Skip to main content
NCBI home page
Frontiers in Pharmacology logoLink to Frontiers in Pharmacology
. 2026 Jul 13;17:1880997. doi: 10.3389/fphar.2026.1880997

Integrative bioinformatics analysis identifies APOE as a candidate link between lipid dysregulation and macrophage activation in inborn errors of metabolism

Rahmat Dani Satria 1, Riswan Hadi Kusuma 1,2, Nazhipah Isnani 3, Usi Sukorini 1,4, Tri Ratnaningsih 1,4, Elizabeth Henny Herningtyas 1,4, Umi S Intansari 1,4, Windarwati 1,4, Nur Imma Fatimah Harahap 1,2, Primadhy Harris Syaifullah 1,4, Kharisa Rasikhatul Hikmatul 1, Litia Lucretia Bodiwan 1, Richardus Wisnandito Prabasaktya 1, Nicholas Matthew Santoso 1, Muhammad Arkaan Putra Aprilia Darmawan 1, Eko Mugiyanto 5, Bambang Ardianto 6, Lalu Muhammad Irham 7, Chiou-Feng Lin 8,9, Chia-Ling Chen 8,10,11,12, Wirawan Adikusuma 13,*
PMCID: PMC13402178  PMID: 42555484

Abstract

Background

Inborn errors of metabolism (IEM) are a heterogeneous group of genetic disorders characterized by metabolic dysregulation and high mortality. Despite extensive genetic discoveries, the molecular mechanisms underlying severe disease progression remain incompletely understood. Increasing evidence suggests that host metabolic states, particularly lipid dysregulation, can profoundly influence immune cell function and may contribute to impaired responses against intracellular pathogens.

Methods

We integrated genomic data from four high-confidence databases (GWAS Catalog, CTD, Open Targets, and ClinVar) and identified 1,288 IEM-associated genes. Disease ontology enrichment, multi-parametric functional annotation, and cell-type enrichment analyses were performed using Open-XGR, WebGestalt, and Tabula Sapiens. Bioinformatics findings were interpreted alongside clinical and cytomorphological observations from a pair of one-year-old twins with suspected IEM and macrophage activation syndrome. No patient-specific genetic or molecular validation of APOE was available.

Results

Functional annotation identified APOE as the highest-scoring gene across seven biological categories, with enrichment in lipid metabolism- and macrophage-related pathways. Cell-type enrichment demonstrated an overrepresentation of IEM-associated genes in macrophage-related reference signatures, including datasets annotated as M2 macrophages. APOE was the only gene consistently represented across macrophage, M2 macrophage, protein–lipid complex, and vasculature-associated macrophage categories. Clinical evaluation of the twin cases showed persistent hyperbilirubinemia, elevated transaminases, cholestasis, and macrophage activation. Bone marrow and cerebrospinal fluid cytology demonstrated lipid-laden monocyte/macrophage morphology. These clinical observations are concordant with lipid-handling abnormalities in macrophages but do not establish an APOE-dependent mechanism or macrophage polarization state.

Conclusion

This study identifies APOE as a highly prioritized candidate gene associated with lipid metabolism pathways and macrophage-related signatures in IEM. These findings support the hypothesis that APOE-associated lipid dysregulation may contribute to macrophage dysfunction and warrant further experimental validation.

Keywords: APOE, bioinformatics, biomarkers, inborn errors of metabolism, macrophage activation syndrome

1. Introduction

Inborn errors of metabolism (IEM), also referred to as inherited metabolic disorders (IMD), are caused by disruptions in biochemical pathways resulting from deficiencies in enzymes, cofactors, or transporters (Mordaunt et al., 2020; Balakrishnan, 2021). These disorders constitute a complex and heterogeneous group of rare diseases characterized by diverse clinical manifestations and significant morbidity and mortality (Mordaunt et al., 2020; Balakrishnan, 2021; Den Hollander et al., 2025). According to the International Classification of Inherited Metabolic Disorders (ICIMD), 1,546 IEMs have been identified to date, and this number continues to increase with advancements in analytical technologies that elucidate the molecular and metabolic basis of previously undefined conditions (Den Hollander et al., 2025). Although each disorder is individually rare, their combined incidence is considerable, estimated to occur in approximately one in every 800 to 2,000 live births (Balakrishnan, 2021; Den Hollander et al., 2025).

IEMs result from genetic variations that impair the function of proteins essential for cellular metabolism, leading to either a deficiency in energy production or the accumulation of pathological macromolecules (Mordaunt et al., 2020; Barbosa-Gouveia et al., 2021). Substrates that are normally metabolized may accumulate, causing disease-specific clinical manifestations (Mordaunt et al., 2020; Barbosa-Gouveia et al., 2021). The clinical presentation of IEMs varies widely, even among individuals with the same disorder. Some conditions display distinctive features that facilitate diagnosis, while others manifest with nonspecific or subtle symptoms and are often identified incidentally through laboratory testing (Mordaunt et al., 2020; Balakrishnan, 2021; Barbosa-Gouveia et al., 2021). A high index of clinical suspicion is essential for accurate diagnosis, as many IEMs are amenable to treatment, and early detection can significantly reduce mortality (Mordaunt et al., 2020; Balakrishnan, 2021).

Recent advances in bioinformatics have facilitated the integration of genomic databases with functional genomics resources, providing deeper insights into disease mechanisms and enabling the identification of factors contributing to mortality in IEM (Mordaunt et al., 2020; Barbosa-Gouveia et al., 2021; Zandl-Lang, 2024). In this study, we combined population-level bioinformatics analyses with descriptive clinical and cytomorphological observations from twin patients with suspected IEM. The clinical cases were included to provide phenotypic context for the bioinformatics findings; however, patient-specific genetic or molecular evidence linking APOE to the observed phenotype was unavailable. Therefore, the aim of this study is to identify distinctive molecular and phenotypic signatures of IEM that may serve as reliable indicators for accurate diagnosis and treatment planning.

2. Methods

2.1. Study design

The analytical workflow in this study is shown in Figure 1. Genes linked to IEM were first retrieved from four high-confidence repositories: the NHGRI-EBI GWAS Catalog, the Comparative Toxicogenomics Database (CTD), the Open Targets Platform, and ClinVar. After merging datasets and removing duplicates, 1,288 unique genes were retained. To confirm that this gene set reflected IEM-related biology, disease ontology enrichment was conducted via the Open-XGR platform. Gene sets were then subjected to functional annotation. Subsequent bioinformatics analyses were integrated: (1) cell-type and tissue-specific expression profiles from the Human Cell Landscape, Cellular Component, and Tabula Sapiens; and (2) significantly expressed genes highlighted in the Manhattan plot. Joint interpretation of the bioinformatics outputs identified APOE as a candidate associated with lipid metabolism and macrophage-related reference signatures. The clinical and cytomorphological findings were evaluated as phenotypic context for macrophage activation and lipid accumulation but were not considered evidence of APOE causality or of a defined M2 polarization state in the patients.

FIGURE 1.

Flowchart illustrating a bioinformatics workflow starting with data extraction from four gene databases, identifying 1,288 genes, followed by analyses including Open-XGr, functional annotation, and enrichment analysis. The workflow leads to bioinformatics and genetics analysis revealing abnormal macrophage activation, focusing on ApoE, bone marrow, and cerebrospinal fluid, supported by icons, charts, and network diagrams.

Open in a new tab

Comprehensive workflow of the bioinformatics and genetic analysis for inborn errors of metabolism (IEM). Genes associated with IEM were extracted from multiple databases resulting in a total of 1,288 genes. Disease ontology enrichment analysis using the Open-XGR platform confirmed that these genes were highly enriched in IEM-related pathways. Functional annotation and enrichment analyses were then performed using cell-related databases such as the Human Cell Landscape, Cellular Component, and Tabula Sapiens. Integration of bioinformatics findings and descriptive clinical observations prioritized APOE as a candidate gene associated with lipid metabolism and macrophage-related signatures in IEM.

2.2. Disease ontology enrichment

Disease ontology enrichment was conducted using Open-XGR (Bao et al., 2023; accessed 20 May 2025) to evaluate the biological validity of the integrated gene list. The analysis used curated disease ontology annotations and hypergeometric testing with multiple-testing correction. Overrepresentation of inherited metabolic disorder terms and related metabolic phenotypes supported the relevance of the gene set for subsequent functional scoring and mechanistic evaluation.

2.3. Data extraction of IEM-Associated genes

An extensive data-mining approach was undertaken to systematically identify and consolidate genes implicated in the pathogenesis of IEMs. Four well-established and high-confidence bioinformatics repositories, each recognized for their comprehensive coverage of genetic and disease-associated information, were accessed on 20 May 2025. These included the GWAS Catalog (https://www.ebi.ac.uk/gwas/home) (Sollis et al., 2023), the Comparative Toxicogenomics Database (CTD; https://ctdbase.org/) (Davis et al., 2023), the Open Targets Platform (https://www.targetvalidation.org/) (Ochoa et al., 2023), and ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) (Landrum et al., 2020). For each database, searches were performed using the standardized keyword “inborn errors of metabolism,” ensuring consistent retrieval across data sources. The GWAS Catalog was utilized to extract genes significantly associated with IEMs based on genome-wide association studies, while the CTD and Open Targets Platform were leveraged to integrate evidence from curated toxicogenomic relationships and target-disease associations, respectively. ClinVar was further employed to obtain clinically validated variants linked to metabolic disorders. To enhance data reliability, a relevance filtering procedure was implemented within the Open Targets Platform, whereby only genes with an association score greater than 0.3 were retained, reflecting moderate to strong confidence in disease relevance (Adikusuma et al., 2023; Adikusuma et al., 2025). After extraction, all gene lists from the four databases were merged, and duplicate entries were removed to generate a non-redundant, high-confidence gene compendium comprising 1,288 genes. This curated dataset serves as a foundational and integrative resource for subsequent analyses.

2.4. Functional annotation scoring of IEM genes

To prioritize genes with potential biological relevance to IEM, a multi-domain functional annotation and scoring framework was applied. Seven complementary annotation domains were integrated to capture functional, cellular, phenotypic, and disease-related evidence: (1) Gene Ontology Biological Process, representing the biological programs and cellular processes in which a gene participates; (2) Gene Ontology Cellular Component, describing its subcellular or extracellular localization; (3) Gene Ontology Molecular Function, reflecting its biochemical, catalytic, or binding activities; (4) DisGeNET, providing curated and literature-supported evidence of gene–disease associations; (5) the Human Cell Landscape, providing tissue- and cell-type expression context for candidate genes; (6) the Human Phenotype Ontology, linking genes to clinically defined phenotypic abnormalities; and (7) the Kyoto Encyclopedia of Genes and Genomes, providing pathway-level information on gene involvement in metabolic, signaling, and disease-associated networks.

All enrichment analyses were performed using WebGestalt 2024, with statistical significance defined as a false discovery rate-adjusted q value <0.05 (Elizarraras et al., 2024). Each gene received a cumulative annotation score based on the number of criteria fulfilled. For a given annotation domain, a gene was assigned a score of 1 if it was represented in at least one significantly enriched term within that domain and a score of 0 if it was not represented. A cumulative score of ≥5 was selected as an operational threshold to identify genes supported by at least five of the seven domains. Genes meeting this threshold were designated high-priority IEM-associated candidate genes. This cutoff was intended for exploratory candidate prioritization and was not derived from a validated statistical classifier or predictive probability. Higher scores indicate broader cross-domain annotation support and should not be interpreted as direct evidence of mechanistic involvement, disease causality, or clinical risk.

2.5. Clinical case assessment

Clinical data from one-year-old twins with clinically suspected inborn errors of metabolism (IEM) were reviewed between July and October 2025. Comprehensive molecular characterization, including identification of causative genetic variant(s), was not available at the time of analysis. Likewise, definitive biochemical characterization of the underlying metabolic pathway defect could not be established. Patient-specific APOE genotyping, targeted sequencing, whole-exome sequencing, APOE transcript or protein measurements, and APOE immunostaining were not performed. Therefore, the clinical assessment was descriptive and was not designed to establish an APOE-associated diagnosis or a direct relationship between APOE and the observed macrophage phenotype. Laboratory profiles included complete blood counts, liver function tests, lipid panels, and serum biochemistry. Peripheral blood hematology profiles were analyzed using the Sysmex XN-1000 analyzer, while serum biochemical parameters were measured using the Cobas Pro 1 system. Cytospin cerebrospinal fluid (CSF) analysis was performed to identify monocyte infiltration within the central nervous system. Bone marrow analysis was conducted by a hematology specialist. A differential count was performed on 500 cells, and inter-observer agreement was assessed using Cohen’s kappa coefficient between two independent experts. Available metabolic characterization was limited to routine biochemical and lipid profiles. Disease-specific enzyme assays, comprehensive metabolomic or lipidomic profiling, and confirmatory biochemical testing for a defined IEM subtype were not available.

2.6. Statistical analysis

All gene extractions were performed using the NHGRI-EBI GWAS Catalog (https://www.ebi.ac.uk/gwas/home), the Comparative Toxicogenomics Database (CTD; https://ctdbase.org/), the Open Targets Platform (https://www.targetvalidation.org/), and ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/). The Manhattan plot was generated using RStudio version 2024.09.0 + 375 (RStudio, Boston, MA, USA). Functional annotation was performed using the WEB-based GEne SeT AnaLysis Toolkit (WebGestalt; https://www.webgestalt.org), covering seven different annotation categories. Enrichment analyses were conducted using two databases, WebGestalt and Open-XGR (http://www.openxgr.com). Gene network analysis was carried out through Enrichr (https://maayanlab.cloud/enrichr-kg). All database searches and bioinformatics analyses were performed on 20 May 2025.

3. Results

3.1. Disease ontology enrichment analysis of IEM-Associated genes

A total of 1,288 candidate genes were identified through extraction from the GWAS Catalog, CTD, Open Targets Platform, and ClinVar (Supplementary Table 1). To confirm that this integrated dataset was truly related to inborn errors of metabolism (IEMs), disease ontology enrichment analysis was performed using Open-XGR (Figure 2). This step was necessary because gene lists aggregated from heterogeneous sources may include genes associated with unrelated conditions.

FIGURE 2.

Bubble scatterplot showing the relationship between Z-score and log base ten of FDR for gene sets linked to metabolic disorders, with bubble size representing gene count. Five labeled disease ontology terms—'inherited metabolic disorder', 'disease of metabolism', 'lysosomal storage disease', 'carbohydrate metabolic disorder', and 'amino acid metabolic disorder'—display highest Z-scores and significant FDR values. Accompanying table lists the Term IDs, disease names, Z-scores, odds ratios with confidence intervals, and FDR values for these terms, highlighting strong statistical associations.

Open in a new tab

Disease ontology enrichment analysis of genes associated with inborn errors of metabolism (IEMs). Bubble plot showing the top enriched disease ontology terms identified by Open-XGR from 1,288 genes extracted from four databases. Bubble size indicates the number of genes per term. The strongest enrichment was observed for inherited metabolic disorder, confirming that the integrated gene set is highly associated with IEM.

The analysis demonstrated strong enrichment for multiple categories of metabolic disorders. The highest enrichment was seen for inherited metabolic disorder, with a Z-score of 18.6, an odds ratio (OR) of 4.45, and an FDR of 2.20 × 10−58. Similar enrichment was observed for disease of metabolism terms (Z-score 17.9; OR 4.1; FDR 2.00 × 10−56) and specific subtypes, including lysosomal storage disease, carbohydrate metabolic disorder, and amino acid metabolic disorder. These results confirm that the integrated gene set is strongly aligned with IEM-related pathways.

3.2. Functional annotation prioritization reveals key genes implicated in IEM

A multi-parametric functional annotation and scoring framework was applied to prioritize genes with potential biological relevance to IEM (Supplementary Table 2; Figure 3). Seven annotation categories were incorporated: (1) essential biological processes; (2) subcellular localization; (3) molecular function; (4) DisGeNET disease associations; (5) Human Cell Landscape expression specificity; (6) Human Phenotype Ontology (HPO) annotations; and (7) KEGG pathway memberships. Annotation analyses were conducted using WebGestalt 2024. Each gene received a cumulative score based on the number of fulfilled criteria. Genes with composite scores ≥5 were designated as high-priority IEM-associated candidate genes, reflecting strong functional convergence and mechanistic relevance.

FIGURE 3.

Panel A displays a heatmap listing gene symbols in rows and functional categories in columns, with blue squares indicating gene-category associations and total counts on the right. Panel B shows a bar chart with counts labeled above each bar, indicating the distribution of gene-category associations from zero to seven. Panel C presents a pie chart dividing gene associations among KEGG, Biological Process, Cellular Component, Molecular Function, Disease DisGeNet, Human cell landscape, and Human phenotype ontology, with corresponding numbers labeled for each segment.

Open in a new tab

Multi-parametric functional annotation and scoring of 1288 IEM-associated genes (A) Table showing the integration of seven annotation features, including biological process, cellular component, molecular function, DisGeNET, Human Cell Landscape, Human Phenotype Ontology, and KEGG pathway memberships. Genes with higher cumulative scores indicate greater biological relevance (B) Distribution of genes by total annotation score, with 58 genes achieving scores ≥5 (C) Pie chart summarizing the number of enriched terms across the seven annotation categories. APOE and APOA2 obtained the highest composite scores, indicating representation across all seven annotation domains.

The integration results identified APOE, followed by APOA2, as the top-scoring genes, each fulfilling all seven annotation categories (Figure 3A). This prioritizes APOE as a pleiotropic candidate represented across multiple metabolic and regulatory annotation domains relevant to IEM. Score distribution analysis showed a right-skewed pattern (Figure 3B). Most genes (n = 562) scored <1, indicating limited annotation support, while only 58 genes achieved scores ≥5 and were designated high-priority IEM-associated candidate genes. This stratification highlights a focused subset of biologically relevant candidate genes. Across the seven annotation domains, KEGG pathways and HPO contained the highest number of enriched terms (n = 60), followed by biological processes (n = 57) and Human Cell Landscape (n = 45) (Figure 3C). The broad functional coverage emphasizes the multifaceted involvement of IEM-related genes across molecular, cellular, and phenotypic levels.

3.3. Cell-type enrichment analysis of the 1,288 IEM-Associated genes

Several studies have discussed the relationship between macrophages and inborn errors of metabolism (Escolà-Gil et al., 2021; Ricci et al., 2024; Ferreira et al., 2025). For that reason, we want to determine which cell types these 1,288 genes are involved in forming. This is to investigate what defects occur in cells in inborn errors of metabolism. It may potentially be useful as a diagnostic marker for IEM.

Our cell-type enrichment analysis indicates that the 1,288 IEM-associated genes are predominantly linked to macrophage-related signatures, including reference datasets annotated as M2 macrophages (Figure 4). However, these enrichment results should not be interpreted as direct evidence of macrophage polarization states in patient samples. Since M2 macrophages rely on oxidative phosphorylation and lipid metabolic pathways, defects in these processes may impair macrophage polarization and function in IEM (Remmerie and Scott, 2018; van Eijk and Aerts, 2021; Vassiliou and Farias-Pereira, 2023; Zandl-Lang, 2024; Ferreira et al., 2025). This finding aligns with recent evidence showing that metabolic gene abnormalities can disrupt immune cell programming (van Eijk and Aerts, 2021; Zandl-Lang, 2024; Ferreira et al., 2025; Mohit et al., 2025). Therefore, macrophage-related pathways may be relevant in a subset of IEMs characterized by lipid accumulation or immune dysregulation and could serve as a promising basis for the development of diagnostic marker. These findings indicate enrichment in macrophage-related reference signatures, including datasets annotated as M2 macrophages, suggesting that metabolic dysregulation may influence immune cell states that are also critical in host responses to intracellular pathogens.

FIGURE 4.

Bubble chart visualizing cell type enrichment with log2 of enrichment ratio on the x-axis and negative log10 of FDR on the y-axis. Larger, darker bubbles such as Macrophage and M2 Macrophage show higher enrichment and significance. Other labeled cell types include Hepatocyte/Endodermal, Fasciculata, Adrenal gland inflammatory, Epithelial, Dendritic, Monocyte, Astrocyte, and Proximal tubule progenitor. A purple color gradient bar at the top left represents bubble size, corresponding to a numerical scale from 20 to 140.

Open in a new tab

Cell-type enrichment analysis of the 1,288 IEM-associated genes. The analysis identified significant overlap between the IEM-associated gene set and macrophage-related reference expression signatures, including a dataset annotated as M2 macrophages. These findings do not establish macrophage abundance, functional activation, or polarization in patient samples. This figure was generated using WebGestalt (https://www.webgestalt.org).

3.4. APOE is prioritized at the intersection of macrophage-related signatures and protein–lipid complexes

To further characterize the cellular features associated with IEM genes, we examined cellular component enrichment. The most significant enrichment was observed for the protein–lipid complex category (Figure 5A), consistent with the known involvement of lipid and protein metabolism in IEMs (Mordaunt et al., 2020; Balakrishnan, 2021; Zandl-Lang, 2024; Den Hollander et al., 2025).

FIGURE 5.

Panel A shows a horizontal bar chart displaying enrichment ratios for various cellular components, with the protein-lipid complex having the highest enrichment. Panel B presents a Venn diagram of gene overlap among macrophages, macrophage M2, and protein-lipid complexes, highlighting APOE, APOC1, and MSR1 as shared genes. Panel C illustrates a network diagram connecting vasculature-macrophage genes to related genes and cell types, with APOE emphasized. At the bottom, a scale graphic indicates 33 M1 macrophage and 35 M2 macrophage related genes, suggesting comparative analysis.

Open in a new tab

Cellular component enrichment and macrophage subtype-associated gene overlap in IEMs (A) Cellular component enrichment analysis of the 1,288 IEM-associated genes showing highest significant enrichment in the protein–lipid complex (B) Venn diagram showing the overlap between genes associated with macrophages, M2 macrophages, and the protein–lipid complex. A total of 35 genes correspond to M2 macrophage differentiation and 33 to M1 macrophage function. The larger number of genes involved in M2 macrophage reference signatures suggests that M2-associated transcriptional signatures were slightly more represented than M1-associated signatures within the enrichment dataset. Three genes (APOE, APOC1, and MSR1) are shared across macrophage subsets and the protein–lipid complex, indicating an overlap between macrophage-associated reference signatures and protein–lipid complex annotations (C) Cell-type enrichment validation using the Tabula Sapiens single-cell atlas. Of the 1,288 genes queried, only 12 genes were specifically associated with the vasculature–macrophage population. APOE appears as the only consistently overlapping gene across macrophage subsets and protein–lipid complex categories, supporting its potential relevance to IEM-related macrophage biology.

Next, we assessed overlapping genes associated with macrophage and M2 macrophage reference signatures. Thirty-five genes were linked to M2 macrophage differentiation, and 33 to M1 macrophage differentiation. The enrichment of genes associated with M2 macrophage reference signatures suggests that lipid-handling macrophage programs may be relevant to IEM pathophysiology. When these were intersected with protein–lipid complex genes, APOE, APOC1, and MSR1 emerged as shared genes (Supplementary Table 3, Figure 5B).

To assess cross-resource consistency, we performed an independent cell-type enrichment analysis using the Tabula Sapiens single-cell atlas. Of the 1,288 genes, only 12 mapped specifically to vasculature-associated macrophages (Figure 5C). APOE was the only gene consistently overlapping across macrophages, M2 macrophages, protein–lipid complex and vasculature-macrophages. This cross-analysis overlap further supports the prioritization of APOE as a candidate associated with macrophage-related and lipid-handling pathways. Consistent with this, APOE also received the highest functional annotation score (Figure 3A).

To further assess APOE’s genetic significance, we examined IEM-related GWAS summary statistics (Supplementary Table 4). Chromosomal mapping revealed two notable loci: ZPR1 (chromosome 11) and APOE (chromosome 19) (Figure 6). Although ZPR1 is implicated in a rare developmental disorder (Ito et al., 2018; Quintana et al., 2025), it received no annotation score in our analysis and lacks established mechanistic links to IEM. In contrast, APOE demonstrated consistent relevance across all analytical layers in this study, providing an additional, independent layer of association supporting its prioritization for further investigation.

FIGURE 6.

Manhattan plot showing the -log10 of p-values for genetic variants across chromosomes, with significant associations labeled, and a strong signal at chromosome 19 for APOE highlighted by a dashed arrow.

Open in a new tab

Manhattan plot of genome-wide association analysis for IEM. The horizontal dashed line indicates the genome-wide significance threshold (P < 5 × 10−8). Variants surpassing this threshold are considered to have significant associations with IEM. Notably, two genomic loci demonstrated the strongest signals: ZPR1 located on chromosome 11 and APOE located on chromosome 19. However, only APOE demonstrated consistent relevance across all analytical layers in this study, providing an additional, independent layer of association supporting its prioritization for further investigation.

Based on the multiple analytical approaches performed in this study, APOE emerged as the highest-priority candidate gene associated with protein–lipid metabolism pathways and macrophage-related signatures that may contribute to macrophage lipid-handling dysfunction. The importance of APOE is further supported by its high level of significance in the Manhattan plot derived from the GWAS dataset. This finding is consistent with previous studies that have similarly reported APOE’s involvement in key IEM-related processes. These findings nominate APOE for evaluation in future studies examining its potential utility as a molecular or pathway-level marker in defined IEM subgroups. Given the central role of macrophages as host cells for intracellular pathogens, the identification of APOE as a candidate gene associated with lipid-associated macrophage pathways may have broader implications for host–pathogen interactions. It should be noted that the GWAS-derived associations and functional annotations were obtained from public datasets and were not generated from the twin patients. Accordingly, the prioritization of APOE cannot be interpreted as evidence that either patient carried an APOE variant or had altered APOE expression.

3.5. Twin case studies

The twin cases are presented as descriptive clinical observations and were not genetically linked to APOE. Although their clinical and cytomorphological findings were compatible with macrophage activation and abnormal intracellular substrate accumulation, the underlying IEM subtype remained unconfirmed. In this case, a pair of one-year-old twins presented to the hospital with similar clinical signs and symptoms. The clinical presentation was highly suggestive of an underlying IEM; however, definitive molecular classification was unavailable. At 11 days of age, both infants exhibited elevated total and direct bilirubin levels that did not normalize with phototherapy. Additionally, AST and ALT levels were markedly elevated above the reference ranges during this neonatal period.

By 1 year of age, both twin patients presented in a deteriorating condition requiring intensive care. At this time, the biochemical abnormalities persisted, with AST and ALT levels remaining several-fold higher than normal, accompanied by reduced albumin levels. Both children also continued to exhibit elevated bilirubin concentrations. Bilirubin is a strong clinical indicator of abnormalities associated with inborn errors of metabolism (Ozen et al., 2014; Heubi et al., 2018), as physiologic hyperbilirubinemia typically resolves within 2–3 weeks after birth; however, in these twins, bilirubin levels remained persistently elevated even at 1 year of age. This biochemical pattern reflects ongoing hepatocellular injury and provides strong evidence supporting an underlying inborn metabolic disorder.

In IEM, immune dysregulation occurs within hepatic macrophages, leading to the activation of inflammatory signaling pathways (Ozen et al., 2014; Tarasenko and McGuire, 2017; Roohani and Tacke, 2021). This process disrupts the excretion of conjugated (direct) bilirubin from the liver into the gastrointestinal tract due to inflammation and fibrosis along the biliary excretory pathways caused by aberrant macrophage activity (Ozen et al., 2014; Li et al., 2017; Chen and Zhang, 2023; Horn and Tacke, 2024). As a result, secondary cholestasis develops, which is reflected by elevated GGT levels in both infants. The co-occurrence of cholestasis and hyperbilirubinemia is recognized as one of the hallmarks of secondary hemophagocytic lymphohistiocytosis (HLH) or macrophage activation syndrome (MAS) (Ozen et al., 2014).

On day 10 of the intensive care follow-up period, the condition of one of the twins (Twin A) deteriorated, accompanied by a marked rise in the absolute monocyte count, reaching 4,770 cells/μL. Consequently, on day 12, a bone marrow aspiration was performed in Twin B to establish a definitive diagnosis for suspected MAS (Figure 7). The evidence of MAS was supported by the presence of recurrent, fluctuating fever over several months, elevated serum triglyceride levels, and clear morphology of excessive macrophage activity in the bone marrow (Figure 8). This heightened macrophage activation was further reflected by an increased absolute monocyte count in the peripheral blood of both twins (Figure 7).

FIGURE 7.

Two line graphs compare monocyte count changes over time in twins labeled A and B. Twin A shows a sharp increase in monocyte count near day 12, marked with a red arrow and labeled "Death." Twin B displays varied monocyte counts with blue arrows indicating bone marrow aspiration, etoposide treatment, and CSF analysis, ending with a decline at day 45, marked "Death" with a red arrow.

Open in a new tab

Time-Course of Monocyte Counts and Clinical Interventions in Twin Infants with IEM. The graphs illustrate the temporal changes in absolute monocyte counts (×103/μL) during the intensive care follow-up period in Twin A (left) and Twin B (right). Twin A demonstrated a progressive rise in monocyte count, peaking at 4.77 × 103/μL on day 10, followed by rapid clinical deterioration and death before treatment could be initiated. Twin B underwent bone marrow aspiration on day 12, confirming macrophage activation syndrome. Etoposide therapy was initiated shortly thereafter; although the monocyte count increased to 2.58 × 103/μL on day 24, it subsequently showed a gradual decline. Subsequent cerebrospinal fluid (CSF) analysis revealed monocyte infiltration, preceding neurological deterioration and eventual death on day 44. Arrows indicate key clinical events: bone marrow aspiration, initiation of etoposide therapy, CSF analysis, and time of death.

FIGURE 8.

Four microscopic images of stained tissue samples display clusters of cells with prominent dark purple nuclei and varying granularity, surrounded by lighter-stained cellular structures; scale bars indicate 15 micrometers for reference.

Open in a new tab

Bone marrow aspirate findings demonstrating excessive lipid-laden macrophages. Numerous enlarged macrophages with abundant, finely vacuolated cytoplasm are present, indicating intracellular lipid accumulation. These features are consistent with lipid-laden macrophages exhibiting prominent cytoplasmic vacuolization suggestive of intracellular lipid accumulation. All images were obtained from one of the twins, as a bone marrow procedure could not be performed in the other due to clinical deterioration.

Unfortunately, Twin A did not survive. The infant died shortly after the monocyte count peaked at 4,770 cells/μL and before treatment could be initiated, as the definitive diagnosis had not yet been established. Once the bone marrow aspirate confirmed macrophage activation syndrome (MAS), Twin B was treated with corticosteroids and etoposide. Although the monocyte count initially increased to 2,580 cells/μL on day 24 of follow-up, it subsequently declined with ongoing therapy. Twin B’s condition later deteriorated due to progressive neurological involvement, characterized by reduced consciousness and recurrent seizures. Cerebrospinal fluid (CSF) analysis demonstrated infiltration of activated monocytes (Figure 9). The patient ultimately died as a result of persistent, refractory seizures, representing a severe neurological complication secondary to macrophage infiltration into the CSF.

FIGURE 9.

Four microscopic images show mononuclear cells with granular, purple-stained cytoplasm and irregular nuclei, each annotated with a 15 micrometer scale bar. Below, a table summarizes cerebrospinal fluid analysis: colorless, clear appearance, cell count of nine cells per microliter, zero percent polymorphonuclear cells, one hundred percent mononuclear cells, total protein of 0.001 grams per liter. Differential cytology identifies one hundred percent activated monocytes. Interpretation states mononuclear predominance with activated monocytes.

Open in a new tab

Cytological analysis of cerebrospinal fluid demonstrating infiltration of activated, lipid-laden monocytes/macrophages in Twin B. Cytospin preparation of cerebrospinal fluid (CSF) from Twin B showing abundant activated monocytes/macrophages with vacuolated cytoplasm consistent with lipid-laden macrophage morphology. These findings indicate mononuclear predominance with infiltration by activated, lipid-laden macrophages, representing central nervous system involvement secondary to macrophage activation syndrome (MAS) and correlating with the patient’s progressive neurological deterioration and refractory seizures.

The increased presence of macrophages in the bone marrow aspirate indicates an intensified effort by the body to eliminate accumulating abnormal metabolic substrates. The enlarged vacuolated macrophages may reflect altered intracellular substrate handling; however, their lipid content, functional state, and polarization phenotype were not directly characterized (Remmerie and Scott, 2018; Wang et al., 2021; Schilperoort et al., 2023; Vassiliou and Farias-Pereira, 2023; Ferreira et al., 2025). Under physiological conditions, excessive intracellular lipid accumulation activates transcription factors such as PPARγ and LXR agonists, driving macrophages toward an adaptive M2 phenotype (Wang et al., 2021; Vassiliou and Farias-Pereira, 2023). However, in inborn errors of metabolism (IEM), excessive buildup of lipid substrates leads to macrophage lysosomal dysfunction, impairing lipid efflux and disrupting normal resolution pathways (van Eijk and Aerts, 2021; Koncz et al., 2023; Schilperoort et al., 2023; Yang et al., 2023). Consequently, macrophages may acquire a dysfunctional lipid-laden phenotype, potentially associated with impaired lysosomal processing and altered lipid handling. The presence of enlarged vacuolated macrophages in bone marrow aspirates (Figure 8) is consistent with intracellular lipid accumulation.

In Twin B, etoposide was initiated due to its well-established in vivo ability to suppress the excessively activated macrophages characteristic of HLH or MAS. A recent meta-analysis demonstrated that etoposide is effective in reducing cytokine storm activity and limiting the proliferation of hyperactivated inflammatory cells, particularly when administered in combination with corticosteroids (Gao et al., 2025). Despite its systemic efficacy, etoposide has poor penetration across the blood–brain barrier (Wei et al., 2021). Despite systemic treatment, subsequent CSF cytology demonstrated activated monocytes/macrophages. Limited central nervous system penetration of etoposide may have contributed to incomplete control of CNS inflammation; however, this possibility was not directly evaluated. This central nervous system involvement ultimately led to severe neurological complications, resulting in the infant’s death.

4. Discussion

Inborn Errors of Metabolism (IEM) arise from genetic variations that disrupt the function of proteins required for cellular metabolism, leading to impaired energy production or the accumulation of pathological macromolecules (Mordaunt et al., 2020; Balakrishnan, 2021). Most IEMs are caused by defects in a single gene encoding a specific enzyme within a metabolic pathway. When metabolism is blocked due to a dysfunctional enzyme, the system attempts to reach a new adaptive steady state that differs from normal physiological homeostasis. Tracer-lipidomics has demonstrated specific disturbances in lipid biosynthesis and remodeling pathways across various IEMs, including the accumulation of abnormal lipids and altered lipid flux (Zandl-Lang, 2024).

Although the enrichment analysis identified macrophage and M2 macrophage-associated signatures, these findings should not be interpreted as evidence of a fixed M2 polarization state. Macrophage activation exists along a dynamic continuum rather than a strict M1/M2 dichotomy. In the present study, the term “M2-like” refers only to bioinformatic enrichment of lipid-handling macrophage signatures and not to experimentally confirmed alternative macrophage polarization. The clinical syndrome observed in the twins is consistent with macrophage activation syndrome, a hyper-inflammatory macrophage-driven condition. The lipid-laden morphology in bone marrow and CSF is therefore interpreted as evidence of macrophage lipid-handling dysfunction rather than proof of anti-inflammatory M2 macrophage predominance.

Based on the findings of this study, we propose a hypothesis-generating framework that may explain how APOE-associated lipid dysregulation could contribute to macrophage dysfunction and disease progression in IEM (Figure 10). Although the proposed model integrates findings from the current study with published literature, several downstream processes depicted in Figure 10, including impaired autophagy, ferroptosis, fibrosis, thrombosis, and intracellular pathogen persistence, were not directly evaluated. Therefore, these mechanisms should be regarded as speculative and hypothesis-generating rather than experimentally demonstrated within the present study. APOE is a major apolipoprotein that regulates lipid metabolism. It is expressed in multiple tissues, including the liver, brain, spleen, kidneys, gonads, adrenal glands, and macrophages. The APOE gene is located on chromosome 19q13.2 and encodes a 317-amino-acid apolipoprotein E precursor (Khalil et al., 2021). Under normal physiological conditions, APOE is incorporated into VLDL and chylomicrons to facilitate lipid clearance from the plasma and acts as a ligand for LDL receptors (LDLR) on hepatocytes. Macrophages also secrete APOE and take up APOE-containing lipoproteins to maintain lipid homeostasis in peripheral tissues. In individuals with APOE mutations or variants, the affinity of APOE for the LDL receptor is reduced, leading to impaired clearance of chylomicron and VLDL remnants by the liver. As a result, plasma lipoprotein levels—particularly total cholesterol and triglycerides—become elevated. Excessive lipid levels upregulate macrophage-activating surface markers (CD11b, CCR2, CD63) and stimulate macrophages to take up large quantities of lipids. Lipid-laden macrophages subsequently transform into foam cells, which accumulate within the circulation and tissues (Ferreira et al., 2025). This mechanism also occurs in several IEM conditions, such as APOE gene mutations, genetic defects affecting macrophage lysosomes, or disruptions in intracellular cholesterol transport proteins, all of which lead to the buildup of abnormal lipid substrates within macrophages (Remmerie and Scott, 2018; van Eijk and Aerts, 2021). Cholesterol accumulation in macrophages activates transcription factors including PPARγ, LXRs, and RXRs, which then increase the expression of ABCA1 and ABCG1, key regulators of cholesterol efflux and scavenger receptors. In addition to promoting cholesterol efflux transporters, these transcription factors also induce the expression of apolipoproteins such as APOE and APOC, which function as receptors for cholesterol (Remmerie and Scott, 2018; van Eijk and Aerts, 2021). However, during chronic lipid accumulation, defects in lipid efflux emerge and APOE becomes abnormally enriched on remnant particles, resulting in dysfunctional lipid remodelling (Remmerie and Scott, 2018).

FIGURE 10.

Diagram illustrates the progression from ApoE-related lipid pathway alteration in a person to chronic lipid overload, formation of lipid-laden macrophages, lysosomal stress, dysfunctional immunosuppressive phenotype, resulting fibrosis and thrombosis, and multi-organ dysfunction affecting the brain, heart, liver, and lungs.

Open in a new tab

Hypothesis-generating model of the potential relationship among APOE-associated lipid pathways, macrophage lipid handling, and disease manifestations in IEM. The model integrates the bioinformatics findings of the present study with mechanisms proposed in previously published literature. APOE was prioritized at the intersection of lipid metabolism and macrophage-related reference signatures, whereas the clinical cases demonstrated macrophage activation and vacuolated macrophage morphology. The proposed relationships among altered APOE-related lipid handling, macrophage dysfunction, lysosomal stress, inflammation, fibrosis, thrombosis, and multi-organ involvement were not directly tested in the twin patients. The model should therefore be interpreted as a framework for future experimental investigation rather than an established causal pathway.

The accumulation of cholesterol and oxidized lipids within macrophages induces ER stress, mitochondrial dysfunction, increased ROS, and impaired mitophagy, thereby altering the cell’s metabolic profile. Although oxidative stress is typically associated with inflammatory activation (M1), in conditions of chronic lipid overload such as IEM, lipid-derived ligands (e.g., oxidized low-density lipoproteins [ox-LDLs]) activate transcriptional receptors such as PPARγ/LXR agonists, causing macrophages to shift toward an M2-like phenotype focused on lipid handling and tissue repair (Remmerie and Scott, 2018; Wang et al., 2021; Vassiliou and Farias-Pereira, 2023). In addition, other studies have reported that high APOE expression suppresses ferroptosis through activation of the PI3K/AKT1 pathway and promotes macrophage polarization toward an M2 phenotype in PTC tumors (Li et al., 2025). Although macrophages shift toward an M2 phenotype, defective lipid efflux causes these cells to become lipid-laden macrophages. This leads to the emergence of pathogenic M2 macrophages: cells that continue to express M2-associated genes but lose essential functional capacities. Lipid-laden macrophages display a unique metabolic profile distinct from classical M1 and M2 macrophages. During efferocytosis, these macrophages accumulate lipids within lysosomes, which are difficult to degrade, leading to lysosomal stress. To compensate, cells activate the MiT–TFE pathway to induce lysosomal biogenesis and autophagy. However, the excessive lipid load causes autophagy failure. Accumulation of p62 occurs as a consequence of impaired autophagy, further contributing to immune suppression. Meanwhile, PPARγ/LXR signaling, normally activated during efferocytosis, remains active but becomes skewed toward suppressing inflammation (inhibiting M1 phenotype) rather than restoring metabolic balance. This results in an immunosuppressive yet dysfunctional immune phenotype (Van Eijk and Aerts, 2021; Schilperoort et al., 2023). Defective efferocytosis leads to the accumulation of cellular debris, which is recognized as DAMPs, triggering chronic inflammation, fibrosis, and tissue necrosis. M2-like lipid-laden macrophages continue to express tissue-repair mediators; however, under abnormal metabolic conditions, their effects become maladaptive, promoting pro-fibrotic and pro-thrombotic activity. TGF-β increases due to PPARγ/LXR activation and false resolution signals from partial efferocytosis. Elevated TGF-β induces fibroblast activation, collagen synthesis, and progressive fibrosis in organs such as the liver, spleen, kidneys, and heart (Yang et al., 2023). Additionally, interactions between M2-like lipid-laden macrophages and fibrin or coagulation factors such as thrombin via PAR-1 activation enhance the expression of TGF-β, VEGF, and PAI-1, sustaining a pro-fibrotic and pro-thrombotic microenvironment. Consequently, coagulation is no longer protective but becomes a driver of chronic inflammation and tissue thrombosis (Koncz et al., 2023). The combination of these failures may contribute to the development of multi-organ dysfunction, thereby increasing mortality risk in IEM.

This study revealed that APOE is a principal gene exhibiting significant differential expression and strong associations with pathways related to lipid handling, immune modulation, and cellular stress response (Khalil et al., 2021; Van Eijk and Aerts, 2021; Theobald et al., 2024; Zandl-Lang, 2024; Ferreira et al., 2025). In the setting of IEM, chronic lipid accumulation gives rise to a maladaptive phenotype known as pathogenic M2 lipid-laden macrophages (Remmerie and Scott, 2018; van Eijk and Aerts, 2021; Wang et al., 2021; Vassiliou and Farias-Pereira, 2023; Ferreira et al., 2025). These cells maintain the expression of reparative mediators but lose critical homeostatic functions, including efferocytosis, resulting in an immunosuppressive and dysfunctional phenotype (Van Eijk and Aerts, 2021; Schilperoort et al., 2023) This maladaptive state triggers profibrotic and prothrombotic signaling pathways, contributing to persistent inflammation, fibrosis, and coagulation abnormalities (Koncz et al., 2023; Yang et al., 2023). The sustained presence and activity of these macrophages promote chronic tissue injury and multi-organ dysfunction, ultimately increasing mortality risk in patients with IEM (van Eijk and Aerts, 2021; Koncz et al., 2023; Schilperoort et al., 2023; Yang et al., 2023). These findings were further supported by morphological examination of bone marrow smears and cerebrospinal fluid, which demonstrated an overexpression of the lipid-laden macrophage phenotype. The convergence of bioinformatics prioritization and cytomorphological observations generates the hypothesis that APOE-related lipid pathways may be relevant to macrophage lipid-handling abnormalities in selected IEM contexts. However, the two evidence streams remain indirect: the bioinformatics analyses were not patient-specific, and the clinical cases were not evaluated for APOE genotype, expression, or protein abundance. The present findings therefore do not establish that APOE caused the observed macrophage phenotype.

Beyond its role in metabolic disease, the findings of this study may have broader implications for host immune responses in conditions involving intracellular pathogens. Macrophages serve as primary host cells for a wide range of intracellular bacteria, and their functional state critically determines the balance between pathogen clearance and persistence (Kaufmann, 2001; Flannagan et al., 2009). The lipid-laden, functionally impaired macrophage phenotype identified in this study may represent a metabolically reprogrammed state with reduced antimicrobial capacity, including impaired phagolysosomal function, altered autophagy, and diminished oxidative responses (Kaufmann, 2001; Russell et al., 2010). Such alterations could create a permissive intracellular environment that favors pathogen survival and persistence.

In this context, APOE-associated lipid dysregulation may contribute to a host microenvironment that is less effective in controlling intracellular infections. Lipid accumulation within macrophages can be exploited by intracellular pathogens as a nutrient source or as a strategy to evade immune responses (Russell et al., 2010; Cumming et al., 2018). For example, lipid-rich foamy macrophages are a well-recognized niche that supports the persistence of Mycobacterium tuberculosis within host tissues (Peyron et al., 2008; Russell et al., 2010). These observations suggest that metabolic conditions such as IEM may indirectly influence infection susceptibility and treatment outcomes through host immune reprogramming. Therefore, the APOE–macrophage axis identified in this study should be viewed as a potential candidate pathway for future investigation into host-directed therapeutic strategies aimed at restoring macrophage function and improving intracellular pathogen clearance (Zumla et al., 2016).

This study has several limitations. First, its observational and hypothesis-generating design does not permit causal inference regarding the relationships among APOE, lipid dysregulation, macrophage activation, and disease progression. The clinical component was also limited to two genetically related twin patients from a single family. These cases cannot be considered fully independent observations and may not represent the substantial genetic and phenotypic heterogeneity of IEM. Accordingly, the clinical findings should not be generalized to the broader IEM population. A further limitation is that the twin cases lacked definitive molecular and biochemical classification. Patient-specific genetic testing, including APOE genotyping or broader sequencing, was not available, and the specific metabolic pathway underlying disease progression could not be established. The observed clinical and cytomorphological phenotypes therefore cannot be directly attributed to APOE-associated mechanisms. The proposed APOE–macrophage relationship should instead be interpreted as a hypothesis derived from integrative bioinformatics analyses and considered alongside descriptive clinical observations, rather than as a disease-specific or experimentally established mechanism.

The pathological findings were primarily based on cytomorphological assessment of bone marrow and cerebrospinal fluid specimens. Although vacuolated, lipid-laden macrophage morphology was observed, macrophage immunophenotyping, lipid-specific characterization, APOE transcript or protein measurement, and APOE immunostaining were not performed. Consequently, the macrophage polarization state, functional activity, and tissue-level involvement of APOE could not be directly determined. No functional perturbation experiments were conducted to examine whether altered APOE expression or activity affects macrophage lipid handling, lysosomal function, autophagy, inflammatory signaling, or cell polarization. Similarly, several downstream processes included in the proposed model, such as intracellular pathogen persistence, ferroptosis, fibrosis, thrombosis, and multi-organ dysfunction, were not directly evaluated. These processes are therefore presented as literature-supported and biologically plausible hypotheses rather than experimentally demonstrated pathways in the present study. Finally, the study did not include an independent clinical validation cohort. Although the bioinformatics findings were examined across multiple annotation resources and single-cell reference datasets, such cross-database convergence does not constitute replication in an independent patient population. Future studies should validate these findings in larger cohorts of unrelated, genetically and biochemically characterized patients and incorporate patient-specific sequencing, comprehensive metabolic profiling, APOE expression and protein analyses, macrophage immunophenotyping, and functional experiments.

5. Conclusion

This integrative analysis prioritizes APOE as a gene associated with lipid metabolism pathways and macrophage-related reference signatures in IEM. The clinical cases independently demonstrated macrophage activation and vacuolated macrophage morphology consistent with intracellular lipid accumulation; however, they did not provide patient-specific evidence linking APOE to these findings or establish an M2 polarization state. The convergence of the bioinformatics and clinical observations supports the hypothesis that altered lipid-handling pathways may be associated with macrophage dysfunction in specific IEM contexts. This relationship remains associative and requires validation in molecularly characterized patient cohorts through macrophage immunophenotyping, APOE expression and protein analyses, and functional perturbation studies. Accordingly, the APOE–macrophage relationship should be regarded as a hypothesis for further investigation rather than an established disease mechanism or therapeutic target.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This bioinformatics analysis was supported by a Doctoral Research Grant from the Faculty of Medicine, Public Health, and Nursing, Universitas Gadjah Mada, as stated in the “Surat Perjanjian Pelaksanaan Penelitian Hibah Dana Masyarakat FK-KMK UGM 2025” under contract number 4019/UN1/KU/PP/PT.01.03/2026.

Footnotes

Edited by: Michal Letek, University of León, Spain

Reviewed by: Jesús Balsinde, CSIC, Spain

Anliang Xia, First Affiliated Hospital of Anhui Medical University, China

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

Ethics statement

The studies involving humans were approved by Medical and Health Research Ethics Committee (MHREC), Faculty of Medicine, Public Health, and Nursing, Universitas Gadjah Mada–Dr. Sardjito General Hospital, Yogyakarta, under ethical approval number KE-FK-1696-EC-2025. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.

Author contributions

RS: Investigation, Conceptualization, Writing – review and editing, Funding acquisition, Writing – original draft, Formal Analysis, Visualization, Methodology, Data curation. RK: Writing – review and editing, Formal Analysis, Visualization, Data curation. NI: Writing – review and editing, Data curation, Methodology, Formal Analysis, Visualization. US: Data curation, Writing – review and editing, Investigation, Formal Analysis. TR: Visualization, Formal Analysis, Data curation, Investigation, Writing – review and editing. EH: Formal Analysis, Validation, Writing – review and editing, Data curation, Investigation. UI: Validation, Formal Analysis, Data curation, Writing – review and editing, Investigation. WW: Formal Analysis, Data curation, Investigation, Writing – review and editing. NH: Formal Analysis, Validation, Data curation, Writing – review and editing, Investigation. PS: Investigation, Validation, Data curation, Writing – review and editing, Formal Analysis. KH: Writing – review and editing, Investigation, Validation, Formal Analysis, Visualization, Data curation. LB: Validation, Data curation, Formal Analysis, Writing – review and editing, Investigation. RP: Investigation, Formal Analysis, Data curation, Validation, Writing – review and editing. NS: Investigation, Writing – review and editing, Data curation, Formal Analysis. MD: Formal Analysis, Investigation, Data curation, Writing – review and editing. EM: Investigation, Visualization, Formal Analysis, Validation, Data curation, Writing – review and editing, Methodology. BA: Data curation, Writing – review and editing, Investigation. LI: Formal Analysis, Validation, Methodology, Writing – review and editing, Data curation, Investigation. C-FL: Investigation, Validation, Supervision, Data curation, Writing – review and editing. C-LC: Validation, Writing – review and editing, Methodology, Data curation, Supervision, Investigation. WA: Validation, Data curation, Formal Analysis, Conceptualization, Supervision, Writing – review and editing, Investigation, Methodology.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI was used exclusively for language editing and improvement of grammar, clarity, and readability.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar.2026.1880997/full#supplementary-material

Supplementaryfile4.xls (395KB, xls)
Supplementaryfile1.xlsx (47.6KB, xlsx)
Supplementaryfile2.xlsx (59.3KB, xlsx)
Supplementaryfile3.xlsx (10.4KB, xlsx)

References

  1. Adikusuma W., Firdayani F., Irham L. M., Darmawi D., Hamidy M. Y., Nopitasari B. L., et al. (2023). Integrated genomic network analysis revealed potential of a druggable target for hemorrhoid treatment. Saudi Pharmaceutical Journal SPJ The Official Publication Saudi Pharm. Soc. 31, 101831. 10.1016/J.JSPS.2023.101831 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Adikusuma W., Firdayani F., Irham L. M., Santri I. N., Phiri Y. V. A., Pitaloka D. A. E., et al. (2025). Unveiling hepatitis B biomarkers through genomic network analysis. Egypt. J. Med. Hum. Genet. 26 (126), 110. 10.1186/S43042-025-00739-Y [DOI] [Google Scholar]
  3. Balakrishnan U. (2021). Inborn errors of metabolism—approach to diagnosis and management in neonates. Indian J. Pediatr. 88, 679–689. 10.1007/s12098-021-03759-9 [DOI] [PubMed] [Google Scholar]
  4. Bao C., Wang S., Jiang L., Fang Z., Zou K., Lin J., et al. (2023). OpenXGR: a web-server update for genomic summary data interpretation. Nucleic Acids Research 51, W387–W396. 10.1093/NAR/GKAD357 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barbosa-Gouveia S., Vázquez-Mosquera M. E., González-Vioque E., Álvarez J. V., Chans R., Laranjeira F., et al. (2021). Utility of gene panels for the diagnosis of inborn errors of metabolism in a metabolic reference center. Genes. 12, 1262. 10.3390/genes12081262 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chen J., Zhang S. (2023). The role of inflammation in cholestatic liver injury. J. Inflamm. Res. 16, 4527–4540. 10.2147/JIR.S430730 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cumming B. M., Addicott K. W., Adamson J. H., Steyn A. J. C. (2018). Mycobacterium tuberculosis induces decelerated bioenergetic metabolism in human macrophages. Elife 7, e39169. 10.7554/eLife.39169 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Davis A. P., Wiegers T. C., Johnson R. J., Sciaky D., Wiegers J., Mattingly C. J. (2023). Comparative Toxicogenomics database (CTD): update 2023. Nucleic Acids Research 51, D1257–D1262. 10.1093/NAR/GKAC833 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. den Hollander B., Hoytema van Konijnenburg E. M. M., Hewitson B., van der Meijden J. C., Balfoort B. M., Winter B., et al. (2025). The metabolic treatabolome and inborn errors of metabolism knowledgebase therapy tool: do not miss the opportunity to treat. J. Inherit. Metabolic Dis. 48, e12835. 10.1002/jimd.12835 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Elizarraras J. M., Liao Y., Shi Z., Zhu Q., Pico A. R., Zhang B. (2024). WebGestalt 2024: faster gene set analysis and new support for metabolomics and multi-omics. Nucleic Acids Research 52, W415–W421. 10.1093/NAR/GKAE456 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Escolà-Gil J. C., Rotllan N., Julve J., Blanco-Vaca F. (2021). Reverse cholesterol transport dysfunction is a feature of familial hypercholesterolemia. Curr. Atheroscler. Rep. 23, 29. 10.1007/s11883-021-00928-1 [DOI] [PubMed] [Google Scholar]
  12. Ferreira E. A., Oud M. M., van der Crabben S. N., Versloot M., Goorden S. M. I., van Karnebeek C. D. M., et al. (2025). Inherited dyslipidemic splenomegaly: a genetic macrophage storage disorder caused by disruptive apolipoprotein E (APOE) variants. Genes. 16, 289. 10.3390/GENES16030289 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Flannagan R. S., Cosío G., Grinstein S. (2009). Antimicrobial mechanisms of phagocytes and bacterial evasion strategies. Nat. Rev. Microbiol. 7, 355–366. 10.1038/nrmicro2128 [DOI] [PubMed] [Google Scholar]
  14. Gao T., Suolitiken D., Yang C., Wu C., He L., Wang Y. (2025). Assessing the effectiveness of etoposide treatment in adult haemophagocytic lymphohistiocytosis: a systematic review and meta-analysis. Clin. Exp. Med. 25 (125), 58. 10.1007/S10238-025-01570-W [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Heubi J. E., Setchell K. D. R., Bove K. E. (2018). Inborn errors of bile acid metabolism. Clin. Liver Dis. 22, 671–687. 10.1016/J.CLD.2018.06.006 [DOI] [PubMed] [Google Scholar]
  16. Horn P., Tacke F. (2024). Metabolic reprogramming in liver fibrosis. Cell Metab. 36, 1439–1455. 10.1016/J.CMET.2024.05.003 [DOI] [PubMed] [Google Scholar]
  17. Ito Y. A., Smith A. C., Kernohan K. D., Pena I. A., Ahmed A., McDonell L. M., et al. (2018). A ZPR1 mutation is associated with a novel syndrome of growth restriction, distinct craniofacial features, alopecia, and hypoplastic kidneys. Clin. Genetics 94, 303–312. 10.1111/CGE.13388 [DOI] [PubMed] [Google Scholar]
  18. Kaufmann S. H. E. (2001). How can immunology contribute to the control of tuberculosis? Nat. Rev. Immunol. 1, 20–30. 10.1038/35095558 [DOI] [PubMed] [Google Scholar]
  19. Khalil Y. A., Rabès J. P., Boileau C., Varret M. (2021). APOE gene variants in primary dyslipidemia. Atherosclerosis 328, 11–22. 10.1016/j.atherosclerosis.2021.05.007 [DOI] [PubMed] [Google Scholar]
  20. Koncz G., Jenei V., Tóth M., Váradi E., Kardos B., Bácsi A., et al. (2023). Damage-mediated macrophage polarization in Sterile inflammation. Front. Immunol. 14, 1169560. 10.3389/fimmu.2023.1169560 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Landrum M. J., Chitipiralla S., Brown G. R., Chen C., Gu B., Hart J., et al. (2020). ClinVar: improvements to accessing data. Nucleic Acids Research 48, D835–D844. 10.1093/NAR/GKZ972 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Li M., Cai S. Y., Boyer J. L. (2017). Mechanisms of bile acid mediated inflammation in the liver. Mol. Aspects Medicine 56, 45–53. 10.1016/J.MAM.2017.06.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li Z., Li M., Sun S., Bin Y., Zuo S., Huo R., et al. (2025). APOE modulates ferroptosis to drive macrophage polarization toward the M2 type and enhance PTC migration and invasion. Immunobiology 230, 152900. 10.1016/J.IMBIO.2025.152900 [DOI] [PubMed] [Google Scholar]
  24. Mohit, Verma S., Laxmi V., Maury J. (2025). Metabolic syndrome and inborn errors of immunity: a plausible connection. Immunol. Lett. 276, 107068. 10.1016/J.IMLET.2025.107068 [DOI] [PubMed] [Google Scholar]
  25. Mordaunt D., Cox D., Fuller M. (2020). Metabolomics to improve the diagnostic efficiency of inborn errors of metabolism. Int. J. Mol. Sci. 21 (4), 1195. 10.3390/ijms21041195 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Ochoa D., Hercules A., Carmona M., Suveges D., Baker J., Malangone C., et al. (2023). The next-generation open targets platform: reimagined, redesigned, rebuilt. Nucleic Acids Research 51, D1353–D1359. 10.1093/NAR/GKAC1046 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ozen S., Dai A., Coskun E., Oztuzcu S., Ergun S., Aktekin E., et al. (2014). Importance of hyperbilurubinemia in differentiation of primary and secondary hemophagocytic lymphohistiocytosis in pediatric cases. Mediterr. J. Hematol. Infect. Dis. 6, e2014067. 10.4084/MJHID.2014.067 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Peyron P., Vaubourgeix J., Poquet Y., Levillain F., Botanch C., Bardou F., et al. (2008). Foamy macrophages from tuberculous patients’ granulomas constitute a nutrient-rich reservoir for M. tuberculosis persistence. PLoS Pathog. 4, e1000204-. 10.1371/journal.ppat.1000204 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Quintana K., Hutchison M., Wong C., Alkhouri R., Hardy L., Clericuzio C., et al. (2025). Congenital alopecia, hypoplastic kidneys, growth restriction, growth hormone resistance, and liver fibrosis: confirmation of a new syndrome caused by biallelic variants in ZPR1. Am. Journal Medical Genetics. Part A 200, 162–170. 10.1002/AJMG.A.64203 [DOI] [PubMed] [Google Scholar]
  30. Remmerie A., Scott C. L. (2018). Macrophages and lipid metabolism. Cell. Immunol. 330, 27–42. 10.1016/J.CELLIMM.2018.01.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ricci S., Sarli W. M., Lodi L., Canessa C., Lippi F., Dini D., et al. (2024). HLH as an additional warning sign of inborn errors of immunity beyond familial-HLH in children: a systematic review. Front. Immunol. 15, 1282804. 10.3389/fimmu.2024.1282804 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Roohani S., Tacke F. (2021). Liver injury and the macrophage issue: molecular and mechanistic facts and their clinical relevance. Int. J. Mol. Sci. 22, 7249. 10.3390/IJMS22147249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Russell D. G., VanderVen B. C., Lee W., Abramovitch R. B., Kim M., Homolka S., et al. (2010). Mycobacterium tuberculosis wears what it eats. Cell Host Microbe 8, 68–76. 10.1016/j.chom.2010.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Schilperoort M., Ngai D., Sukka S. R., Avrampou K., Shi H., Tabas I. (2023). The role of efferocytosis-fueled macrophage metabolism in the resolution of inflammation. Immunol. Reviews 319, 65–80. 10.1111/IMR.13214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sollis E., Mosaku A., Abid A., Buniello A., Cerezo M., Gil L., et al. (2023). The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Research 51, D977–D985. 10.1093/NAR/GKAC1010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tarasenko T. N., McGuire P. J. (2017). The liver is a metabolic and immunologic organ: a reconsideration of metabolic decompensation due to infection in inborn errors of metabolism (IEM). Mol. Genetics Metabolism 121, 283–288. 10.1016/J.YMGME.2017.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Theobald H., Bejarano D. A., Katzmarski N., Haub J., Schulte-Schrepping J., Yu J., et al. (2024). Apolipoprotein E controls Dectin-1-dependent development of monocyte-derived alveolar macrophages upon pulmonary β-glucan-induced inflammatory adaptation. Nat. Immunol. 25 (6), 994–1006. 10.1038/s41590-024-01830-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. van Eijk M., Aerts J. M. F. G. (2021). The unique phenotype of lipid-laden macrophages. Int. J. Mol. Sci. 22, 4039. 10.3390/IJMS22084039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Vassiliou E., Farias-Pereira R. (2023). Impact of lipid metabolism on macrophage polarization: implications for inflammation and tumor immunity. Int. J. Mol. Sci. 24, 12032. 10.3390/IJMS241512032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wang C., Ma C., Gong L., Guo Y., Fu K., Zhang Y., et al. (2021). Macrophage polarization and its role in liver disease. Front. Immunol. 12, 803037. 10.3389/FIMMU.2021.803037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wei H. J., Upadhyayula P. S., Pouliopoulos A. N., Englander Z. K., Zhang X., Jan C. I., et al. (2021). Focused ultrasound-mediated blood-brain barrier opening increases delivery and efficacy of etoposide for glioblastoma treatment. Int. J. Radiat. Oncol. Biol. Phys. 110, 539–550. 10.1016/j.ijrobp.2020.12.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Yang H., Cheng H., Dai R., Shang L., Zhang X., Wen H. (2023). Macrophage polarization in tissue fibrosis. PeerJ 11, e16092. 10.7717/PEERJ.16092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zandl-Lang M. (2024). Tracing the lipidome in inborn errors of metabolism. Biochimica et Biophysica Acta (BBA) - Mol. Cell Biol. Lipids 1869, 159491. 10.1016/J.BBALIP.2024.159491 [DOI] [PubMed] [Google Scholar]
  44. Zumla A., Rao M., Wallis R. S., Kaufmann S. H. E., Rustomjee R., Mwaba P., et al. (2016). Host-directed therapies for infectious diseases: current status, recent progress, and future prospects. Lancet Infect. Dis. 16, e47–e63. 10.1016/S1473-3099(16)00078-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementaryfile4.xls (395KB, xls)
Supplementaryfile1.xlsx (47.6KB, xlsx)
Supplementaryfile2.xlsx (59.3KB, xlsx)
Supplementaryfile3.xlsx (10.4KB, xlsx)

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

Articles from Frontiers in Pharmacology are provided here courtesy of Frontiers Media SA

RESOURCES