Monoclonal T-cell expansions in asymptomatic individuals and in patients with large granular leukemia consist of cytotoxic effector T cells expressing the activating CD94:NKG2C/E and NKD2D killer cell receptors

Abstract

We have analyzed the phenotype, cytokine profile, and mitotic history (telomere length) of monoclonal T-cell expansions in 5 CD3 ؉ T-cell large granular lymphocyte (TLGL) leukemia patients by fluorescence activated cell sorting (FACS) and single-cell polymerase chain reaction (PCR). We confirm that the common phenotype of TLGL leukemia is CD3 ؉ CD8 ؉ CD45RA ؉ CD27 ؊ CD94 ؉ (CD57 ؉ ). Interestingly, the C-type lectin-like type killer cell receptor CD94 was invariably associated with the activating form of its signaltransducing molecule NKG2. Furthermore, when judged by criteria such as interferon gamma (IFN-␥)/tumor necrosis factor (TNF) production, expression of granzyme, FasL, and NKG2D, the TLGL cells had all the features of a cytotoxic effector T cell. Telomere shortening in TLGL cells was in the normal range for CD8 ؉ T cells, indicating that they had not divided significantly more than chronically stimulated CD8 ؉ T cells in healthy individuals. In 25 of 27 controls, cells with a TLGL phenotype occurred at low (1%-3%) frequencies. However, in the other 2 individuals (ages 28-36 years), large stable (> 3 years) monoclonal expansions of CD3 ؉ CD8 ؉ CD45RA ؉ CD27 ؊ CD57 ؉ CD94 ؉ NKG2C ؉ were found which rendered these controls phenotypically indistinguishable from TLGL leukemia patients. We believe that the TLGL clonopathy, rather than being IntroductionLymphoproliferative disorders of large granular lymphocytes (LGLs) are classified as lymphoid neoplasms. 1,2 They are divided into 2 distinct entities: approximately 10% to 20% are of the natural killer (NK) lineage (CD3 Ϫ CD16 ϩ ), whereas the others consist of mature CD3 ϩ T cells expressing either an ␣␤ or a ␥␦ T-cell receptor (TCR). NK-LGL leukemia is by far the more aggressive form. Most patients have more than 10 ϫ 10 9 /L LGLs in their blood and a significant percentage dies within 2 months after diagnosis. 3 By contrast, CD3 ϩ T-cell LGL (TLGL) leukemia that affects mainly elderly patients progresses in a much milder way. Lymphocytosis is usually modest 4 and many patients may even remain asymptomatic. 3,[5][6][7] Moreover, clinical complications consist primarily of neutropenia or autoimmune features. 8 Therefore, it remains under debate whether this T-cell clonopathy of variable significance 9 should be considered as a malignancy or rather as a secondary reactive phenomenon. 6 Apart from the low malignancy of TLGL leukemia, there are more indications that the clonal expansion is the result of a perhaps rather common dysregulation of CD8 ϩ T cells. In fact, large oligoclonal expansions of CD8 ϩ T cells are quite frequent in elderly people. 10 Such clones that may be maintained by cytokines rather than by antigen 11 are of a phenotype similar to that of normal CD8 ϩ effector T cells of which the clonal size is still controlled through apoptosis. The latter is also true for the TLGL clone. In the majority of patients, these cells are CD8 ϩ CD45RA ϩ CD27 Ϫ , 6,7,12,13 express Fas and FasL,14,15 and may or may not express CD57. 6,7 Therefore, they share m...

The raft marker GM1 identifies functional subsets of granular lymphocytes in patients with CD3+ lymphoproliferative disease of granular lymphocytes

“…As a consequence, a small amount of activated cells should be detected among the clonal population they found, as the expression of the persisting stimulation. 28 Our results seems to validate this hypothesis, by demonstrating the expression of the raft marker GM1 bright , identifying activated cells, expressed in a small but well-defined percentage of GL in all patients with LDGL studied. In addition, since the percentage of GM1 bright GL and the lymphocytosis are extremely stable during time in all the patients analyzed (see Figure 3), our results point to a role of these activated cells in the biology of GL proliferation, possibly involved in the renewal of circulating GL.…”

“…It is not clear whether the lack of CD28, together with the presence of activating NK receptors (namely CD94/NKG2C and NKG2D), which are commonly demonstrated in GL of CD3 þ LDGL patients, could indeed be related to the peculiar pattern of rafts' expression. 28 It is interesting to note that CD94 has been reported to bind HLA-E antigens 29,30 and that this interaction is particularly efficient in the recognition of some viral antigens, namely CMV, EBV and influenza viruses. 31 It can be speculated that the multiple activating signals released by different activating receptors (CD94/NKG2C, NKG2D) 28,32 might be involved in the expression of GM1 bright on the GL surface, although it should be noted that the percentage of cases characterized by the expression of CD94 includes only a fraction of patients under study and NKG2D was not expressed in two cases.…”

“…28 According to this hypothesis, in patients with LDGL the accumulating lymphocytes should be considered as a clonal population likely recognizing viral antigens, which persists in vivo after the clearance of antigens. As a consequence, a small amount of activated cells should be detected among the clonal population they found, as the expression of the persisting stimulation.…”

Expression of inhibitory KIR is confined to CD8+ effector T cells and limits their proliferative capacity

“…FACS re-analysis (>98% cells positive for GL183) and limited dilution cloning (10/ 10 growing clones were positive for GL183) confirmed the purity of the sorted cells. cDNA was prepared from 3,000-5,000 sorted cells and spectratyping was performed as described [43].…”

Bone marrow biopsy related haemorrhage and low molecular weight heparin

“…Antigen stimulation of LGL effector cells and triggering of effector molecules may dysregulate immune mechanisms and thereby cause clinical problems (Bigouret et al, 2003). In line with the idea of antigen-activated cytotoxic effector cells, we have previously shown that T-LGL proliferations often display a polyclonal to oligoclonal Vb repertoire, with many cases showing one or more dominant clones with restricted Vb usage or even clear monoclonal Vb usage (Langerak et al, 2001).…”