- Top of page
- Materials and methods
Nasal natural killer (NK)/T cell lymphoma (NNKTL) is associated with Epstein–Barr virus (EBV). The present study analysed gene expression patterns of the NNKTL cell lines SNK6, SNK1 and SNT8, which are positive for EBV and latent membrane protein (LMP)-1, using a complementary DNA array analysis. We found that CD70 was specifically expressed in SNK6 and SNT8. Reverse transcription polymerase chain reaction and flow cytometric analyses confirmed that CD70 was expressed in all 3 NNKTL cell lines, but not in the other EBV-positive NK-cell lines. In vitro studies showed that NNKTL cell lines proliferated, in a dose-dependent fashion, in response to exogenous soluble CD27, which is the ligand for CD70. In NNKTL patients, we confirmed that the CD70 was expressed on the lymphoma cells in NNKTL tissues and that soluble CD27 was present in sera at higher levels as compared to healthy individuals. Finally, complement-dependent cytotoxicity assay showed that anti-CD70 antibody mediated effective complement-dependent killing of NNKTL cells and the affected target CD70 expression on the cells. These results suggest that CD70 acts as a functional receptor binding to soluble CD27, resulting in lymphoma progression and that immunotherapy using anti-CD70 antibody may be a potential candidate for treatment for NNKTL.
Nasal natural killer (NK)/T cell lymphoma (NNKTL), has distinct epidemiological, clinical, histological and aetiological features. NNKTL is clinically characterized progressive necrotic lesions in the nasal cavity and a poor prognosis caused by rapid progression (Harabuchi et al, 1996; Jaffe et al, 1996). Original cells of NNKTL are reported to be NK- or γ∂ T cell lineages, both of which express the NK-cell marker, CD56 (Harabuchi et al, 1996; Nagata et al, 2001). Regarding aetiological factors, since we first indicated the presence of Epstein–Barr virus (EBV) DNA, EBV-oncogenic proteins and clonotypic EBV genome in NNKTL, EBV is thought to play a role in lymphomagenesis (Harabuchi et al, 1990, 1996; Minarovits et al, 1994).
The biological characteristics of NNKTL have become gradually clearer following the establishment of the EBV-positive cell lines ‘SNK6’ and ‘SNT8’ from primary lesions (Nagata et al, 2001). We previously showed that NNKTL cells produce several cytokines and chemokines such as γ-interferon (IFNγ), interleukin (IL) 9, IL10 and IFNγ-inducible protein (IP)10, which play roles in the proliferation and invasion of the cells in an autocrine manner (Nagato et al, 2005; Takahara et al, 2006; Moriai et al, 2009). Furthermore, we have recently shown that environmental monocytes attracted by IP10 enhance proliferation of the NNKTL cells in a cell-to-cell contact manner (Ishii et al, 2012).
Histological characteristics of NNKTL include angiocentric and polymorphous lymphoreticular infiltrates, which are called polymorphic reticulosis (Harabuchi et al, 1996; Harris et al, 2000). The infiltrating cells in the NNKTL tissue contain many types of cells, including tumour cells as well as inflammatory cells, such as granulocytes, monocytes, macrophages, lymphocytes and plasma cells. Such inflammatory cell infiltration is likely to be caused by chemotactic effects of a wide variety of chemokines including IL8, Mig and IP10, which were reported to be produced by NNKTL cells (Ohshima et al, 2004; Moriai et al, 2009). The inflammatory cells, and the chemokines and cytokines that they produce, are likely to influence proliferation, survival and migration of the lymphoma cells.
In order to determine which genes are expressed specifically in NNKTL, the present study compared the gene expression profiles of NNKTL cell lines to those from the other cell lines using complementary DNA (cDNA). We found that CD70 was strongly expressed in NNKTL cell lines (SNK6 and SNT8) as compared to non-NNKTL cells (NK92) and peripheral blood mononuclear cells (PBMC) from healthy individuals.
CD70 is a member of the tumour necrosis factor (TNF) superfamily (Bowman et al, 1994; Hintzen et al, 1995). CD70 expression is normally restricted in normal cells to a small subset (10%) of activated B-cells, activated T cells and dendritic cells (Hintzen et al, 1994; Tesselaar et al, 2003). The only known receptor of CD70, CD27, is expressed on the surface of memory B-cells (Klein et al, 1998), most T cells (de Jong et al, 1991) and NK-cells (Sugita et al, 1992). Ligation of CD70 to its receptor CD27 induces a signal transduction pathway, resulting in activation and proliferation of B-cells and T cells (Garcia et al, 2004; Dang et al, 2011). With regard to haematopoietic malignancies, CD70 expression has been reported in 50% of B-cell chronic lymphocytic leukaemia, 33% of follicular lymphoma and 71% of diffuse large cell lymphoma (Davi et al, 1998; Lens et al, 1999), some cases of T cell lymphoma (Zambello et al, 2000) and a case of chronic active EBV infection-associated T cell lymphoma (Shaffer et al, 2011a). However, the functional role of CD70 on haematopoietic malignancies is not yet fully understood.
CD27 is known to be cleaved from in activated B-cells or in T cells after triggering of the T cell receptor (TCR)/CD3 complex (Hintzen et al, 1991; Bohnhorst et al, 2002), resulting in the formation of a soluble form of CD27 in serum (Loenen et al, 1992). Elevated serum levels of soluble CD27 have been reported in several autoimmune diseases (Font et al, 1996; Bohnhorst et al, 2002) and B-cell malignancies (van Oers et al, 1993). However, whether soluble CD27 binding to CD70 has any biological role in such diseases remains to be determined.
As described above, given that CD70 is more highly expressed in normal tissues and is widespread in various malignancies, CD70 has been known to be an attractive target for immunotherapies. Investigations to exploit CD70 as a cancer target have led to the identification of potential antibody-based clinical candidates. Both unconjugated antibodies and antibody-drug conjugates targeting CD70 have been tested in animal models of human cancers (Israel et al, 2005; McEarchern et al, 2007; Grewal, 2008). However, there has been no report regarding anti-CD70 antibody therapy for NNKTL.
In the present study, we found that CD70 was specifically expressed in NNKL cell lines and that it played a role in cell growth by binding to soluble CD27. Moreover, we confirmed that the lymphoma cells expressed CD70 in the NNKTL tissues and that soluble CD27 was present at higher levels in sera. Finally, we showed in vitro that the anti-CD70 antibody could mediate effective complement-dependent killing of NNKTL cells and the effects target CD70 expressed on the cells. These data suggest that CD70 may play a role in lymphoma proliferation by binding to soluble CD27 and that immunotherapy using anti-CD70 antibody may be a potential candidate for the treatment of NNKTL.
- Top of page
- Materials and methods
In the present study, analysis of gene expression using cDNA array revealed that CD70 mRNA is expressed at a much higher level in NNKTL cells (SNK6 and SNT8) than in PBMC of healthy volunteers and NK92 non-NNKTL cells (Fig 1). RT-PCR analysis confirmed that CD70 mRNA was expressed in SNK6, SNT8, KAI3 and Raji cells, but not in YT and NK92, Jurkat, MOLT4 and PEER cells (Fig 2). It is notable, in EBV-positive cell lines, that CD70 expression was detected in KAI3 and Raji cells, which were also LMP-1-positive, however, the expression never detected in YT and NK92 cells, which were LMP-1-negative. In the literature, Burkitt lymphoma lines containing the most stringent form of EBV latency (type I), in which EBV-determined nuclear antigen-1 (EBNA-1) is the only viral protein produced, do not express CD70 (Israel et al, 2001). However, Burkitt lymphoma lines with type III latency, EBV-immortalized lymphoblastoid cell lines, nasopharyngeal carcinomas and Hodgkin lymphoma and chronic active EBV infection-associated T cell lymphoma, all of which are LMP-1-positive, were reported to express CD70 (Herbst et al, 1996; Israel et al, 2001; Shaffer et al, 2011a). In vitro study previously showed that CD70, which was absent from the parental line, was expressed in virtually all LMP-1-transfected epitherial cells (Niedobitek et al, 1992). It is suggested, on the basis of the data, that LMP-1 may induce CD70 expression in EBV-positive malignancies.
Aside from EBV-positive malignancies, CD70 expression has been reported on 50% of B-cell chronic lymphocytic leukaemia cases, 33% of follicular lymphomas and 71% of diffuse large cell lymphomas (Davi et al, 1998; Lens et al, 1999) and some cases of T cell lymphoma (Zambello et al, 2000). However, the functional role of CD70 in malignancy is not fully understood. Several reports described its role in tumour immunity, but they were not uniform. It was reported that deliberate expression of CD70 on tumour cells stimulated both NK and T cell immunity in lymphoma and glioma models (Kelly et al, 2002). On the contrary, CD70 expression was reported to protect the tumour from lysis by CD27-expressing cytotoxic T cells in some glioblastomas (Wischhusen et al, 2002). On the other hand, Lens et al (1999) demonstrated that some B-cell leukaemia cell lines, which expressed CD70 as well as CD27, could proliferate vigorously in response to anti-CD70 monoclonal antibody due to an agonistic signal delivered via CD70, suggesting that CD70 can operate as receptor inducing a signal transduction pathway, thus contributing to the progression of these B-cell malignancies.
The present study clearly showed that NNKTL cell lines (SNK6, SNK1 and SNT8) that were positive for CD70, but negative for CD27 (Figs 1 and 2), could proliferate vigorously in response to exogenous soluble CD27 dose-dependently, but CD70-negative NK cell lines could not (Fig 3A). We further showed that the exogenous soluble CD27-dependent cell growth of NNKTL cell lines was inhibited completely by administration with anti-CD27 neutralizing antibody (Fig 3B). These findings suggest that CD70 acts as a functional receptor that binds to soluble CD27, inducing cell growth. In NNKTL patients, we showed that CD70 was expressed on CD56-positive lymphoma cells in NNKTL tissues (Fig 4A). We further found that soluble CD27 was present in sera at significantly higher level as compared to healthy individuals (Fig 4B). The findings observed in patients' samples support that soluble CD27 also acts as a growth factor for NNKTL cells in vivo.
Elevated serum levels of soluble CD27 have been reported in diseases characterized by abnormalities in B-cell differentiation and activation including autoimmune diseases (Font et al, 1996; Bohnhorst et al, 2002) and B-cell malignancies (van Oers et al, 1993). However, the biological role of soluble CD27 has not been extensively studied. It was initially thought that soluble CD27 would compete with the membrane-bound CD27 receptor for binding to CD70, thus blocking the CD27–CD70 pathway (Hintzen et al, 1991). Such reaction thereby results in creating a feedback mechanism that would hamper immune responses. On the contrary, soluble CD27 has been recently demonstrated as an enhancer of immune responses; soluble CD27 induced IgG production from antigen-primed B-cells (Dang et al, 2011) as well as expression of CD40LG and APRIL for B-cell activation (Ho et al, 2008). This is the first report to show that soluble CD27 binding to CD70 has a biological role for cell growth of NNKTL cells.
Although this study detected significantly higher levels of soluble CD27 in sera of NNKTL patients, we cannot explain completely how soluble CD27 increased in sera of NNKTL patients. Soluble CD27 is cleaved from the extracellular portion of the surface CD27 receptor presenting in a wide variety of immune cells, such as most T cells (de Jong et al, 1991), memory B-cells (Klein et al, 1998) and NK-cells (Sugita et al, 1992). The microenvironmental feature of NNKTL is characterized as polymorphous lymphoreticular infiltrates containing many types of immune cells, such as granulocytes, monocytes, macrophages, T and B-cells and plasma cells (Harabuchi et al, 1996; Harris et al, 2000). Such infiltrating immune cells are caused by the chemotactic effects of chemokines, including IL8, Mig and IP10, secreted by NNKTL cells (Ohshima et al, 2004; Moriai et al, 2009) and are likely to release soluble CD27. Recently, matrix metalloproteases (MMPs) were reported to induce cleavage of soluble CD27 (Kato et al, 2007). Interestingly, we recently found, on the cDNA array and RT-PCR analyses, that NNKTL (SNK6) cells produce high levels of MMPs (data not shown). In addition, some cytokines, such as IL2, IL9, IL10 and IFNγ, which are secreted by NNKTL cells (Nagato et al, 2005; Takahara et al, 2006), may possibly induce CD27 cleavage. In the NNKTL tissue microenvironment, a positive feedback loop of interaction between lymphoma cells and infiltrating immune cells may contribute to lymphoma progression, i.e. NNKTL cells produce a wide variety of cytokines, chemokines and MMPs, which induce chemotactic reaction as well as soluble CD27 release of such immune cells, resulting in proliferation of lymphoma cells by binding of soluble CD27–CD70 on the NNKTL cells. We have already shown that microenvironmental monocytes, attracted by IP10, which is secreted by NNKTL cells (Moriai et al, 2009), enhance proliferation of NNKTL cells by cell contact-dependent interaction through membrane-bound IL15 (Ishii et al, 2012).
Recently other roles of the CD27–CD70 pathway on tumour progression have been reported. Claus et al (2012) demonstrated that CD70 expression in tumours is a negative prognostic factor and correlates with increased regulatory T cell accumulation, which promotes tumour progression by CD70 on the tumour cell, stimulating CD27 on the regulatory T cells. Schurch et al (2012) showed that stimulation of the TNF family receptor, CD27, on leukaemic stem-like cells of chronic myeloid leukaemia, where CD27 is a receptor for the CD70 ligand, is involved in leukaemic stem-like cell proliferation. Further studies that knockdown CD70 in NNKTL cells will be needed to define the mechanism via which CD70 engagement on NNTKL cells by soluble CD27 may stimulate cell proliferation.
The restricted expression pattern of CD70 in normal tissues and its widespread expression in various malignancies makes it an attractive target for cytotoxic T cell therapy as well as antibody-based therapy. Recently, Shaffer et al (2011b) demonstrated that CD70-specific T cells killed CD70-positive lymphoma cell lines by IFNγ and IL2 secretion and that adoptively transferred CD70-specific T cells induced sustained regression of established murine xenografts, suggesting that CD70-specific T cells may be a promising immunotherapeutic approach for CD70-positive malignancies. In the present study, we clearly showed in vitro that the anti-CD70 antibody can mediate effective complement-dependent killing of NNKTL cells and the effects target for CD70 expressed on the cells (Fig 5). Israel et al (2005) also showed that an anti-CD70 antibody mediated complement-dependent killing of CD70-positive Burkitt lymphoma cells in vitro. They further showed in severe combined immunodeficiency (SCID) mice that anti-CD70 antibody strikingly inhibited the growth of CD70-positive Burkitt lymphoma cells (Israel et al, 2005). McEarchern et al (2007) showed that the administration of an engineered anti-CD70 monoclonal antibody significantly prolonged the survival of SCID mice bearing CD70-disseminated human non-Hodgkin lymphoma xenografts. Although further experiences in animal models will be needed before clinical use, it is suggested, on the basis of our in vitro data together with these previous results in SCID mice, that the administration with anti-CD70 antibodies may be useful for the treatment of NNKTL patients.
In conclusion, this study clearly showed that NNKTL cells expressed CD70 possibly due to the ability of LMP-1, but did not express its only receptor, CD27. We further showed that CD70 played a role for cell growth by binding to soluble CD27. In NNKTL patients, we confirmed that the lymphoma cells expressed CD70 in the tissues and that soluble CD27 was present at higher levels in sera. Finally, we showed in vitro that the anti-CD70 antibody could mediate effective complement-dependent killing of NNKTL cells and the effects target for CD70 expressed on the cells. We conclude that NNKTL expresses CD70, which plays a role in lymphoma proliferation by binding to soluble CD27, and that immunotherapy using anti-CD70 antibody may be a potential candidate for treatment for NNKTL