Skip to main content
NCBI home page
Journal of Experimental & Clinical Cancer Research : CR logoLink to Journal of Experimental & Clinical Cancer Research : CR
. 2026 Feb 3;45:61. doi: 10.1186/s13046-026-03653-2

AKR1B10 reprograms neutrophils by histone lactylation to foster immune evasion in KRASG12C mutation colorectal cancer liver metastasis

Weiwei Li 1,#, Wenkang Yuan 1,#, Zihao Du 1,#, Xiangyu Wang 3,#, Daoyue Wang 5,6, Songlin Xing 1, Tingting Shen 7, Canliang Lu 1, Jiale Chen 1, Anhai Yu 1, Xinyu Jiang 1, Shiwei Zhang 1, Shuhao Zheng 1, Xiaowen Feng 2,4, Tianqi Wang 1, Jieliang Zuo 8, Jinhong Chen 3,✉, Chao Zhang 1,✉, Xuefu Wang 2,4,✉, Chong Zhang 1,2,✉
PMCID: PMC12955136  PMID: 41630031

Abstract

Background

The KRASG12C mutation is one of the special mutation types in patients with colorectal cancer liver metastasis (CRLM). Although several small molecule inhibitors specifically targeting KRASG12C mutation have been developed, they have only shown limited clinical benefits for CRLM patients. Thus, alternative approaches are still needed.

Methods

We screened out the differentially expressed gene Aldo–keto reductase family 1 member B10 (AKR1B10) between the KRASG12C mutation and wild-type CRLM through RNA sequencing, and characterized the tumor microenvironment (TME) changes of the KRASG12C mutation CRLM using multi-omics analysis. The role of AKR1B10 in the TME and its progression of KRASG12C mutation CRLM was confirmed by in vitro and in vivo experiments, and the molecular mechanism of lactate on neutrophils reprogramming was detected by immunofluorescence, western blot and Chip-qPCR.

Results

AKR1B10 was highly expressed in the KRASG12C mutation CRLM and was associated with a poor prognosis. Mechanistically, AKR1B10 promotes the recruitment of neutrophils in the TME by CXCL8/CXCR2 pathway. Meanwhile, AKR1B10 could promote the production of lactate by regulating crucial glycolytic enzymes. The increased lactate accumulation in the TME promoted histone lactylation of neutrophils, which induced PD-L1 transcription and prompted the reprogramming of neutrophils to an immunosuppressive phenotype.

Conclusion

AKR1B10 facilitated immune evasion of KRASG12C mutation CRLM by recruiting and reprogramming neutrophils to remodel the immunosuppressive TME, providing a potential therapeutic target for KRASG12C mutation CRLM patients.

Graphical Abstract

graphic file with name 13046_2026_3653_Figa_HTML.jpg

Supplementary Information

The online version contains supplementary material available at 10.1186/s13046-026-03653-2.

Keywords: KRASG12C mutation, Colorectal cancer liver metastasis, Tumor microenvironment, Neutrophils infiltration, Lactylation

Background

Colorectal cancer (CRC) is the third most prevalent cancer worldwide and the second leading cause of cancer-related deaths globally [1]. Distant metastasis is the primary reason for the high mortality rate among CRC patients, with approximately 70% of them developing colorectal cancer liver metastasis (CRLM) [2]. The KRAS gene is one of the most frequently mutated oncogenes in CRC, and the KRASG12C mutation, a specific subtype, is responsible for about 3% to 4% of CRLM cases [3]. Despite advancements in systemic treatment and surgical techniques, the prognosis for KRAS mutation CRLM patients remains poor [4]. Particularly, KRAS mutations decrease sensitivity to immunotherapies, such as anti-PD-1 treatment, and further limit the range of drug treatment options for patients [5]. Therefore, it is extremely urgent to explore the potential mechanisms of KRAS mutation CRLM patients and develop effective treatment strategies.

In recent years, the mechanism by which KRAS mutations reshape the immune landscape of the tumor microenvironment (TME) has drawn significant study attention [6]. However, in the specific TME of liver metastasis from CRC, the effects of KRASG12C mutations on regulating the composition of immune cells and the progression of tumor metastasis, as well as treatment of tumor response, remains unclear. Several studies have indicated that in cases with KRASG12C mutations, the infiltration of CD8+ T cells within metastatic tumor is reduced, whereas the infiltration of neutrophils is increased [7, 8]. As a principal constituent of the TME, CD8+ T cells have been verified by our previous study to be correlated with a favorable prognosis in patients with CRLM [9]. As the most abundant type of innate immune cells, neutrophils could rapidly migrate to the TME once recruited by chemokines [10]. Previous studies have demonstrated that neutrophils possess dynamic functional plasticity and can be reprogrammed via tumor-derived signals. Tumor-associated neutrophils could inhibit the function of T cells in a PD-L1 dependent fashion and promote tumor growth and metastasis by enhancing immunosuppressive TME [11, 12]. Meanwhile, in our previous study, we revealed that high neutrophil-to-lymphocyte ratio in CRLM patients was closely related to poor long-term survival [13]. However, there is still a lack of relevant study on whether intrinsic genetic or epigenetic alterations in liver metastasis of KRASG12C mutation CRLM could recruit and reprogram neutrophils.

Aldo–keto reductase family 1 member B10 (AKR1B10) mainly expressed in the gastrointestinal system and functions as a reductase to exert cellular detoxification effects [14]. Furthermore, AKR1B10 is involved in metabolic pathways and affects immune responses and inflammatory diseases [15]. Some studies have also reported the important role of AKR1B10 in regulating the activity of mutant KRAS [16]. The significant function of AKR1B10 in malignant tumors has also garnered extensive attention. It is commonly posited that AKR1B10 is significantly up-regulated in malignant tumors including hepatocellular carcinoma (HCC) [17], cholangiocarcinoma [18], and lung cancer [19], while it is paradoxically down-regulated in gastric cancer (GC) and CRC [15]. In contrast, there is widespread controversy regarding its role in CRC. It is worth noting that studies have shown that the expression of AKR1B10 is relatively low in the early stage of CRC, but up-regulated significantly as the cancer progresses [20]. Currently, the function of AKR1B10 in cellular metabolism and its potential correlation with the progression of KRASG12C mutation CRLM remain ambiguous.

In this study, we uncovered that the AKR1B10 level is up-regulated in the KRASG12C mutation CRLM, and AKR1B10 recruit neutrophils by up-regulating the CXCL8 chemokine. Importantly, the lactate accumulation induced by AKR1B10 increased the histone lactylation level of PD-L1 promoter and consequently promoted PD-L1 transcription in neutrophils. Overall, our studies revealed the molecular mechanism by which AKR1B10 promotes immune escape by enhancing the recruitment and reprogramming of neutrophils and inhibiting the cytotoxicity of CD8+ T cells, providing a potential therapeutic strategy for patients with KRASG12C mutation CRLM.

Materials and methods

Cell culture and generation of the KRASG12C mutation cells

CT26 was purchased from the American Type Culture Collection. The cell line was identified by STR analysis and confirmed to be negative for mycoplasma contamination. KRASG12C mutation cells were constructed from wild-type CT26 cells expressing KRAS using CRISPR-Cas9 RNP and Oligo (provided by Haixing Bioscience) containing expression cassettes for hSpCas9 and chimeric guide RNA. Briefly, two guide RNA sequence of ATGGTTGGAGCTGATGGCGTAGG and CTTGTGATGGTTGGAGCTGATGG were selected through the http://crispr.mit.edu website. According to the manufacturer's instructions (ThermoFisher Scientific), CRISPR-Cas9 RNP and Oligo containing guide RNA sequences were electrotransfected into cells using Neon™ transfection system. To determine the presence of mutations in KRASG12C gene targeted clones, the genomic sequences near the gRNA target sites were sequenced (Figure S1). All cells were cultured in RPMI 1640 medium supplemented with 10% FBS (Gibco) and 1% penicillin and streptomycin (Beyotime) at 37 ℃ in 5% CO2.

Patients and tissue samples

The liver metastasis tissues of CRLM were obtained from patients who underwent radical surgery at the First Affiliated Hospital of Anhui Medical University and Huashan Hospital Fudan University. KRAS G12C status was determined by Next-Generation Sequencing. The clinicopathological characteristics of CRLM patients are presented in Table S1. All patients underwent routine laboratory tests of peripheral blood and imaging examinations (CT scan or MRI). The total tumor volume (TTV) changes in response to immunotherapy were calculated through CT scans before and after receiving immunotherapy to evaluate its therapeutic effect. Briefly, TTV was determined as the product of the voxel volume of all CRLM present in the liver and the number of their segmented voxels [21]. This study strictly adhered to the ethical review requirements and was all approved by the ethical review committees of each institution. All subjects signed written informed consent forms, and this procedure fully complies with the ethical norms stipulated in the research protocol.

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA of various types of cells and tissues was isolated by TRIzol reagent (Invitrogen, 15,596–018). The cDNA synthesis was performed with PrimeScript RT Master Mix (Takara, RR036). qRT-PCR was conducted using TB Green Premix Ex Taq (Takara, RR820) in LightCycler 480 System (Roche, Switzerland). The primer sequences are shown in the Table S2.

Western blot

Tissues and cells were lysed in lysis buffer supplemented with PMSF (1 mM) and protease and phosphatase inhibitors. Proteins were separated by 10% or 12.5% SDS-PAGE and transferred on nitrocellulose membranes. After incubated with primary antibodies overnight, the membranes were incubated with secondary antibodies for 1 h. The protein complexes were visualized with enhanced chemiluminescence reagents. The list of antibodies are shown in Table S3.

Short hairpin RNA (shRNA) construction and transfection

The shRNA targeting AKR1B10 (shAKR1B10#1, shAKR1B10#2 and shAKR1B10#3) and negative control (shNC) were inserted into the lentiviral vector. Transfection was carried out as directed by the manufacturer. Briefly, tumor cells were spread at a density of 50%. After the cells were incubated overnight and adhered, lentivirus culture medium was added to the plate for 48 h. Then, the infected cells were transferred to fresh medium containing 5 µg/mL puromycin for the selection of stable clones. The sequences of shRNAs are shown in Table S2.

Animal model induction and treatment

All animal experiments were performed on 5–8 weeks male BALB/c mice (GemPharmatech Co.,Ltd., Jiangsu, China). After inhalation anesthesia with isoflurane, 1 × 106 luciferase-labeled tumor cells in 100 µL PBS were injected into the spleen of each mice to establish the liver metastasis model. Liver metastasis was monitored in vivo by bioluminescence imaging weekly for 4 weeks. Briefly, mice were anesthetized and intraperitoneally injected with D-Luciferin (150 mg/kg body weight, Beyotime, China) and then images were obtained using the IVIS Tanon machine. The mice were randomly assigned to the designated treatment group after confirmation of liver metastasis. Treatment of mice included: IgG isotype control), 2 µg intraperitoneal (i.p.) every 2 days; anti-CXCL8 antibodies (Selleck, A2524), 1 mg i.p. every 2 days; CXCR2 inhibitor, SB225002 (Selleck, S7651), 10 mg/kg i.p. every 2 days; anti-Ly6G antibodies (Selleck, A2158), 5 mg/kg i.p. every 2 days; anti-CD8 antibodies (Selleck, A2102), 5 mg/kg i.p. every 2 days; anti-PD-1 antibodies (Selleck, A2122), 10 mg/kg i.p. every 3 days; daily gavage with 0.5% methylcellulose (vehicle group); AKR1B10 inhibitor, Epalrestat (Selleck, S2035) diluted in 0.5% methylcellulose, 50 mg/kg intragastrical (i.g.) every day; and combined therapies: anti-Ly6G antibodies + anti-CD8 antibodies; anti-Ly6G antibodies + anti-PD-1 antibodies and Epalrestat + anti-PD-1 antibodies. Four weeks after tumor cells injection, mice were euthanized and subsequent experiments were conducted. All experimental procedures have been approved by the Animal Ethics Committee of Anhui Medical University.

Flow cytometry

To analyze immune cells infiltrating the tumor tissue, tumors were dissected and cut into small pieces, and then shaken and digested at 37 ℃ for 30 min in RPMI 1640 medium containing 2% FBS, 0.5 mg/ml collagenase Ⅳ (Sigma, C5138), and 0.5 µg/mL DNase I (Roche). After filtration with 70 µm cell strainers, the single cell suspensions were used for further flow cytometry. For cell surface proteins staining, single cell suspensions were incubated with indicated antibodies for 30 min at 4 ℃ in the dark. For intracellular cytokine staining, cells were cultured in RPMI 1640 complete medium and stimulated with monensin (5 µg/mL), Ionomycin (1 µg/mL) and PMA (30 ng/mL) for 4 h. Subsequently, the cultured cells were collected, washed and stained with surface markers at 4 ℃ for 30 min. Subsequently, after fixation and permeability (eBioscience, 88–8824-00) for 30 min, the cells were incubated with antibodies against intracellular cytokine markers at 4 ℃ for 30 min. The list of antibodies is presented in Table S4.

Chromatin immunoprecipitation (ChIP)-qPCR

ChIP assay was performed using SimpleChIP Enzymatic Chromatin IP kit (Cell Signaling Technology, 9003S). Briefly, freshly isolated neutrophils were cultured at a density of 3 × 106 cells in 6-well plates with or without additional lactate for 24 h. Subsequently, cross-linking of proteins-DNA was carried out with 37% formaldehyde for 10 min and quenched with 125 mM glycine. Chromatin was sonicated to appropriate fragments with the use of sonication, and then incubated with corresponding ChIP grade antibodies or anti-IgG (Cell Signaling Technology, 3900) antibodies overnight at 4 ℃ with rotation. The samples then incubated with protein G magnetic beads, followed by washing and elution. qRT-PCR was performed to detect eluted target protein-bound DNA. Primer sequences are listed in the Table S2.

Hematoxylin & eosin (H&E) and Immunohistochemical (IHC) staining

H&E staining was performed as described previously. Briefly, the liver metastasis tissues of CRLM were fixed with 10% neutral buffered formalin and embedded in paraffin. Then, paraffin sections with a thickness of 4 µm were dewaxed, hydrated, and then stained with hematoxylin for 5–8 min and counterstained with eosin for 2–3 min for H&E staining. The area of liver metastatic burden was determined on H&E stained slides. Liver metastatic burden was calculated as the percentage metastatic area of total liver area evaluated. IHC staining was performed on the formalin-fixed and paraffin-embedded CRLM tissue sections. Briefly, the sections were baked at 65 ℃ for 2 h and then dewaxed and hydrated. Antigen retrieval was performed in microwave with citrate buffer (PH = 6.0) for 30 min and endogenous peroxidase activity was blocked using 3% H2O2 for 15 min. Subsequently, the sections were blocked with 5% bovine serum albumin (BSA). The sections were incubated with primary antibodies at 4 ℃ overnight and then with appropriate secondary antibodies labeled with horseradish peroxidase (HRP) followed by 4,6-diamino-2-phenylindole (DAPI) staining for nuclei visualization. For quantification, the sections were evaluated by two independent pathologists who had no knowledge of the patient's clinical information. The Information of corresponding antibodies are listed in Table S3.

Multiplex IHC (mIHC) and immunofluorescence (IF) staining

To evaluate the localization and abundance of multiple immune cells in CRLM tissues, mIHC was performed in serial paraffin sections from each CRLM patient using the Opal 7-Colour IHC Kit (PerkinElmer) as previously described [9]. Briefly, after dewaxing, hydration, antigen retrieval, and endogenous peroxidase activity blocking treatment, sections were incubated overnight with two panels of primary antibodies, including CK, CD8, CD4, FoxP3, PD-1 and CD20; CK, CD3, CD68, CD56, PD-L1 and CD66b. Then, the sections were incubated with corresponding HRP-conjugated secondary antibodies and dyes with different fluorescence signals (Opal540, Opal570, Opal620, Opal650 and Opal690), and the nuclei were stained with DAPI.

For cells IF, the freshly isolated neutrophils were cultured on sterile glass slides treated with polylysine for 24 h under different conditions. Subsequently, the cells were fixed with 4% paraformaldehyde. For tissues IF, multiple serial paraffin sections with a 4 μm thickness were used for subsequent processing. Similar to the IHC staining of the tissue, the serial paraffin sections underwent dewaxing, hydration, antigen retrieval and subsequent endogenous peroxidase activity blocking treatments. After being blocked with 5% BSA, the tissues samples were incubated with primary antibodies overnight at 4 ℃. Then the samples were incubated with the corresponding HRP-conjugated secondary antibodies and fluorescent dyes for 1 h, followed by DAPI staining for nuclei visualization. Fluorescent images were taken with a fluorescent microscope. The Information of corresponding antibodies are listed in Table S3.

Patient derived tumor organoids (PDTOs) culture

The tumors were obtained by surgical resection and further organoid culture was carried out within 24 h. Briefly, tumor tissue of liver metastasis from CRC were minced into small pieces and incubated in a digestive medium containing RPMI 1640 medium with collagenase IV (Sigma‐Aldrich, C9407), hyaluronidase (Solarbio, h8030), DNase I, RHOK inhibitor Y‐27,632 (Sigma‐Aldrich, Y0503) and penicillin–streptomycin on an orbital shaker for 30 min at 37 ℃. After digestion, the dissociated tissues were centrifuged and resuspended by TrypLE Express (Gibco, 12,604,013) and then filtered through a 100 μm cell filter to dissociate into single cells. The isolated cells were embedded in Matrigel and plated in a 24-well plate. After the Matrigel balls were polymerized, 1 mL of organoid culture medium was added for cultivation, and fresh medium was replaced every 3 days.

Statistical analysis

Data were presented as individual data points and the mean ± SD except were stated otherwise. To evaluate the significance between two groups, Student's t tests or Wilcoxon signed rank tests were employed. In instances involving multiple groups, statistical analysis utilized either one-way or two-way analysis of variance. Overall survival (OS) and disease-free survival (DFS) were analyzed using the Kaplan–Meier analysis, and correlation analyses were performed using Pearson’s correlation tests. GraphPad Prism 9 or R (version4.5.0) were used for all statistical analyses, and the statistical significance was as follows: ns, not significant; *P < 0. 05; **P < 0. 01; ***P < 0.001; ****P < 0.0001.

Results

AKR1B10 correlates with poor prognosis of KRASG12C mutation CRLM

To explore the correlation between KRASG12C mutation and survival prognosis of CRLM patients, a total of 152 KRASG12C mutation and 277 wild-type CRLM patients were included in this study. We found that KRASG12C mutation was positively correlated with poorer DFS and OS of CRLM patients (Fig. 1A-B). To identify potential therapeutic targets for KRASG12C mutation CRLM patients, we collected liver metastatic tumor tissues from KRASG12C mutation and wild-type CRLM patients, and screened for differentially expressed genes through RNA-seq. The results indicated that the expression of AKR1B10 was significantly up-regulated in the liver metastatic tumor tissues of KRASG12C mutation CRLM (Fig. 1C). The subsequent qRT-PCR experiment further verified this result (Fig. 1D). Meanwhile, compared to the matched adjacent tissues, the level of AKR1B10 protein was significantly up-regulated in liver metastatic tumor tissues, especially in the KRASG12C mutation CRLM (Fig. 1E). Furthermore, IHC staining revealed that in liver metastatic tumor tissues, the level of AKR1B10 in the KRASG12C mutation group was significantly higher than that in the wild-type group, while there was no significant difference in the corresponding adjacent normal tissues (Fig. 1F-H).

Fig. 1.

Fig. 1

Open in a new tab

AKR1B10 up-regulated in KRASG12C mutation CRLM correlates with poor prognosis. A, B Kaplan–Meier diagrams illustrating DFS and OS of patients with KRASG12C mutation or wild-type CRLM. C Volcano plots represent the differentially expressed genes in RNA-seq analysis of KRASG12C mutation and control wild-type CRLM liver metastatic tumor tissues. D, E Protein and mRNA levels of AKR1B10 in KRASG12C mutation and KRAS wild‐type CRLM samples. F AKR1B10 immunostaining within tumor tissue and adjacent normal tissues in a representative human CRLM case carrying KRASG12C mutation compared to a KRAS wild‐type case. G, H Quantification of AKR1B10 within tumor tissues and adjacent normal tissues in KRASG12C mutation (n = 152 patients) and KRAS wild‐type cases (n = 277 patients). I Kaplan–Meier diagrams illustrating DFS and OS of patients with KRASG12C mutation or wild-type CRLM in different AKR1B10 expression groups. J Multivariate regression analyses forest plot for CRLM patients with KRASG12C mutation. K Representative CT scans of target lesions pre- and post-treatment and corresponding TTV changes in patients with CRLM expressing different AKR1B10. L, M Protein and mRNA levels of AKR1B10 in CT26K and CT26 cells. N, O Representative fluorescence images (N), gross images and HE images (O) of liver metastasis in the CT26K and CT26 groups of CRLM models. P, Q Liver weight (P) and percentage of liver metastasis foci in the liver (Q) of CT26K and CT26 groups mice. R, S Protein and mRNA levels of AKR1B10 in CT26K and CT26 groups of CRLM models tumor samples

To evaluate the relationship between AKR1B10 and prognosis of patients, liver metastasis samples from KRASG12C mutation and wild-type CRLM patients were classified into high-expression AKR1B10 group and low-expression AKR1B10 group, respectively, according to the intensity of IHC staining. The Kaplan–Meier analysis indicated that the DFS and OS of CRLM patients in the AKR1B10 high-expression group were lower than those in the low-expression group (Fig. 1I). In addition, the expression of AKR1B10 was closely associated with number of liver metastases, tumor grade and the pN stage (Table S5). Multivariate COX regression analyses demonstrated that AKR1B10 serves as an independent prognostic factor for KRASG12C mutation CRLM (Fig. 1J). Furthermore, we collected and compared the CT result of patients with CRLM before and after undergoing immunotherapy, and found that patients with high expression of AKR1B10 exhibited a relatively larger TTV and less favorable treatment outcomes (Fig. 1K).

Then to clarify the relevant mechanism by which KRASG12C mutation promotes the progression of CRLM, the cDNA expressing KRAS G12C was stably transfected into the CT26 cell line expressing wild-type KRAS to establish cells expressing KRAS G12C (designated as CT26K). Meanwhile, CT26 cells transfected with the corresponding empty vector were utilized as the control (hereinafter referred to as CT26). We demonstrated that the level of AKR1B10 protein and mRNA in CT26K cells was elevated compared to those in CT26 cells (Fig. 1L-M). Subsequently, we generated a CRLM model using BALB/c mice, and found that compared to the CT26 group, the CT26K group notably facilitated liver metastasis (Fig. 1N-Q). Compared to the CT26 group, the mRNA and protein levels of AKR1B10 in liver metastases in the CT26K group were increased, which consisted with IHC results (Fig. 1R-S). Together, these findings suggested that AKR1B10 participated in the progression of KRASG12C mutation CRLM.

AKR1B10 generates an immunosuppressed TME that drives CRLM progression

To investigate the immune heterogeneity between KRASG12C mutation and wild-type CRLM, we further performed cluster analysis focusing on immune cells in the single-cell sequencing database. During this process, 4,789 qualified cells were identified in KRAS mutation CRLM case (Figure S2, S3) and 6,615 qualified cells in wild-type CRLM case (Figure S4, S5). The major cell populations were revealed by canonical marker genes examination (Figure S6-S9). Compared with patients with wild-type CRLM, the presence of KRAS mutation was significantly associated with an increased enrichment level of neutrophils (Fig. 2A-C). Subsequently, we conducted a comparative analysis of the differences in immune infiltration from bulk RNA-seq data of 126 KRAS mutation and 157 wild-type CRLM patients. Similar to the results of single-cell cluster analysis, the infiltration of neutrophils in liver metastases of KRAS mutation CRLM cases was increased compared to wild-type cases (Figure S10A-B). And worthly, the pearson analysis revealed a significant negative correlation between CD8+ T cells and neutrophils in KRAS mutation CRLM samples (Figure S10C).

Fig. 2.

Fig. 2

Open in a new tab

AKR1B10 generates an immunosuppressed TME that drives CRLM progression. A, B Uniform manifold approximation and projection (UMAP) plot showing the major cell types in KRAS mutation (A) and wild-type (B) CRLM. Clusters denoted by color were labeled with inferred cell types. Single cell data from HRR066493 and HRR066502. C Proportion bar graphs showing the population proportions of the indicated cell 10 types. D The immune cell composition of TME was detected using mIHC in CRLM tissue samples using two panels of 11 antibody markers for further analysis. E Representative mIHC images of two panels in KRASG12C mutation and wild-type CRLM tissue samples. Scale bar, 50 µm. F Quantification of the density of CD8+ T cells and CD66b+ neutrophils in KRASG12C mutation and wild-type CRLM cases. G Pearson's correlation analysis between CD8+ T cells density and CD66b+ neutrophils density in CRLM tissues. H Representative plots and cell proportion of neutrophils and CD8+ T cells in shNC and shAKR1B10 groups of liver metastatic tumors. I-K Representative plots and statistical analysis of surface PD-1, TIM-3 (I), and IFNγ and TNFα production (J) and CD69 (K) of CD8+ T cells from the indicated mice. L PD-L1 levels on tumor cells and tumor-infiltrating neutrophils from the indicated mice

To further corroborate our findings, we detected liver metastatic tumor tissue samples from 20 patients with KRASG12C mutation and 20 patients with wild-type CRLM by using mIHC staining (Fig. 2D). Seven immune cell markers (CD4+ T cells, CD8+ T cells, FOXP3+ Treg cells, CD20 + B cells, CD56 + NK cells, CD66b + neutrophils and CD68 + macrophages) were simultaneously detected in adjacent sections of CRLM tumor tissues (Fig. 2E). We observed that compared to the wild-type, the TME of KRASG12C mutation CRLM showed a significant reduction in CD8+ T cells infiltration, while CD66b+ neutrophil infiltration increased (Fig. 2F). In addition, the infiltration of CD20+ B cells, CD56+ NK cells and CD68+ macrophages were also relatively reduced in KRASG12C mutation CRLM tumors (Figure S11A-E). Correlation analysis revealed that CD8+ T cells density was inversely correlated with CD66b+ neutrophils density in KRASG12C mutation CRLM (Fig. 2G).

To elucidate the biological effects of AKR1B10 on CRLM, we stably knocked down AKR1B10 in CT26K cells using three different shRNAs, and selected shAKR1B10#1 for subsequent experiments (Figure S12A-B). CCK-8 assays showed AKR1B10 knockdown significantly reduced the cell proliferation ability of CT26K cells (Figure S12C). Meanwhile, wound-healing and transwell assays revealed that AKR1B10 facilitated the migration and invasion of CT26K cells (Figure S12D-E). Subsequently, to investigate the crucial role of AKR1B10 in CRLM, we established the CRLM mouse model using control CT26K cells (shNC) and AKR1B10-knockdown CT26K cells (shAKR1B10), respectively. Compared to the shNC group, knockdown of AKR1B10 in CT26K cells significantly inhibited liver metastasis (Figure S12F-G). Consistent with these results, the liver weight and the proportion of liver metastasis also decreased (Figure S12H-I).

In light of the substantial disparities in the relative proportions of CD8+ T cells and neutrophils within liver metastatic tumor tissues between KRASG12C mutation and wild-type CRLM patients, we further employed flow cytometry to detect CD8+ T cells and neutrophils in the liver metastatic tumor tissues of mice in the shNC group and the shAKR1B10 group, and found that the knockdown of AKR1B10 augmented the CD8+ T cells proportion in liver metastatic tumors and concurrently diminished the infiltration of neutrophils within the tumors (Fig. 2H). In addition, knockdown of AKR1B10 also reduced the PD-1 and TIM-3 expression on CD8+ T cells in liver metastatic tumors, accompanied by an increase in the production of IFNγ and TNFα (Fig. 2I-J). However, this process did not affect the activation level of CD8+ T cells, and there was no significant difference in the percentage of CD69+ CD8+ T cells between the two groups (Fig. 2K). Notably, we observed that the expression of PD-L1 on neutrophils in liver metastatic tumors of the shNC group was higher than that of the shAKR1B10 group. Moreover, the PD-L1 expression on neutrophils was significantly higher than that on tumor cells (Fig. 2L). Accordingly, these results indicated that AKR1B10 derived tumor progression by reprogramming the immunosuppressive TME of KRASG12C mutation CRLM via influencing CD8+ T cells and neutrophils.

AKR1B10 promotes neutrophil recruitment through the CXCL8/CXCR2 axis

To explore the molecular mechanism of AKR1B10 mediated immunosuppression, RNA-seq was performed on CT26K cells with AKR1B10 knockdown. KEGG enrichment analysis indicated that the pathway with the highest enrichment score was the down-regulated cytokine-cytokine receptor interaction pathway (Fig. 3A). Subsequently, qRT-PCR experiments were conducted on the differential genes with the most significant changes among them. It was found that knockdown of AKR1B10 resulted in the most obvious down-regulation of CXCL8, followed by TGFβ and CXCL1 (Fig. 3B). Furthermore, neutrophil chemotaxis assay also demonstrated that CT26K cells significantly weakened their ability to actively recruit neutrophils after knockdown of AKR1B10 (Fig. 3C).

Fig. 3.

Fig. 3

Open in a new tab

AKR1B10 promotes neutrophil recruitment through the CXCL8/CXCR2 axis. A CT26K cells with AKR1B10 knockdown were analyzed via RNA-seq and KEGG enrichment analysis. B RT-qPCR detected expression of the 6 most pronounced differential genes in the cytokine pathway. C Representative plots and statistical analysis of neutrophil chemotaxis assay. D Schematic diagram of the establish of the CRLM models and the corresponding treatment. E, F Representative fluorescence images (E) and statistical analysis (F) of different CRLM model groups. G Representative gross images and HE images of different CRLM model groups. H, I Liver weight (H) and percentage of liver metastasis foci in the liver (I) of different CRLM model groups. J, K Representative plots and cell proportion of CD8+ T cells (J) and neutrophils (K) of different CRLM model groups. L-N Representative plots and statistical analysis of IFNγ production (L), and surface PD-1 (M), TIM-3 (N) of CD8+ T cells of different CRLM model groups. O Representative plots and statistical analysis of surface PD-L1 of neutrophils of different CRLM model groups

We further discovered that CXCL8/CXCR2 are the top five communication axes of typical interactions between tumor cells and neutrophils in KRAS mutation CRLM (Figure S13A). In addition, for CXCL8, we conducted numerous matching screenings in neutrophils to search for various types of receptors that are compatible with its expression. The results also verified the interaction between CXCR2 and CXCL8 in neutrophils and tumor cells (Figure S13B). To further investigate the specific neutrophil subpopulations that interaction with tumor cells through the CXCL8/CXCR2 axis during the liver metastasis process of KRAS mutation CRLM, we identified four distinct clusters (neutrophil_1-4) based on the scores of the gene sets (Figure S14A). Notably, the neutrophil_3 was found to be significant involved in the CXCL8/CXCR2 communication between tumor cells and neutrophils (Figure S14B-D). Moreover, we found that the PD-L1 expression was higher in neutrophil_3 (Figure S14E-F). Neutrophil _3 has also been found to have the highest interaction strength with CD8+ T cells through the PD-1/PD-L1 axis (Figure S14G). Together, these results indicated neutrophil_3 in KRAS mutation CRLM patients to be a new immunosuppressive subpopulation that may potentially mediate tumor resistance to anti-PD-1 therapy.

To further determine the neutrophil recruitment effect of CXCL8 during liver metastasis, we conducted functional experiments using the anti-CXCL8 neutralizing antibodies (Anti-CXCL8) and CXCR2 inhibitor (SB225002) (Fig. 3D), and found that blocking either CXCL8 or CXCR2 could reduce the burden of liver metastatic tumors. Nevertheless, compared to the shAKR1B10 group, the shNC group mice were more significantly affected by the progression of liver metastasis after being treated with anti-CXCL8 or SB225002, resulting in a significant reduction in the proportion of liver weight and tumor area to liver metastasis (Fig. 3E-I). Moreover, anti-CXCL8 and SB225002 largely increase CD8+ T cells infiltration and reduce neutrophil infiltration (Fig. 3J-K). It is worth noting that the impaired cytotoxicity of CD8+ T cells in liver metastatic tumors could be partially rescued by administration of Anti-CXCL8 or SB225002 (Fig. 3L-N). However, there was no significant difference in the expression of PD-L1 on neutrophils among each group (Fig. 3O). Collectively, AKR1B10 could promote neutrophils recruitment via the CXCL8/CXCR2 axis, consequently suppressing the CD8+ T cells mediated anti-tumor immune response.

Neutrophils favor immune escape by impeding CD8 + T cells

To evaluate whether the tumor-promoting function of neutrophils depends on CD8+ T cells, we used neutralizing antibody against Ly6G (Anti-Ly6G) and/or CD8 (Anti-CD8) in the CRLM mouse model (Fig. 4A). The depletion of neutrophils significantly reduced the progression of liver metastases in the shNC group, but this phenomenon could be reversed by eliminating CD8+ T cells. The effect of CD8 clearance on tumor progression in the shNC group was less pronounced compared to that in the shAKR1B10 group. This observation consistent with above discovery that cytotoxic CD8+ T cells were substantially suppressed in the tumors of the shNC group (Fig. 4B-F). These results indicated that the tumor-promoting effect of neutrophils in liver metastases of CRLM was mainly mediated by inhibiting CD8+ T cells.

Fig. 4.

Fig. 4

Open in a new tab

Neutrophils favor immune escape by impeding CD8+ T cells. A Schematic diagram of the establish of the CRLM models and the corresponding treatment with Ly6G neutralizing antibody, CD8 neutralizing antibody, the combination and IgG as a control. B, C Representative fluorescence images (B) and statistical analysis (C) of different CRLM model groups. D Representative gross images and HE images of different CRLM model groups. E, F Liver weight (E) and percentage of liver metastasis foci in the liver (F) of different CRLM model groups. G The relative strength of PD-L1/PD-1 interaction between different subpopulations of neutrophils and CD8+ T cells. H Schematic diagram of the establish of the CRLM models and the corresponding treatment with Ly6G neutralizing antibody, anti-PD-1 antibody, the combination and IgG as a control. I, J Representative fluorescence images (I) and statistical analysis (J) of different CRLM model groups. K Representative gross images and HE images of different CRLM model groups. L, M Liver weight (L) and percentage of liver metastasis foci in the liver (M) of different CRLM model groups

Subsequently, as previous results revealed that the levels of PD-L1 on neutrophils and PD-1 on CD8+ T cells in liver metastatic tumors had changed (Fig. 2L, Fig. S14G), we assumed that the PD-1/PD-L1 axis between neutrophils and CD8+ T cells may play a vital role in the dysregulation of CD8+ T cells in AKR1B10-mediated KRASG12C mutation CRLM. Next, we randomly divided mice injected with shNC and shAKR1B10 cells into four groups, including IgG group, Anti-Ly6G group, anti-PD-1 antibody (Anti-PD-1) group and Anti-Ly6G plus Anti-PD-1 group, respectively (Fig. 4G), and found that compared to the IgG group, fewer metastatic lesions were observed in the Anti-Ly6G group and Anti-PD-1 group. Notably, mouse treatment with Anti-Ly6G plus Anti-PD-1 could further inhibit tumor metastasis (Fig. 4H-L).

AKR1B10 promotes lactate and expression of PD-L1 on neutrophils through H3K9/H3K18 lactylation

Lactate could induce the PD-L1 expression on monocytes and play a vital role in the progression of various cancers [22]. Meanwhile, previous studies have confirmed that AKR1B10 could promote acquired resistance to pemetrexed in lung cancer-derived brain metastasis by enhancing glycolysis and lactate formation [23]. Therefore, we focused on the relationship between AKR1B10 and lactate. The results indicated that after knocking down the AKR1B10, the lactate concentration in the tumor conditioned medium decreased significantly (Fig. 5A). In addition, in the mouse CRLM model, the lactate level in the liver metastatic tumor tissues of the shAKR1B10 group was also significantly lower than that of shNC group (Fig. 5B). Subsequently, qRT-PCR demonstrated that AKR1B10 knockdown could decrease the mRNA expression of certain crucial glycolytic enzymes, including glucose transporter 1 (GLUT1), hexokinase 2 (HK2) and lactate dehydrogenase A (LDHA) (Fig. 5C).

Fig. 5.

Fig. 5

Open in a new tab

AKR1B10 promotes lactate and PD-L1 expression in neutrophils through H3K9/H3K18 lactylation. A, B Lactate concentrations in tumor conditioned medium (A) or CRLM model tumor tissues (B). C The mRNA levels of certain crucial glycolytic enzymes in CT26K cells changed after knockdown of AKR1B10. D Representative flow cytometry images and histogram showing the PD-L1 expression on neutrophils after stimulated by shNC and shAKR1B10 cells supernatant and supplementing additional lactate. E–G Representative flow cytometry images and histogram showing PD-L1 expression on neutrophils stimulated by supernatant derived from CT26K cells treated with or without LDHA inhibitor (E), HK2 inhibitor (F) or MCT inhibitor (G). H-I After culturing neutrophils for 24 h at different lactate concentrations, the neutrophils histone pan lactylation (Kla) (H) and site-specific histone lactylation (I) were analyzed by western blot. J-L After neutrophils were cultured for 24 h at different lactate concentrations, the neutrophils histone Kla (J), H3K9 (K) and H3K18 (L) specific histone lactylation were detected by immunofluorescence. Scale bar, 5 μm. M–N Neutrophils were cultured in culture medium supplied with or without lactate for 24 h and subjected to ChIP-qPCR using H3K9 (M) or H3K18 (N) lactylation antibody

To investigate the correlation between tumor-associated lactate and the PD-L1 expression in neutrophils, we gathered the conditioned culture media of shNC and shAKR1B10 cells for subsequent neutrophil culture. Compared to the shNC group, the PD-L1 expression on neutrophils cultured in shAKR1B10 cells conditioned medium was significantly decreased, but this phenomenon could be restored by supplementing additional lactate (Fig. 5D). Considering that the levels of LDHA and HK2 in CT26K cells were markedly down-regulated subsequent to AKR1B10 knockdown, we added GSK2837808A (an LDHA inhibitor) and Lonidamine (a HK2 inhibitor) to the cell culture medium, respectively, and the corresponding conditioned medium was collected for neutrophils culture. Similar to the previous results, the PD-L1 expression on neutrophils cultured in the conditioned medium with LDHA inhibitor or HK2 inhibitor was decreased, and this phenomenon was more obvious in the shNC group (Fig. 5E-F). Furthermore, the monocarboxylate transporter (MCT) are lactate transporter receptors on neutrophils [24]. Therefore, we applied α-cyano-4-hydroxycinnamic acid (an MCT inhibitor) to verify the effect of lactate on PD-L1, and found that MCT inhibitor could inhibit the PD-L1 expression on neutrophils induced by CT26K conditioned medium (Fig. 5G).

Since AKR1B10 could enhance lactate production in CRLM and lactylation represented a novel form of post-translational modification mediated by lactate [25], we further investigated whether lactate could drive lactation-mediated expression of PD-L1 on neutrophils. Neutrophils were cultured at various lactate concentrations, and we discovered that the additional lactate could increase the pan lactylation (Kla) of histone in neutrophils (Fig. 5H). Subsequently, we assessed the lactylation level of specific residue on histone in neutrophils. We demonstrated that additional lactate could significantly induced the lactylation levels of H3K9 and H3K18 in neutrophils (Fig. 5I). Meanwhile, IF staining analysis also supported the above results (Fig. 5J-L). Previous studies have shown that alterations in histone modifications could affect the transcriptional activation and inhibition of target genes [26]. To elucidate whether histone lactylation of neutrophils influences tumor immunity through the regulation of PD-L1 gene expression, we further conducted ChIP-qPCR experiments. We found that under the condition of high lactate level culture in neutrophils, the lactylation level of H3K9 was notably enriched in the promoter regions of PD-L1 (Fig. 5M). In addition, the enrichment of H3K18 lactylation was also increased in neutrophils (Fig. 5N). Together, these results suggested that the lactate regulates expression of PD-L1 in neutrophils through H3K9/H3K18 lactylation on the PD-L1 promoter.

AKR1B10 inhibitor suppresses liver metastasis and enhances the response of immunotherapy

Based on the key role of AKR1B10 in the TME of CRLM, we randomly divided the KRASG12C mutation CRLM mouse model into Vehicle, Epalrestat (a reported specific inhibitor of AKR1B10), Anti-PD1 and Epalrestat plus Anti-PD1 groups to further investigate its potential clinical application (Fig. 6A). As expected, both the Epalrestat group and the Anti-PD1 group partially inhibited the progression of tumor metastasis. Moreover, the treatment of epalrestat combined with anti-PD1 exhibited a strong inhibitory effect on liver metastasis, indicating a synergistic effect (Fig. 6B-H).

Fig. 6.

Fig. 6

Open in a new tab

AKR1B10 inhibitor suppresses liver metastasis and enhances the response of immunotherapy. A Schematic diagram of the establish of the CRLM models and the corresponding treatment with Epalrestat, anti-PD-1 antibody, the combination and vehicle as a control. B, C Representative fluorescence images (B) and statistical analysis (C) of different CRLM model groups. D Representative gross images and HE images of different CRLM model groups. E, F Liver weight (E) and percentage of liver metastasis foci in the liver (F) of different CRLM model groups. G, H Representative plots and statistical analysis of IFNγ production (G), and surface PD-1 (H) of CD8+ T cells of different CRLM model groups. I Schematic diagram. The liver metastases of KRASG12C mutation CRLM were minced into small pieces, cultured to form PDTOS and then co-cultured with CD8+ T cells and neutrophils isolated from peripheral blood of these patients. J H & E staining and IHC staining of Ki67, CDX2, β-catenin, pan-CK and CK20 on CRLM organoids and corresponding primary tumors. (T, primary tumors; O, CRLM organoids). K, L Representative images (K) and quantification (L) of PDTOs treated by indicated therapy

To extend these findings to patient models, we established PDTOs from tumor samples of 3 patients with KRASG12C mutation CRLM (Fig. 6I). H&E staining showed that PDTOs presented the same histopathological features and patient‐specific heterogeneous morphology as their corresponding tumors (Fig. 6J). Next, we detected the protein expressions of important molecular markers and found that the expression patterns of Ki67, β-catenin, pan-CK, CK20 and CDX2 in CRLM PDTOs and parental tumors were consistent (Fig. 6J), which was consistent with previous studies [27]. Subsequently, we co-cultured the successfully cultivated PDTOs with isolated neutrophils and CD8+ T cells in 3D culture plates (Fig. 6K). We observed that the combined administration of epalrestat and anti-PD-L1 treatment notably suppressed organoid growth compared to treatment alone (Fig. 6L).

Discussion

At present, the most widely studied mode of action of KRASG12C mutation involves the intracellular mechanism of human cancer pathogenesis, especially the emergence of the KRASG12C inhibitor Sotorasib has attracted great attention [28, 29]. However, for patients with KRASG12C mutation metastatic CRC, Sotorasib has only shown limited clinical benefits, with an objective response rate of only 7.1% [30]. There are relatively few studies on how KRASG12C mutation promote immune escape in liver metastasis of CRC. Consequently, understanding how KRASG12C mutation reprogram the TME during liver metastasis may provide another strategy for targeting KRASG12C mutation CRLM. Here, we revealed that AKR1B10 participated in the maintenance of an immunosuppressive TME in KRASG12C mutation CRLM, and that targeting AKR1B10 could reactivate anti-tumor immunity and exhibit synergistic effects with anti-PD-1 therapy to induce tumor regression.

Noteworthily, the expression of AKR1B10 exhibits variability even across distinct regions and stages of the same tumor. Several studies have demonstrated that AKR1B10 exhibits high expression level during the early to middle stages of HCC, but its expression is relatively low in advanced HCC [31]. In GC and CRC, the expression of AKR1B10 presents an upward trajectory from the early stage to the advanced stage [20, 32]. This contradictory phenomenon also exists in CRLM. Previous studies reported that AKR1B10 suppresses the metastasis of CRC through REDOX dependent post-translational modification of the PP2A/c-Myc axis [33]. However, in our study, we revealed the pro-tumor role of AKR1B10 in KRASG12C mutation CRLM. We found that AKR1B10 was significantly overexpressed in KRASG12C mutation CRLM and closely related to the poor prognosis of CRLM patients. This phenomenon may be related to the involvement of the KRASG12C mutation and the establishment of immunosuppressive TME. Previous studies have further indicated that knocking down AKR1B10 could down-regulation of active form KRAS and its downstream CRAF/MEK/ERK pathway as well as significant up-regulation of E-cadherin [16]. As an essential enzyme in the metabolism of carbonyls, retinal and farnesal/geranylgeranial, AKR1B10-mediated metabolic events played an important role in KRAS mutated tumors. It may be due to the influence of AKR1B10 on farinoid/geraniol involved in protein prenylation, while the activation of KRAS mutants requires protein prenylation [34]. In addition, we also observed that in the TME of KRASG12C mutation CRLM, the infiltration of neutrophils increased, concurrently accompanied by the reduction in an abundance of CD8+ T cells. Then, we further discovered that neutrophils were the main executors of immune escape in KRASG12C mutation CRLM. AKR1B10 inhibited the cytotoxicity CD8+ T cells by recruiting neutrophils and reprogramming neutrophils to an immunosuppressive phenotype. These results provided additional evidence regarding the manner in which epigenetic dysregulation in KRASG12C mutation tumor cells gives rise to immune evasion.

A notable feature of KRAS mutation tumors is the increased level of lactate [35]. Our study found that AKR1B10 could promote the production of lactate by regulating crucial glycolytic enzymes such as LDHA and HK2, which leads to the lactate concentration accumulation in the TME of liver metastasis from KRASG12C mutation CRC. As a metabolic byproduct, lactate plays an important role in the TME. Tumor derived lactate could prolong the lifespan of neutrophils, promote the PD-L1 expression on neutrophils [36, 37], and also sensitize tumor specific cytotoxic CD8+ T cells to activation-induced cell death in the TME [5]. Lactylation, as a newly discovered post-translational modification, directly links lactate metabolism to gene regulation [38]. Lactylation modification can facilitate tumor metabolic reprogramming and regulate the malignant phenotype of tumors by modifying key effector proteins [39]. For instance, in acute myeloid leukemia, STAT5 promotes expression of PD-L1 by promoting H4K5lac to drive immunosuppression [40]. H3K18lac has also been proven to act as a direct epigenetic transactivator for PD-L1 in HCC [41]. Nevertheless, most studies have focused on the lactylation of tumor cells, while there is still a lack of relevant studies on the effects of lactylation on immune cells within the TME, especially neutrophils. In our study, we observed that with the increase of lactate concentration, the expression level of PD-L1 on neutrophils gradually increased, accompanied by an increase in histone lactylation level on neutrophils. Subsequently, we further discovered that the PD-L1 promoter region of neutrophils was significantly enriched with lactylation of H3K9 and H3K18 sites, thereby enhancing the transcription of PD-L1 and inhibiting the function of CD8+ T cells. In addition, AKR1B10 could promote neutrophil recruitment through the CXCL8/CXCR2 pathway, thereby further facilitating the formation of neutrophil-mediated immunosuppressive TME. These discoveries clarified a novel interaction between neutrophils metabolic reprogramming and immune regulation, and provided a new approach for the remodeling of the immunosuppressive TME in KRASG12C mutation CRLM.

Conclusions

In conclusion, our study revealed that in the KRASG12C mutation CRLM, AKR1B10 directly augmented the recruitment and reprogramming of neutrophils, consequently facilitating tumor progression and immune escape through the inhibiting the cytotoxicity of CD8+ T cells. Mechanistically, we found that AKR1B10 could up-regulate the CXCL8/CXCR2 pathway to promote neutrophil recruitment. Meanwhile, AKR1B10 promoted the formation of lactate in the TME, which inducing lactylation at the H3K9 and H3K18 sites of neutrophils and enhancing the PD-L1 expression, and thereby reprogramming neutrophils to an immunosuppressive phenotype. Our study not only provides a novel molecular mechanism of KRASG12C mutation CRLM, but also offers clinical insights for potentially therapeutic inhibition of AKR1B10 and PD-1/PD-L1 axis to prevent KRASG12C mutation CRLM.

Supplementary Information

Supplementary Material 1. (2MB, pdf)

Acknowledgements

Not applicable.

Abbreviations

CRLM

Colorectal cancer liver metastasis

AKR1B10

Aldo–keto reductase family 1 member B10

TME

Tumor microenvironment

CRC

Colorectal cancer

HCC

Hepatocellular carcinoma

GC

Gastric cancer

TTV

Total tumor volume

qRT-PCR

Quantitative real-time polymerase chain reaction

ChIP

Chromatin immunoprecipitation

H&E

Hematoxylin & eosin

IHC

Immunohistochemical

BSA

Bovine serum albumin

HRP

Horseradish peroxidase

DAPI

4,6-Diamino-2-phenylindole

mIHC

Multiplex IHC

IF

Immunofluorescence

PDTOs

Patient derived tumor organoids

OS

Overall survival

DFS

Disease-free survival

GLUT1

Glucose transporter 1

HK2

Hexokinase 2

LDHA

Lactate dehydrogenase A

MCT

Monocarboxylate transporter

Authors’ contributions

The studies were designed by WW L, WK Y, ZH D and XY W. Experiments were performed by WW L, WK Y and ZH D. Data analysis was carried out by DY W, TT S, CL L, JL C, AH Y, XY J.SW Z,SL X, SH Z, XW F, TQ W and JL Z. The manuscript was written by WW L and WK Y, and revised by JH C, C Z (Chao Zhang),XF W and C Z (Chong Zhang). All authors revised and approved the manuscript.

Funding

This work was supported by the University Research Project of Anhui Province (Grant nos 2025AHGXZK40773), the National Natural Science Foundation of China (Grant nos 82303475), the Inflammation and Immune Mediated Diseases Laboratory of Anhui Province (Grant nos IMMDL202406), the Quality Enhancement Project for New Era Education of Education Department of Anhui Province (Grant nos 2024zyxwjxalk048), the National Health Commission Capacity Building and Continuing Education Center (Grant nos GWJJMB202510022056) and the University Research Project of Anhui Province (Grant nos 2022AH040161).

Data availability

The datasets supporting the conclusions of this article are included within the article and its additional files. Bulk RNA-Seq data were from GEO database, and single cell data were from The National Genomics Data Center (https://ngdc.cncb.ac.cn). Additional data related to this article will be shared upon reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

Informed consent was provided by all patients, and the Ethical Committee of the First Affiliated Hospital of Anhui Medical University (Anhui, China) (approval number: Quick-PJ 2022–14-33) and Huashan Hospital Fudan University (Shanghai, China) (approval number: KY2022-946). All mouse experiments were approved by the Animal Ethics Committee of Anhui Medical University (approval number: LLSC-20230579).

Consent for publication

All authors have agreed with publishing this manuscript.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Weiwei Li, Wenkang Yuan, Zihao Du and Xiangyu Wang contributed equally to this work.

Contributor Information

Jinhong Chen, Email: jinhongch@hotmail.com.

Chao Zhang, Email: chaozhang@fy.ahmu.edu.cn.

Xuefu Wang, Email: wangxuefu@ahmu.edu.cn.

Chong Zhang, Email: zhangchong@fy.ahmu.edu.cn.

References

  • 1.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229–63. [DOI] [PubMed] [Google Scholar]
  • 2.Chandra R, Karalis JD, Liu C, Murimwa GZ, Voth Park J, Heid CA, et al. The colorectal cancer tumor microenvironment and its impact on liver and lung metastasis. Cancers. 2021;13(24):6206. [DOI] [PMC free article] [PubMed]
  • 3.Zhu G, Pei L, Xia H, Tang Q, Bi F. Role of oncogenic KRAS in the prognosis, diagnosis and treatment of colorectal cancer. Mol Cancer. 2021;20:143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Buisman FE, Giardiello D, Kemeny NE, Steyerberg EW, Höppener DJ, Galjart B, et al. Predicting 10-year survival after resection of colorectal liver metastases; an international study including biomarkers and perioperative treatment. Eur J Cancer (Oxford, England : 1990). 2022;168:25–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Liu H, Liang Z, Cheng S, Huang L, Li W, Zhou C, et al. Mutant KRAS Drives Immune Evasion by Sensitizing Cytotoxic T-Cells to Activation-Induced Cell Death in Colorectal Cancer. Adv Sci (Weinheim, Baden-Wurttemberg, Germany). 2023;10:e2203757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zhu G, Pei L, Ye D, Tang Q, Xu H, Bi F. Targeting KRAS in colorectal cancer immunotherapy: rationale, challenges and future directions. Biochim Biophys Acta. 2025;1880:189382. [DOI] [PubMed] [Google Scholar]
  • 7.Lou E, Xiu J, Baca Y, Saeed A, Prakash A, Gholami S, et al. Differential landscape of immune evasion in oncogenic RAS-driven primary and metastatic colorectal cancers. Mol Ther Oncolytics. 2024;32:200786. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zhou Y, Kuang Y, Wang C, Yu Y, Pan L, Hu X. Impact of KRAS mutation on the tumor microenvironment in colorectal cancer. Int J Biol Sci. 2024;20:1947–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Zhang C, Wang XY, Zuo JL, Wang XF, Feng XW, Zhang B, et al. Localization and density of tertiary lymphoid structures associate with molecular subtype and clinical outcome in colorectal cancer liver metastases. J Immunother Cancer. 2023;11(2):e006425. [DOI] [PMC free article] [PubMed]
  • 10.Haider C, Hnat J, Wagner R, Huber H, Timelthaler G, Grubinger M, et al. Transforming growth factor-β and Axl induce CXCL5 and neutrophil recruitment in hepatocellular carcinoma. Hepatology (Baltimore, MD). 2019;69:222–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Wang TT, Zhao YL, Peng LS, Chen N, Chen W, Lv YP, et al. Tumour-activated neutrophils in gastric cancer foster immune suppression and disease progression through GM-CSF-PD-L1 pathway. Gut. 2017;66:1900–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Eruslanov E, Nefedova Y, Gabrilovich DI. The heterogeneity of neutrophils in cancer and its implication for therapeutic targeting. Nat Immunol. 2025;26:17–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Wang XY, Zuo JL, Fu H, Zhang C, Fan QQ, Zhang B, et al. The impact of neoadjuvant therapy on the prognostic value of preoperative neutrophil-to-lymphocyte ratio for colorectal liver metastases: a multi-center cohort study. Int J Colorectal Dis. 2025;40:10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Maccari R, Ottanà R. In search for inhibitors of human Aldo-Keto reductase 1B10 (AKR1B10) as novel agents to fight cancer and chemoresistance: current state-of-the-art and prospects. J Med Chem. 2025;68:860–85. [DOI] [PubMed] [Google Scholar]
  • 15.Shen Y, Qiu A, Huang X, Wen X, Shehzadi S, He Y, et al. AKR1B10 and digestive tumors development: a review. Front Immunol. 2024;15:1462174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Zhang W, Li H, Yang Y, Liao J, Yang GY. Knockdown or inhibition of aldo-keto reductase 1B10 inhibits pancreatic carcinoma growth via modulating Kras-E-cadherin pathway. Cancer Lett. 2014;355:273–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Wang Z, Pei Y, Li W, Zhang J, Liu J. Clinical value of AKR1B10 in hepatocellular carcinoma: a systematic review and meta-analysis. PLoS One. 2022;17:e0279591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Cai J, Cui Z, Zhou J, Zhang B, Lu R, Ding Y, et al. Mettl3 promotes glycolysis and cholangiocarcinoma progression by mediating the m6A modification of AKR1B10. Cancer Cell Int. 2022;22:385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Suvilesh KN, Manjunath Y, Nussbaum YI, Gadelkarim M, Raju M, Srivastava A, et al. Targeting AKR1B10 by drug repurposing with Epalrestat overcomes chemoresistance in non-small cell lung cancer patient-derived tumor organoids. Clin Cancer Res. 2024;30:3855–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Liu C, Shi L, Li W, Huang Z, Wang S, Xu P, et al. AKR1B10 accelerates the production of proinflammatory cytokines via the NF-κB signaling pathway in colon cancer. J Mol Histol. 2022;53:781–91. [DOI] [PubMed] [Google Scholar]
  • 21.Michiel Zeeuw J, Wesdorp NJ, Ali M, Bakker AJJ, Voigt KR, Starmans MPA, et al. Prognostic value of total tumor volume in patients with colorectal liver metastases: A secondary analysis of the randomized CAIRO5 trial with external cohort validation. Eur J Cancer (Oxford, England : 1990). 2024;207:114185. [DOI] [PubMed] [Google Scholar]
  • 22.Liu Q, Liu Z, Zhang X, Zeng A, Song L. Revisiting of cancer immunotherapy: insight from the dialogue between glycolysis and PD-1/PD-L1 axis in the tumor microenvironment. Int J Biol Sci. 2025;21:1202–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Duan W, Liu W, Xia S, Zhou Y, Tang M, Xu M, et al. Warburg effect enhanced by AKR1B10 promotes acquired resistance to pemetrexed in lung cancer-derived brain metastasis. J Transl Med. 2023;21:547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Fiaschi T, Marini A, Giannoni E, Taddei ML, Gandellini P, De Donatis A, et al. Reciprocal metabolic reprogramming through lactate shuttle coordinately influences tumor-stroma interplay. Can Res. 2012;72:5130–40. [DOI] [PubMed] [Google Scholar]
  • 25.Zhang D, Tang Z, Huang H, Zhou G, Cui C, Weng Y, et al. Metabolic regulation of gene expression by histone lactylation. Nature. 2019;574:575–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Rho H, Hay N. Protein lactylation in cancer: mechanisms and potential therapeutic implications. Exp Mol Med. 2025;57:545–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Mo S, Tang P, Luo W, Zhang L, Li Y, Hu X, et al. Patient-derived organoids from colorectal cancer with paired liver metastasis reveal tumor heterogeneity and predict response to chemotherapy. Adv Sci (Weinheim, Baden-Wurttemberg, Germany). 2022;9:e2204097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Isermann T, Sers C, Der CJ, Papke B. KRAS inhibitors: resistance drivers and combinatorial strategies. Trends Cancer. 2025;11(2):91–116. [DOI] [PubMed]
  • 29.Wu J, Xie K, Zhang Y, Zhang W, Cheng R, Zhang Y, et al. RASON promotes KRAS(G12C)-driven tumor progression and immune evasion in non-small cell lung cancer. J Exp Clin Cancer Res. 2025;44:106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Hong DS, Fakih MG, Strickler JH, Desai J, Durm GA, Shapiro GI, et al. KRAS(G12C) inhibition with sotorasib in advanced solid tumors. N Engl J Med. 2020;383:1207–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Xie C, Ye X, Zeng L, Zeng X, Cao D. Serum AKR1B10 as an indicator of unfavorable survival of hepatocellular carcinoma. J Gastroenterol. 2023;58:1030–42. [DOI] [PubMed] [Google Scholar]
  • 32.Liu YY, Liu YW, Huang GK, Hung KC, Lin YH, Yeh CH, et al. Overexpression of AKR1B10 predicts poor prognosis in gastric cancer patients undergoing surgical resection. Current oncology (Toronto, Ont). 2022;30:85–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wu X, Huang S, Gao J, Huang S, Chen L, Zhao Z, et al. AKR1B10 dictates c-Myc stability to suppress colorectal cancer metastasis via PP2A nitration. Sci Adv. 2025;11:eadv6937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Klimpel A, Stillger K, Wiederstein JL, Krüger M, Neundorf I. Cell-permeable CaaX-peptides affect K-Ras downstream signaling and promote cell death in cancer cells. FEBS J. 2021;288:2911–29. [DOI] [PubMed] [Google Scholar]
  • 35.Kerk SA, Papagiannakopoulos T, Shah YM, Lyssiotis CA. Metabolic networks in mutant KRAS-driven tumours: tissue specificities and the microenvironment. Nat Rev Cancer. 2021;21:510–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Deng H, Kan A, Lyu N, He M, Huang X, Qiao S, et al. Tumor-derived lactate inhibit the efficacy of lenvatinib through regulating PD-L1 expression on neutrophil in hepatocellular carcinoma. J Immunother Cancer. 2021;9(6):e002305. [DOI] [PMC free article] [PubMed]
  • 37.Xu P, Zhang X, Chen K, Zhu M, Jia R, Zhou Q, et al. Tumor cell-derived microparticles induced by methotrexate augment T-cell antitumor responses by downregulating expression of PD-1 in neutrophils. Cancer Immunol Res. 2023;11:501–14. [DOI] [PubMed] [Google Scholar]
  • 38.Zong Z, Zhang L, Zhou F. Lactylation in cancer: advances, challenges, and future perspectives. Cancer Res. 2025;85:3192–5. [DOI] [PubMed] [Google Scholar]
  • 39.Hou X, Hong Z, Zen H, Zhang C, Zhang P, Ma D, et al. Lactylation in cancer biology: Unlocking new avenues for research and therapy. Cancer Commun (London, England). 2025;45(11):1367–406. [DOI] [PMC free article] [PubMed]
  • 40.Huang ZW, Zhang XN, Zhang L, Liu LL, Zhang JW, Sun YX, et al. STAT5 promotes PD-L1 expression by facilitating histone lactylation to drive immunosuppression in acute myeloid leukemia. Signal Transduct Target Ther. 2023;8:391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ding CH, Yan FZ, Xu BN, Qian H, Hong XL, Liu SQ, et al. PRMT3 drives PD-L1-mediated immune escape through activating PDHK1-regulated glycolysis in hepatocellular carcinoma. Cell Death Dis. 2025;16:158. [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

Supplementary Material 1. (2MB, pdf)

Data Availability Statement

The datasets supporting the conclusions of this article are included within the article and its additional files. Bulk RNA-Seq data were from GEO database, and single cell data were from The National Genomics Data Center (https://ngdc.cncb.ac.cn). Additional data related to this article will be shared upon reasonable request to the corresponding author.

Articles from Journal of Experimental & Clinical Cancer Research : CR are provided here courtesy of BMC

RESOURCES