Butyrophilins (BTN) are a family of proteins encoded by 7 genes: BTN1A1, BTN2A1, BTN2A2, BTN2A3, BTN3A1, BTN3A2, BTN3A3. Since BTN2A3 is a pseudogene, there are only six functional BTNs [1]. They belong to the superfamily of immunoglobulins and structurally resemble the regulatory B7 family of molecules [2,3]. BTNs are strongly expressed in lymphoid tissues, including B cells and T cells [2]. Among them, BTN1A1 is rarely expressed in leukocytes, whereas BTN2A2 is exclusively expressed in these cells. All other BTNs are widely expressed in different tissues, both healthy and cancerous [1]. So far, BTN expression has been primarily studied in solid tumours [4]. For instance, low expression of BTN3A3 is a negative prognostic factor in non-small cell lung cancer [5]. Similarly, higher expression of BTN3A1 in bladder cancer indicates better overall survival [6].

BTN2A1, BTN3A1, BTN3A2, and BTN3A3 are involved in γδ T cell stimulation, while BTN1A1 and BTN2A2 exert inhibitory effects [3]. BTN2A1 and BTN3A1 are crucial for the phosphoantigen-dependent activation of human Vδ2 cells [3,7]. Even though the function of BTN3A1 had been recognised for some time, only the discovery of BTN2A1’s importance was a breakthrough [8]. Under physiological conditions, the phosphoantigen level in normal human cells remains low. However, in numerous infections or neoplastic processes, the concentration of phosphoantigens increases. Phosphoantigens bind the intracellular B30.2 domain of BTN3A1, which leads to its heterodimerisation or heteropolymerisation with BTN2A1 and possibly also BTN3A2. This complex is then recognized by the Vγ9Vδ2 TCR unit, resulting in Vδ2 activation [9–11].

γδ T cells constitute a small fraction of total T cells (approx. 2%–5%), yet they are a crucial part of the immune system due to their ability to rapidly respond to infections, e.g., tuberculosis, and cancer. While they are essential in normal immunosurveillance, γδ T cells may also be implicated in the pathogenesis of various diseases, including multiple sclerosis or asthma [12–14]. They are typically divided into Vδ1, Vδ2, and Vδ3 based on the variable fragment of the T cell receptor (TCR) δ they use [15]. Each subset recognises distinct, partially conserved antigens: Vδ1 primarily binds to self-antigens such as MIC-A (MHC class I polypeptide–related sequence A) or MIC-B (MHC class I polypeptide–related sequence B), Vδ2 responds to phosphoantigens, whereas the antigen targets of Vδ3 remain poorly characterised [16–19].

γδ T cells, especially the Vδ2 ones, are exhausted and dysfunctional in the course of chronic lymphocytic leukaemia (CLL) [20]. While human γδ T cells, both Vδ1 and Vδ2 ones, from healthy donors recognise and potently kill CLL cell lines like MEC-1 in vitro, those derived from CLL patients show severely limited cytotoxic activity [21,22]. CLL patients have previously been divided based on the proliferative capacity of Vδ2 cells into responders and non-responders; the non-responders are more likely to have unmutated IGVH (immunoglobulin heavy chain variable region), an important negative prognostic marker in CLL [23–25]. Notably, IGVH mutation status is of clinical significance. While the classical front-line therapy consists of a fludarabine, cyclophosphamide, and rituximab regimen, patients with unmutated IGVH are less likely to respond well to treatment and fail to achieve long-term remissions. Additionally, they are more frequently affected by severe toxicities, including treatment-related myeloid neoplasms [26]. Various mechanisms underlying impaired proliferation were proposed and tested, but the expression pattern of BTNs in CLL has never been investigated.

We hypothesised that BTNs are significantly dysregulated in CLL patients and that their altered expression contributes to γδ T cell dysfunction, which in turn may affect disease progression and patient survival. To test our assumptions, we extracted the expression data of BTN genes from publicly available datasets and, after normalisation, conducted comparative analyses.

In this study, we identified significant dysregulation of BTN gene’ expression in CLL. Moreover, this dysregulation had significant clinical implications, as higher BTN expression was linked to longer OS.

BTNs are crucial for the phosphoantigen-mediated activation of the Vδ2 subset of γδ T cells. As previously mentioned, γδ T cells in CLL are exhausted and functionally impaired [20]. To date, we lack any mechanistic explanation for this phenomenon. The dysregulation of BTN2A1 and BTN3A1 and their significantly lower expression in CLL are probably at least partially responsible for the γδ T dysfunction. Indeed, as recently proved by Cano et al., the expression of BTN2A1 on cancerous cells is necessary for Vδ2-mediated cytotoxic response [41]. Moreover, their findings indicate that the strength of this response is directly proportional to the surface expression of BTN2A1 on target cells [41].

Interestingly, it is a high expression of BTN3A1 that has negative prognostic value in some solid tumours like glioblastoma, pancreatic ductal adenocarcinoma, or oesophagal cancer [42–44]. On the other hand, according to Liang et al., in pan-cancer analysis of BTN3A1, its high expression is mostly a predictor of longer overall survival, e.g., in lung, breast or gastric cancer; at the same time, the negative prognostic value of BTN3A1 was noted only in testicular germ tumours and low-grade gliomas [45]. This is in line with our observations that BTN3A1 expression in PBMCs is decreased in CLL patients, even more so in those with negative prognostic markers (IGVH unmutated). More importantly, we found that higher BTN3A1 expression is a predictor of longer overall survival in CLL. Indeed, Malinowska et al. reported a case of acute myeloid leukaemia with translocations that involved the BTN3A1 gene and caused loss of heterozygosity; they concluded that it may significantly contribute towards leukaemic cell malignancy [46]. Interestingly, BTN3A2 and BTN3A3 demonstrate the most pronounced separation of survival curves; however, the underlying reasons for this remain unclear. Current literature does not provide an explanation, suggesting that further studies are needed to explain this phenomenon.

Similarly to our observations, BTN2A1, BTN3A1, BTN3A2, and BTN3A3 were significantly downregulated in breast cancer [47]. Moreover, Ren et al. also noted better overall survival in patients with high BTN2A1, BTN3A1, BTN3A2, and BTN3A3 [47]. This is also in line with our observations and further confirms their potential value and implications. Although each BTN has its own functions, they closely cooperate and synchronised expression seems necessary for proper function, most importantly for γδ T activation [48]. Thus, it is also important to analyse the expression of all BTN genes concurrently in the same samples. It may also be valuable to assess the Epstein-Barr virus (EBV) infection in the same samples. EBV is known to regulate the expression of BTN3A1 and BTN2A1 on the one hand and negatively affect the overall survival of CLL patients on the other [49–51]. Both BTN2A2 and BTN1A1 have suppressive potential and inhibit T cell activation and proliferation [52]. Interestingly, we have not observed any impact of BTN2A2 expression on OS, which is contrary to the observations in other tumours, e.g., glioma [53].

While BTN2A1 is critical for γδ T activation and cytotoxic response against cancer, it is also implicated in the transition of macrophages from pro-inflammatory M1 to suppressive M2 phenotype [41,54]. Moreover, Kreneur et al. noted that with the use of blocking anti-BTN2A1 antibodies, they were able to induce a significant shift back from M2 to M1 [54]. While this may be beneficial in some cases, it would also significantly lower the potential of γδ T cells; thus, caution is needed, especially in cases like CLL, where BTN2A1 is already lowered. While anti-BTN2A1 antibodies may have limited potential in CLL, they could be useful in tumours with higher BTN expression, like glioblastoma [55]. Additionally, γδ T cell infiltration is a positive prognostic marker in various solid tumours, e.g., head and neck cancer, where γδ T infiltration is directly proportional to the expression of BTNs in tumour cells [56].

Finally, the pattern of BTN1A1 expression is worth noting. As mentioned in the introduction, BTN1A1 expression is negligible in leukocytes. Our data indicate that, indeed, there is no expression of BTN1A1 in T lymphocytes, but in B cells and total PBMCs, it remains low, yet detectable.

Although our findings contribute to a better understanding of CLL, some limitations must be discussed. The current study was based on a re-analysis of publicly available datasets, which enabled the inclusion of a relatively high number of samples from patients with potentially different sociodemographic backgrounds. At the same time, this approach substantially affected the amount of other data available—the majority of datasets had only very limited clinical information available. The survival analysis is based on a large dataset that includes patients from the USA, Spain, and Germany treated over an extended period. While this diverse dataset allows for the inclusion of a high number of patients with varied sociodemographic backgrounds, it is important to acknowledge that differences in treatment protocols across countries may influence overall survival and introduce potential bias in our results. Another key limitation of our analysis is the use of bulk PBMCs and bone marrow samples, which do not account for variations in CLL cell load among patients. Given that PBMCs in CLL patients are largely composed of malignant B cells, while those from healthy controls contain a diverse mix of immune populations, direct comparisons may be affected by this cellular imbalance. Ideally, the CLL load should be taken into consideration. However, current dataset limitations prevent precise adjustments for CLL cell load. Future validation using isolated cell populations or adjusting for CLL burden would provide a more accurate assessment, and we aim to address this in upcoming studies. Nevertheless, we obtained encouraging results that implicate the need for further studies on fresh CLL samples with concomitant assessment of EBV infection and γδ T cells.

A strong dysregulation of BTNs in CLL samples indicates their potential clinical importance. This is further corroborated by the survival analysis results. Patients with lower expression of BTNs had significantly shorter OS. Moreover, patients with unmutated IGVH, thus those with worse prognosis, demonstrated markedly lower expression of BTNs as well. Notably, the expression of BTN genes in PBMCs had a significant impact on overall survival, highlighting their potential prognostic value. Since BTN expression in PBMCs can be easily assessed at diagnosis using a simple qPCR assay, it could potentially be incorporated into clinical practice to enhance risk stratification and improve patient management. Given the established role of BTN2A1 and BTN3A1 in γδ T cell activation, their downregulation may contribute to γδ T cell dysfunction, potentially affecting immune surveillance and disease progression. Further research is needed to validate these results in prospective cohorts and to explore the underlying mechanisms driving BTN dysregulation in CLL and their importance for the function of γδ T cells.