Genetically predicted expression of EP300, LDHC, and TKT and risk of neurodegenerative diseases

Neurodegenerative diseases such as Alzheimer’s disease (AD), Parkinson’s disease (PD), multiple sclerosis (MS), and amyotrophic lateral sclerosis (ALS) are increasing in global burden. Emerging studies suggest that cellular metabolism and epigenetic mechanisms—especially lactylation—play roles in disease pathogenesis. Histone lysine lactylation, a recently discovered epigenetic mark driven by lactate metabolism, is enriched in activated microglia and aged brain tissues [1, 2]. However, whether altered expression of lactylation-related genes causally contributes to neurodegeneration remains unknown.

To address this, we performed a two-sample Mendelian randomization (MR) analysis using genetically predicted gene expression to test the causal effects of lactylation-related genes on neurodegenerative disease risk (Fig. 1A). We curated 44 lactylation-related genes from recent reviews and retained 29 with valid cis-eQTL instruments from eQTLGen (n = 31,684) (Supplementary Table 1). Summary-level outcome data were obtained from large GWASs for AD, PD, MS, and ALS. Instrument strength was ensured using F-statistics ≥ 10, and linkage disequilibrium clumping was applied (r² < 0.1). We used inverse-variance weighted (IVW) as the primary method, supplemented by MR-Egger and weighted median approaches. Bonferroni correction was applied (α = 0.0033), and sensitivity tests included heterogeneity (Cochran’s Q), pleiotropy (Egger intercept and MR-PRESSO), and leave-one-out analyses.

Fig. 1.

(A) Overview of the Mendelian randomization framework. (B) Circular heatmap of MR results across lactylation-related genes and four neurodegenerative outcomes. The plot displays IVW odds ratios (ORs), P-values, and results from MR-Egger, weighted median, heterogeneity (Q), and MR-PRESSO for each gene–disease pair. Color indicates effect size direction and magnitude. (C) Forest plot of Bonferroni-significant MR associations. Six gene–disease associations showed significance (P < 0.0031), with OR and 95% confidence intervals estimated using IVW. Sensitivity results (Q_P, Egger intercept, and MR-PRESSO) are listed alongside. (D) Scatter plots showing SNP-level causal estimates for the associations of EP300–AD, LDHC–PD, LDHC–MS, and TKT–MS using five MR methods. (E) Leave-one-out sensitivity analysis confirming the robustness of MR estimates for EP300–AD, LDHC–PD/MS, and TKT–MS

Fifteen genes remained after harmonization (Supplementary Table 1). Among 15 × 4 gene-disease tests (Fig. 1B, Supplementary Tables 2–5), six associations passed Bonferroni correction. Higher genetically predicted expression of EP300 was associated with increased AD risk (OR = 1.08, 95% CI: 1.03–1.13, P = 0.0011), LDHC with PD (OR = 1.08, P = 0.0022) and MS (OR = 1.11, P = 0.0006), and TKT with MS (OR = 1.09, P = 0.0003) (Fig. 1C). Directionally concordant positive slopes across IVW, MR-Egger, and weighted median are shown in the SNP–effect scatter plots (Fig. 1D). Sensitivity analyses supported robustness: heterogeneity was limited (Q P ≥ 0.05 except SLC16A4 and SMARCA4 where P = 0.02), no directional pleiotropy was detected (Egger-intercept P ≥ 0.05), MR-PRESSO global tests were non-significant (P ≥ 0.05), and leave-one-out analyses showed no single-SNP dependence (Fig. 1E).

EP300, encoding a histone acetyltransferase (KAT3B), has known roles in chromatin remodeling and neuronal survival. Dysregulated EP300 expression in AD is linked to loss of miR-132, enhanced Tau acetylation, and cognitive decline in experimental models [3]. Our MR results provide genetic evidence consistent with a causal contribution of EP300 to AD. LDHC, involved in lactate/pyruvate interconversion, has no established link to PD or MS, but its metabolic role suggests a connection to glycolysis-driven neuroinflammation [4]. Similarly, TKT is a key enzyme in the pentose phosphate pathway, implicated in redox balance and myelin maintenance, potentially relevant to MS. These findings provide novel mechanistic hypotheses connecting metabolic-epigenetic crosstalk and neurodegeneration.

Our study has several strengths: strong instruments, cross-method consistency, large GWAS datasets, and robust sensitivity analyses. However, limitations include the use of summary-level data and the lack of functional validation. Notably, our eQTL instruments were derived from whole-blood datasets rather than brain-specific tissues, and formal colocalization analysis was not performed. Future work should assess brain or immune-specific QTLs, explore colocalization, and validate EP300, LDHC, and TKT in disease models.

In conclusion, this MR study provides genetic evidence consistent with a potential causal contribution of EP300, LDHC, and TKT to the risk of AD, PD, and MS. These genes may represent targets for mechanistic studies and therapeutic intervention.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Houwen Zhang: present idea, perform MR analysis and manuscript writing. Chunrong Li: manuscript writing and evaluate the quality of MR. Bin Xu and You Wu: data collection and figure drawing. Fangzheng Cao: final approval.

Funding

This study was supported by National Natural Science Foundation of China (NSFC) - Young Scientists Fund (Grant 82205012 to Y.W.), the Natural Science Foundation of Zhejiang Province (Grant LMS25H290004 to Y.W., Grant LQ23H090019 to C.R.L.), the Medical Health Science and Technology Project of Zhejiang Province (Grant 2024KY675 to C.R.L.), and the Traditional Chinese Medicine Science and Technology Initiative of Zhejiang Province (Grant 2025ZL335 to Y.W., Grant 2024ZL291 to C.R.L., Grant 2026ZL0407 to F.Z.C.).

Data availability

Exposure summary statistics are available from the eQTLGen consortium. Outcome GWAS can be accessed via the GWAS Catalog/IEU OpenGWAS using the accession IDs GCST90027158, GCST90027164, ieu-b-7, and ieu-b-18.

Declarations

Ethics approval and consent to participate

This study used publicly available, de-identified summary statistics only. No new human participants or animals were involved; therefore, institutional review board approval and informed consent were not required.

Consent for publication

Not applicable.

Competing interests

The authors have no conflicts of interest to disclose.