Abstract

Background

   Fatty acids serve as a crucial energy source for tumor cells during the
   progression of chronic lymphocytic leukemia (CLL). The present study
   aims to elucidate the characteristics of fatty acid metabolism (FAM) in
   CLL, construct a related prognostic score, and investigate the
   regulatory role and mechanisms of FAM in CLL development.

Methods

   Bulk RNA sequencing data from CLL patients and healthy controls were
   analyzed to identify differentially expressed fatty acid metabolic
   genes. FAM-score was constructed using Cox-LASSO regression and
   validated. Single-cell RNA sequencing was used to analyze the
   expression of key FAM genes in CLL immune cell subsets and investigate
   cellular communication. Functional assays, including cell viability,
   drug sensitivity, and oxygen consumption assays, were performed to
   assess the impact of fatty acid oxidation (FAO) inhibition on CLL
   cells.

Results

   Three FAM-related genes (LPL, SOCS3, CNR1) were identified with
   independent prognostic significance to construct the risk score. The
   FAM-score demonstrated superior prognostic performance compared to the
   Binet stage and was associated with established clinical prognostic
   markers. Single-cell analysis revealed distinct expression patterns of
   LPL, SOCS3, and CNR1 across CLL immune cell subsets. Cellular
   communication analysis highlighted the regulatory role of distinct B
   cell and Treg subsets in the CLL microenvironment. CLL patients with
   high FAM-score displayed distinct immune infiltration patterns, with
   increased FAO pathway activity. Inhibition of FAO reduced CLL cell
   viability, synergistically enhanced the efficacy of the PI3K inhibitor
   idelalisib.

Conclusion

   The present study constructed a prognostic risk score based on FAM gene
   expression, revealing related immune phenotypic differences and
   exploring the regulatory role of FAO in CLL development. Targeting
   fatty acid metabolism potentially modulates the CLL immune
   microenvironment and synergistically enhances the efficacy of PI3K
   inhibitors.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s40364-025-00753-7.

   Keywords: Chronic lymphocytic leukemia, Fatty acid metabolism,
   Immunophenotypes, PI3K inhibitors

Introduction

   Chronic lymphocytic leukemia (CLL) is a relatively indolent hematologic
   malignancy of mature CD19^+CD5^+ B cells. CLL cells exhibit increased
   dependence on free fatty acid oxidation (FAO) for energy compared to
   normal B cells [[41]1]. Fatty acid metabolism has implications for drug
   resistance in CLL, with ibrutinib resistance linked to FAO and abnormal
   phospholipid levels associated with chemotherapy resistance
   [[42]2–[43]5]. However, the gene expression patterns related to fatty
   acid metabolism in CLL have not been systematically studied.

   Extensive research has shown that metabolism reprograming, especially
   fatty acid metabolism (FAM), shapes an immune suppressive
   microenvironment that promotes tumor progression [[44]6–[45]8].
   Multi-omics analysis in colorectal cancer patients has highlighted the
   impact of lipid metabolism on immunogenicity [[46]6]. Correspondingly,
   the fatty acid transport pathway is directly involved in the
   pathological activation of neutrophils in a variety of tumors such as
   lymphoma, lung cancer, and pancreatic cancer [[47]7]. Moreover, tumor
   cells and CD8^+ T cells showed different metabolic adaptability to
   obesity in a mouse cell model of colorectal cancer, resulting in
   changes in fatty acid distribution in high-fat diet tumors, thereby
   impairing the infiltration and function of CD8^+ T cells [[48]8]. As a
   lymphoproliferative disease with the perturbation of the immune system,
   the pathogenesis and progression of CLL are highly dependent on the
   immune microenvironment. Nevertheless, it has not been entirely
   elucidated whether alterations in fatty acid metabolism in CLL perturb
   immune cell fate and function.

   In this study, we integrated bulk and single-cell transcriptomic data
   from CLL patients to comprehensively characterize the role of FAM in
   CLL. Based on the FAM-specific transcriptomic signature, we developed a
   prognostic risk model FAM-score containing 3 genes (LPL, SOCS3, and
   CNR1). Mechanistically, these genes are mainly related to the FAO and
   lipogenesis [[49]9–[50]11]. Moreover, the relationship between the CLL
   immunophenotype and hub genes related to FAM was investigated combined
   with single-cell transcriptome analysis of CLL patients. Meanwhile,
   this study found that the inhibition of FAO impeded the survival of CLL
   cells and enhanced their sensitivity to PI3K inhibitor idelalisib,
   paving the way for the modulation of fatty acid metabolism to improve
   CLL therapy.

Methods

Patient specimens

   The study was approved by the Medical Ethics Committee of Shandong
   Provincial Hospital, and each patient gave the informed consent. CLL
   was diagnosed according to the updated International Workshop on
   Chronic Lymphocytic Leukemia (IWCLL) criteria [[51]12]. CLL primary
   cells were extracted and isolated from the peripheral blood of 5
   first-treatment CLL patients from Shandong Provincial Hospital’s
   Department of Hematology as previously published [[52]13, [53]14]. In
   addition, peripheral blood mononuclear cells from 5 primary CLL
   patients and 2 healthy controls were processed for 10x Genomics
   single-cell RNA sequencing (scRNA-seq).

Data acquisition

   The fatty acid metabolism-related genes were defined as genes enriched
   in fatty acid metabolic pathways based on the Molecular Signatures
   Database (MSigDB, [54]https://www.gsea-msigdb.org/gsea/msigdb). Bulk
   and single-cell transcriptome data of CLL patients were obtained from
   the Gene Expression Omnibus database (GEO;
   [55]https://www.ncbi.nlm.nih.gov/geo/) and International Cancer Genome
   Consortium (ICGC; [56]https://dcc.icgc.org/). To establish and validate
   the FAM-score, the RNA-seq data, scRNA-seq data and clinical prognosis
   information of 791 CLL patients from multiple cohorts were
   incorporated.

Construction and validation of the risk score model base on the fatty acid
metabolic genes

   The differentially expressed genes (DEGs) with significant differences
   in expression levels between CLL and normal samples were calculated
   using DESeq2 [[57]15]. The characteristic genes of fatty acid
   metabolism in CLL are defined as the intersection of DEGs and the set
   of fatty acid metabolism-related genes. Statistical significance was
   assigned to genes with a |logFC| of more than 1.2.

   The Cox-Least absolute shrinkage and selection operator (LASSO)
   regression model was performed in ICGC cohort. The risk score was
   calculated as: Inline graphic . The “Coef” represents regression
   coefficients, and “ExpGene” is the expression values of genes from the
   prognostic risk score model. Finally, the reliability and applicability
   of the prognostic risk score model were further validated by
   Kaplan-Meier analysis and time-dependent ROC curve in the external
   validation set. For a deeper understanding of the relationship between
   the risk score and clinical features, clinical correlation analysis was
   conducted among patients in different subgroups. CLL patients in the
   ICGC cohort were categorized by age, sex, Binet stage, IGHV mutation
   status, TP53 mutation status and disease status at the end of
   follow-up.

Expression signatures and metabolic activity in single-cell transcriptomes

   For visualization purposes, dimensional reduction was carried out using
   tSNE. The analysis of intercellular communication was performed with
   CellChat. Furthermore, the activity of fatty acid metabolic pathways in
   each single cell sample was calculated by UCell [[58]16]. The knnDREMI
   (conditional-Density Resampled Estimate of Mutual Information) was
   applied to estimate the functional relationship of hub genes expression
   to other genes across the dynamic range of expression [[59]17].

Metabolic and immune function enrichment analysis associated with FAM-score

   The fatty acid metabolism pathways activity was assessed by
   single-sample gene set enrichment analysis (ssGSEA). Subsequently, the
   CIBERSORT deconvolution algorithm was employed to analyze the cellular
   composition of the samples based on gene expression profiles [[60]18].
   Moreover, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and
   Genomes (KEGG) analysis were conducted to decipher the primarily
   enriched signaling pathways and biological functions between the high-
   and low-risk groups in the training set [[61]19].

Cell metabolism assay

   The measurement of extracellular oxygen consumption was conducted using
   the Extracellular Oxygen Consumption Assay Kit (Abcam, Cambridge,
   England). Following the manufacturer’s instructions, the Fatty Acid
   Oxidation Assay Kit (Abcam, Cambridge, England) and the Extracellular
   Oxygen Consumption Assay Kit were used to measure fatty acid oxidation.
   The initial slope of fluorescence intensity was calculated from the
   same linear portion of the curve (0–10 min) in the Fatty Acid Oxidation
   assay using GraphPad Prism.

RNA isolation and quantitative real-time PCR

   Total RNA was purified with SPARKeasy cell RNA kit (SparkJade, Jinan,
   China) and reversed transcribed into cDNA using a reverse transcription
   kit (Agbio, Changsha, China). Real-time PCR was performed in a Light
   Cycler 480 II Real-Time PCR system (Roche Diagnostics, Basel,
   Switzerland) using SYBR Green (Agbio, Changsha, China). Real-time PCR
   of each sample was performed in triplicate. Results were obtained using
   the sequence detection software Light cycler 480 and analyzed using
   GraphPad Prism version 9.0 statistical software.

Analysis of cell proliferation and apoptosis

   1 × 10^4 cells were placed into each well of a 96-well culture plate
   and treated with the relevant drug doses. Following incubation periods
   of 24, 48, and 72 h, 10 μl of CCK-8 (DOJINDO, Kumamoto, Japan) solution
   was added to each well. The contents were mixed thoroughly, and
   absorbance was measured at 450 nm after 2 h.

   1 × 10^6 cells were inoculated into each well of a six-well plate and
   treated with the appropriate drug doses. After 24 h, the cells were
   collected and washed with 1×PBS. For apoptosis analysis, the cells were
   resuspended in 100 μL of binding buffer, and 5 μL of 7-AAD and Annexin
   V-PE were added to each tube. The samples were incubated at room
   temperature in the dark for 15 min, followed by analysis using the
   ATTUNE NXT Flow Cytometer (Thermo Fisher, Massachusetts, USA).

Statistical analysis

   All statistical analyses were accomplished via R software (Version
   4.2.3, [62]http://www.R-project.org). The comparison of each
   Kaplan-Meier curve contained in this study was completed by the
   log-rank test. The differences of the immune infiltrates in the
   different subgroups and FAM-score in various clinicopathologic
   parameters were detected by the Wilcoxon test. Univariate and
   multivariate Cox regression analyses were exploited to screen out the
   OS-related hub genes and the independent prognostic indicators of OS.
   Statistical significance was determined as p < 0.05.

Results

Dissecting FAM-related genes expression in CLL by integrating bulk and
single-cell transcriptomics

   Differential expression analysis of CLL cells and healthy B cells
   identified genes with altered expression, particularly those related to
   FAM, based on the hypothesis that FAM dysregulation contributes to CLL.
   To identify potential prognostic markers or therapeutic targets by
   pinpointing metabolic vulnerabilities in CLL, differential expression
   analysis identified 34 DEGs associated to fatty acid metabolism in
   expression arrays of 180 CLL blood samples and 32 sorted CD19^+ B cell
   samples from healthy donors in [63]GSE50006 (Figure [64]S1A). To
   investigate the prognostic value of these altered metabolic genes, 34
   FAM-related DEGs were analyzed in the ICGC cohort (training set) using
   univariate Cox regression, identifying 10 prognostic genes (p < 0.05;
   Figure [65]S1B). LASSO-Cox regression analysis was then performed to
   narrow the number of genes. Finally, 3 hub genes (LPL, SOCS3, CNR1)
   with independent prognostic significance were selected to construct the
   robust risk score named FAM-score (Figure [66]S1C-F).

   To further explore the role of FAM in CLL immune regulation, the
   expression characteristics of LPL, SOCS3, and CNR1 at the single-cell
   level were analyzed. First, we compared the expression of these hub
   genes in CLL patients and healthy controls (Fig. [67]1A). In healthy
   control samples, LPL is predominantly expressed on CD4^+ T cells,
   whereas in CLL patients, LPL is expressed at relatively higher levels
   on B cells (Fig. [68]1B). SOCS3 tends to be expressed on CD4^+ T cells,
   but its expression levels in CLL patients are lower than in healthy
   controls (Fig. [69]1C). CNR1, on the other hand, exhibits exceedingly
   low expression, with only minimal presence observed in B cells.
   However, CNR1 is also slightly expressed in CD8^+ T cells in CLL
   patients (Fig. [70]1D). Furthermore, the expression signature of hub
   genes in CLL samples at different time points supports these findings
   (Figure [71]S1G-J). The t-SNE profiles revealed cell subpopulation
   specificity in the expression of key genes in CLL, suggesting that
   fatty acid metabolism potentially induces differences between immune
   cell types.

Fig. 1.

   [72]Fig. 1
   [73]Open in a new tab

   Expression characteristics of FAM at the single-cell level. (A) The
   expression of these hub genes LPL, SOCS3, and CNR1 in CLL patients and
   healthy controls. (B) The expression characteristics of LPL. (C) The
   expression characteristics of SOCS3. (D) The expression characteristics
   of CNR1

Alterations in FAM induce immune dysfunction in CLL patients

   To determine the function and mechanism of FAM-related genes in the
   immune microenvironment of CLL patients, single-cell transcriptome data
   from CLL patients were further subdivided into different immune cell
   clusters (Fig. [74]2A). The expression characteristics of LPL, CNR1,
   and SOCS3 were then investigated within these clusters (Fig. [75]2B-D).
   Cellular communication analysis revealed that exhausted B cells,
   regulatory T cells (Tregs), and a population of unswitched memory B
   cells exert broad regulatory influence within the CLL immune
   microenvironment (Fig. [76]2E-G). Notably, SOCS3 expression was highly
   concentrated within the Treg population and, furthermore, within the
   effector memory CD8^+ T cell population targeted by these Tregs. This
   localized SOCS3 expression underscores its crucial role in modulating
   immune responses within these specific immune cell subsets in CLL
   patients.

Fig. 2.

   [77]Fig. 2
   [78]Open in a new tab

   Expression characteristics of FAM genes in immune cell subsets. (A)
   t-SNE visualization of immune cell clusters from CLL patients. (B-D)
   The expression characteristics of LPL, SOCS3, and CNR1 in immune cell
   clusters from CLL patients. (E-F) The interaction relationship between
   immune cells in CLL patients. The thickness of the connecting line
   indicates the strength of the interaction. (G) Incoming communication
   patterns of immune cells in CLL patients

   These findings are corroborated by bulk RNA sequencing data in the ICGC
   cohort, which revealed distinct immune infiltration patterns between
   the two FAM-score subgroups (Figure [79]S2B-C). High FAM-score CLL
   patients exhibited significantly decreased levels of activated CD4^+ T
   cells, CD8^+ T cells, γδ T cells, and dendritic cells, concurrent with
   an expansion of M2 macrophages, plasma cells, follicular helper T
   cells, and activated NK cells. These findings demonstrate that CLL
   patients exhibit two distinct immune phenotypes based on differential
   expression of genes related to fatty acid metabolism.

   Mechanistically, GO enrichment analysis of the bulk RNA-seq data linked
   the FAM-score to pathways involved in cell fate determination and
   specification, foam cell differentiation, and unsaturated fatty acid
   metabolism (Figure [80]S2D). Furthermore, KEGG analysis showed
   enrichment of immune cell differentiation, activation, and immune
   response regulation pathways in the high-risk subgroup (Figure
   [81]S2E). These alterations in fatty acid metabolism pathways may thus
   mediate immune cell differentiation and function, contributing to CLL
   development and progression.

Enhanced fatty acid oxidation drives CLL progression through the immune
regulation

   While fatty acid metabolic reprogramming is known to influence tumor
   progression in various cancers by shaping an immunosuppressive
   microenvironment, the specific impact of altered fatty acid metabolism
   on immune cell fate and function in CLL remained unclear [[82]6–[83]8].
   Fatty acid metabolism encompasses the breakdown of fatty acids for
   energy production and their synthesis for storage or molecular building
   blocks. To explore how fatty acid metabolism modulates the CLL immune
   phenotype, the scMetabolism algorithm was applied to quantify fatty
   acid metabolic activity from single-cell transcriptomes. This analysis
   revealed that degradation pathways were more active than biosynthesis
   pathways in CLL cells (Fig. [84]3A-B). Furthermore, fatty acid
   β-oxidation and cellular response to fatty acids pathways were
   significantly more active in CLL patients than in healthy controls
   (Fig. [85]3C). To identify alterations of fatty acid metabolism in CLL
   patients with elevated FAM-scores, ssGSEA analysis of the FAM-related
   pathways was implemented on the ICGC cohort (Table [86]S1). Consistent
   with previous studies, multiple pathways related to fatty acid
   oxidation, synthesis, and transport showed significantly elevated
   transcriptional activities in the high-risk subgroup (Figure [87]S2A).
   To validate our transcriptomic findings, we measured extracellular
   oxygen consumption and fatty acid oxidation in peripheral blood
   mononuclear cells from CLL cell lines, primary CLL patient samples, and
   healthy controls (Fig. [88]3C-D, F-J). CLL cells exhibited
   significantly higher extracellular oxygen consumption and fatty acid
   oxidation compared to healthy controls (Fig. [89]3E, K).

Fig. 3.

   [90]Fig. 3
   [91]Open in a new tab

   CLL fatty acid metabolism is characterized by increased fatty acid
   oxidation. (A-B) Activity of fatty acid biosynthesis and degradation
   pathways at the single-cell level. (C) Comparison of the activities of
   major fatty acid metabolic pathways in CLL patients (CLL) and healthy
   controls (Normal). (D) Extracellular oxygen consumption assay in MEC1
   cells, EHEB cells, and normal peripheral blood mononuclear cells. (E)
   Extracellular oxygen consumption assay in CLL primary cells, and normal
   peripheral blood mononuclear cells. (F) Comparison of extracellular
   oxygen consumption levels between CLL cells and normal peripheral blood
   monocytes. (G-L) Fatty acid oxidation assay in CLL cells and normal
   peripheral blood mononuclear cells. (M) Comparison of fatty acid
   oxidation activity between CLL cells and normal peripheral blood
   monocytes. NS: not significant; *: p < 0.05; **: p < 0.01; ***:
   p < 0.001

   Subsequently, the reciprocal molecules of the key genes were identified
   in the single-cell transcriptome to investigate the mechanism of
   FAM-induced differences of CLL immunophenotype (Fig. [92]4A-C). The
   distribution of gene DREMI coefficient related to CNR1, LPL and SOCS3
   were shown in Fig. [93]4D, and the thresholds for interacting genes are
   > 0.5, > 1 and > 1.5 respectively. Consistent with the results of
   functional enrichment analysis in bulk RNA-sequencing data, the
   interacting genes of CNR1, LPL, and SOCS3 are all enriched in the
   immune cell activation pathway and its regulatory pathways
   (Fig. [94]4E). These results suggest that within the context of fatty
   acid metabolism reprogramming, immune regulation might serve as a
   potential mechanism to promotes the progression of CLL.

Fig. 4.

   [95]Fig. 4
   [96]Open in a new tab

   Identification of interacting molecules of FAM-related genes in the CLL
   single-cell transcriptome. (A) Heatmap of expression of
   CNR1-interacting molecules in CLL single-cell transcriptome. (B)
   Heatmap of expression of SOCS3-interacting molecules in CLL single-cell
   transcriptome. (C) Heatmap of expression of LPL-interacting molecules
   in CLL single-cell transcriptome. (D) The distribution of gene DREMI
   coefficient related to CNR1, LPL and SOCS3. (E) KEGG pathway enrichment
   analysis of FAM-related gene interactors

Inhibition of fatty acid oxidation in CLL cells impedes survival and
proliferation

   Both bulk and single-cell RNA sequencing analyses demonstrated
   significant upregulation of FAO pathways in high-risk CLL patients. To
   further investigate the effects of targeted regulation of fatty acid
   metabolism on CLL cells, we inhibited FAO in CLL cells by blocking
   fatty acid entry into mitochondria using perhexiline, according to
   published literature [[97]20, [98]21]. The MEC1, EHEB, and primary CLL
   cells co-cultured with perhexiline (0 μM, 2 μM, 4 μM, and 8 μM)
   exhibited a concentration-dependent decrease in viability across
   various durations, as shown by cell viability assays (Fig. [99]5A).
   Furthermore, co-culturing MEC1 and EHEB cells with the PI3K inhibitor
   idelalisib in the presence or absence of 4 μM perhexiline revealed an
   enhanced sensitivity of CLL cells to idelalisib upon FAO inhibition
   (Fig. [100]5B-F). Flow cytometric analysis confirmed a significant
   increase in apoptosis in CLL cells following FAO inhibition
   (Fig. [101]5G). In line with our results from integrated bulk and
   single-cell RNA analyses, treating CLL cells with Perhexiline to
   inhibit FAO revealed a downregulation of LPL and CNR1 gene expression
   compared to untreated controls (Fig. [102]5H-I). Conversely, a
   significant upregulation of SOCS3 expression was observed
   (Fig. [103]5J).

Fig. 5.

   [104]Fig. 5
   [105]Open in a new tab

   Fatty acid oxidation inhibition impeded CLL cells and enhanced
   sensitivity to PI3K inhibitors. (A-C) Cell proliferation analysis of
   MEC1, EHEB cells and primary cells from CLL patients treated with
   perhexiline. (D) Mechanisms of fatty acid metabolism regulating the
   development of CLL. (E-F) Sensitivity assay for the PI3K inhibitor
   idelalisib in MEC1 and EHEB cells treated with perhexiline. (G)
   Apoptosis analysis of MEC1 and EHEB cells treated with FAO inhibitors
   perhexiline. Symbols on the top represent significant difference. (H)
   The relative transcription levels of hub genes in MEC1 and EHEB cells
   treated with FAO inhibitor perhexiline at a concentration of 4 μ M for
   24 h. NS: not significant; *: p < 0.05; **: p < 0.01; ***: p < 0.001;
   ****: p < 0.0001

   Furthermore, analysis of the FAM-score’s relationship with clinical
   features in the ICGC cohort revealed significant differences across
   subgroups, with the high-risk group associated with age > 60, male
   gender, unmutated IGHV, advanced Binet stage, and disease progression
   (p < 0.05; Figure [106]S3A-E). Further analysis of cytogenetic and
   transcriptomic data from 109 CLL patients showed lower FAM-score
   (p < 0.05; Figure [107]S3F-G). Kaplan-Meier analysis confirmed the
   FAM-score’s prognostic value across different sexes, ages, and IGHV
   mutation statuses (p < 0.05; Figure [108]S4A-F), though significance
   was observed only in TP53 wild-type patients, possibly due to limited
   TP53-mutated samples (Figure [109]S4G-H). These findings suggest that
   FAM-related genes, acting through the fatty acid oxidation pathway,
   play a crucial role in CLL cell survival and proliferation.
   Furthermore, inhibiting fatty acid oxidation synergistically enhances
   the efficacy of PI3K inhibitor in CLL.

Discussion

   The present study identified a novel prognostic risk score based on
   fatty acid metabolic genes and has elucidated the immunophenotypic
   differences in CLL related to the FAM-related genes. Integrated
   analysis of transcriptomic data and cellular assays demonstrates that
   increased fatty acid oxidation is a prominent feature of CLL metabolic
   reprogramming. Furthermore, pharmacological inhibition of FAO, the most
   significantly altered FAM pathway in the high-risk CLL subgroup,
   enhances CLL cell sensitivity to PI3K inhibitors. Collectively, these
   findings demonstrate that the observed reprogramming of fatty acid
   metabolism, specifically the increased FAO pathway in CLL cells, drives
   their immunosuppressive phenotype and resistance to PI3K inhibitors,
   thereby promoting the development of CLL (Fig. [110]5D).

   LPL, CNR1, and SOCS3 were identified as key regulators of fatty acid
   metabolism and used to construct the FAM-score, whose expression levels
   exhibit significant differences between samples of CLL and healthy
   controls. LPL is aberrantly overexpressed in chronic lymphocytic
   leukemia cells, and facilitates the use of fatty acids for energy
   production in CLL cells, contributing to their survival and
   proliferation [[111]4, [112]22–[113]24]. It is regulated by B-cell
   receptor signaling and STAT3, and its expression is associated with
   adverse prognosis [[114]2, [115]25, [116]26]. SOCS3 regulates FAO and
   energy metabolism through the leptin signaling pathway [[117]11,
   [118]27, [119]28]. In CLL, the expression of the SOCS3 is suppressed
   and significantly associated with leukemia progression [[120]29,
   [121]30]. CNR1, which encodes cannabinoid receptor 1, plays a
   significant role in regulating lipid metabolism by mediating
   inflammatory responses and oxidative stress through the NF-κB signaling
   pathway [[122]31–[123]33]. In CLL, CNR1 is overexpressed compared to
   normal B cells and is associated with adverse prognosis
   [[124]34–[125]36]. These genes are directly or indirectly implicated in
   FAO, which is consistent with our findings. In comparison with healthy
   controls, the FAO pathway demonstrates heightened activity in patients
   diagnosed with CLL, particularly within high-risk subgroups, suggesting
   that the reprogramming of fatty acid metabolism might change the
   survival of CLL patients.

   The FAM-score consistently predicts clinical outcome in diverse CLL
   patient profiles, except for cases involving TP53 mutations. Existing
   studies have highlighted that TP53 mutations mediate the disorder of
   fatty acid metabolism and participate in the lipolysis reaction of
   adipocytes [[126]37–[127]39]. However, due to the limited number of
   TP53 mutation cases (only 16 cases) in this study, we cannot
   definitively conclude the ineffectiveness of FAM-score in CLL patients
   with TP53 mutations. Therefore, further studies are needed to clarify
   the relationship between FAM reprogramming and TP53 mutations in CLL
   patients.

   The importance of crosstalk between the spectrum of immune cells and
   CLL cells in the tumor microenvironment has been highlighted in
   previous studies [[128]40]. The present study revealed that expression
   signatures of genes involved in fatty acid metabolism effectively
   manipulate the immunophenotype of CLL patients. Earlier work have
   demonstrated the impact of metabolic factors on immune cell
   differentiation and function [[129]8, [130]41–[131]46]. Consistent with
   these findings, the present study highlights distinct immune cell
   subsets affected by FAM, particularly FAO, further supporting the
   opinion that fatty acid metabolism regulates immune cell
   differentiation. In the present study, single-cell sequencing of CLL
   patients revealed significant immune cell subset-specific expression of
   the fatty acid metabolism (FAM) genes LPL, SOCS3, and CNR1. This
   distinct expression pattern was further associated with the
   immunosuppressive phenotype observed in the high-risk patient subgroup
   identified through immune infiltration analysis. The abundance of DCs
   responsible for activating T cells was significantly reduced in CLL
   patients with increased FAM-score, and essential effector T cell
   subsets (such as activated CD4^+ T cells, CD8^+ T cells, and γδ T
   cells) were concomitantly reduced, collectively revealing that
   immunosuppressive features are associated with elevated FAM-score.

   The PI3K signaling pathway materially influences cell proliferation,
   survival, and metabolism in response to external cues, abnormal
   activation of which is a common oncogenic event in human cancers
   [[132]47–[133]50]. PI3Kδ inhibitors, such as idelalisib, have been
   approved by the FDA for the treatment of relapsed chronic lymphocytic
   leukemia, follicular non-Hodgkin’s lymphoma, and small lymphocytic
   lymphoma, demonstrating their clinical significance in these
   hematological malignancies [[134]47]. Positioned at the intersection of
   metabolism and immunity, the PI3K pathway has intriguing implications.
   For instance, in a mouse model of gastric cancer, PI3K inhibitors
   targeting the PI3K-AKT-mTOR pathway reduced free fatty acids
   accumulation and enhanced the efficacy of immune checkpoint (PD-1)
   inhibitors [[135]42]. Similarly, using the PI3Kα inhibitor CYH33 in
   pancreatic cancer mouse models enhances tumor microenvironment fatty
   acid metabolism, activating CD8^+ T cells and triggering anti-tumor
   immunity [[136]51]. In the present study, inhibition of FAO in CLL
   cells effectively reinforced the antitumor effects of idelalisib. The
   observed synergy between these two approaches underscores the critical
   interplay between metabolic reprogramming, particularly FAO, and PI3K
   signaling in CLL, thus strongly supporting further investigation of
   combination therapies targeting both pathways to overcome resistance
   and improve treatment outcomes.

Conclusion

   In summary, the present study has identified a robust clinical
   prognostic score based on fatty acid metabolic genes. Despite its
   limitations, our study contributes to the understanding of fatty acid
   metabolism reprogramming in CLL. We found that CLL cells dependence on
   FAO regulates the immune function, which contributes to both CLL cells
   proliferation and the establishment of an immunosuppressive
   microenvironment. Additionally, inhibition of FAO in CLL cells enhances
   the therapeutic efficacy of PI3K inhibitor and suppresses cell
   proliferative activity. Our work elucidates the impact of fatty acid
   metabolism on CLL immunophenotype and disease progression. FAM
   modulation might represent a promising immunotherapeutic strategy by
   reversing the immunosuppressive microenvironment, while inhibition of
   FAO enhances PI3K inhibitor efficacy through synergistic effects.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [137]Supplementary Material 1^ (2.6MB, pdf)
   [138]Supplementary Material 2: Table S1^ (611.6KB, xlsx)
   [139]Supplementary Material 3: Table S2^ (137.6KB, xlsx)
   [140]Supplementary Material 4: Table S3^ (98.3MB, xlsx)
   [141]Supplementary Material 5: Table S4^ (459.6KB, xlsx)

Acknowledgements