# Identification of lactylation and its hub genes in contributing immune activation and renal allograft fibrosis by integrative bioinformatics and machine learning

## TL;DR

This study identifies five key genes linked to lactylation that are associated with immune activation and kidney transplant fibrosis, offering potential biomarkers and therapeutic targets.

## Contribution

The study introduces a novel integrative approach combining bioinformatics and machine learning to identify lactylation-related hub genes in renal allograft fibrosis.

## Key findings

- Five lactylation-related hub genes (IKZF1, PDLIM1, S100A11, STAT4, SLC2A3) were strongly associated with renal allograft fibrosis.
- The lactylation-related risk score (LRS) demonstrated robust predictive accuracy in both training and validation cohorts.
- Hub genes showed upregulation in murine models of renal fibrosis and correlated with immune-cell infiltration.

## Abstract

Late graft loss due to chronic renal allograft fibrosis remains a major challenge after kidney transplantation. Excessive immune-cell activation is a key driver of allograft fibrosis; however, the underlying mechanisms remain incompletely understood. Recent studies have implicated lactylation, a post-translational protein modification derived from lactate, in immune activation. Nonetheless, the role of lactylation in renal allograft fibrosis has not been systematically explored.

Transcriptomic datasets from kidney transplant recipients with and without interstitial fibrosis/tubular atrophy (IFTA) were obtained from the GEO database. Differentially expressed genes were intersected with lactylation-related genes (LRGs) to identify differentially expressed LRGs (DELRGs). Functional enrichment analyses were performed to explore associated biological processes and pathways. Weighted gene co-expression network analysis (WGCNA) combined with multiple machine-learning algorithms was used to screen for hub genes. A lactylation-related risk score (LRS) was constructed and validated across independent cohorts, and its predictive performance was evaluated by receiver operating characteristic (ROC) analysis. Single-nucleus RNA sequencing (snRNA-seq) data from allograft biopsies (GSE195718) were processed with Seurat and Harmony for clustering and annotation; cell type–specific hub LRG expression and lactylation scores were profiled. Two murine renal fibrosis models were established to validate the expression of hub genes and to assess their associations with immune-cell infiltration.

We identified five hub LRGs—IKZF1, PDLIM1, S100A11, STAT4 and SLC2A3—that were strongly associated with renal allograft fibrosis. These genes were closely linked to pathways related to lactate metabolism, immune activation and oxidative stress. The LRS based on these genes showed robust predictive accuracy in both the training and validation cohorts. In addition, snRNA-seq of allograft biopsies localized hub LRGs predominantly to immune-lineage and stromal clusters with higher lactylation scores in IFTA samples; concordantly, immune-infiltration analyses revealed significant positive correlations between hub LRGs and multiple immune-cell subsets. Furthermore, these hub genes were upregulated in murine models of renal fibrosis.

This study identified five lactylation-related hub genes that are closely associated with immune-cell infiltration and exhibit strong predictive performance, suggesting their potential as diagnostic biomarkers and therapeutic targets in renal allograft fibrosis.