

---

---

Posted by buchanle on Friday, May 15, 2026 in May 2026.

Tags: familial hypercholesterolemia, gene ontology, genetic variants, LDLR, mutation spectrum