Abstract
Background.Diffuse large B cell lymphoma (DLBCL) represents the most common lymphoma subtype (30%-40% lymphomas). Gene expression profiling (GEP) identifies two main subtypes of DLBCL defined by their postulated cell of origin (COO). Activated B cell-like DLBCL (ABC DLBCL) is characterized by activation of B-cell receptor signaling and the nuclear factor kB pathway: these patients respond worse to standard R-CHOP but better to novel agents (ibrutinib or lenalidomide) than patients with DLBCL of the germinal-center-type (GCB DLBCL). We have previously reported an upregulation of the transcriptional factor ETS1 in up to 25% of DLBCL (Bonetti, Testoni et al. Blood, 2013). ETS1 is involved in many biologic processes, including B cell differentiation. ETS1 undergoes a series of post-translational modifications, and, in particular, ETS1 phosphorylation at Threonine 38 (p-ETS1) is a marker for ETS1 activation. Here, we present the impact of p-ETS1 in DLBCL cellular models and in a large series of clinical specimens.
Patients and Methods. Levels of p-ETS1 (ab59179, Abcam) and, as controls, of ETS1 (C-20, Santa Cruz Biotechnology), ERK (93, Santa Cruz Biotechnology), p-ERK-Tyr204 (p-ERK; 7383, Santa Cruz Biotechnology) and IRF4 (4964, Cell Signaling Technology) were analyzed in cell lines derived from ABC (n.=7), GCB (n.=8) and type 3 (n.=4) DLBCL. p-ETS1 was examined by immunohistochemistry on clinical specimens: cases were defined as positive if > 10% of the neoplastic cells nuclei were p-ETS1 positive.
Results.p-ETS1 was detected in ABC, not in GCB DLBCL cell lines (100% vs 0%, P<0.05), but also in 3/4 Type 3 (75%). All the cell lines expressed ETS1 and ERK. The ABC marker IRF4 was expressed in all ABC (100%), 1/4 type 3 (25%) and 0/8 GCB cell lines (0%). p-ERK was expressed in 6/7 ABC (86%), 2/4 Type 3 (50%) and 0/8 GCB (0%). Since ETS1 is a transcription factor and p-ETS1 is marker for its activation, we looked at the pattern of expression between nucleus and cytoplasm in five ABC DLBCL cell lines. p-ETS1 was present predominantly in the nucleus of all cell lines, while total ETS1 was in both cytoplasm and nucleus in 4/5, and only in the nucleus in 1/5 (TMD8). To evaluate the mechanisms sustaining p-ETS1, two ABC DLBCL cell lines (U2932, TMD8) were exposed to the PI3K-delta inhibitor idelalisib (1 μM), the BTK inhibitor ibrutinib (0.5 μM) and the MEK inhibitor pimasertib (0.5 μM) after stimulation or no stimulation with anti-IgM. In both cell lines, inhibition of BTK or MEK decreased the baseline levels of p-ETS1, while only inhibition of MEK was able to inhibit the increase in p-ETS1 induced by IgM stimulation. PI3K-delta inhibition only lead to a minimal reduction of the baseline p-ETS1 levels. Similar changes were seen for p-ERK. To understand the clinical significance of our findings, we assessed p-ETS1 expression in DLBCL biopsies. First, on a small series of 14 DLBCL cases classified for their COO using the Hans algorithm, p-ETS1 was mostly detected at nuclear level. The percentage of p-ETS1 positive cells was higher in ABC (no.= 7) than in non-ABC (no.= 7) (P 0.023), in agreement with the preclinical data. We then extended the analysis to a cohort of GEP-classified 315 de novo DLBCL cases treated with R-CHOP regimen collected as part of The International DLBCL Rituximab-CHOP Consortium Program Study (Xu-Monette et al, Blood 2013). The presence of p-ETS1 was further confirmed to be more frequent in ABC than in GCB cases: 79% (123/155) vs 57% (91/160) (P < 0.001). The p-ETS1 positive GCB DLBCL presented an inferior progression free survival (PFS) than p-ETS1 negative cases (P 0.034), while no association with outcome was detected in ABC DLBCL.
Conclusions. Our data identify ETS1 phosphorylation at threonine 38 as i) associated with the ABC DLBCL phenotype; ii) associated with poor PFS in GCB DLBCL; iii) downstream to BCR signaling and amenable to pharmacological interventions with BTK and MEK inhibitors.
Rossi:Gilead: Honoraria, Research Funding; Abbvie: Honoraria; Janseen: Honoraria.
Author notes
Asterisk with author names denotes non-ASH members.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal