AUTHORS
He Tan, PhD, Yegnasew Teferi, PhD2, Giorgio Napolitani3, Stella Bouziana, MD, PhD4, Paul Buckley, PhD5, Liron Barnea Slonim6, Robbert Hoogeboom7, Piers E M Patten, MB ChB PhD8
AFFILIATIONS
1 King's College London, London, United Kingdom
2 King's College London, London, United Kingdom
3 King's College London, London, United Kingdom
4 King's College Hospital, London, United Kingdom, King's College London, London, United Kingdom
5 King's College London, London, United Kingdom
6 King's College Hospital, London, United Kingdom
7 King's College London, London, United Kingdom
8 King's College Hospital, London, United Kingdom, King's College London, London, United Kingdom
ABSTRACT
The transcription factor T-bet (TBX21) has been reported to suppress proliferation of chronic lymphocytic leukemia (CLL) cells, with cases showing expression in peripheral blood (PB) CLL cells having improved overall survival. Paradoxically, T-bet can be induced by microenvironmental stimuli that lead to NF-κB signaling, a pathway central to activation and proliferation in CLL. To address this apparent paradox, we assessed T-bet expression across CLL compartments and its relationship to proliferation and tumor microenvironment (TME) stimuli.
T-bet was assessed in CLL and non-Hodgkin lymphoma cell lines, in cells from PB, spleen and lymph node (LN) from Eμ-TCL1 mice, in primary human CLL cells obtained from PB and LN fine-needle aspirates (FNA) and in human formalin-fixed paraffin-embedded (FFPE) LN tissue. T-bet expression and cell phenotypes were analyzed using qPCR (2?ΔCt vs HPRT), conventional and imaging flow cytometry, confocal immunofluorescence, and mass cytometry (CyTOF). In vitro, T-bet protein induction was determined following IFNγ, CpG and/or IgM stimulation in primary human PBMC cultures and following stimulation using irradiated CD40L-expressing fibroblasts ± IL21.
T-bet expression in CLL lines (MEC-1, PCL-12) was benchmarked against T-bet+ MYD88-mutated ABC-DLBCL lines (OCI-LY3, HBL-1), and T-bet- GCB-DLBCL and Burkitt lymphoma (Ramos). At mRNA and protein level, T-bet was highly expressed in MEC-1 (2?ΔCt 0.234; 1650 gMFI), comparable to OCI-LY3 (2?ΔCt 0.225; 2403 gMFI). T-bet was detectable in PCL-12 (2?ΔCt 0.039; 952 gMFI), similar to HBL-1 (2?ΔCt 0.002; 1051 gMFI). T-bet was undetectable in GCB-DLBCL lines and Ramos (2?ΔCt non-detectable; ~500–600 gMFI). T-bet in CLL lines reached levels comparable to MYD88-mutated lines despite lacking MYD88 mutations.
In the Eμ-TCL1 model, CLL B cells showed detectable, consistent T-bet. Human PB CLL cells had detectable T-bet (~100 gMFI, n=11, above FMO controls; p<0.01), but lower than internal T-cell and NK-cell controls (~1.7-fold and ~5.5-fold higher than CLL cells, respectively). In contrast, LN flow cytometry revealed a distinct T-bet-high CLL B-cell population (T-BetHI). Across 7 paired samples, T-BetHI frequency was higher in LN (8–9%) than PB (2–3%) (p<0.05). Imaging flow cytometry confirmed nuclear T-bet co-localizing with DAPI in LN cells, whereas PB T-BetHI often showed non-nuclear signal. T-BetHI cells were confirmed in FFPE LN sections from 8 independent CLL patients (2–4.5% per patient).
T-BetHI LN CLL cells were larger, blast-like and more proliferative than T-BetLO. In LN CLL cells (n=5), T-bet was highest in G2 (~1100 gMFI), lower in G1 (~600 gMFI), and lowest in G0 (~350 gMFI; all p<0.05). CyTOF analysis showed T-BetHI had higher Ki67, BCL-2 and PD-1 (all p<0.05) and a trend to lower CXCR4/CXCR5 vs T-BetLO. In FFPE LN tissue, CD20+ T-BetHI were more likely to be Ki67+ than T-BetLO (OR=29.2, p<0.0001, paired Wilcoxon p=0.016, n=8).
To assess what signals could contribute to T-bet induction, CLL PBMCs were cultured in vitro. In 3-day cultures (n=11), CpG increased T-bet gMFI over medium (~123–139 ΔgMFI); IFNγ + CpG gave the strongest induction (~215–230 ΔgMFI), whereas IFNγ alone had limited effect (~68–105 ΔgMFI). CpG upregulated the IFNγ receptor (T-bet-high IFNγ-receptor-high cells 5.0% medium vs 63.1% CpG). IgM alone, or combined with IFNγ or CpG, did not increase T-bet (all ns, n=11), indicating BCR engagement alone is an insufficient signal. In all conditions, ΔT-bet did not differ between U-CLL and M-CLL. 6-day co-cultures with CD40L-expressing fibroblasts increased T-bet vs CD40L- controls (n=10; ~450 gMFI; p<0.01). CD40L+IL-21 gave stronger induction (p<0.001; ~2000 gMFI), with T-bet highest in G2 fraction.
T-bet expression in CLL is variable and dynamic, with highest levels in a distinct T-BetHI population within LN, sites of known tumor proliferation. These T-BetHI cells were enriched for Ki67 and markers of activation and apoptosis resistance (BCL-2, PD-1), consistent with a proliferative, tissue-activated state. In vitro, T-bet could be increased in PB CLL cells by microenvironmental signals (IFNγ, TLR9/CpG, CD40L/IL21), indicating induction is context-dependent rather than intrinsic to a CLL clone. Together, these findings suggest T-bet marks a tissue-conditioned, proliferative CLL state induced by microenvironmental signaling, and raise the possibility that T-bet acts as a negative-feedback regulator restraining malignant expansion.
DISCLOSURES
He Tan, PhD: No relevant conflicts of interest to declare.