Abstract
Dysfunctional T-cells associated with human tumors can be identified utilizing multi-parameter flow cytometry and studied for intracellular signalling checkpoints using phospho-flow or time-of-flight mass spectrometry. Whether tumor-induced hypo-responsive T-cells in patients with myeloma are anergic, exhausted or senescent has not yet been determined. Anergic T-cells are hypo-responsive with low cytokine production and induce tolerance to protect the host from autoimmune disease. T-cell exhaustion is usually associated with chronic viral infection such as CMV but also with some cancers. Both exhausted and anergic T-cells express PD-1, LAG-3, Tim-3, CD160 and CD28. Hypo-responsive T-cells associated with melanoma have the phenotype of exhausted T-cells which can be reactivated using targeted immune check point blockade, resulting in clinical benefit. Senescent T-cells accumulate in an oligoclonal manner with ageing and chronic antigen exposure. Senescence can be characterised by telomere shortening, the expression of CD57, KLRG-1, CD160 and the absence of CD28. However, not all senescent cells have shortened telomeres as telomere independent senescence involving the p21-p53 and p16-pRb pathways and senescence associated secretory phenotype (SASP) T-cells have been described. The aim of this study was to determine whether tumor-induced hypo-responsive T-cells in patients with myeloma are anergic, exhausted or senescent and to identify targetable checkpoints to restore T-cell function.
Cytotoxic T-cell clones (CD57+ CD28- TCRVβ restricted) determined by BetaMark TCRVβ analysis were present in 51% of patients with myeloma (n=264), are protective (OS χ2=6.2; p<0.01) and maintain high cytokine secretion. CFSE proliferation assays of flow sorted T-cell clones demonstrated a significant lack of proliferation after stimulation with MACS iBeads (n=12; p<0.0001). They failed to respond in vitro to cytokines such as IL- 2, IL-12, IL-15 and OX40. In contrast, the T-cell clones were proliferative in a group of long term myeloma survivors in whom such T-cells are universally present suggesting that the impaired proliferation may be reversible. Geneset enrichment analysis of mRNA expression from Affymetrix U133 plus 2.0 arrays demonstrated pleiotropic dysfunction in multiple intracellular signalling pathways involved with proliferation and T cell inactivation. Utilizing phospho-flow techniques we demonstrated reduced phosphorylated ERK levels in clonal vs non-clonal T-cells (n=8; p<0.002) as well as elevated levels of phosphorylated Smad2/3 (n=10; p<0.02) and Bcl-xL (n=7; p<0.04) .These findings suggest that the ERK pathway, associated with proliferation, was suppressed, the TGFβ pathway was activated and that the dysfunctional T-cell clones have a survival advantage due to apoptotic resistance. In contrast, normal SHP-2 and phosphorylated ZAP-70 levels suggested there was no defect in upstream TCR signaling.
Clonal T-cells lacked CD28, PD-1, CTLA-4, LAG-3, CD160 and Tim-3 expression and therefore are neither exhausted nor anergic. However, telomere assessment by qPCR demonstrated telomere T/S ratios were normal length for age and not significantly less than autologous CD57+TCR-Vβ- cells. T-cell clones also had normal p38 expression suggesting that the senescence was not due to TNF-α inhibiting the p38/MAPK pathway.
The cytotoxic T-cell clones in patients with myeloma have the characteristics of telomere-independent senescence. The normal telomere length suggests that the hypo-responsiveness is induced independent of telomeres, and is therefore potentially reversible, though the pleiotropic nature of this dysfunction might require the targeting of multiple intracellular signalling pathways. The lack of PD-1 and CTLA-4 expression suggests that immune checkpoint blockade in myeloma would be unable to reverse the T cell dysfunction.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.