Nebivolol prevents exhausted T cells and enhances cytotoxicity against MCF-7 breast cancer cells in a β2-adrenergic receptor-dependent manner.

Introduction
T cell exhaustion is a dysfunctional state that arises from sustained antigenic stimulation, as seen in chronic infections and cancer [1], which diminishes the cytolytic capacity of CD8+ cytotoxic T cells (Tc) and is correlated with poor clinical outcomes [2, 3]. Although CD8+ T cell exhaustion is well characterized, the exhaustion dynamics of CD4+ helper T cells (Th) and their interplay with Tc cells remain less understood. Exhausted T cells are defined by a combination of functional and phenotypic features, including reduced effector cytokine secretion (e.g. TNF and IFN-γ), impaired cytotoxicity, and the upregulation of inhibitory immune checkpoint receptors such as PD-1 and TIM-3 [4]. These changes are accompanied by extensive transcriptional and metabolic reprogramming. A principal regulator in this context is the nuclear factor of activated T cells (NFAT), particularly NFATc1, which accumulates in the nucleus under chronic stimulation [5] and drives the expression of exhaustion-associated genes, including thymocyte selection-associated HMG BOX (TOX), which renders T cells dysfunctional [6]. CD38 is elevated in terminally exhausted T cells and contributes to metabolic dysfunction and reduced effector capacity, particularly within the tumor microenvironment (TME) [7, 8]. CD8 T cell exhaustion was documented in a variety of viral infections and tumor models using PD-1, TIM-3, CD38, cytokine production, and cytotoxicity as reliable metrics to evaluate exhaustion status [9–15].
These exhaustion-driven impairments pose a major obstacle to the success of cancer immunotherapies such as immune checkpoint blockade and adoptive cell transfer strategies, including chimeric antigen receptor (CAR) T cell therapy. Although checkpoint blockade aims to reverse T cell dysfunction by disrupting inhibitory signaling, and CAR-T cells are engineered to directly target tumor antigens, their efficacy remains limited due to persistent TME-induced exhaustion, especially in solid tumors [16]. Among solid tumors, breast cancer is distinguished by a particularly complex and immunosuppressive tumor microenvironment, which reveals a high proportion of exhausted T cells, contributing to impaired antitumor immunity and resistance to immunotherapy [17]. The MCF-7 cell line was derived from breast ductal carcinoma and has been used extensively in cancer research [18]. Considering the limitations of current immunotherapies, understanding the regulatory mechanisms underlying T cell exhaustion is essential to designing new strategies that restore durable T cell function in cancer.
Solid tumors such as prostate, pancreatic, and breast cancers stimulate the growth of sympathetic nerves, resulting in local noradrenaline release and sustained β-adrenergic receptor (AR) activation within the TME [19, 20]. Sympathetic signaling is associated with tumor progression, metastasis, and apoptosis resistance [21–23]. In murine cancer models, β-adrenergic signaling, particularly through β2-AR, impaired CD8+ effector function, promoted metabolic dysfunction, and exacerbated T cell exhaustion, effects mitigated by β2-AR-deficient mice (ADRB2−/−) or a pan β-AR blockade [24, 25].
Nebivolol, though clinically classified as a β1-AR blocker, is now well recognized as a β2-AR biased agonist, preferentially engaging β-arrestin signaling over canonical cAMP/PKA pathways [26]. This pharmacological profile provides a rationale for testing its effects on T cell function, as biased β2-AR signaling may attenuate the immunosuppressive outcomes of canonical β2-AR activation and instead promote alternative signaling supportive of T cell activity. Beyond its cardiovascular applications in hypertension, angina, and heart failure, recent studies have shown that nebivolol modulates T helper cell differentiation and suppresses NF-κB activity in human memory CD4+ T cells via β2-AR signaling [27, 28]. Despite this evidence, the impact of nebivolol on T cell exhaustion and cytotoxicity against cancer cells has not been investigated. Here, we demonstrate that nebivolol decreases T cell exhaustion marker expression and restores cytotoxic activity in both CD4+ and CD8+ T cells. These findings suggest that nebivolol could be further explored as an immunomodulatory target for reinvigorating exhausted T cells in cancer therapy.

Materials and methods

T cell isolation and culture
Peripheral blood mononuclear cells (PBMCs) were collected from healthy adult participants following a published protocol [27], after obtaining informed consent from the participants. The study was approved by the Concordia University Human Research Ethics Committee (certificate 30009292) and conducted according to the Declaration of Helsinki guidelines. Individuals under 18 years, those with underlying medical conditions, or those taking disqualifying medications were excluded. Additionally, blood donation was postponed if participants had recently received vaccinations or used recreational drugs within the previous 2 weeks. According to the manufacturer’s instructions, pan (CD3+) T cells were isolated from PBMCs using the EasySep® Human T Cell Isolation Kit (Stemcell Technologies). Purity was evaluated by flow cytometry (FACS Verse, BD Biosciences) using anti-CD3-Peridinin-chlorophyll-Cyanine5.5 (BD Biosciences, clone OKT3) and anti-CD56 (NCAM-1)-Alexa Fluor 488 (BD Biosciences, clone B159) antibodies. The analysis confirmed that CD3+ cells accounted for >97% of the enriched population, and there were negligible CD56+ cells (Supplementary Figure 1).
Purified T cells were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, and 2 mM EL-Glutaplus (Wisent Bioproducts, CA). Cells were incubated at 37°C in a humidified atmosphere with 5% CO2.

T cell activation and repeated stimulation
To replicate chronic antigenic stimulation in vitro, purified human T cells were resuspended at a density of 1.87 × 106 cells/mL and repeatedly activated using 5 µl/mL of ImmunoCult™ Human CD3/CD28/CD2 T cell activator (Stemcell Technologies). T cells underwent repeated stimulation every 48 h over 8 days, with fresh media and treatment conditions replenished at each interval. The conditions were unstimulated cells (No activation), single-round stimulated cells (1× activation), two rounds of stimulated cells (2× activation), and cells subjected to four rounds of activation (4× activation). Stimulation was conducted either in the presence or absence of different β-AR modulators. These included a selective β2-AR agonist (terbutaline sulfate, 10 μM, Sigma Aldrich, USA), a β2-AR biased agonist (nebivolol hydrochloride, 10 μM, Sigma Aldrich, USA), a non-selective β-AR agonist (isoproterenol, 10 μM, Cayman Chemical, USA), and a selective β1-AR antagonist (metoprolol tartrate, 10 μM, Sigma Aldrich, USA). Where indicated, metoprolol was administered 30 min before nebivolol (Met + Neb). A vehicle control (VC) group was included, in which cells received glycerol at an equivalent dilution. Experimental groups were tested in triplicate wells.

Assessment of T cell viability and proliferation
Cell viability was measured at the 8-day endpoint using trypan blue exclusion. T cells were manually counted using a hemocytometer, distinguishing live cells from dead ones based on membrane integrity and reporting viability percentage. For proliferation analysis, T cells were pre-labeled with the cell-permeable dye CFDA-SE (carboxyfluorescein diacetate N-succinimidyl ester) before the initial activation cycle. Following the 8-day culture period, the fluorescence intensity profile was assessed via flow cytometry to evaluate dye dilution across cell divisions. The proliferation index, representing the mean number of divisions within the population of cells that had undergone at least one division, was calculated.

Flow cytometric detection of exhaustion markers
Exhaustion marker profiling was conducted by measuring the surface expression of PD-1, TIM-3, and CD38 using multicolor flow cytometry. Treated T cells were stained with fluorophore-conjugated monoclonal antibodies specific to CD4-Allophycocyanin (BD Biosciences, clone RUO), CD8-Phycoerythrin-Cy7 (BD Biosciences, clone RPA-T8), PD-1 (CD279)—fluorescein isothiocyanate (BD Bioscience, clone MIH4), TIM3 (CD366)—Brilliant Violet™ 421 (BD Bioscience, clone 7D3), and CD38-Phycoerythrin (BD Biosciences, clone HIT2). Samples were incubated with the antibody cocktail at saturating concentration for 30–45 min on ice, protected from light. After staining, cells were washed twice with buffer (1% FBS in phosphate-buffered saline) by centrifugation at 433 × g for 5 min and then resuspended in 500 µl of the same buffer for acquisition. Data acquisition was performed on the BD FACSVerse flow cytometer, recording a minimum of 50,000 events per sample. Each condition was analyzed in technical triplicate. Fluorescence compensation was conducted using single-stained controls. Isotype controls (mouse IgG1κ conjugated to FITC, PE, and BV421) were included to define non-specific background and set gating thresholds appropriately (Supplementary Figure 2). FlowJo v10 software (BD Biosciences) was used to analyze the data. Exhaustion phenotypes were evaluated within CD4+ and CD8+ T cells by applying a Boolean gating to quantify PD-1+CD4+ and PD-1+CD8+ subsets, followed by analysis of TIM-3+CD38+ co-expression within these populations.

Quantitative real-time PCR (RT-qPCR) analysis of TOX and ADRB2 transcripts
Total RNA of treated T cells from different conditions was extracted using PureLink™ RNA Mini Kit (BD Biosciences), and complementary DNA synthesis was performed using the iScript™ Reverse Transcription Supermix (Bio-Rad Laboratories, USA). Quantitative PCR was carried out with TaqMan™ probe-based assays (ThermoFisher Scientific) on targeted TOX (Hs04264584_m1) and ADRB2 (Hs00240532) genes. PPIA (Hs99999904_m1) was used as the endogenous control gene. For each condition, at least two technical replicates were analyzed. Relative expression levels were calculated using the 2−ΔΔCt method. The expression data were normalized to PPIA and expressed as fold change relative to the 1× activation group, which was set to 1.

Nuclear NFATc1 protein quantification by immunoblotting
Nuclear fractions of treated T cells were separated using the NE-PER™ Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific, CA). Protein concentrations were determined using the Bradford assay kit (Bio-Rad Laboratories, USA). Published protocols [27] described that 15 μg of nuclear protein from each sample was resolved on SDS-polyacrylamide gels (SDS-PAGE) and transferred to nitrocellulose membranes. After blocking overnight at 4°C in TBST (0.1% Tween-20 in Tris-buffered saline) containing 5% (wt/vol) non-fat milk (Anatol Spices, Montreal, Canada), membranes were incubated with primary antibodies against NFATc1 (1/500, clone 7A6, Biolegend, CA) and Lamin A/C (1/500, clone LASS2D9, Biolegend, CA) at 4°C for another overnight step. Lamin A/C, a structural component of the nuclear envelope, was used as a nuclear-specific loading control and normalization factor to verify nuclear protein integrity and ensure accurate quantification of nuclear NFATc1 levels. After washing, blots were incubated with appropriate HRP-conjugated secondary mouse antibody (1/1500, BioRad Laboratories, USA) in 5% skim milk at room temperature. Chemiluminescence signals were developed using ECL prime reagent (RayBiotech, USA) and visualized using the ChemiDoc™ XRS+ System (Bio-Rad Laboratories, USA). Densitometric analysis of the immunoblots was performed using Image Lab™ Software (Bio-Rad). NFATc1 band intensity was quantified relative to Lamin A/C, and data were expressed as the NFATc1/Lamin A/C ratio.

Co-culture cytotoxicity assay
Cytotoxicity of treated T cells was assessed using an MCF-7 cell line. Four hours before the tumor cell killing assay, 1 × 105 MCF-7 cells were seeded in each well of a flat-bottom 48-well plate in 200 µl of fresh RPMI-1640 medium supplemented with 10% FBS, 1% penicillin-streptomycin, and 2 mM EL-Glutaplus.
Treated T cells were washed and adjusted to 0.5 × 106 cells/mL (1 × 105 cells per 200 µl of fresh medium per well), establishing a 1:1 effector-to-target (E:T) ratio. Co-cultures were incubated for 24 hours under standard culture conditions. MCF-7 cell death was quantified using LIVE/DEAD™ fixable green dead cell stain fluorescence (Thermo Fisher Scientific, CA) with CD3+ events excluded to ensure only target cells were analyzed.
The relative killing efficiency was determined using the following equation:
In this equation, the viability of treated target cells refers to the survival rate of MCF-7 cells co-cultured with either CD3+ T cells 1× activation or 4× activation with or without treatment. The viability of untreated target cells corresponds to the survival rate of MCF-7 cells cultured without T cells over the same duration.

Cytokine production evaluation
The production of IFN-γ and TNF by T cells was assessed using enzyme-linked immunosorbent assay (ELISA). Cytokine levels were measured in culture supernatants collected at 2-, 4-, and 8-day post-activation, following a 24-hour co-culture of T cells with MCF-7. All ELISA assays were run according to the manufacturer’s protocols. Any measurement falling more than one standard deviation from the group mean was excluded from the dataset.

CRISPR/Cas9-based gene editing knockout of ADRB2 in T cells
Rested CD3+ T cells were activated at a density of 1 × 106 cells/mL using ImmunoCult™ Human CD3/CD28/CD2 T Cell Activator in RPMI-1640 medium supplemented with 10% FBS, 100 U/mL penicillin/streptomycin, and 100 IU/mL recombinant human IL-2 (Fisher Scientific, Ottawa, ON). After 48 h of activation, T cells were washed and resuspended in fresh pre-warmed complete medium at a density of 1 × 106 cells/mL. Cells were rested for an additional 24 h before gene editing to minimize stress-induced responses.
According to the published protocol [27], gene editing involved assembling ribonucleoprotein (RNP) complexes by incubating 100 pmol optimized multi-single-guide RNA (sgRNA) (sequence details in Supplementary Table 1) (Synthego, USA) and 50 pmol of Cas9 protein (ThermoFisher, CA) for 10 minutes at room temperature. A total of 2 × 106 T cells were washed in PBS, resuspended in Buffer R, and combined with the prepared RNP complexes. Electroporation was performed using the Neon™ transfection system with program settings (1600 V, 10 ms, 3 pulses). Following electroporation, T cells were placed in a pre-warmed recovery medium containing 10% FBS, 100 U/mL penicillin/streptomycin, and 400 IU/mL recombinant human IL-2, then incubated at 37°C, 5% CO2 for 72 h to maintain cell viability and functional integrity.
Gene editing efficiency was confirmed via RT-qPCR for ADRB2 mRNA expression and flow cytometry detection of T cell receptor (TCR) alpha constant (TRAC) protein levels as a positive control. Negative controls included non-electroporated cells and cells transfected with non-targeting gRNA (nt-gRNA). Upon validation, both edited and control T cells underwent the 8-day exhaustion protocol.

Statistical analysis
Statistical analyses were performed using one-way or two-way ANOVA, depending on the experimental design. One-way ANOVA was used to assess differences across single-factor experiments, while two-way ANOVA was employed to investigate the combined effect of repeated stimulation and β2-AR treatments. Tukey’s post hoc test was performed in pairwise comparisons between treatment groups. Results are reported as mean ± standard error of the mean (SEM). Significance thresholds were defined as follows: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****). The number of independent biological replicates is provided in the relevant figure legends. All statistical analyses were conducted using GraphPad Prism 9.

Results

Repeated activation causes T cell exhaustion, characterized by diminished cytokine production and impaired cytotoxic function
To establish an in vitro model of T cell exhaustion, purified human CD3+ T cells were repeatedly stimulated with CD3/CD28/CD2 activators every 48 h, with media replenishment, for a maximum of four stimulation cycles over 8 days (Fig. 1a). This experimental design enabled the comparative assessment of T cell exhaustion phenotypes across different activation states, including unstimulated cells (No activation), early activation (1× activation), intermediate activation (2× activation), and repeated activation T cells (4× activation). At the 8-day endpoint, cell viability remained unchanged across activation cycles (Fig. 1b), indicating that repeated stimulation did not induce apoptosis. Moreover, the proliferation index increased after 2× activation, but no change was observed across activation cycles (Fig. 1c). To assess functional exhaustion, IFN-γ and TNF were measured in culture supernatants collected at distinct time points. IFN-γ increased following 1× and 2× activation relative to non-activated controls but declined markedly under 4× activation conditions (Fig. 1d). In contrast, TNF peaked after 1× activation but declined, showing a substantial reduction at 2× activation and becoming nearly undetectable at 4× activation (Fig. 1e). Thus, TNF was more sensitive to being exhausted than IFN-γ when comparing the effects of the 2× activation samples. To evaluate cytotoxic function, T cells from each activation condition were co-cultured with MCF-7 breast cancer cells for 24 hours, and tumor cell viability was assessed. Repeated 4× activation of T cells exhibited a substantial loss of cytotoxicity compared with early activation T cells (Fig 1f and g).

Repeated activation progressively upregulated exhaustion markers PD-1, TIM-3, and CD38 in both CD4+ and CD8+ T cells
To determine the impact of repeated stimulation on exhaustion marker expression, PD-1, TIM-3, and CD38 were analyzed in CD4+ and CD8+ T cells by flow cytometry at the day-8 endpoint. Representative histogram plots gated on CD4+ (Fig. 2a) and CD8+ (Fig. 2e) T cells illustrate expression patterns of PD-1, TIM-3, and CD38 across 1×, 2×, and 4× activation conditions. In CD4+ T cells, PD-1 remained comparable between non-activated and 1× samples, PD-1 showed a slight increase at 2×, and it was further elevated at 4× (Fig. 2b). TIM-3 levels did not differ significantly at 1× or 2× but were significantly upregulated under 4× stimulation (Fig. 2c). CD38 in CD4+ T cells was increased at 2× and significantly higher at 4× activation compared with earlier conditions (Fig. 2d). In CD8+ T cells, PD-1 increased progressively across stimulation conditions, with the highest levels observed at 4× (Fig. 2f). TIM-3 remained unchanged at 1× and 2× but was significantly upregulated following 4× activation (Fig. 2g). Similarly, CD38 was markedly increased at 4× relative to both 1× and 2× (Fig. 2h). Collectively, these results indicate that repeated stimulation leads to increased expression of exhaustion-associated markers, particularly under 4× conditions, in both CD4+ and CD8+ T cell subsets.

Nebivolol reduces exhaustion marker co-expression and improves cytotoxic function in exhausted T cells
To investigate whether nebivolol modulates T cell exhaustion, repeatedly stimulated T cells were treated with nebivolol and compared with terbutaline. As expected, 4× activation significantly increased the frequency of CD4+TIM-3+CD38+ cells compared with 1× activation. Terbutaline, a β2-AR-specific agonist, did not further enhance the exhaustion under 4× stimulation. In contrast, nebivolol significantly reduced the frequency of CD4+TIM-3+CD38+ cells relative to both the 4× condition and the vehicle control, which is used to gauge the effect of the solvent used to dissolve nebivolol.
To determine whether β1-AR contributed to the observed effect, metoprolol, a selective β1-AR blocker, was included as a receptor-selective control as a classic pharmacological approach to distinguish β1- versus β2-AR involvement, not to assess drug–drug interactions. Under these conditions, metoprolol neither altered exhaustion nor interfered with nebivolol’s activity (Fig. 3a and b). Regarding gated CD8+ T cells, similar patterns were observed. The 4× activation increased CD8+TIM-3+CD38+; terbutaline had no effect, nebivolol decreased exhaustion markers, and metoprolol again failed to block nebivolol’s effects (Fig. 3c and d).
Next, T cells were co-cultured with MCF-7 breast cancer cells to evaluate functional consequences. Consistent with their phenotype, 4×-activated T cells exhibited less cytotoxicity relative to 1×-activated T cells. Terbutaline and isoproterenol (a non-selective β1/β2 agonist) did not further increase the cytotoxicity in the 4× condition. Nebivolol treatment, however, significantly restored cytotoxic activity in 4× exhausted T cells, reaching levels comparable to 1× controls (Fig. 4a). At the cytokine level, 4× exhausted T cells produced significantly less TNF than 1×-activated cells. Nebivolol restored TNF secretion, whereas terbutaline and isoproterenol produced only modest, non-significant increases (Fig. 4b). Together, these data show that nebivolol reduces exhaustion marker expression and restores effector function in both CD4+ and CD8+ T cells.

Nebivolol restored ADRB2 expression while reducing TOX and nuclear NFATc1 levels in exhausted T cells
To evaluate the impact of nebivolol on transcriptional regulation, ADRB2 and TOX mRNA and nuclear NFATc1 were measured. Compared with 1× activation, 4× exhausted T cells displayed lower ADRB2. Terbutaline had no effect on ADRB2; in contrast, nebivolol increased ADRB2 expression (Fig. 5a). The 4× exhausted T cells had higher TOX when compared with 1× activation. Terbutaline had no effect on TOX. Nebivolol markedly reduced TOX expression in 4× exhausted T cells to the levels comparable to 1× activation (Fig. 5b). Similarly, the 4× exhausted T cells elevated nuclear NFATc1 when compared with the 1× activation. Terbutaline had no effect on nuclear NFATc1. Nebivolol markedly reduced nuclear NFATc1 expression in 4× exhausted T cells back to the levels seen in the 1× activation (Fig. 5c and d). Thus, nebivolol reinvigorates exhausted T cells as shown by increased ADRB2, and decreased TOX and nuclear NFATc1.

β2-AR deficiency suppressed exhaustion marker co-expression and restored cytotoxic function in exhausted T cells
Although nebivolol is clinically classified as a β1-AR blocker, the lack of effect from metoprolol suggested that its immunomodulatory activity was not mediated through β1-AR. To evaluate the possible contribution of β2-AR, CRISPR/Cas9 was used to disrupt the ADRB2 gene on a genomic level in primary human T cells. The single-exon gene encoding the β2-AR was targeted with a multi-guide RNA strategy in primary human T cells. ADRB2 disruption partially reduced ADRB2 mRNA levels compared with non-electroporated and non-targeting controls (Fig. 6a). TRAC knockout, used as a positive control, demonstrated efficient editing at the protein level compared with the same control groups (Supplementary Figure 3). Surprisingly, ADRB2-disrupted T cells activated 4× displayed reduced exhaustion markers relative to controls. Terbutaline or nebivolol did not further alter exhausted markers in ADRB2-deficient CD4 or CD8-gated T cells (Fig. 6b and c). Functionally, ADRB2-disrupted T cells retained cytotoxic capacity under 4× activation, comparable to 1× activation and drug treatment did not further modify this effect (Fig. 6d). Thus, partial ADRB2 loss was sufficient to prevent the induction of exhaustion in vitro, regardless of drug treatment.

Discussion
T cell exhaustion represents a major barrier to effective antitumor immunity and arises from sustained antigenic stimulation that leads to functional decline, persistent expression of inhibitory receptors, and transcriptional reprogramming [4]. While in vivo studies and patient-derived samples have been instrumental in defining exhausted T cell states [29–31], controlled human in vitro systems remain essential for mechanistic dissection and therapeutic testing. Accordingly, several in vitro protocols have been developed to model exhaustion, including repeated TCR stimulation (e.g. CD3/CD28) over 8–14 days, repeated or continuous peptide antigen exposure in antigen-specific systems, and extended cytokine-supported cultures [32–34]. Across both human and murine models, these approaches consistently induce exhaustion phenotypes, reproducing hallmark features including reduced TNF and IFN-γ secretion, impaired cytotoxicity against tumor cells, and elevated PD-1, TIM-3, and CD38 expression, consistent with in vivo exhaustion profiles [34–37].
Using this framework, we developed a four-cycle repeated stimulation model that reliably induced exhaustion while preserving T cell viability and proliferative capacity, indicating functional rather than activation-induced apoptotic exhaustion. This feature aligns with observations from tumors [38] and chronic infection settings [39], and contrasts with longer stimulation protocols (e.g. ≥13 days) [34], which have been associated with increased cell death driven by metabolic stress and suppressive cytokine environments [40, 41]. Functionally, repeated stimulation resulted in progressive impairment of cytokine production and tumor cell killing, with an earlier decline in TNF compared with IFN-γ, consistent with previously reported hierarchical sensitivity of effector functions during exhaustion [14, 36, 37, 42, 43]. Notably, T cells subjected to only two rounds of stimulation retained robust cytotoxic capacity despite increased inhibitory receptor expression, highlighting a defined transition point beyond which exhaustion predominates. Together, these findings position our model as methodologically consistent with established in vitro exhaustion systems while offering a controlled and reproducible framework for further mechanistic interrogation.
The key innovation of our model is the intentional incorporation of β-AR signaling into this validated exhaustion framework. Unlike all the existing systems, which rely primarily on TCR engagement, co-stimulation, and cytokine modulation, our system introduces a neuroendocrine stress component through β2-AR engagement during chronic T cell activation. β-AR signaling has been recognized as a modulator of immune suppression in the TME, yet the impact of β2-AR signaling pathways, especially those activated by biased agonists, on exhaustion dynamics remains incompletely understood. In a murine breast cancer model, Qiao et al. demonstrated that chronic adrenergic stress drives metabolic dysfunction and promotes an exhausted phenotype in tumor-infiltrating CD8+ T cells, while a β-AR blocker (propranolol) restored effector activity and improved antitumor immunity [25]. More broadly, β-agonist activity within TME has been associated with immunosuppression, as reviewed by Qiao et al. and Eng et al., who highlighted adrenergic signaling as a checkpoint that constrains T cell function and impairs responsiveness to immunotherapy [44, 45].
In our study, terbutaline, a short-acting canonical β2-AR agonist, had no detectable effect on T cell proliferation, exhaustion marker expression, or anti-tumor effector function. This finding is consistent with a few preclinical cancer studies reporting neutral effects of β2-AR agonism on tumor cell proliferation, viability, or invasion in breast cancer cell models [46]. In contrast, pro-tumorigenic effects of β2-AR signaling have been reported in a context-dependent manner, particularly under conditions of sustained or high-intensity adrenergic stimulation, such as chronic stress exposure or treatment with longer-acting β2-AR agonists [47, 48], and in specific tumor settings [49]. Mechanistically, terbutaline and isoproterenol predominantly signal through the canonical Gs–cAMP–PKA pathway [50], which can suppress proximal TCR signaling; however, in our system, this effect may be overridden by strong tri-antibody activation (anti-CD3/CD28/CD2) or limited by the transient signaling profile of terbutaline compared with long-acting β2-AR agonists such as formoterol and salmeterol [51].
Nebivolol, as a β2-AR-biased agonist, has not previously been examined in the context of T cell exhaustion. In our study, nebivolol markedly reduced exhaustion-associated features, enhanced TNF production, and restored cytotoxicity against MCF-7 breast cancer cells. Importantly, these findings are consistent with established clinical and preclinical literature describing nebivolol’s β-arrestin-biased signaling and antioxidant properties, as well as recent cancer-related studies demonstrating that nebivolol suppresses tumor growth and viability in melanoma and carcinoma models, effects not observed with conventional β1-selective blockers [52, 53]. The absence of similar effects with metoprolol further supports that nebivolol’s activity is independent of β1-adrenergic blockade and aligns with its recognized non-canonical β2-AR signaling profile [26–28].
At the transcriptional level, repeated stimulation of T cells led to a marked decrease in ADRB2 expression, consistent with prior observations that prolonged activation and stress reduce β2-AR expression in T cells [27, 54, 55]. Terbutaline did not alter this downregulation, whereas nebivolol restored ADRB2 expression to levels comparable to single stimulation. This suggests that the effect cannot be simply explained by receptor desensitization from repeated agonist exposure, which is a well-documented outcome of sustained β2-AR stimulation [56]. Instead, biased signaling through nebivolol may stabilize ADRB2 expression by preferentially engaging β-arrestin pathways, thereby avoiding the classical internalization and degradation processes [51].
Mechanistically, nebivolol also reduced TOX and nuclear NFAT levels, disrupting key exhaustion-associated transcriptional programs. This contrasts with terbutaline, which showed minimal effects, aligning with previous evidence that canonical β2-AR signaling can induce TOX [57] and calcineurin inhibition to indirectly suppress NFAT activity [58]. Together with our previous finding that nebivolol downregulates NF-κB [27], these results indicate that biased β2-AR signaling exerts a broad reprogramming effect on T cells distinct from both classical agonists and β1-AR antagonists.
To determine whether nebivolol acts on β2-AR, we employed a CRISPR/Cas9-mediated ADRB2 gene disruption approach. Interestingly, ADRB2-deficient T cells failed to develop exhaustion after repeated stimulation, even without drug treatment. This suggests that endogenous β2-AR signaling is required for the acquisition of an exhausted phenotype, potentially due to tonic sensitivity to low levels of catecholamines present in serum or produced by T cells themselves [59]. The paradox that both nebivolol treatment and ADRB2 disruption reduced exhaustion may reflect their distinct mechanisms: nebivolol stabilizes receptor expression and promotes non-canonical β-arrestin pathways, whereas ADRB2 deletion prevents canonical signaling altogether. However, because our editing efficiency was partial, we cannot exclude residual receptor activity or compensatory pathways, and overexpression rescue experiments would be informative in clarifying this relationship.
We also note that while ADRB1 and ADRB2 are the predominant adrenergic receptors expressed in human T cells, ADRB3 is absent in naïve and memory subsets [27], suggesting that nebivolol’s effects are unlikely to involve ADRB3 signaling. In addition, we previously found that ICI-118551 only partially blocks nebivolol [28], likely due to differences in binding site engagement [60]. Our current study emphasizes CRISPR-based disruption as a more direct approach. Pharmacological dissection using GRK–β-arrestin pathway inhibitors such as paroxetine or CMPD101 may further refine understanding of nebivolol’s signaling bias.
Together, these findings support a model in which repeated TCR stimulation induces functional T cell exhaustion, while biased β2-adrenergic signaling selectively disrupts exhaustion-associated transcriptional and functional programs without compromising viability; this integrated model is summarized schematically in Fig. 7. Future studies should extend these findings to subset-specific cytotoxic assays and in vivo tumor models and explore whether targeting biased β2-AR signaling can be leveraged in combination with checkpoint blockade or adoptive cell therapies.

Supplementary Material
uxag018_Supplementary_Data