Increased IL4I1 expression predicts poor survival and modulates the immune microenvironment in acute myeloid leukemia

Abstract

Background

The immunometabolic enzyme Interleukin-4-induced-1 (IL4I1) is implicated in cancer pathogenesis, yet its specific function and clinical relevance in acute myeloid leukemia (AML) remain unclear.

Methods

Comparative analysis of IL4I1 mRNA levels between AML patients and normal controls was performed using the Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases. The Kaplan–Meier survival analysis was conducted to evaluate the prognostic value of IL4I1. Functional insights were derived from analyses of differentially expressed genes (DEGs), Gene Set Enrichment Analysis (GSEA), and Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment. Immune infiltration was evaluated using the ssGSEA, ESTIMATE, quanTIseq and single-cell RNA sequencing (scRNA-seq) analysis. Finally, in vitro and in vivo functional experiments were perfromed to explore the impact of IL4I1 on AML progression and immunoregulation.

Results

IL4I1 expression was significantly elevated in AML compared to normal controls (p = 0.0004) and associated with poorer overall survival (p = 0.003). Bioinformatic analysis revealed that IL4I1 was linked to immune-related pathways—including humoral immune response, leukocyte interactions, and chemokine signaling—and to cellular amino acid metabolism. Its expression correlated with immune cell infiltration and checkpoint molecule expression. Experimentally, IL4I1 promoted leukemia cell proliferation in vitro and in vivo (p < 0.05). Furthermore, silencing IL4I1 suppressed M2 macrophage polarization and reduced secretion of inflammatory factors (p < 0.05).

Conclusions

IL4I1 may serve as a potential biomarker for poor prognosis and an attractive target for immune-based therapeutic interventions in AML.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-026-07814-x.

Introduction

Acute myeloid leukemia (AML) is a malignant disorder of the hematopoietic system, accounting for nearly 70% of all adult acute leukemia cases. Its incidence rises with age, posing a significant threat to human health due to its rapid progression. Although treatments such as chemotherapy, bone marrow transplantation, and targeted therapy have improved complete remission (CR) rates and extended disease-free survival (DFS), outcomes remain suboptimal for many AML patients [1]. Targeted agents, including FLT3 inhibitors (e.g., Sorafenib), BCL-2 inhibitors (e.g., Venetoclax), and IDH2 inhibitors (e.g., Enasidenib), have advanced care, yet primary and acquired drug resistance persist as major challenges. Approximately 20%–40% of patients fail initial therapy, and 50%–70% of those achieving remission later relapse. Therefore, identifying new therapeutic strategies and targets remains urgent. Encouragingly, immunotherapy has emerged as a promising avenue for relapsed/refractory AML (R/R AML) and may help improve overall prognosis.

IL4I1 (Interleukin-4-induced 1) is a secretory enzyme implicated in amino acid metabolism and immune regulatory processes. It was initially named for its induction by interleukin-4 in B cells. IL4I1 is predominantly produced by B and T lymphocytes, monocytes/macrophages, and mature dendritic cells. It suppresses the activation and expansion of human CD4⁺ and CD8⁺ T cells [2] and also modulates B cell differentiation [3]. Furthermore, IL4I1 is expressed in tumor-associated macrophages across various human malignancies [4, 5]. By dampening CD8⁺ T-cell activity, IL4I1 facilitates tumor growth and immune escape.

Functionally, IL4I1 is recognized as a regulator of adaptive immunity, fine-tuning B- and T-cell responses [6]. It activates the aryl hydrocarbon receptor (AHR) by converting tryptophan into indole-3-pyruvic acid (I3P). AHR is a transcription factor originally linked to environmental toxin responses. In glioblastoma, IL4I1-driven AHR activation promotes tumor cell migration, inhibits T-cell proliferation, and correlates with poor patient survival [7]. Experimental models further show that IL4I1 enhances AHR activity and upregulates downstream targets including IDO1 (Indoleamine 2,3-Dioxygenase 1) and TDO2 (Tryptophan 2,3-Dioxygenase 2). Generally, AHR signaling fosters an immunosuppressive milieu by expanding regulatory T cells and impairing cytotoxic T-cell function [8], and IL4I1 likely amplifies this process [9].

Although IL4I1 contributes to immune evasion and worse outcomes in multiple malignancies, its role in acute leukemia—particularly AML—remains poorly defined. Our study aims to clarify the prognostic significance, immunomodulatory function, and mechanistic impact of IL4I1 in AML. This work may establish IL4I1 as a potential prognostic indicator and therapeutic target in AML.

Materials and methods

Data source

Processed RNA-seq data in TPM (Transcripts Per Million) format, and clinical information of acute myeloid leukemia (AML) and 32 other cancer types were acquired from the UCSC XENA database (https://xenabrowser.net/datapages/). These data, encompassing both TCGA tumor samples and GTEx normal tissue samples, had been uniformly reprocessed using the Toil and STAR pipeline [10, 11]. Samples lacking corresponding clinical information were excluded from the analysis. The remaining AML-associated dataset comprised of 173 AML samples from TCGA-LAML and 70 normal controls from GTEx database. According to the clinical information, AML patients were stratified according to the European LeukemiaNet (ELN) risk recommendations for AML [12]. The corresponding clinicopathological data are listed in Table 1. Additionally, we acquired a single-cell RNA sequencing (scRNA-seq) dataset for AML from the GEO database (accession number GSE154109). After filtering samples with poor quality, 3 health control and 7 AML samples were retained for further scRNA-seq analysis.

Table 1.

Association of IL4I1 expression with clinicopathological features in AML patients from TCGA-LAML cohort

Characteristics	Low expression of IL4I1	High expression of IL4I1	p values
n	75	75
Gender, n (%)			0.870
Female	34	33
Male	41	42
Age, n (%)			0.005*
<= 60	52	35
> 60	23	40
FAB classifications, n (%)			<0.001*
M0	8	7
M1	20	15
M2	24	14
M3	11	3
M4	9	20
M5	2	13
M6	0	2
M7	0	1
Cytogenetic risk, n (%)			0.002*
Favorable	23	7
Intermediate/normal	32	50
Poor	18	18
FLT3 mutation, n (%)			0.370
Negative	48	53
Positive	25	20
RAS mutation, n (%)			0.267
Negative	73	68
Positive	2	6
IDH1 mutation (R132), n (%)			0.147
Negative	65	70
Positive	9	4
IDH1 mutation (R140), n (%)			0.547
Negative	69	67
Positive	5	7
IDH1 mutation (R172), n (%)			0.477
Negative	72	74
Positive	2	0
NPM1 mutation, n (%)			0.069
Negative	63	53
Positive	12	21

* p < 0.05

Gene expression and clinical value estimation of IL4I1

Differential gene expression of IL4I1 between AML or other pan-cancer samples and healthy controls was evaluated via the Mann–Whitney U test. 150 AML patients from the TCGA dataset with available IL4I1 expression data were stratified into low- and high-expression groups based on the median IL4I1 mRNA expression, with this cutoff applied consistently in subsequent analyses. Kaplan-Meier (KM) survival curves (https://kmplot.Com/) were generated to assess the prognostic significance of IL4I1 expression in AML patients [13]. Differences between groups stratified by IL4I1 expression level were compared with the log-rank test. A two-sided p-value < 0.05 was considered statistically significant. Univariate and multivariate Cox proportional hazards models were employed to identify independent prognostic factors using survival R package (v3.7.0). Furthermore, we performed transcriptomic profiling to identify DEGs between the IL4I1-low and IL4I1-high cohorts [14].

Protein-protein interaction (PPI) network construction

We utilized the STRING database [15] to identify potential IL4I1-interacting proteins (combined score ≥ 0.900, highest confidence). The interconnectivity among IL4I1 and its interacting protein-encoding genes in AML was visualized using a circos plot generated with R package circlize (v0.4.1).

Functional enrichment analysis

GSEA-based normalization was performed on TCGA RNA-Seq datasets [16]. The clusterProfiler R package (v4.14.3) [17] was applied to conduct GO and KEGG pathway enrichment analysis, aiming to elucidate the potential biological roles of IL4I1. Statistical significance of enrichment results was defined by false discovery rate (FDR)-adjusted q-value < 0.05. The q-values were calculated using the Benjamini-Hochberg method for multiple hypothesis testing.

Single-cell RNA sequencing (scRNA-seq) data analysis

scRNA-seq data were obtained from the GEO database and processed using the Seurat (v5.0.1) R package. Detailly, low-quality cells and potential doublets were filtered out using stringent thresholds: cells with fewer than 200 or more than 5,000 detected genes (nFeature_RNA), or with total UMI counts (nCount_RNA) exceeding 40,000 were excluded. To remove cells exhibiting apoptosis or significant stress, cells with mitochondrial gene content (percent.mt) > 5%, hemoglobin gene expression (percent.hb) > 0.1%, or ribosomal gene expression (percent.ribo) > 30% were also discarded. The gene expression data for each cell were normalized and variance-stabilized using the SCTransform function. To integrate datasets and mitigate technical batch effects, the harmony package (v1.2.1) was applied. Post-integration, highly variable genes were identified for downstream dimensionality reduction. Principal component analysis (PCA) was performed on the scaled data of highly variable genes. The top 22 principal components (PCs, dims = 22), determined by an elbow plot of standard deviations, were used to construct a shared nearest neighbor (SNN) graph. Cells were clustered at a resolution of 0.1 for subsequent major population identification. Cell types were annotated using canonical marker genes from the CellMarker database and relevant literature [18].

Expression patterns of IL4I1 across annotated cell clusters were assessed and visualized via violin plots and feature plots. To investigate its functional role, monocyte subpopulations were stratified into IL4I1-positive (detected expression > 0) and IL4I1-negative (detected expression = 0) groups. DEGs between these two groups were identified using a Wilcoxon rank-sum test in Seurat (with logfc.threshold > 0.25 and adjusted p-value < 0.05). These DEGs were subsequently subjected to GO and KEGG pathway enrichment analyses using the clusterProfiler (v4.14.3) R package to elucidate associated biological processes and pathways.

Immune cell infiltration profiling

Immune cell infiltration levels were quantified using two complementary computational approaches. First, single-sample Gene Set Enrichment Analysis (ssGSEA) was implemented via the GSVA R package (v1.46.0) [19]. This algorithm calculated enrichment scores for 24 predefined immune cell types, based on marker gene sets derived from a published immunogenomic framework [20]. Second, the stromal and immune components within the tumor microenvironment were independently assessed using the ESTIMATE algorithm (v1.0.13) [21] which generates stromal, immune, and combined ESTIMATE scores for each sample.

To evaluate the relationship between IL4I1 expression and the tumor immune contexture, Spearman’s rank correlation analysis was employed. Correlation coefficients (R) and their corresponding p -values were reported, with a two-sided p < 0.05 considered statistically significant. Findings were further validated using an orthogonal deconvolution method, the quanTIseq algorithm via the immunedeconv R package (v2.1.3) [22], which estimates the abundances of 10 immune cell types from bulk transcriptomic data [23].

Cell culture and construction of IL4I1-knockdown M2 type macrophage cell model

The human monocytic THP-1 cell line (Cat. No. GDC0100) was sourced from the China Center for Type Culture Collection. For routine culture, THP-1 cells were incubated in RPMI 1640 medium with 10% fetal bovine serum (Gibco, USA) and 1% penicillin-streptomycin at 37 °C in 5% CO₂. To generate M0 macrophages, THP-1 monocytes were seeded in six-well plates at a density of 1 × 10⁶ cells per well and differentiated with 100 ng/mL phorbol 12-myristate 13-acetate (PMA; MedChemExpress, USA) for 24 h. Subsequently, for M2 macrophage polarization, the differentiated M0 macrophages were maintained in fresh culture medium containing 100 ng/mL PMA along with 20 ng/mL recombinant human IL-4 and 20 ng/mL recombinant human IL-13 (both from MedChemExpress, USA) for an additional 48 h to establish the M2-like macrophages. To establish stable IL4I1-knockdown models, parental cells were transduced with either control (shRNA-NC) or IL4I1-targeting (shRNA-IL4I1) puromycin-resistant lentiviral constructs. Following 72 h of transduction, stable polyclonal populations were selected using 2 µg/ml puromycin. The resulting cell lines were utilized for subsequent functional analyses in cellular proliferation and metastatic potential. Supernatant from cell culture was centrifuged (1000 ×g, 10 min, 4 °C), filtered, and aliquoted for ELISA. Meanwhile, cells were harvested for western blot and RT-qPCR.

RNA extraction and quantitative PCR (RT-qPCR)

TRIzol reagent (Invitrogen, USA) was used to extract total RNA from cells, which was then reverse-transcribed into cDNA using the iScript cDNA Synthesis Kit (Bio-Rad, USA). RT-qPCR was subsequently performed on an ABI 7500 instrument with TB Green^® Premix Ex Taq™ (Tli RNaseH Plus) (TaKaRa, Japan). Primer sequences are listed in Table 2.

Table 2.

Real-time PCR primers

Gene name	Primer sequences (5’ to 3’)
IL4I1	Forward: 5’-GCCAAGACCCCTTCGAGAAAT-3’ Reverse: 5’-CCGATCCTGTTATCTGCCTCC-3’
GAPDH	Forward: 5’-GGAGCGAGATCCCTCCAAAAT-3’ Reverse: 5’-GGCTGTTGTCATACTTCTCATGG-3’
ITGAM (the gene encoding CD11b)	Forward: 5’-GACCTCAGCATCACCTTCA-3’ Reverse: 5’-CCTCACCATCATTTCTCACA-3’
CD68	Forward: 5’-CTTCTCTCATTCCCCTATGGACA-3’ Forward: 5’-GAAGGACACATTGTACTCCACC-3’
CD163	Forward: 5’-CTTGGGACTTGGACGATGCT-3’ Reverse: 5’-GCAGGACAATCCCACAAGGA-3’
MRC1 (the gene encoding CD206)	Forward: 5’-GGGTTGCTATCACTCTCTATGC-3’ Reverse: 5’-TTTCTTGTCTGTTGCCGTAGTT-3’

Western blot

Cell lysates were prepared with lysis buffer for Western and IP assays supplemented with protease cocktail (Beyotime, China). After SDS-PAGE separation and transfer to PVDF membranes, blots were probed with IL4I1 (DF4819, Affinity), GAPDH (ab181602, Abcam), CD11b (ab133357, Abcam), CD68 (ab955, Abcam), CD163 (ab182422, Abcam), CD206 (ab252921, Abcam) and Goat Anti-Rabbit IgG H&L (HRP) (ab205718, Abcam) antibodies.

Cell functional assays

Proliferation kinetics were assessed by daily CCK-8 (Beyotime, China) measurements over a 5-day period. Cells were plated at a density of 10,000 cells per well in 96-well plates, and 10 µL of reagent was administered to each well 24 h prior to each timepoint measurement. Measurement of absorbance at 450 nm was carried out using a Multiskan FC Microplate Photometer (Thermo Fisher Scientific, USA). For migration and invasion assays, THP-1 shRNA-IL4I1 cells and their controls were serum-starved overnight. On the following day, cells were seeded at equal density into 24-well Transwell inserts (Corning, USA) containing serum-free medium. And a 24-hour incubation was carried out at 37 °C under 5% CO₂. Afterward, cells retained on the upper chamber membrane were stained with crystal violet. Images were acquired using an inverted microscope (Olympus, Japan). Cell cycle distribution and apoptosis were assessed with a BD FACSVerse™ flow cytometer (BD, USA) following the manufacturer’s protocols.

In vivo tumor proliferation assay

Xenograft experiments were performed to assess the in vivo tumorigenic potential of THP-1 shRNA-IL4I1 cells and their controls. For this assay, 5 × 10⁶ THP-1 cells were subcutaneously implanted into 5-week-old female BALB/c nude mice (SPF grade; Cavens Laboratory Animal, China). Tumor development was monitored every three days. After three weeks, all mice were humanely euthanized, and the subcutaneous tumors were excised, weighed, measured, photographed, and fixed in 4% paraformaldehyde for further analysis. Tumor volume and mass were recorded. IL4I1, TGF-β1, and IL10 levels in mouse serum and culture supernatants were determined using ELISA kits (CUSABIO, China). All animal procedures were approved by the Ethics Committee of the First Affiliated Hospital of Fujian Medical University.

Statistical analyses

Statistical analyses were conducted using R (v4.2.1) and SPSS (v24.0). Functional enrichment and gene expression visualization were carried out using the clusterProfiler (v4.14.3) and ggplot2 (v3.4.4) R packages, respectively. Associations between IL4I1 expression and clinical characteristics were analyzed using logistic regression. Survival analyses were conducted using the KM method and compared by log-rank tests. Categorical data were analyzed using Chi-square or Fisher’s exact tests. Quantitative measurements were presented as mean ± SD from three or six independent replicates. Comparison between the two groups was performed using Student’s t-test or the Mann–Whitney U test. Statistical significance thresholds were defined as *p < 0.05, **p < 0.01, ***p < 0.001 throughout all analyses.

Results

Elevated expression of IL4I1 in AML correlates with unfavorable prognosis

Expression comparison indicated that IL4I1 was markedly higher expressed in AML patients compared to the normal controls (Fig. 1A, p = 0.0004). KM survival plots (Fig. 1B) revealed that elevated IL4I1 levels were associated with adverse patient outcomes (HR = 1.95, 95% CI: 1.26–3.02, p = 0.003). Patients with lower IL4I1 expression exhibit improved survival time (Fig. 1C), which was supported by a significant link between high IL4I1 expression and unfavorable OS events (p = 0.0211, Fig. 1D). Elevated IL4I1 expression in AML was associated with unfavorable clinicopathological features (Table 1), including advanced age (> 60 years; p = 0.005) and a higher proportion of intermediate/normal cytogenetic risk (p = 0.002). To further evaluate the role of IL4I1 in other cancer type, we conducted a pan-cancer evaluation of IL4I1 expression between malignant and normal tissues across 30 cancer types. Comparative analysis revealed that IL4I1 was significantly upregulated in 27 cancer types (including ACC, BLCA, BRCA, …, UCS) whereas it was downregulated in KICH and TGCT (Supplementary Fig. S1). Univariate Cox regression identified age (HR = 3.321, 95% CI: 2.156–5.116, p < 0.001), cytogenetic risk Intermediate/normal (HR = 2.934, 95% CI: 1.487–5.791, p = 0.002), cytogenetic risk Poor (HR = 4.124, 95% CI: 1.927–8.823, p < 0.001) and IL4I1 expression (HR = 1.952, 95% CI: 1.262–3.021, p = 0.003) as significant prognostic factors (Fig. 1E). However, in multivariate analysis adjusted for gender, race, age, cytogenetic risk, FLT3 mutation, IDH1 mutation, RAS mutation, and NPM1 mutation, high IL4I1 expression was not retained as an independent risk factor for poor prognosis (adjusted HR = 1.557, 95% CI: 0.959–2.526, p = 0.073).

Fig. 1.

Association of the IL4I1 expression levels with clinical outcomes in AML. (A) IL4I1 mRNA expression in AML samples (n = 173) and normal controls (n = 70) were assessed using the TCGA and GTEx database (p = 0.0004). (B) Kaplan-Meier analysis was performed to assess the prognostic impact of IL4I1 in AML, with patients from TCGA-LAML corhort (n = 150) stratified into high- and low-expression groups based on the median IL4I1 mRNA level (HR = 1.95, 95% CI: 1.26–3.02, p = 0.003). (C) Scatter plot showing the distribution of IL4I1 expression levels relative to patient survival status (0: alive; 1: dead). (D) Overall survival analysis further confirms the poor prognosis associated with high IL4I1 expression (p = 0.0211). (E) Cox regression analyses were employed to identify factors associated with prognosis. p values calculated from Mann–Whitney U test (A, D). *p < 0.05, **p < 0.01, ***p < 0.001

Genes associated with IL4I1 in AML

The PPI network illustrated highest-confidence interactions between IL4I1 and ten partner genes (Fig. 2A, combined score ≥ 0.900). The circos plot (Fig. 2B) further demonstrated a positive correlation between IL4I1 and TAT (R = 0.276, p < 0.001 by Spearman), but negative correlations with NIT2 (R = -0.436, p < 0.001) and BCAT1 (R = -0.336, p < 0.001). Correlation analysis identified the top 20 genes most positively correlated with IL4I1 expression (R > 0.3, p < 0.001 by Spearman) (Fig. 2C). Most of which were differentially expressed in AML versus normal controls, specifically, sixteen genes (CD300C, CDH23, CX3CR1, DOK2, GPR132, FGR, FGD2, IL1RN, MVP, LRRC25, LILRB2, SCIMP, TFEB, SLC15A3, VDR, and LILRB1) were significantly upregulated, while UNC119 was downregulated in AML specimens (Supplementary Fig. S2). Additionally, high expression of most of these IL4I1-associated genes (including CD300C, CDH23, CX3CR1, and others) was significantly linked to poor prognosis in AML patients (Supplementary Fig. S3).

Fig. 2.

PPI network and highly correlated genes of IL4I1. (A) PPI network of IL4I1 and its top 10 interacting proteins, as identified from the STRING database (combined score ≥ 0.900, highest confidence). (B) Circos plot illustrating the interconnectivity between IL4I1 and its interacting protein-encoding genes. Ribbon thickness and color intensity represent the strength of the correlation. (C) Correlation heatmap between IL4I1 expression and most positively correlated genes top 20 in the TCGA-AML cohort (R > 0.3, p < 0.001 by Spearman). ***p < 0.001

Immune and metabolism disorder in AML patients with high IL4I1 expression

Transcriptomic analysis identified 2,601 DEGs between IL4I1-high and IL4I1-low patient cohorts, including 1,898 upregulated and 703 downregulated genes (Fig. 3A). GSEA revealed that IL4I1 expression was significantly associated with multiple immune-related pathways, including humoral immune responses, leukocyte interactions, interleukin-10 signaling, NK cell cytotoxicity, as well as chemokine and cytokine signaling pathways (Fig. 3B). Additionally, eighty pathways showed notable enrichment according to KEGG pathway analysis, of which 76 were positively associated with IL4I1 expression, while 4 showed negative associations. IL4I1 expression showed the strongest positive associations with three KEGG pathways: the chemokine signaling pathway, cytokine-cytokine receptor interaction and NK cell cytotoxicity (Fig. 3C). Meanwhile, the four pathways negatively correlated with IL4I1 expression included: pentose and glucuronate interconversions, ascorbate and aldarate metabolism, olfactory transduction and cytochrome P450-mediated drug metabolism (Fig. 3D). Further studies indicated that IL4I1 is associated with cellular amino acid biosynthesis and metabolism, particularly in relation to tryptophan, tyrosine, alanine, aspartate, glutamate, cysteine, methionine and phenylalanine, as revealed by GO (Fig. 3E) and KEGG analyses (Fig. 3F).

Fig. 3.

Characterization and functional annotation of genes associated with IL4I1. (A) Volcano plot depicting DEGs between AML cases stratified by IL4I1 expression levels. (B) GSEA analysis showing the top five terms positively associated with elevated IL4I1 expression. (C, D) KEGG pathway enrichment results from GSEA, highlighting the strongest positive (C) and negative (D) enrichment in relation to IL4I1 expression. (E, F) GO (E) and KEGG (F) enrichment analyses of DEGs. Statistical significance of enrichment results was defined by q-value < 0.05. q values for Fig. 3 (B-F) are provided in Supplementary Table S1

Immune infiltration landscape associated with IL4I1 expression

Analysis using the ssGSEA algorithm revealed that AML patients with high IL4I1 expression had an enriched proportion of dendritic cells, eosinophils, macrophages, neutrophils, T cells, and B cells within the tumor immune microenvironment (Fig. 4A). Moreover, IL4I1 expression showed strong positive associations with infiltration of macrophages (R = 0.522, p < 0.001 by Spearman) (Fig. 4B). Notable associations were also observed with multiple immune cells, including T cells and B cells. The ESTIMATE algorithm further confirmed that IL4I1 expression was significantly positively correlated with both the degree of immune cell infiltration and stromal content (Fig. 4C). The quanTIseq analysis validated significant enrichment of M2 macrophages, monocytes, and CD8⁺ T cells in the IL4I1-high group (Fig. 4D). Additionally, IL4I1 expression positively correlated with the infiltration levels of M2 macrophages and monocytes (R > 0.3, p < 0.001 by Spearman, Fig. 4E).

Fig. 4.

The immune infiltration landscape associated with IL4I1 expression in AML. (A) Immune cell composition in high versus low IL4I1 expression groups, analyzed with ssGSEA algorithm. (B) Spearman correlation between IL4I1 expression and 24 immune cell subtypes (ssGSEA-based). (C) Association of IL4I1 expression with immune and stromal scores derived from TCGA data using ESTIMATE algorithm. (D) Comparative immune cell profiling by IL4I1 expression level, assessed with the quanTIseq algorithm. (E) Spearman correlation between IL4I1 expression level and 10 immune cell subtypes (quanTIseq-based). p values calculated from Mann–Whitney U test (A, D). *p < 0.05, **p < 0.01, ***p < 0.001. ns, not significant. The p values for Fig. 4A and D, along with the R values for Fig. 4E, are provided in Supplementary Table S2

IL4I1 expression and function within the AML microenvironment

Given the observed correlations between immune cell infiltration and IL4I1, we utilized single-cell RNA sequencing (scRNA-seq) to characterize its expression patterns and potential functions across diverse cell populations in the AML microenvironment. Analysis of the GSE154109 dataset revealed that IL4I1 is specifically and highly expressed in bone marrow-derived monocytes from AML patients (n = 8), while its expression is minimal or undetectable in other cell types (Fig. 5A-D). IL4I1 expression in AML monocytes was significantly elevated compared to that in healthy controls (Fig. 5E). To gain deeper functional insights, we stratified monocytes from AML bone marrow specimens into two distinct subpopulations based on IL4I1 expression: IL4I1-positive (IL4I1⁺) and IL4I1-negative (IL4I1⁻) cells, and performed comparative transcriptomic profiling. GO enrichment analysis demonstrated that the DEGs were significantly enriched in the following biological processes: (1) viral response and defense mechanisms; (2) RNA processing and ribosome biogenesis; (3) regulation of mitochondrial function; and (4) modulation of innate immune response (Fig. 5F; Supplementary Fig. S4). Additionally, these genes were implicated in the cellular response to bacterial-derived molecules and the canonical NF-κB signal transduction pathway. These findings suggest that IL4I1 may play a pivotal role in modulating monocyte function in AML through the regulation of immune- and metabolism-related pathways, particularly by influencing critical cellular processes such as antiviral immunity, cellular energy metabolism, and ribosome biogenesis, thereby potentially contributing to the malignant phenotype.

Fig. 5.

Single-cell transcriptomic analysis of IL4I1 in the AML microenvironment. (A) UMAP plot showing the annotation of 7 major cell types in GSE154109 dataset from the GEO database. (B) Bar plot comparing the proportional abundance of each cell type between AML patients (n = 8) and healthy controls (n = 4). (C) UMAP visualization of IL4I1 expression across the different cell clusters. (D) IL4I1 demonstrates predominant expression within the monocyte population. (E) Differential expression of IL4I1 in monocytes from AML patients versus healthy controls. (F) Significantly enriched GO terms related to biological processes. Statistical significance of enrichment results was defined by q-value < 0.05

KEGG pathway analysis further indicated that IL4I1-associated genes might be involved in the pathogenesis of several malignancies, including AML, chronic myeloid leukemia (CML), bladder cancer, and small cell lung cancer (Fig. 6). Moreover, the analysis suggested a potential role for IL4I1 in the mechanism of “transcriptional misregulation in cancer.” These bioinformatic insights strongly support the hypothesized functional significance of IL4I1 in AML pathogenesis. Furthermore, the DEGs were also engaged in the regulation of apoptosis and cell cycle progression, antigen processing and presentation, the tricarboxylic acid (TCA) cycle, chemokine signaling pathways, and the IL-17 signaling pathway, underscoring their involvement in key metabolic and inflammatory processes.

Fig. 6.

KEGG pathway enrichment analysis based on single-cell data. Significantly enriched KEGG pathways for DEGs between IL4I1-posivtive and IL4I1-negative monocyte subpopulations. Key cancer-related pathways, including “Transcriptional misregulation in cancer”, “Chronic myeloid leukemia” and “Acute myeloid leukemia” are highlighted. Statistical significance of enrichment results was defined by q-value < 0.05

IL4I1 expression is associated with immune checkpoints

To elucidate the immunomodulatory role of IL4I1, a correlation analysis of its expression with a panel of immune checkpoint (IC) genes was further performed based on TCGA dataset. IL4I1 expression showed significant positive correlations with several critical immune checkpoint genes, including PDCD1, PDCD1LG2, CTLA4, LAG3, TGFB1, CD86, and IL10 (Fig. 7A) (all p < 0.05). Among these, PDCD1 was downregulated in AML, while CD274, PDCD1LG2, CTLA4, and LAG3 were markedly upregulated (Fig. 7B-F) (all p < 0.001).

Fig. 7.

Correlation between IL4I1 and immune checkpoint gene expression. (A) Heatmap showing the correlation coefficients between IL4I1 and a panel of 19 immune checkpoint genes. (B-F) Comparison of mRNA expression levels of key immune checkpoint genes (B: PDCD1, C: CD274, D: PDCD1LG2, E: CTLA4, F: LAG3) between normal and AML samples. Spearman correlation analysis was performed to assess co-expression relationships between genes (A). p values calculated from Mann–Whitney U test (B-F). ***p < 0.001

IL4I1 facilitates the progression of AML cells in vitro

Successful knockdown of IL4I1 in THP-1 cells was confirmed by Western blot and RT-qPCR (Fig. 8A, B). Functional assays demonstrated that IL4I1 depletion significantly inhibited cell proliferation (Fig. 8C), impaired migration and invasion capabilities (Fig. 8D-F), promoted apoptosis (Fig. 8G, H), and induced cell cycle arrest at the G0/G1 phase (Fig. 8I, J). Collectively, these findings indicate that IL4I1 enhances the proliferative and metastatic capabilities of AML cells in vitro.

Fig. 8.

Impact of IL4I1 knockdown on the AML cell phenotype. (A, B) Validation of stable IL4I1 knockdown at the protein (A) and mRNA (B) levels in THP-1 cells. (C) Cell proliferation curves assessed by CCK-8 assay. (D-J) The migration (D, E) and invasion (D, F) abilities of the cells in each group were assessed by transwell assays, while apoptosis (G, H) and cell cycle (I, J) analyses were conducted via flow cytometry. The data were presented as means ± SEMs of three independent experiments and analyzed by a two- tailed Student’s t-test. **p < 0.01, ***p < 0.001. p-values for comparisons are provided in Supplementary Table S3

IL4I1 in M2-like macrophages, and its silencing inhibits M2 polarization in AML

The potential role of IL4I1 in mediating cross-talk between tumor-associated macrophages (TAMs) and AML cells and its influence on macrophage polarization were evaluated. Investigation into macrophage polarization markers revealed a strong positive correlation between IL4I1 and prototypical M2 markers (TGFB1, IL10, CD163, MS4A4A, and CSF1R) in TCGA datasets (Fig. 9A-C). In contrast, IL4I1 showed no consistent relationship with key M1 markers (IL12A, NOS2, PTGS2, and TNF). These proofs suggested that IL4I1 might affect M2 polarization. To confirm this, we differentiated THP-1 cells into M0 and M2 macrophages. WB and RT-qPCR analysis not only verified the successful differentiation but also revealed that IL4I1 expression was highest in the M2 phenotype (Fig. 9D-F). Knockdown of IL4I1 in M2-polarized macrophages led to reduced expression of M2 markers (CD163, CD206) and decreased secretion of IL10 and TGF-β1 (Fig. 9G-K).

Fig. 9.

IL4I1 promotes M2-like macrophage polarization in AML. (A) Heatmap analysis reveals the association between IL4I1 expression and markers of M1 and M2 macrophages in TCGA databases. (B, C) Scatter plots confirming strong positive correlations between IL4I1 and TGFB1 (B) and IL10 (C) expression in AML samples. (D-F) Comparative analysis of protein and mRNA expression levels for markers (CD11b, CD68, CD163, CD206 and IL4I1) of THP-1, M0 and M2 macrophages by WB and RT-qPCR. (G-I) M2-like macrophages transfected with either control siRNA (NC) or IL4I1-targeting siRNA (siIL4I1) were analyzed for CD163, CD206, and IL4I1 expression using WB and RT-qPCR. (J, K) ELISA measuring the concentrations of IL10 (J) and TGF-β1 (K) in the culture supernatants of the M2-like macrophage. The data were presented as means ± SEMs of three independent experiments and analyzed by a two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001. ns, not significant. p-values for comparisons are provided in Supplementary Table S3

IL4I1 facilitates the progression of AML cells in vivo

In the mouse xenograft model, tumors formed by IL4I1-knockdown THP-1 cells exhibited significantly slower growth, reduced final tumor volume and weight (Fig. 10A-C) compared to the control group. Mouse body weights were comparable between groups (Fig. 10D). Hematoxylin and eosin staining of tumors from the shRNA-NC group revealed variably sized nuclei with dense chromatin in tumor cells (Fig. 10E), whereas tumors from the shRNA-IL4I1 group showed increased areas of necrosis. These findings indicate that IL4I1 promotes AML progression in animal models. Moreover, serum levels of IL4I1, IL10, and TGF-β1 were significantly lower in the shRNA-IL4I1 group (Fig. 10F-H).

Fig. 10.

Knockdown of IL4I1 impairs the proliferation of THP-1 xenograft tumors. (A) Representative photographs of subcutaneous tumors resected from the shRNA-NC and shRNA-IL4I1 groups at the endpoint. (B-D) Dynamic changes in tumor volume (B), final tumor weight (C), and mouse body weight (D) over the 21-day experimental period. (E) Representative H&E staining of tumor sections from both groups. (F-H) Serum levels of IL4I1 (F), IL10 (G), and TGF-β1 (H) in mice were measured by ELISA. The data were presented as means ± SEMs and analyzed by a two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001. p-values for comparisons are provided in Supplementary Table S3

Discussion

Prognostic significance and expression landscape of IL4I1 in AML

IL4I1 is recognized for its roles in immunoregulation and tumorigenesis across various cancers, largely through its function in amino acid metabolism and AHR pathway activation [6, 7, 24]. Consistent with its documented oncogenic role across both hematological and solid tumors [25–27], our integrated analysis of TCGA and GTEx data revealed significant IL4I1 upregulation in AML compared to healthy controls (p = 0.0004). Pan-cancer analysis further indicated its elevated expression in numerous solid tumors. In univariate Cox regression analysis, high IL4I1 expression was associated with adverse cytogenetic profiles, advanced age, and poorer overall survival (HR = 1.952, p = 0.003). However, in multivariate analysis adjusting for established prognostic factors, IL4I1 did not retain independent prognostic significance. This suggests that its prognostic association may be mediated through its correlation with high-risk disease features (e.g., AML with adverse cytogenetics and advanced age) rather than as a primary independent driver. The limited sample size and incomplete clinical annotations in the TCGA-LAML cohort may also contribute to this lack of statistical independence. Nonetheless, the consistent trend warrants validation in larger, risk stratified cohorts. Moreover, the PPI network and circos plot illustrated the complex relationships among IL4I1 and its interacting protein-encoding genes in AML. By mediating amino acid metabolism, these enzymes encoded by IL4I1-linked genes contribute to establishing an immunosuppressive metabolic network in tumors [28]. These findings suggest the potential of IL4I1 and its interaction network as therapeutic targets in AML, a possibility worthy of further exploration.

IL4I1 in immune and metabolic reprogramming

Our functional enrichment analyses delineate a dual role for IL4I1 in immune modulation and metabolic rewiring. GSEA linked IL4I1 expression to humoral immune responses, leukocyte interactions, interleukin-10 signaling, and chemokine/cytokine pathways. Experimentally, serum levels of IL4I1, IL-10, and TGF-β1 were lower in IL4I1-knockdown xenograft models. This aligns with evidence that IL4I1 can synergize with IL-10 and TGF-β1 to co-regulate immunosuppressive networks [29, 30]. Concurrently, GO/KEGG analyses associated IL4I1 with the metabolic regulation of amino acids such as tryptophan, phenylalanine, and tyrosine [9], supporting its established enzymatic function in the I3P-AHR axis [7].

We further investigated the correlation between IL4I1 expression and immune infiltration. Our analysis revealed significant associations between IL4I1 levels and the abundance of various immune cells, including T-cell subsets, dendritic cells, and macrophages. This aligns with reports that IL4I1 can drive Treg proliferation [7]. Single-cell RNA sequencing localized high IL4I1 expression to bone marrow-derived monocytes in AML. Functional enrichment of these monocytes highlighted involvement in tumor-associated pathways. Key enriched processes included the regulation of apoptosis and the cell cycle—a finding subsequently validated in our cellular experiments. Additionally, pathways related to T-helper cell differentiation, antigen presentation, and innate immune signaling (such as Toll-like and NOD-like receptor pathways) were prominent. Collectively, these findings suggest IL4I1 may shape the AML immune microenvironment by influencing both adaptive and innate immunity while linking these processes to metabolic pathways.

IL4I1 as a mediator of immune suppression and macrophage polarization

IL4I1 may contribute to an immunosuppressive microenvironment by interacting with key immune checkpoint pathways. We found strong positive correlations between IL4I1 expression and immune checkpoint genes (R > 0.3), consistent with pan-cancer observations [31]. The PD-L1/PD-1 axis is a critical mechanism for immune evasion in AML [32], and our data suggest IL4I1 may be part of this regulatory network. For patients with high-risk, refractory AML harboring TP53 mutations, emerging strategies that combine immune checkpoint inhibitors with other agents are being explored [33], showing encouraging potential. Furthermore, IL4I1 is implicated in T-cell exhaustion, a process driven by persistent antigen exposure and cytokines like IL-10 and TGF-β [34]—pathways in which IL4I1 is involved according to our GSEA.

An attractive finding by us is the role of IL4I1 in polarizing tumor-associated macrophages toward an M2-like, pro-tumor phenotype in AML. IL4I1 expression correlated with M2 markers (TGFB1, IL10, CD163), and its knockdown in vitro suppressed M2 polarization, reducing marker expression and cytokine secretion. This aligns with its documented role in M2 polarization in glioma [5] and clear cell renal cell carcinoma [35]. Taken together, IL4I1 appears to foster an immunosuppressive microenvironment in AML by enhancing checkpoint expression, promoting T-cell dysfunction, and skewing TAMs toward an M2-like phenotype.

Therapeutic implications and future directions

The treatment landscape for AML has evolved with novel targeted agents, enabling a more personalized treatment approach [36]. However, drug resistance and relapse remain major challenges [1]. Immunotherapy offers promise, yet its application in AML is limited by disease heterogeneity, an immunosuppressive bone marrow niche, and a lack of leukemia-specific antigens [37, 38]. Functionally, our in vitro and in vivo experiments revealed that IL4I1 promotes the proliferation of leukemia cells. These findings therefore nominate IL4I1 as a candidate therapeutic target in this context. Its involvement in immune checkpoint regulation, metabolic reprogramming, and M2 macrophage polarization suggests that inhibiting IL4I1 could counteract multiple immunosuppressive mechanisms. Preclinical studies support targeting the IL4I1-AHR axis [7, 24]. Moreover, IL4I1 plays a pivotal role in mediating ibrutinib resistance in lymphoma [25]. Future research should investigate IL4I1 inhibition in AML models and explore its synergy with existing immunotherapies, such as immune checkpoint inhibitors [32], antibody-drug conjugate [39] or chimeric antigen receptor T-cell (CAR-T) [40, 41]. Prioritizing the characterization of the IL4I1-high tumor microenvironment, developing targeted inhibitors, and identifying predictive biomarkers will be crucial next steps.

Limitations

We acknowledge that this study has certain limitations. First, the retrospective nature and sample size of the TCGA cohort may introduce bias. Second, while we identified associations between IL4I1 and immune processes, the precise mechanistic interactions require further experimental dissection. Third, we did not stratify analyses by specific molecular subtypes due to data constraints, which may affect the generalizability of our findings. Addressing these in future studies will clarify the translational potential of IL4I1-directed strategies.

Conclusions

In summary, elevated IL4I1 expression is markedly associated with unfavorable prognosis in AML patients and may contribute to promote leukemia cell proliferation while inhibiting apoptosis. Additionally, IL4I1 may disrupt immune homeostasis within the tumor microenvironment. The findings underscore the potential role of IL4I1 as both a prognostic biomarker and a target for immune-based therapeutic strategies in AML.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

IL4I1

Interleukin-4-induced 1

AML

Acute myeloid leukemia

TCGA

The Cancer Genome Atlas

GTEx

Genotype-Tissue Expression

GSEA

Gene Set Enrichment Analysis

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

GSVA

Gene Set Variation Analysis

CR

Complete remission

DFS

Disease-free survival

R/R AML

Relapsed/refractory acute myeloid leukemia

AHR

Aryl hydrocarbon receptor

I3P

Indole-3-pyruvic acid

IDO1

Indoleamine 2,3-Dioxygenase 1

TDO2

Tryptophan 2,3-Dioxygenase 2

DEGs

Differentially expressed genes

ESTIMATE

Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data

RT-qPCR

Reverse Transcription Quantitative PCR

WB

Western Blot

KM

Kaplan-Meier

PPI

Protein-Protein Interaction

TAMs

Tumor-associated macrophages

IC

Immune checkpoint

TISCH

Tumor Immune Single- cell Hub

CAR-T

Chimeric Antigen Receptor T-Cell Immunotherapy

Author contributions

All authors contributed to the manuscript. JLH and LYY contributed to study design and manuscript writing. JYC, LYY and GLL contributed to material preparation and experiments. ZYZ, JFL, JMC, YQL, XQZ, APY, QJC, JFD, PC contributed to data collection and analysis. All authors read and approved the final manuscript.

Funding

This work was supported by the Natural Science Fundation of Fujian Province, Grant/Award Number: 2023J01598; the Joint Funds for the Innovation of Science and Technology, Fujian Province, Grant/Award Number: 2025Y9178; and the Fujian Provincial Health Technology Project, Grant/Award Number: 2023QNA024.

Data availability

The data derived and analyzed in the present research are openly accessible in TCGA database (https://genome-), GTEx database (https://www.gtexportal.org/), GSEA (https://www.gsea-msigdb.org/gsea/msigdb/collections.jsp/), STRING database (http://string-db.org/), GEO database (https://www.ncbi.nlm.nih.gov/geo/), CellMarker database (http://117.50.127.228/CellMarker/), Tumor Immune Single-cell Hub (TISCH) database (http://tisch.comp-genomics.org/). The R code for the bioinformatics analyses can be accessed in the GitHub repository at: https://github.com/olwwan/IL4I1-bioinformatic-analysis.

Declarations

Animal studies

All animal experiments were performed in accordance with institutional ethical guidelines and were approved by the Ethics Committee of the First Affiliated Hospital of Fujian Medical University.

Informed consent

N/A.

Registry and the registration No. of the study/trial

N/A.

Conflict of interest

The authors declare no competing interests.