Upregulation of TCF1 and BCL11B in CD8+ effector T cells predicts favorable response to ibrutinib in patients with chronic lymphocytic leukemia Authors: Ge Liu, Taotao Liang, Hao Ai, Shan Wang, Songke Yin, Xiaojiao Wang, Hongfei Wu, Qian Wang, Hongmian Zhao, Xiaojuan Liu, Xiaodong Luv, Jia Liu, Qingsong Yin Published: 2026 ABSTRACT AbstractThe Bruton’s kinase inhibitor ibrutinib has become a cornerstone in the treatment of chronic lymphocytic leukemia (CLL). Previous studies reported an association between upregulation of TCF1 and BCL11B transcription factors with favorable survival outcomes in CLL patients. This study aimed to investigate the correlation of TCF1 and BCL11B with response to ibrutinib. Four GEO databases were used to examine the transition of CD8 + T cell subsets and the expression of TCF1 and BCL11B in CLL patients undergoing ibrutinib treatment, and the findings were validated in a cohort of 98 CLL patients. Briefly, ibrutinib treatment increased the proportion of effector T cell-like (Teff-like) cells and upregulated TCF1 and BCL11B in CD8 + T cells. We identified a subset of TCF1 + BCL11B+ Teff-like cells with high cytotoxic activity score that were almost undetectable prior to treatment but started to emerge following ibrutinib treatment. Importantly, these changes correlated with favorable response to ibrutinib, where chromatin accessibilities of TCF7 (encoding TCF1) and BCL11B were increased and TCF1 interacted with BCL11B to restore T cell immunity, but not in patients with poor response to ibrutinib. These findings suggest that the emergence and increase of BCL11B + TCF1 + CD8+ Teff cells likely reflect reconstitution of T cell immunity induced by ibrutinib.Subject terms: Predictive markers, Chronic lymphocytic leukaemia Introduction The Bruton’s tyrosine kinase (BTK) inhibitor ibrutinib improves progression-free survival (PFS) and overall survival (OS) compared with chemotherapy/chemoimmunotherapy regimens in patients with chronic lymphocytic leukemia (CLL)1,2, and thus has become a cornerstone in the treatment of CLL. Accumulating data suggests that, in addition to direct targeting B cell receptor (BCR) signaling to kill CLL-B cells, ibrutinib also regulates tumor microenvironment and T cell immunity through BTK-independent pathways3–5. Previous studies showed that ibrutinib increases T cell repertoire diversity in CLL patients6 and restores CD8 + T cell function by downregulating inhibitory receptors3,7,8. Ibrutinib pretreatment has been shown to improve the viability and proliferation of chimeric antigen receptor T (CAR-T) cells with low efficacy in CLL patients9,10 while reducing PD1 expression, thereby enhancing clinical responses of CAR-T cell therapy11. The cytotoxic activity of CD8 + T cells depends on their differentiation status12. Acute antigen stimulation drives the differentiation of CD8 + T cells into a large number of effector T (Teff) cells3,4. Once the antigens are cleared, a small proportion of Teff cells differentiate into memory T cells that rapidly generate functional Teff cells upon subsequent antigen exposure13. However, sustained exposure to antigens from chronic viral infections or cancer induces T cell exhaustion, which in turn limits the efficacy of T cell-mediated immunotherapy3,14. T cell differentiation and exhaustion are controlled by a variety of transcription factors (TFs) and chromatin regulators. T cell factor 1 (TCF1, encoded by the TCF7 gene) is a hub TF that sustains the stem-like property of T cells and controls the differentiation of stem-like T cells into effector or exhausted T cells by coordinating with other related TFs15,16. TCF1 + T cells possess potent cytotoxicity and have the potential to differentiate into terminal T cells17. Infiltration of TCF1 + PD1 + T cells into solid tumor tissues correlates with tumor regression and responses to anti-PD1 therapy18–20. In a recent study, the use of PI3Kδ/γ inhibitors during CAR-T cell preparation enhanced the expression of TCF1 and improved the efficacy of CAR-T cells21. However, the clinical use of PI3Kδ/γ inhibitors is restricted due to T cell-mediated autoimmunity22. B cell leukemia/lymphoma 11B (BCL11B), a target gene of TCF1 in T cell development in the thymus15,23, maintains peripheral effector T cell differentiation, activation, and expansion15,23,24. Co-expression of TCF1 and BCL11B has been shown to be associated with improved survival outcomes in CLL patients25. In the current study, we examined the potential correlation between the TCF1 and BCL11B expression and the response to ibrutinib in patients with CLL. Gene Expression Omnibus‌‌ (GEO) databases were analyzed to investigate the transition of CD8 + T cell subsets in CLL patients throughout the course of ibrutinib therapy. The findings were then validated in a cohort of 98 patients with CLL undergoing ibrutinib monotherapy at the authors’ centers. The study design is shown in Fig. 1. Flow chart of the study. The initial bioinformatics analysis of the transition of CD8 + T cell subsets from CLL patients during ibrutinib therapy was performed using single-cell and bulk RNA-seq from GEO databases, and then validated with CLL patients treated with ibrutinib monotherapy from the authors’ center by FCM, Western blot, and CUT&Tag-seq assays. CLL: chronic lymphocytic leukemia; GEO: gene expression omnibus; PB: peripheral blood. This flow chart was created with BioRender.com (https://BioRender.com/fq2uqr1). Results CLL-B cells and T cell subsets prior to versus after ibrutinib treatment A total of 98 CLL patients treated with ibrutinib in a real-world retrospective cohort were analyzed, including 50 newly diagnosed (ND) patients and 48 relapsed or refractory (R/R) patients. With a median follow-up of 58 months (range, 8–165 months), the overall response rate (ORR) was 89% (92% in ND patients and 81% in R/R patients, respectively). In ND patients, neither median PFS nor OS was reached (Fig. 2A). In R/R patients, median PFS and OS were 44 months (95%CI, 26 to not reached) and 132 months (95%CI, 121 to not reached). CLL-B cells and T cells prior to versus after ibrutinib treatment. A PFS and OS of ND and R/R patients treated with ibrutinib. B Volcano plot illustrating the upregulated or downregulated DEGs prior to versus after ibrutinib in B cells (left) and GO BP analysis of downregulated DEGs (right). C Volcano plot showing the upregulated or downregulated DEGs prior to versus after ibrutinib in T cells (left) and GO BP analysis results of upregulated DEGs (right). BP biological processes, DEG differentially expressed genes, GO Gene Ontology, R/R relapsed or refractory, ND newly diagnosed. In the GSE254718 dataset (10 T-cell samples and 11 CLL B-cell samples), a total of 936 upregulated and 1734 downregulated genes were found in B cells after 6-month ibrutinib monotherapy; downregulated genes were mainly involved in the B cell receptor (BCR) signaling pathway (e.g., BTK, SYK, and LYN), immunoglobulin-mediated immune response, B cell homeostasis, activation and proliferation (Fig. 2B). A total of 4249 upregulated and 1551 downregulated genes were found in T cells; upregulated genes were mainly involved in intracellular signal transduction, T cell receptor signaling pathway, T cell differentiation, positive regulation of canonical Wnt signaling pathway, and NF-κB signal transduction (Fig. 2C). Both TCF1 and BCL11B were among the upregulated differentially expressed genes (DEGs) in T cells following ibrutinib therapy. Percentage of Teff-like T cells prior to versus after ibrutinib therapy Unsupervised clustering analysis grouped CD8 + T cells into nine distinct clusters, as visualized in a heatmap of the top 20 upregulated DEGs per cluster (Supplementary Fig. 1), using single-cell RNA sequencing (scRNA-seq) data from the GSE111014 dataset (peripheral blood mononuclear cells (PBMCs) collected from 4 CLL patients at three time points: prior to, 30, and 120 days after ibrutinib monotherapy). These clusters were annotated and color-coded based on expression of well-established canonical T cell subset marker genes (Fig. 3A)26,27. The Teff-like cell cluster was characterized by high expression of effector markers (GNLY and GZMB) and lack of naïve T cell or memory markers (TCF7 and SELL), including conventional Teff, CD53+ Teff, and CDC42+ Teff (the latter two were identified by high expression of CD53 and CDC42, respectively, both of which ranked among the top 20 DEGs). The central memory T cell (Tcm) cluster was characterized by high expression of TCF7, SELL, and GZMK, and lack of effector markers (GNLY, GZMB, and GZMA). The effector memory T cell (Tem) cluster was characterized by low expression of effector markers compared to Teff-like cells and slightly high expression of naïve marker genes. The transitional memory T cell (TTM) cluster is an intermediate state between Tcm and Tem cells27. The tissue-resident memory T cell (Trm) cluster was characterized by high expression of ITGAE (encoding CD103). Accordingly, this Trm cluster with high MKI67 expression was refined as a proliferating Trm-like subset (ITGAE + MKI67 + ). The exhausted T cells (Tex) cluster was characterized by high expression of LAG3 and HAVCR2 (encoding Tim3). The γδ T cell cluster was characterized by high TRDC expression. Ibrutinib therapy increases Teff-like population in favorable responders. A UMAP projection of CD4+ and CD8 + T-cell subsets identified by cluster-specific gene expression in scRNA-seq data from 4 CLL patients (left). Average expression levels and cell expression proportions of T cell marker genes in the indicated clusters. Color: average expression levels; size: percentage expression of selected genes (middle). UMAP projection of nine CD8 + T cell clusters. Each dot corresponds to a single cell colored according to each cluster (right). B Dynamic changes of nine CD8 + T cell clusters after 0, 30, and 120 days of ibrutinib treatment. C The stacked bar chart and line chart illustrate the changes in proportions of various T cell subsets at three time points. Consistent with the previous studies3,6,7, CD8 + T cells were predominant in the peripheral blood of CLL patients before treatment, and mainly consisted of memory T cells and terminally differentiated T cells, with not naïve T cells were absent (Supplementary Fig. 2). The CD4/CD8 ratio was inverted at baseline, and gradually reversed after nine months of ibrutinib treatment. Meanwhile, the percentage of Tex increased during the first month of therapy and decreased rapidly thereafter (Fig. 3B). The percentage of memory T cells (including Tem and Tcm subsets) decreased, especially Tem, from 10.3% at baseline to 4.6% at 120 days of therapy, while Trm remained steady. The percentage of Teff-like cells increased from 22.3% at baseline to 73.4% after four months of ibrutinib therapy (Fig. 3C). Differentiation of memory T cells into Teff-like cells after ibrutinib therapy scRNA-seq data from four patients with complete remission (CR) on ibrutinib monotherapy revealed distinct DEG profiles across nine CD8 + T cell clusters; the top 5 DEGs for each cluster are shown in Fig. 4A. The enrichment analysis of Gene Ontology (GO) biological processes (BPs) of DEGs in Teff-like clusters indicated activated immune response-related pathways (e.g., T cell receptor signaling pathway, antigen receptor-mediated signaling pathway, immune response-activating cell surface receptor signaling pathway) prior to treatment, and then progressively enhanced activation from day 30 to day 120. In contrast, the effector function pathways (e.g., cell killing, leukocyte-mediated cytotoxicity, and alpha-beta T cell activation) were relatively weak at baseline, increased at 30 days of treatment, and then declined towards baseline by day 120. Most notably, the activation level of T cell differentiation-associated pathways was low at baseline but showed significant enhancement by 120 days of ibrutinib treatment. Oxidative phosphorylation and canonical NF-κB signal transduction also demonstrated increased activation following ibrutinib treatment (Fig. 4B). Ibrutinib therapy drives the differentiation of memory T cells into Teff-like cells. A Volcano plot showing DEGs across nine CD8 + T cell clusters. B Functional transformation of Teff-like cells from day 0 to day 120 following ibrutinib treatment. C Cell developmental trajectory analysis for the six clusters of CD8 + T cells as inferred by Monocle2. Cell expression profiles in a two-dimensional independent space. Black line: the main diameter path of the minimum spanning tree; dot: individual cell colored by cluster (left plot) or by pseudotime (right plot). D The distribution of T cells expressing TCF7 or BCL11B on cell fate trajectory and the expression levels of TCF7 or BCL11B in six CD8 + T cell clusters. E The dynamic changes of six clusters of CD8 + T cells from CLL patients within cell fate trajectory during ibrutinib treatment. F UMAP visualization of the T cell differentiation trajectory inferred by Slingshot. The black smooth curves represent the main lineages, with arrows indicating the direction of differentiation. G Pseudotime mapping on the UMAP embedding. The teal arrow (Lineage 1) highlights the primary differentiation path toward the terminal effector phenotype. Gray dots represent cells not assigned to the inferred lineages. H Density plot showing the distribution of pseudotime values across CD8 + T-cell subsets involved in Lineage 1. I Expression dynamics of key memory-associated (GZMK and IL7R) and cytotoxic (GNLY and GZMB) genes along pseudotime in Lineage 1. Cell fate trajectory analysis using Monocle2 suggested that memory T cells differentiated mainly into Teff cells (Fig. 4C). TCF7 and BCL11B expression levels were elevated in Tcm cells at the root of trajectory, and diminished along the path to Tex and Teff cells (Fig. 4D). Trajectory analysis of six T cell subsets throughout ibrutinib treatment revealed a significant decrease in memory T cells and a gradual increase in Teff-like cells over three time points, with a temporary rise in Tex cells at 30 days (Fig. 4E), aligning with Fig. 3C in gene expression profile clustering analyses. The differentiation trajectory from memory T cells to Teff cells was confirmed by both Monocle3 (Supplementary Fig. 3) and Slingshot (Fig. 4F–H) analyses, which tracked the pseudotemporal expression dynamics of key genes associated with memory and cytotoxicity (Fig. 4I). TCF1/BCL11B expression and proportion of TCF1 + BCL11B+ Teff-like cells In multi-parameter FCM analysis (Fig. 5A), the proportion of memory T (CD8 + CD45RO + ) cells in CLL patients prior to ibrutinib treatment (Pre-ibr) was significantly higher than that in healthy controls (P = 0.0009) (Supplementary Fig. 4). The median proportion of TCF1 + BCL11B + T cells in CD8 + T cell population was 6.60% (95% CI: 4.45–8.25) in Pre-ibr patients (n = 16) versus 36.51% (95% CI: 30.50–49.38) in healthy controls (n = 10) (P < 0.0001) (Fig. 5B). The proportion of TCF1 + BCL11B + T cells in CD8 + T cell population was 10.54% (95% CI: 6.11–12.92) in patients treated with ibrutinib for <1 year (n = 8) (P = 0.0257 versus the Pre-ibr level) and 11.41% (95% CI: 7.23–20.29) in patients treated with ibrutinib for >1 year (n = 11) (P = 0.0016 versus the Pre-ibr level). Elevated expression of TCF1 and BCL11B following ibrutinib therapy was verified at the mRNA and protein levels (Supplementary Fig. 5A–C, Supplementary Fig. 6). Ibrutinib therapy upregulates the expression of TCF1 and BCL11B and increases the proportion of TCF1 + BCL11B+ Teff-like cells. A Flow cytometry gating strategy and CD8 + T cell subset definition. B Representative FACS contour plots (left) and the statistical analysis of the percentage of TCF1 + BCL11B + T cells in CD8 + CD45RO + T cells from Pre-ibr (n = 16), Post-ibr (n = 8 for <1 year after ibrutinib; n = 11 for >1 year), and Ibr-resis (n = 3) patients and HCs (n = 10) (right). C Representative FACS contour plots (left) and the statistical analysis of the percentage of TCF1 + BCL11B + T cells in Teff-like cells from Pre-ibr (n = 16) and Post-ibr (n = 8 for <1 year after ibrutinib; n = 11 for >1 year) patients and HCs (n = 10). D Representative FACS contour plots (left) and the change in the percentage of GrB+perforin+ T cells in Teffs from Pre-ibr and Post-ibr patients (n = 4) and HCs (n = 4) (right). E The change of the percentage of TCF1 + PD1 + T cells before (n = 16) and after (n = 16) ibrutinib treatment. Error bar: SD. HCs healthy controls, Ibr-resis ibrutinib resistance, Post-ibr after ibrutinib therapy, Pre-ibr prior to ibrutinib therapy. ns: P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. In addition to Tcm (CD45RO + CD62L + ), TCF1 and BCL11B were co-expressed in a small subset of Teff-like cells (CD8 + CD45RA + CD62L-). Compared to 0.91% (95% CI: 0.41–1.33) in Pre-ibr patients (n = 16), the median percentage of TCF1 + BCL11B+ Teff-like cells was 2.39% (95% CI: 0.70–4.31) in patients treated with ibrutinib for <1 year (n = 8), and 2.9% (95% CI: 1.32–4.76) in patients treated with ibrutinib for >1 year (n = 11) (P = 0.0276 and 0.0004 versus the Pre-ibr level, respectively). There was a statistically non-significant trend for a higher proportion of TCF1 + BCL11B+ Teff-like cells in patients who were exposed to ibrutinib for >1 versus <1 year (Fig. 5C). The proportion of TCF1 + BCL11B + T cells decreased to the baseline level once disease progressed (n = 3). The percentage of TCF1 + BCL11B + T cells was considerably lower in Tex cells (PD1+Tim3 + ) than in non-exhausted T cells (non-Tex: PD1-Tim3-) (Supplementary Fig. 7). The expression of granzyme B and perforin (GrB+perforin + ) detected by FCM (n = 4) was increased in CD8 + T cells following ibrutinib treatment (Fig. 5D), and the percentage of TCF1 + PD1 + T cells increased significantly (n = 16) (Fig. 5E), indicating enhanced T cell cytotoxic activity. A variety of immunosuppressive molecules, including Tim3, CTLA4, and PD1, were downregulated after ibrutinib treatment, with a decreasing proportion of PD1 + CD8 + T cells (Supplementary Fig. 5D, E). In scRNA-seq data analysis, a small subset of Teff-like cells co-expressing TCF1 and BCL11B were barely detectable prior to treatment but emerged at 30 days of ibrutinib in CLL patients who achieved CR after 2–3 years of ibrutinib (n = 4) (Fig. 6A). Notably, this co-expression population showed significant increase in Teff-like cells by 120 days (P < 0.05 versus baseline) (Fig. 6B). However, in one patient with progressive disease (PD) on ibrutinib28, the proportion of TCF1 + BCL11B+ Teff-like cells increased after 3 months of ibrutinib treatment, then decreased to baseline levels at 27 months as the disease progressed (Supplementary Fig. 8). TCF1 + BCL11B + CD8+ Teff-like cells exhibit robust cytotoxic activity. A Scatter plot of dynamic changes of CD8 + T cells expressing TCF7 or/and BCL11B. B Line graph depicting the dynamic changes of the proportion of TCF1 + BCL11B+ cells in Teff-like, Tem, Tex, Trm cells during ibrutinib treatment. C Dynamic changes of cytotoxicity score in CD8 + T cells, Teff-like cells, TCF1 + BCL11B + CD8 + T cells, and TCF1 + BCL11B+ Teff T cells during ibrutinib treatment. D Dynamic changes in cytotoxicity score across CD53+ Teff, Teff, and CDC42+ Teff cells during ibrutinib treatment. HCs healthy controls, Teff-like effector T cell-like, Tem, effector memory T cell, Tex exhausted T cell, Trm tissue resident memory T cell. Error bar: SD. ns: P > 0.05; *P < 0.05; **P < 0.01; ****P < 0.0001. Cytotoxicity scores from healthy individuals (GSE298578 dataset) were included as healthy controls (Supplementary Fig. 9). Cytotoxicity profiling demonstrated that ibrutinib significantly enhanced T cell effector function and expanded the Teff-like subset in patients with CLL. Cytotoxicity score in total CD8 + T cells increased significantly by 30 days (P < 0.0001 versus baseline) and by 120 days (P < 0.0001 versus day 30). The Teff subset displayed significantly higher cytotoxicity than bulk CD8 + T cells at both 30 and 120 days (Both P < 0.0001). Trends for increasing cytotoxicity score were also observed in TCF1 + BCL11B + CD8 + T cells and TCF1 + BCL11B+ Teff subsets (Fig. 6C). Cytotoxicity score increased progressively in Teff-like cells following ibrutinib treatment (Fig. 6D). TCF1/BCL11B interaction and T cell immunity Assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq) demonstrated lower chromatin accessibility at the TCF7 and BCL11B loci in four CLL patients at baseline versus healthy controls, a gradual increase during ibrutinib treatment towards normal levels upon CR, and then a decrease to baseline level upon subsequent PD (Fig. 7A). TCF1 and BCL11B synergize to maintain T cell immunity during ibrutinib treatment. A ATAC results showing TCF7 and BCL11B chromatin accessibility in healthy controls (left) and CLL patients who achieved CR (middle) and PD (right) on ibrutinib. B The average signal heatmap relative to the TSS with ±2 Kb flanking regions. The heatmap was obtained from a CUT&Tag assay for TCF1 (left). Representative UCSC browser track of the BCL11B locus in chromatin accessibility of CUT&Tag assay for TCF1 and average TSS expression (right). C The average signal heatmap relative to the TSS with ±2 Kb flanking regions. The heatmap was obtained from a CUT&Tag assay for BCL11B (left). Representative UCSC browser track of the TCF7 locus in chromatin accessibility of CUT&Tag assays for BCL11B and average TSS expression (right). D Volcano plot showing the differential gene peaks detected by CUT&Tag assay for TCF1 in Pre-ibr and Post-ibr samples (left); Enrichment GO BP analysis results of upregulated TCF1 target genes (right). E Volcano plot showing differential peak genes detected by CUT&Tag assay for BCL11B in Pre-ibr and Post-ibr samples (left); Enrichment GO BP analysis results of upregulated BCL11B target genes (right). BP biological process, CR complete remission, GO gene ontology, PD progressive disease, Post-ibr after ibrutinib therapy, Pre-ibr prior to ibrutinib therapy, TSS transcription start site. Cleavage under targets and tagmentation sequencing (CUT&Tag-seq) analysis demonstrated significant increases in the number of TCF7 and BCL11B binding peaks across various genomic functional regions following ibrutinib treatment (Supplementary Fig. 10). TCF1 binding to BCL11B was enriched at the TSS (transcription start site) of the BCL11B gene upon CR versus the baseline (Fig. 7B). Reciprocally, BCL11B binding was enriched at the TSS region of the TCF7 gene (Fig. 7C). GRAMM rigid docking produced a total of 10 effective conformations, with the first-ranked conformation showing the strongest binding energy of -38.4 kcal/mol between TCF1 and BCL11B (Supplementary Fig. 11), indicating that the two target proteins can form a stable interaction on their surfaces. Based on the surface model diagram, TCF1 and BCL11B are in contact and bonded, with the complex’s stability enhanced through hydrogen bonds, supporting the direct interaction between TCF1 and BCL11B. The TCF1 CUT&Tag-seq dataset showed 7537 differentially enriched TCF1-binding peaks, including 3378 upregulated genes (e.g., CD69, FOXP1, STAT6, SIRT1, and IRF4) and 3026 downregulated genes (e.g., BCL2, IKZF2, and IL2). Enriched BPs included signaling transduction, cell differentiation, cell surface receptor signaling pathway, and positive regulation of response to stimulus (Fig. 7D). The BCL11B CUT&Tag-seq dataset showed 6,325 differentially enriched BCL11B-binding peaks: 3,257 increased bindings (e.g., IRF4, ICOS, FOXP1, SIRT1, and IFNGR1) and 2147 decreased occupancies (e.g., IKZF2, DKK1, NOS2, and CNTN4). Enriched BPs included cell differentiation, immune response, signaling transduction, positive regulation of cell communication, and synapse assembly (Fig. 7E). The BPs enriched by TCF1 and BCL11B target genes showed considerable overlap and primarily involve effector function of CD8 + T cells. Discussion The key finding in the current study is upregulation of TCF1 and BCL11B, as well as an increased proportion of Teff-like cells following ibrutinib treatment in CLL patients. Teff-like cells were derived primarily from memory T cells. Notably, a small subset of Teff-like cells co-expressing TCF1 and BCL11B was identified and exhibited a high cytotoxic activity score. At a mechanistic level, chromatin accessibility at TCF7 and BCL11B binding sites was low prior to ibrutinib treatment and increased upon treatment response to ibrutinib. Enhanced interaction between TCF7 and BCL11B also seemed to contribute to restoring T cell immunity upon ibrutinib treatment. Accumulating evidence suggests that ibrutinib regulates T cell immunity through off-target effects, such as inhibiting IL-2 inducible T-cell kinase (ITK) and blocking the cytokine network3–6. The current study demonstrated that ibrutinib therapy affects both the number and function of T cells. In brief, CD8+ Tex cells were gradually eliminated, leading to the gradual normalization of the inverted CD4/CD8 ratio after about a year. Conversely, the proportion of Teff-like cells and their cytotoxicity increased significantly, suggesting that ibrutinib treatment reshapes T cell immunity. Additionally, consistent results across multiple trajectory inference methods and the shift in canonical gene-expression patterns provide supportive evidence for a differentiation pathway from Tcm to Teff cells. We found a notable upregulation of TCF1 and BCL11B, along with decreased inhibitory receptor expression upon ibrutinib treatment response, but sharp decreases in TCF1 and BCL11B upregulation upon subsequent PD. Of note, we identified a small subset of Teff-like cells co-expressing TCF1 and BCL11B in CLL patients who achieved CR on ibrutinib treatment. These findings are consistent with our previous observation that the upregulation of TCF1 and BCL11B positively correlated with improved PFS and OS in CLL patients25. Consistent with previously reported roles of TCF1 and BCL11B in maintaining stem cell-like characteristics and T cell effector function29,30, cell fate trajectory analysis in this study indicated strong differentiation potential in T cells expressing both TCF1 and BCL11B. TCF1 + BCL11B + CD8 + T cells exhibited a gradually increasing cytotoxicity score following ibrutinib therapy, particularly TCF1 + BCL11B+ Teff-like cells, raising the possibility that Teff-like cells co-expressing TCF1 and BCL11B may represent a specific subset of T cells, precursor effector T cells31, responsible for durable anti-tumor activity following ibrutinib therapy. TCF1 is a key TF that maintains the self-renewal and anti-tumor activity of stem-like CD8 + T cells through interaction with other molecules15,16. In the current study, chromatin accessibility at TCF7 and BCL11B binding sites gradually increased in patients with a favorable response to ibrutinib. Increased chromatin accessibility and mutual interaction between TCF1 and BCL11B were accompanied by increased expression of downstream target genes involved in T cell differentiation, effector function, and signal transduction, e.g., FOXP1, IFNGR1, IL2, and ICOS32,33. Interestingly, FOXP1 was found among the upregulated DEGs by both TCF1 and BCL11B. This finding is consistent with the role of FOXP1 in promoting differentiation of memory T cells, reducing T cell exhaustion, and maintaining long-lasting anti-tumor immunity in T cell-mediated cellular immunotherapy33. These findings shed light on the roles of TCF1 and BCL11B in ibrutinib-induced reconstitution of T cell immunity, as they collaboratively facilitate the differentiation of functional Teff-like cells. In conclusion, the current study revealed an association between upregulation of TCF1 and BCL11B in Teff-like cells with favorable responses to ibrutinib treatment in CLL patients. In addition to the implication in developing novel treatments for CLL, TCF1 + BCL11B+ Teff cells could be used to evaluate, and possibly predict the effectiveness of ibrutinib treatment, and perhaps therapeutic interventions targeting T cells in general. Methods Study design GEO databases were used to investigate the transition of CD8 + T cell subsets in CLL patients during the course of ibrutinib therapy. Findings were then validated in a cohort of CLL patients undergoing ibrutinib monotherapy at the authors’ centers. The study design is shown in Fig. 1. Sample acquisition and clinical data This is an observational study with longitudinal follow-up of a subset of CLL patients undergoing ibrutinib monotherapy at the authors’ centers from March 2017 to August 2024. The diagnosis was based on the revised guidelines of the International Workshop on Chronic Lymphocytic Leukemia/National Cancer Institute1. The ORR included CR, partial remission, and partial remission with increased lymphocytes. Patients’ characteristics are detailed in Supplementary Table 1. All ethical regulations relevant to human research participants were followed. This study was reviewed and approved by the Ethics Committee of the Affiliated Cancer Hospital of Zhengzhou University (Henan Cancer Hospital). Informed consent was waived by the ethics committee. Bulk RNA-seq datasets The bulk RNA-seq GEO datasets (GSE254718) were analyzed34. Gene expression profiles of T cells and CLL-B cells were derived from 10 and 11 CLL patients, respectively, each providing a comparison of DEGs before and after 6 months of ibrutinib monotherapy. The DAVID database was used for the GO analysis of DEGs35,36. scRNA-seq datasets scRNA-seq data for PBMCs were retrieved from three GEO databases. The GSE111014 database37 included 13 scRNA-seq PBMC samples in four CLL patients who achieved CR after 2–3 years of ibrutinib; total number of cells included in the analysis was 15,007, 13,908, and 14,398 prior to, 30, and 120 days after ibrutinib monotherapy, respectively. The GSE152469 database28 included three scRNA-seq PBMC samples from one CLL patient with PD after 27 months of ibrutinib treatment; total cell numbers were 4468, 3975, and 5288 prior to, 90 days, and 27 months, respectively. The GSE298578 database included three scRNA-seq PBMC samples from three healthy individuals. Quality control, data processing, and determination of cell types Cells with <200 detected genes or >15% of the reads mapped to mitochondrial genes were removed from further analysis using the R package “Seurat” (v5.2.1)38. Only genes expressed in >10 cells were included in the analysis. PBMCs were annotated based on average expression of the following marker genes: CD3E, CD3D, CD3G, and CD247 for T cells; CD19, CD79A, CD79B, and CD5 for CLL cells; CD19, CD79A, and CD79B for B cells; FCGR3A, NKG7, and NCAM1 for natural killer cells; LYZ, CD14 and FCER1A for monocytes; PPBP and PF4 for platelets. T cells were clustered into subsets using the Seurat “subset” function. CD4+ (CD4) and CD8+ (CD8A, CD8B) T cells were initially identified. A total of 8150 CD8 + T cells were acquired and analyzed. CD8 + T cells were grouped into 9 cell types based on marker genes and auxiliary annotations concerning the “MonacoImmuneData” dataset through “SingleR” (v2.4.1)39. TCF1 + BCL11B+ cells were defined as expression of both TCF7 and BCL11B at a level of >1. Differential gene expression and pathway enrichment analysis GO enrichment analysis was conducted using the R package “clusterProfiler” (v4.10.1)40. The “FindMarkers” function in Seurat (v5.2.1) with the parameters “min.pct = 0.1” and “logfc.threshold = 0.25” was used to identify DEGs in 9 types of CD8 + T cells. Then, the “enrichGO” function was used for pathway enrichment analysis of DEGs in Teff-like cells. Reconstructing cell differentiation trajectories The R package “Monocle2” (v2.26.0) was used to reconstruct the pseudotime trajectory of CD8 + T cell subset differentiation. Raw counts were normalized using the estimateSizeFactors and estimateDispersions functions (default parameters). The differentialGeneTest function identified the top 1000 most variable genes (ranked by Q-value) for pseudotime analysis. Dimensionality reduction and trajectory inference were conducted using DDRTree. To further validate the pseudotime results from Monocle2, the R packages “Monocle3” (v1.4.26) and “Slingshot” (v2.16.0)41 were additionally applied to infer trajectories using independent computational frameworks. Expression enrichment scores of cytotoxic activity The PercentageFeatureSet function in the Seurat package (v5.2.1)38 was used to calculate the cytotoxicity score based on the following 29 genes: CXCL13, PRF1, GZMH, GZMB, GZMA, BATF, LYST, TIGIT, LAG3, HAVCR2, PDCD1, PRDM1, RBPJ, SLAMF7, CCL4, TNFSF10, TBX21, NFATC1, HIF1A, TNFRSF4, CXCR6, PTPN7, EWSR1, UBXN1, PSMB9, LY6E, CCNDBP1, IRF9, and LCP213,42. Flow cytometry Multi-parameter FCM was conducted on a BD FACS LSRFortessa 4 L (BD Biosciences, San Jose, CA, USA). Cells were stained with anti-human antibodies: CD3-APC/A750 (A94680, Beckman Coulter), CD4-KrO (A96417, Beckman Coulter), CD8-APC/A700 (B49181, Beckman Coulter), Tim3-BV421 (345008, Biolegend), PD1-APC (379207, Biolegend), CD45RO-ECD (IM2712U, Beckman Coulter), CD45RA-PE/Cy7 (B10821, Beckman Coulter), CD62L-PE/Cy5 (304808, Biolegend), BCL11B-FITC (ab123449, Abcam), TCF1-PE (14456, Cell Signaling Technology), Granzyme B-FITC (560211, BD Pharmingen), and Perforin-PE (556437, BD Pharmingen). Matching isotype mAbs were purchased from BD Biosciences. Data analysis was performed using FlowJo (v.10.5.3). Patient characteristics (23 CLL patients before and after one year of ibrutinib treatment and 10 healthy controls) are shown in Supplementary Table 2. Western blot assay PBMCs were lysed in RIPA buffer (P0027, Beyotime) containing protease inhibitors (P002, NCM) for Western blot of TCF1/TCF7 (2203, Cell Signaling Technology) or BCL11B (ab28448, Abcam), using beta-actin (AB0035, Abways) as the internal control. A Chemiluminescent Imaging System (SCG-W3000, Servicebio) was used for signal visualization. Image analysis was conducted using ImageJ software. The assay was conducted in triplicate. RNA extraction and RT-qPCR Total RNA from PBMCs was extracted and reverse-transcribed into cDNA using commercial kits from Solarbio (R1200 and RP1105). RT-qPCR was carried out on the 7500 Fast Real-Time PCR System (Applied Biosystems, USA) using 2× SYBR Green PCR Mastermix (SR1110, Solarbio). GAPDH was used as an internal reference. Primer sequences are shown in Supplementary Table 3. Transcript level was quantified using the 2-ΔΔCT (cycle threshold) method. Patient characteristics (21 CLL patients before and after one year of ibrutinib treatment) are shown in Supplementary Table 4. ATAC-seq analysis ATAC-seq datasets were obtained from GSE11101337. The dataset comprised 35 CD8 + T cell samples from 7 patients with CLL after 0–280 days of ibrutinib therapy and 3 healthy control subjects. The bigwig ATAC-seq data were visualized using the UCSC Genome Browser’s trackhub. CUT&Tag-seq assay CUT&Tag-seq was performed to detect the interaction between TCF7 and BCL11B. Briefly, PBMCs were sorted using CD3 microbeads (130-117-038, Miltenyi Biotec) to derive CD3 + T cells with >95% purity and >90% activity. Aliquots of CD3 + T cells (approximately 200,000 cells) were washed twice in 1.5 mL of wash buffer for the next steps. The following antibodies were used for CUT&Tag-seq assay: BCL11B (ab28448, Abcam) and TCF1 (2203, Cell Signaling Technology). Amplified libraries were obtained using the Hyperactive Universal CUT&Tag Assay kit for Illumina (Vazyme, TD903) and sequenced by Seqhealth Technology Co. (Wuhan, China). MACS2 software was used with a q-value threshold of <0.05 for peak calling and screening. The UCSC browser was used to visualize genome-wide, normalized signal coverage tracks. The GO analysis of upregulated genes in CUT&Tag-seq was conducted using the GO website (https://geneontology.org/). The characteristics of 4 patients with CLL are shown in Supplementary Table 5. Protein-protein docking Protein structures of TCF1 and BCL11B were retrieved from the UniProtKB database. The GRAMM analysis was conducted using the default settings for protein-protein docking. The primary evaluation metric was binding energy (with a standard threshold of <−4 kcal/mol, where lower values indicate more stable binding). Properties such as interaction surface area, hydrogen bonds, and key amino acid residues were analyzed to identify the best docking conformation. PDBePISA was used to calculate binding free energy, and PyMOL 3.1 was employed for visualization. Statistics and reproducibility Survival outcomes were analyzed using the Kaplan–Meier method followed by a log-rank test. For parametric continuous variables, inter-group comparisons were performed using Student’s t test for independent samples. Nonparametric longitudinal analyses between paired time points were conducted using the Wilcoxon signed-rank test. The results are reported as mean ± standard deviation (SD). scRNA-seq data were analyzed using R (v. 4.3.3) (R Core Team, Vienna, Austria)43. Analytical scripts and code are available upon request from the corresponding authors. P < 0.05 (2-sided) was considered statistically significant. R, SPSS v.19.0 (SPSS, Chicago, IL) and GraphPad Prism 9 (GraphPad Software Inc, La Jolla, CA) were used for analyses and graphing. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Reporting Summary Acknowledgements This study was supported by Zhongyuan Sci-Tech Innovation Leading Talents (244200510008), Henan Province Science and Technology Research and Development Program (232102311123, 242102310233), National Natural Science Foundation of China (82200222), Clinical Study and Transfer Project of Henan Cancer Hospital (ITTA2412), and Henan Clinical Research-oriented Doctor Program (HNCRD202510). Author contributions Q.S.Y. designed the study, analyzed the data, and wrote and revised the manuscript. G.L. and T.T.L. analyzed the data, performed the statistical analysis, and wrote the manuscript. S.W., X.D.L., and X.J.W. conducted FCM analyses and Western blot experiments. H.A., S.K.Y., H.F.W., Q.W., H.M.Z., J.L., and X.J.L. participated in patient management. G.L. and T.T.L. carried out bioinformatics analysis. G.L., T.T.L., and H.A. contributed equally to this work. All authors read and approved the final manuscript. Peer review Peer review information Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: Johannes Stortz. Data availability All data supporting the findings described in this manuscript are either shown in the article and in the Supplementary Information or available from the corresponding author upon request. Source data are provided as a Source Data file. CUT&Tag-seq data for CD3 + T cells from CLL patients treated with ibrutinib at the authors’ center were deposited in GSA-Human (accession number HRA017731) at the Chinese National Genome Science Data Center. Other sequencing data were obtained from publicly available sources. The scRNAseq analysis was performed on GSE111014 and GSE152469 datasets of PBMCs from CLL patients treated with ibrutinib. The bulk RNAseq analysis was performed on the GSE254718 dataset. ATAC-seq analysis was performed on GSE111013. Code availability All custom code used for data processing and analysis in this study is publicly available on GitHub and permanently archived in Zenodo44 10.5281/zenodo.19597054), ensuring reproducibility of all analyses reported in this work. 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. These authors contributed equally: Ge Liu, Taotao Liang, Hao Ai. Supplementary information The online version contains supplementary material available at 10.1038/s42003-026-10201-3. References Associated Data Supplementary Materials Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Reporting Summary Data Availability Statement All data supporting the findings described in this manuscript are either shown in the article and in the Supplementary Information or available from the corresponding author upon request. Source data are provided as a Source Data file. CUT&Tag-seq data for CD3 + T cells from CLL patients treated with ibrutinib at the authors’ center were deposited in GSA-Human (accession number HRA017731) at the Chinese National Genome Science Data Center. Other sequencing data were obtained from publicly available sources. The scRNAseq analysis was performed on GSE111014 and GSE152469 datasets of PBMCs from CLL patients treated with ibrutinib. The bulk RNAseq analysis was performed on the GSE254718 dataset. ATAC-seq analysis was performed on GSE111013. All custom code used for data processing and analysis in this study is publicly available on GitHub and permanently archived in Zenodo44 10.5281/zenodo.19597054), ensuring reproducibility of all analyses reported in this work.