ALBERT

Description: Large granular lymphocyte (LGL) leukemia results from clonal expansion of CD3+ cytotoxic T-lymphocytes (CTL) or CD3-natural killer (NK) cells. Chronic antigen stimulation promotes long-term survival of LGL leukemia cells through constitutive activation of multiple survival pathways, leading to global dysregulation of apoptosis. Clinical manifestations of LGL leukemia include neutropenia, anemia and rheumatoid arthritis. Treatment for LGL leukemia patients relies on immunosuppressives such as methotrexate and is not curative. No standard therapy has been established. We reported that nuclear factor kappa B (NF-kB) is central to the leukemic LGL survival network, but the mechanisms of constitutive NF-kB activation in LGL leukemia are undefined. TNF-related apoptosis-inducing ligand (TRAIL) is a potent inducer of apoptosis and activates NF-kB via binding to TRAIL receptor (TR) 1, 2 and decoy receptor 2 (DcR2). DcR2 is uniquely unable to transduce downstream death signals yet retains NF-kB transduction activity. The mechanisms of TRAIL expression and regulation in LGL leukemia are unknown. Thus the current study investigates these mechanisms and their potential therapeutic applications, with the use of NF-kB inhibitors ixazomib and bortezomib, in LGL leukemia. Methods: LGL leukemia cell lines TL1 (T-LGL) and NKL (NK-LGL), peripheral blood mononuclear cells (PBMC) from LGL leukemia patients, and PBMC from normal donors were studied. NF-kB DNA binding activity was determined by EMSA. Results were confirmed using probe-based NF-kB (p50/65) transcription factor assay and immunocytochemistry (ICC). Serum TRAIL levels were detected by ELISA. Cellular TRAIL expression was determined by real-time PCR, western blot and ICC. DcR2 and Mcl-1 siRNA knock-down was performed with electroporation. Flow cytometry was used to detect TR 1-3 and DcR2 expression and apoptosis. Results: The average serum levels of TRAIL in LGL leukemia patients were nearly 4-fold higher than normal (NL) control values (p ≤ 0.0001). Data from RT-PCR (p ≤ 0.04), western blot and ICC revealed that LGL leukemia cells were the major source of TRAIL overexpression. Identical expression levels of TR1, 2 and 3 were observed in PBMC from LGL leukemia patients and from NL controls. Like normal PBMC, LGL leukemia cells were resistant to TRAIL-induced apoptosis. In contrast, the expression frequency of DcR2 was at least 4-fold greater in LGL leukemia PBMC compared to NL control, and it correlated to the percentage of circulating LGL leukemia cells. We found that TRAIL activates NF-kB and NF-kB downstream target genes, including TRAIL and McL-1, in LGL leukemia samples. To confirm that TRAIL is responsible for constitutive NF-kB activation in LGL leukemia, T-LGL leukemia PBMC were treated with pooled sera from 3 each of either NL controls or T-LGL leukemia patients. Leukemia sera increased NF-kB activity on EMSA, and this effect was completely blocked by TRAIL neutralizing antibody. DcR2 siRNA knockdown specifically decreased RelA and NF-kB1 (p105/p50) levels in TL1 and NKL cells. Mcl-1 siRNA mediated increased apoptosis in the same cell lines. Likewise, ixazomib and bortezomib facilitated leukemia-selective apoptosis in LGL leukemia cell lines and in patient PBMC, via inhibition of NF-kB activity and of downstream targets (ixazomib, p≤0.0001; bortezomib, p≤0.03). Additionally, caspase-3 and PARP cleavage were observed in LGL leukemia cells treated with ixazomib or bortezomib. Serum TRAIL levels in LGL leukemia patients were significantly lower in methotrexate responders versus non-responders, corresponding with reduced NF-kB DNA binding activity and increased absolute neutrophil counts, indicative of treatment response. Conclusion: These data indicate that expression of DcR2 and constitutive activation of NF-kB are responsible for TRAIL resistance in leukemic LGLs. TRAIL triggers prolonged NF-kB activation via interaction with DcR2, and activated NF-kB in turn promotes further TRAIL production in leukemic LGLs, creating a TRAIL autocrine regulatory loop. Inhibition of NF-kB activity with ixazomib and bortezomib interrupts this loop, impairs expression of Mcl-1 and induces apoptosis of leukemia cells. Our preclinical findings provide a solid framework for clinical evaluations of ixazomib and bortezomib in the treatment of LGL leukemia. Disclosures No relevant conflicts of interest to declare.

Description: Large granular lymphocyte (LGL) leukemia results from chronic expansion of cytotoxic T cells or natural killer (NK) cells. Apoptotic resistance resulting from constitutive activation of survival signaling pathways is a fundamental pathogenic mechanism. Recent network modeling analyses identified platelet-derived growth factor (PDGF) as a key master switch in controlling these survival pathways in T-cell LGL leukemia. Here we show that an autocrine PDGF regulatory loop mediates survival of leukemic LGLs of both T- and NK-cell origin. We found high levels of circulating PDGF-BB in platelet-poor plasma samples from LGL leukemia patients. Production of PDGF-BB by leukemic LGLs was demonstrated by immunocytochemical staining. Leukemic cells expressed much higher levels of PDGFR-β transcripts than purified normal CD8+ T cells or NK cells. We observed that phosphatidylinositol-3-kinase (PI3 kinase), Src family kinase (SFK), and downstream protein kinase B (PKB)/AKT pathways were constitutively activated in both T- and NK-LGL leukemia. Pharmacologic blockade of these pathways led to apoptosis of leukemic LGLs. Neutralizing antibody to PDGF-BB inhibited PKB/AKT phosphorylation induced by LGL leukemia sera. These results suggest that targeting of PDGF-BB, a pivotal regulator for the long-term survival of leukemic LGLs, may be an important therapeutic strategy.

Description: Persistently expanded large granular lymphocytes (LGL) are found in bone marrow failure disorders such as pure red cell aplasia, myelodysplastic syndromes, aplastic anemia, paroxysmal nocturnal hemoglobinuria (PNH), and in LGL leukemia. Many of the characteristics of BMF diseases overlap and it is postulated they share a fundamental pathogenetic link. LGL leukemia represents the best characterized disease of LGL, and was used as a model to study pathogenic LGL proliferation in bone marrow failure diseases. LGL leukemia cells are characteristically resistant to Fas-regulated death, despite expressing high levels of Fas receptor and Fas ligand (FasL). Elevated FasL may cause the neutropenia that is observed in many LGL leukemia patients. A poorly understood feature of LGL leukemia is serorecognition of a discreet epitope derived from the HTLV envelope BA21 region. To query the common pathogenesis in all of the BMF disorders, FasL expression and BA21 seroreactivity levels were obtained for patients with bone marrow failure disease (n=87). The serum FasL levels as determined by antigen capture were inversely correlated with neutrophil counts regardless of disease subtype. A positive correlation was noted between FasL expression and lymphocytosis, basophile counts, and serum chloride mEgL (Pearson correlation p values 0.0621, 0.0224, 0.0465). Negative correlations for FasL included neutrophil count and serum albumin levels (p= 0.0333 and 0.0657 respectively). Initial data from an assay we designed to obtain standardized anti-BA21 IgG data revealed a strong negative correlation for BA21 IgG and FasL (−0.3668) for the PNH cohort. An additional moderately positive Pearson correlation for FasL and BA21 IgG was found only for the LGL leukemia group (0.1813), although all BMF groups demonstrated elevated IgG recognition of BA21 (n=78). In addition, FasL and BA21 IgG levels were most elevated in female patients. Our results suggest that FasL and BA21 IgG levels correlate with certain laboratory features of hematopoeitic inhibition as well as with patient gender. A potential utility in discriminating between bone marrow failure disease subtypes and determining treatment response potentials could also be realized with further study.