The induction of immunogenic cell death by type II anti‐CD20 monoclonal antibodies has mechanistic differences compared with type I rituximab

The anti-CD20 monoclonal antibody (mAb) rituximab has improved outcomes for patients across a range of B-cell malignancies and catalysed the development of a number of other novel anti-CD20 mAb, including the type II anti-CD20 antibody GA101 (Obinutuzumab). GA101 is glycoengineered to enhance antibody-dependent cellular cytotoxicity (ADCC) and binds to the CD20 antigen in a fundamentally different way to rituximab to initiate direct Programmed Cell Death (PCD) (Niederfellner et al, 2011) through a novel pathway that is non-apoptotic and dependent on actin rearrangement, lysosomal permeabilisation and reactive oxygen species (ROS) generation via NADPH oxidase and independently of glutathione pathway (Honeychurch et al, 2012). We were interested to investigate whether this type of PCD is 'immunogenic'. Immunogenic cell death is associated with the release of damage-associated molecular patterns (DAMPs), such as heat-shock proteins (HSPs), HMGB1 and ATP which are able to promote effective T-cell priming (Apetoh et al, 2007). Recent data has suggested a 'vaccination' effect following anti-CD20 mAb therapy with the generation of T-cell responses which can protect against tumour rechallenge (Hilchey et al, 2009; Abes et al, 2010; Braza et al, 2011; Wahlin et al, 2011). It remains unknown how targeting CD20 contributes to the activation of tumour-specific T-cells, and induction of immunogenic cell death may be a potential mechanism by which anti-CD20 antibodies actuate cellular immunity. In the present study we investigated the potential immunogenicity of malignant B-cells dying by PCD induced by type II anti-CD20 antibodies. Our data demonstrate that mAb-induced PCD is accompanied by the release of significant levels of DAMPs, enhances dendritic cell maturation and subsequent T-cell activation. PCD induced by type II anti-CD20 mAb leads to rapid permeabilization of the plasma membrane and cell lysis (Fig 1A, Fig S1). Loss of plasma membrane integrity is associated with the release of intracellular contents, many of which can function as endogenous danger signals (Basu et al, 2000; Griffith & Ferguson, 2011). This led us to investigate whether type II mAb-induced PCD was accompanied by the release of molecular determinants capable of functioning as immune adjuvants (see Supplemental Methods). For comparison, DAMP release was also assessed following induction of necrosis, apoptosis or treatment with type I anti-CD20 mAb, which induce very little direct PCD. Treatment of Raji cells with GA101 led to significant release of HSP60, HSP90 and HMGB1 into culture supernatant at levels similar to that seen in Raji cells undergoing necrosis (Fig 1B). In contrast, apoptosis induced by the chemotherapy agent mitoxantrone led to release of HMGB1 but not HSP, as reported previously (Apetoh et al, 2007); whilst DAMP release was not observed following treatment with type I mAb correlating with low levels of induced cell death. DAMP release was not unique to GA101 as it was also seen with other type II antibodies, such as tositumomab (Fig 1, Fig S1b), suggesting this is a general phenomenon of direct type II PCD. To confirm that DAMP release was the consequence of a regulated cell death pathway, we determined whether pharmacological inhibition of biochemical events triggered by CD20-ligation including actin reorganization, and ROS generation (Honeychurch et al, 2012) prohibited DAMP release. Release of HSP90 and HMGB1 was blocked following inhibition of cell death by treatment with actin and NADPH oxidase inhibitors (Fig 1C, Fig S2). However, inhibition of apoptosis through the pan-caspase inhibitor QVD or over-expression of BCL2 had no effect on the release of HSPs and HMGB1 following type II mAb treatment (Fig S3a–b). GA101 treatment also led to significant release of ATP (P < 0·01) (Fig 1D), another critical signal of immunogenic cell death (Ghiringhelli et al, 2009) (Fig S3c) which could be inhibited by blocking the PCD pathway through either actin inhibition or NADPH oxidase inhibition. Similar profiles of DAMP release were seen in other lymphoma cell lines including Daudi (Fig 1E) and SUDHL4 (Fig S3d). No differences in the levels or pathway of PCD or DAMP release was seen between type II CD20 antibodies GA101 and tositumomab (data not shown). As type I anti-CD20 mAbs (such as rituximab or ofatumamab/2F2) do not induce significant PCD and therefore DAMP release, we investigated whether DAMP release could be induced by other mAb effector pathways, such as complement dependent cytotoxicity (CDC). Experiments looking at HMGB1 levels showed that rituximab- and 2F2 (Ofatumumab)- mediated CDC also led to the liberation of DAMP (Fig S4a–c). However, cell death and release of HMGB1 appeared to occur through a different pathway as it was independent of actin rearrangement (Fig S4d–e) and vacuolar ATPases (data not shown). Having determined that anti-CD20 mAb therapy can lead to DAMP release upon induction of cell death we investigated whether levels of DAMPs released following GA101 antibody treatment were sufficient to impact upon cellular immune effectors. We generated primary dendritic cells (DC) from a number of healthy donors and investigated what impact the supernatants had upon DC maturation. When immature DC were cultured for 2 d with GA101-treated Raji supernatant, induction of DC maturation was seen with a statistically significant upregulation of CD86 and CD83 (Fig 2A–C). Conversely, culture with supernatant from cells treated with control mAb (Herceptin), which does not induce PCD or DAMP release, failed to induce DC maturation. GA101 supernatant-matured DC were able to induce an allo-mixed lymphocyte reaction (Fig 2D,E) with T-cell proliferation approaching that seen with lipopolysacharide (LPS)-matured DC whereas no T-cell proliferation was induced with control supernatant-matured DC. Enhancing anti-tumour T-cell responses is potentially an important strategy for improving anti-CD20 mAb therapy. To date, there is little data investigating the mechanisms by which CD20 antibody therapy may induce cellular immune responses although it has recently been shown that GA101 is able to potentiate the killing of lymphoma cells by γδ T-cells (Braza et al, 2011). Here, we demonstrate that PCD induced by type II anti-CD20 mAbs, such as GA101 is a form of immunogenic cell death characterized by the release of DAMPs, such as HSP90, HMGB1 and ATP, which can induce dendritic cell maturation and subsequent T-cell proliferation. In summary, these novel findings provide a potential mechanism whereby type II anti-CD20mAbs may potentially enhance host immune response and induce durable tumour remissions. Whether this immunogenic cell death will be important in improving anti-CD20 mAb efficacy and priming a cellular immune response in vivo currently remains unknown and further preclinical and clinical studies are required to elucidate this. We are grateful to all our collaborators for providing the various reagents used in this study. GA101 was provided by C. Klein, Roche Glycart AG and 2F2 by Professor Mark Cragg (University of Southampton). This work was supported by grants from Cancer Research UK (CRUK) and Leukaemia and Lymphoma Research. This work was supported in part by research funding from Roche Glycart AG to TMI. EC, JH, TMI initiated the studies and designed the research. EC, LS, SD, MM, WA and JH performed experiments. All authors analysed the data. EC wrote the manuscript. TMI and JH edited the manuscript. TMI has received honorarium for consultancy work with Hoffman La Roche. The remaining authors declare no competing financial interests. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.