Zhao Lab

Chong Yang & Toshio Suda

The programmed cell death protein 1 (PD-1) -programmed death-ligand 1 (PD-L1) axis is emerging as an immune checkpoint that regulates anti-tumour immune responses against solid tumours and haematological malignancies[1] , but its function in T cell acute lymphoblastic leukaemia (T-ALL) leukaemogenesis remains unknown. 
Accumulative studies have demonstrated that tumours are composed of hierarchically heterogeneous populations of cells, with cancer stem cells representing a distinct subset that dwells at the apex of the hierarchy with self-sustaining capability[2] . In haematological malignancies, studies on acute myeloid leukaemia (AML) have identified a panel of cell-surface markers that are associated with leukaemia stem cells (LSCs) in patients with AML and mouse models of xenotransplantation[3] . However, the LSC concept has not been well characterized and remains debatable in the context of human T-ALL. One pressing question is whether the leukaemia-initiating cells emerge from populations with self-renewal potential, such as haematopoietic stem and progenitor cells (HSPCs), or whether they are derived from already-differentiated populations with newly acquired stemness. Some studies have suggested that genetic or epigenetic abnormalities in immature T cells might result in the acquisition of the self-renewal capability.[4] Distinct from the normal haematopoietic system, leukaemic cells seem to exhibit a unique self-renewal capability with a differentiation block; hence, a patient-derived xenograft assay is essential for the identification of LSCs. Therefore, more in-depth investigations of the origin, functions and regulatory mechanisms of LSCs in T cell malignancies deserve urgent attention. In this issue of Nature Cell Biology, Xu et al.[5] report that PD-1 marks a distinct population of LSCs in T-ALL that can potentially be targeted in T-ALL therapy. 
Xu et al. found that bone marrow cells expressing PD-1 reside at the apex of the T-ALL hierarchy and possess stem-cell and leukaemia-initiating characteristics. Although PD-1 is often involved in immune surveillance, the authors clarified the intrinsic PD-1 signalling in LSCs and demonstrated that PD-1-positive (PD-1+ ) LSCs were enriched with quiescence, leukaemia stemness, drug-resistance and gene sets related to the NOTCH1 -MYC pathway. Mechanistically, they proposed that NOTCH1 signalling and MYC activity are enhanced in PD-1+ LSCs to maintain LSC stemness and initiate T-ALL progression. As a result of active NOTCH1 pathways in PD-1+ LSCs, PD-1 expression and signalling are well maintained and subsequently protect the LSCs from activated T cell receptor (TCR)-induced apoptosis (Fig. 1). 
The present study directed our attention to the emerging role of PD-1 in various haematological malignancies. As part of the CD28 family, PD-1 is a checkpoint protein that is remarkably expressed on the surfaces of all activated T cells, as well as monocytes, macrophages, B cells, dendritic cells and natural killer cells. The major regulatory functions of PD-1 include the reduction of harmful inflammatory responses during tissue damage and the evasion of cancer cells to avoid immune responses. Notably, PD-1 disrupts the cooperative TCR -major histocompatibility complex (MHC) -CD8 trimolecular interaction to inhibit the TCR signalling that is responsible for T cell activation[6] . The binding partners of PD-1 consist of PD-L1 and PD-L2, with PD-L1 primarily expressed on tumour cells and responsible for tumour immune modulation. Thus, the induced antitumour immune responses can be reversed by blocking these immunosuppressive machineries, with inhibitors of PD-1 -PD-L1 representing promising candidates to facilitate a positive immune response to kill tumours[7]. 
Recently, increased PD-1 and PD-L1 levels were observed in patients with myeloid malignancies, such as myelodysplastic syndromes (MDS) and AML. Although both in vitro and in vivo preclinical studies have indicated beneficial outcomes of PD-1 -PD-L1 blockade in treating these malignancies, the current clinical outcomes of PD-1 -PD-L1 combination therapy in MDS and AML are mixed. However, several clinical trials have suggested that PD-1 -PD-L1 blockade potentially exerts a synergistic effect when combined with hypomethylating or chemotherapeutic agents and yields promising outcomes in treatment-naive patients[8]. 
Similarly, the effects of PD-1 -PD-L1 blockade against adult T cell leukaemia/lymphoma (ATLL) caused by human T lymphotropic virus type 1 remain controversial owing to the complexity of the disease. Immune evasion caused by structural variations that disrupt the PD-L1 3′ untranslated region (3′-UTR) was reported in ATLL cohorts[9] . It has therefore been suggested that careful evaluation of the ATLL subtype, PD-L1 3′-UTR disruption and mechanisms that underlie drug resistance to PD-1-targeted therapy is required to predict ATLL patient responses and ensure positive safety profiles after treatment[10]. 
Although PD-1-targeted immunotherapy is yet to be validated for T-ALL treatment, previous studies reported high expression of PD-1 (also known as PDCD1) and PD-L1 (CD274) mRNA in patients with T-ALL[11]. Notably, in human T-ALL, the hallmark of disease onset is chromosomal translocation that involves TCR loci on severalchromosomal regions. Another well-established molecular abnormality in T-ALL is the NOTCH1 mutation that results in its aberrant activation. More specifically, mutation in the NOTCH1 heterodimerization domain leads to its cleavage and release of the intracellular form of NOTCH1 protein (ICN1), which is subsequently translocated to the nucleus to regulate transcriptional regulatory complexes containing the DNA-binding protein RBPJ (also known as CSL or CBF1)[12]. This subsequently triggers upregulation of NOTCH target gene expression — including the MYC oncogene, which is known to be involved in cell cycle regulation and cell metabolism — and ultimately leads to T-ALL cell growth, proliferation and leukaemia-initiating cell activity[4,13]. 
In line with previous findings about NOTCH1 activities in T-ALL, the study by Xu et al. demonstrated that increased NOTCH1 signalling was also observed in PD-1+ LSCs and showed transcriptionally regulated and high PD-1 expression in T-ALL cells. Moreover, prominent SHP1 and SHP2 activity in PD-1+ LSCs compromised TCR-activation signals; thus, PD-1 -PD-L1 blockade suppressed T-ALL progression, as the cells lost protection from PD-1 against TCR-signalling-induced apoptosis (Fig. 1). 
Another notable finding is the prominent self-renewal ability of PD-1+ LSCs that expressed not only HSPC markers (that is, c-KIT and SCA1), but also lymphoid lineage proteins, including TCR and IL-7Rα. Notably, even a single PD-1+ LSC was able to bulk generate T-ALL cells, rendering it responsible for disease initiation. Taking the heterogeneous nature of stem cells into consideration, follow-up studies may consider the adoption of a single-cell RNA-sequencing (scRNA-seq) approach to delineate PD-1+ LSC subpopulations, as well as the identification of markers and unique functions associated with PD-1+ LSCs subtypes (such as maintaining stemness or drug resistance). In addition, although autocrine PD-L1 was proposed to have a protective role against TCR-induced cell death in this study, a niche-specific PD-L1 effect might be of interest for future LSC investigations as PD-L1 is also expressed in bone marrow stromal cells, which are often recognized as the microenvironment for HSCs and bone marrow-derived LSCs. 
Importantly, PD-1+ LSCs were shown to be enriched with leukaemic stemness and drug-resistant genes, as well as to contribute to leukaemia recurrence. As LSCs are often considered to be the primary reason for chemotherapeutic drug resistance and leukaemia relapse[14], targeting LSCs could be a potential strategy to improve the long-term survival of patients with T-ALL. Further studies are required to clarify the mechanisms for the relapse of patients that is potentially caused by PD-1+ LSCs. Moreover, the preclinical investigation in this study adopted combination therapy using PD-1 blockade and several chemotherapy agents and demonstrated improved survival of mice engrafted with mouse or human T-ALL cells. These results indicate potential demand for clinical validation of PD-1-targeted immunotherapy to treat patients with T-ALL by eliminating leukaemia-initiating cells from the root.

Fig. 1: Regulatory network involving the PD-1–PD-L1 axis and NOTCH signalling in bone-marrow-derived LSCs in T-ALL.
Before T cell development in the thymus, LSCs emerge in the bone marrow that are marked by high expression of PD-1 as well as c-KIT, SCA1, FLK2 and T cell lineage markers — including TCR and IL-7Rα. NOTCH signalling triggers the release of ICN1 from the plasma membrane and its translocation into the nucleus, and interacts with various transcription factors and cofactors to activate transcription of target genes such as MYC and PDCD1 (which encodes PD-1). Increased PD-1 binds to autocrine PD-L1 and protects the LSCs from overactivated TCR-signalling-induced cell death. Enhanced MYC expression results in the maintenance, self-renewal and disease-initiating capacities of LSCs. AKT, protein kinase B; CLP, common lymphoid progenitor; DN, CD4–CD8– double negative; DP, CD4+CD8+ double positive; LT-HSC, long-term HSC; MPP, multipotent progenitor; PI3K, phosphatidylinositol 3-kinase; ST-HSC, short-term HSC; SP, CD4+ or CD8+ single positive.

References
1. Ferris, R. Cancer 119, E1 -E3 (2013).
2. Kreso, A. & Dick, J. E. Cell Stem Cell 14, 275 -291 (2014).
3. Ding, Y., Gao, H. & Zhang, Q. Stem Cell Investig 4, 19 (2017).
4. Tan, S. H., Bertulfo, F. C. & Sanda, T. Front. Oncol. 7, 218 (2017).
5. Xi, X. et al. Nat. Cell Biol. https://doi.org/10.1038/s41556-022-01050-3 (2023).
6. Li, K. et al. Nat. Commun. 12, 2746 (2021).
7. Alsaab, H. O. et al. Front. Pharmacol. 8, 561 (2017).
8. Yang, X. et al. Exp. Hematol. Oncol. 11, 11 (2022).
9. Kataoka, K. et al. Nature 534, 402 -406 (2016).
10. Jalili-Nik, M. et al. Intl. Immunopharmacol. 98, 107870 (2021).
11. Yang, K. et al. Pathol. Res. Pract. 215, 265 -271 (2019).
12. Weng, A. P. et al. Science 306, 269 -271 (2004).
13. Sanchez-Martin, M. & Ferrando, A. Blood 129, 1124 -1133 (2017).
14. Venton, G. et al. Blood Cancer J. 6, e469 (2016).