Page 17 - 2021_05-Haematologica-web

T-cell dysfunction in CLL
inhibitory receptors with their ligands, excretion of immunosuppressive cytokines by the tumor cells and per- sistent antigen exposure to the T cells.48 Together this results in reduced effector function and a dysfunctional transcriptional program, the latter of which is driven by differential transcription factor activation and dictates cell fate and function.49,50 Research into T-cell exhaustion and dysfunction, which is almost exclusively focused on CD8+ T cells,26,51 led to identification of transcription-factor dynamics that determine these aberrant T-cell pheno- types.
Development of an exhausted and dysfunctional state allows T cells to stay alive, possibly retaining some resid- ual function, although incapable of effectively eradicating the tumor or pathogen.42 Since studies have shown that early cancer-induced T-cell dysfunction is still a plastic dif- ferentiation state,27 reversing or preventing the dysfunc- tional state is the ‘holy grail’ for immunotherapy. Although, transcriptional dynamics of CLL-induced T-cell dysfunction have rarely been studied so far, we argue that several of these mechanisms can be extrapolated to CLL T-cells.
T-cell factor-1 (TCF-1) is a fate-defining transcription factor,52 important for self-renewal and therefore crucial in naive and memory T-cell subsets. It functions as a tran- scription factor and histone deacetylase,53 meaning that it can both activate and suppress gene expression. Chen et al.52 showed that early in CD8+ T-cell differentiation after chronic infection, TCF-1 initiates development of precur- sor exhausted T cells by antagonizing effector T-cell differ- entiation. Loss of TCF-1 in stem-cell like precursor exhausted cells is irreversible in vivo and these cells differ- entiate into terminally exhausted and dysfunctional cells. In human cancer studies, TCF-1 expression on tumor-infil- trating lymphocytes (TIL) is positively correlated to ICB responses,54,55 as TCF-1+ T cells tend to be more persistent than those that have lost TCF-1 expression; a key concept relevant for CAR T-cell production.
T-cell exhaustion is accompanied by global chromatin remodeling, details of which are discussed in the section Epigenetics as regulator of T-cell function. Until recently, the molecular cause of chromatin remodeling was unclear, however, several groups have shown an essential role for the transcription factor thymocyte selection-associated high mobility group box protein (TOX) in terminal T-cell exhaustion.56–60 As is often the case with transcription fac- tors, TOX may play a more versatile and species-specific role than initially thought.61 A recent study showed that TOX+CD8+ T cells exist in the circulation of most humans,62 especially effector memory cells tend to express TOX. Still, human studies also confirmed the involvement of TOX in T-cell exhaustion, as HIV- and HCV-specific CD8+ T cells show high levels of TOX58,62 as well as human TIL.56 Dynamic assessment of TOX expression in T cells during CLL progression or the production of CAR T-cells could indicate the level of exhaustion in these cells.
The tight relationship between two T-box transcription factors, T-bet and Eomes, has been linked to specific check- points in T-cell differentiation.63,64 T-bet is well-known for its role in CD4+-Th1 differentiation by establishing IFNγ expression. In CD8+ T cells, T-bet is similarly involved in effector functions. In addition, T-bet regulates PDCD1, the gene encoding PD-1, expression through an exhaustion- specific enhancer.65 However, T-bet expression is not sus- tained in terminally exhausted T cells; upon loss of TCF-1
expression, a transition from T-bet to Eomes expression takes place in the final stages of differentiation.52,66
Eomes is known for its role in CD8+ T-cell differentia- tion and effector function,50 and is involved in early mem- ory formation in CD8+ cells.67 For these reasons, Eomes is an essential transcription factor for anti-tumor immunity.68 In exhausted cells, Eomes often co-expresses with inhibitory receptors such as PD-1 and TIM-3,69 and is ele- vated in TIL.68,70 Loss of Eomes also leads to increased TCF-1 expression and reduced expression of TOX,68 sug- gesting this could be a target to improve stemness of T cells and prevent terminal exhaustion.
Whilst expression patterns of TCF-1 and TOX are large- ly unexplored in T cells from CLL patients, we can specu- late on their role. The T-cell compartment of CLL patients displays a skewing towards differentiated effector cells and this is likely reflected by a reduction in TCF-1 expres- sion because of its prominent role in self-renewal of naive and memory cells.53,71 A clue to whether TCF-1 plays a role in CLL T cells comes from the TCL1 mouse model; a widely accepted model for human CLL including the T- cell dysfunction.69,72,73 Targeting SLAMF6, a surrogate marker for TCF-1, resulted in a reduced leukemic burden and a reduction in exhausted T cells.74 The role of TCF-1 in self-renewal is an important factor for CAR T-cell devel- opment as persistence and stemness are beneficial to out- come.75 Pre-selection for TCF-1+ cells prior to CAR T-cell infusion, or expansion conditions that induce TCF-1 expression, could be a strategy to improve outcome.
Currently, no studies have found differential TOX expression in CLL T cells neither in patient samples nor in the TCL-1 model.35,69,76 It is likely that CLL T cells have not reached the terminally exhausted state that is associated with high, TOX expression,58 which is also evident by their residual cytokine production. Especially loss of IFNγ expression is known to occur very late in development of exhaustion.42 Possibly, a small subset of CLL T cells are ter- minally exhausted but cannot be picked up by bulk RNA sequencing methods. High-dimensional flow cytometry and single-cell sequencing technologies on both LN and PB compartments could provide new insights into this mat- ter.
High expression of T-bet has been reported in CD8+ T cells of CLL patients, which coincided with an elevated IFNγ expression.24 Since T-bet expression decreases upon differentiation into terminal T-cell exhaustion, there is likely effector skewing rather than true terminal T-cell exhaustion in CLL (Figure 2). With regards to T-bet expres- sion in CD4+ T cells, Roessner et al.31 found increased lev- els in CLL patients compared to healthy controls. The accumulation of T-bet expressing CD4+ T cells was also confirmed in the TCL1 mouse model, however, reducing their number had no impact on CLL development. Thus, the mechanistic role of T-bet in this model is not clear yet. Eomes has been studied in the TCL1 mouse model as well. Knock-out of Eomes in CD8+ and CD4+ T cells result- ed in impaired tumor control, concluding that this tran- scription factor seems essential for adaptive immunity against CLL cells.77 However, this needs to be confirmed in human CLL studies.
Overall, whilst T-bet and Eomes expression patterns have been studied in CLL T cells, the patterns of TCF-1 and TOX expression are relatively unknown in CLL T cells, especially in CD4+ T cells. Since T-cell function arises from a dynamic interplay between transcription factors, it
haematologica | 2021; 106(5)
1237