Expression And Significance Of Unfolded Protein Response Related Signal Molecules In DLBCL

The unfolded protein response (UPR) is fundamentally a cyto-protective response, which mitigates normal and unusual levels of endoplasmic reticulum (ER) stress to promote cell survival and growth by enhancing the capacity of the secretory apparatus and by altering transcriptional and translational programs of the secretory pathway. Recently reports described that UPR activation plays a critical role in normal B cell development and lymphomagenesis. UPR-related molecule PRDM1 is a tumor-suppressor gene whose loss contributes to lymphomagenesis by blocking plasma cell differentiation in ABC-DLBCL. Paradoxically, some published data demonstrated that UPR regulate cytoprotective autophagy and survival in Eμ-Myc B-cell lymphoma. Here, we want to assess the influence of UPR pathway and target genes expression on development and prognosis of DLBCL. We hope our experimantal results will furnish the theoretic foundation and experimental evidence for further researching on DLBCL genotyping and exploring its potential as a pharmacodynamic biomarker.We first focused on 2 major UPR sensors (IREla and PERK) phosphorylation status by IHC analysis in 154 de novo DLBCL biopsies and 30 normal lymphoid tissues. The specimens from DLBCL patients and donor normal lymphoid tissues were divided into 4 staining groups:negative, low (10%-30%), moderate (30%-80%) and high (>80%). We detected that moderate/high staining of IREla-pS724 (121 of 154 cases,78.57%) and PERK-pT981 (103 of 154 cases,66.88%) were more frequent in cases of DLBCL of either GCB or non-GCB/ABC subtype than in the normal lymphoid tissues (11 of 30 cases,36.67%, and 7 of 30 cases,23.33% respectively). We also found that DLBCL patients with moderate/high staining of IREla-pS724 and/or PERK-pT981 had significantly worse overall survival (P=0.0054 and P=0.0386, respectively). We extended these results by combined analyzing some UPR activated features including ATF6 and its cleaved form (ATF6-p50), BiP, IRE1α, PERK and their phosphorylation status in vitro in a panel of unstimulated DLBCL lines. Peripheral circulating B lymphocytes from healthy donors were analyzed as controls. Immunoblotting demonstrated that basal expression of ATF6, BIP, IREla and PERK protein were elevated in some DLBCL lines than in normal B cells. We also detected moderately increased phosphorylation levels of IRE1α and PERK, and very modest levels of ATF6 cleaved form (ATF6-p50) in the same DLBCL lines compared with controls. Although the activation of UPR (including ATF6, IREla, PERK and its substrate eIF2a) was persistently increased in DLBCL lines up to 36 hours with treatment of TQ but no apoptotic cell death (AnnexinV PE-7AAD method) occured in TG-induced DLBCL lines. Due to activation of the UPR sensors (ATF6, IRE1α and PERK) were compatible with secondary activation of ER stress and induced autophagy or apoptosis, we then assessed whether UPR activation in DLBCL cells should have released from the inhibitory ER chaperone BiP. HEK 293 cells displaying sensitivity to apoptosis induced by Thapsigargin (TG) were used as control. Both Co-IP and Confocal microscopy analysis show no dissociation of either IREla-pS724/BiP complex or PERK-pT981/BiP or ATF6/Bip in unstimulated DLBCL lines and TG-induced lines. These results fully support our hypothesis that a chronic yet non-lethal UPR activation presented in DLBCL cells without apparent signs of ER stress. Next, we generated DLBCL lines (OCI-LY10 and SU-DHL4) that stably express lenti-shRNA targeting IREα, ATF6 or PERK, and observed that TG-inducedapoptotic cell death occured in DLBCL lines with IRE1α orATF6 depletion, but not in PERK deletion lines. However, PERK knockdown could partially rescued TG-induced apoptosis in DLBCL lines with IRE1α or ATF6 shRNA. Immunoblot also showed that the expression levels of pro-aopototic ATF4 and CHOP protein were higher in DLBCL lines with IRE1α or ATF6 shRNAthan in DLBCL lines with IRE1α/PERK or ATF6/PERK shRNA co-transfection after induction of ER stress by TG induced apoptosis.Together, our results conclude a substantial but variable basal activation of UPR-related components in clinical DLBCL cases and cell lines without apparent signs of ER stress. We also demonstrate that UPR activation is critical for survival of DLBCL cells undergoing TG-induced ER stress, indicating that persistent activation of the pathways is an important adaptive mechanism to ER stress in DLBCL cells, and suggest that interruption of IREla and ATF6 signaling may be useful in combination with drugs that induce ER stress to kill DLBCL cells. Furthermore, the elevated levels of UPR activation correlate with poor DLBCL patient prognosis, which has implications for therapeutic interventions.