Mystery of the missing target

In this issue of Blood, Fox et al report on their investigations of CD8⁺ T-cell targeting of extranodal natural killer (NK)/T lymphoma.¹  T cells that had been engineered to recognize the Epstein-Barr virus (EBV) latency membrane protein 2 (LMP2) killed NK lymphoma cell lines. This is not surprising, except that that the NK tumor cell lines did not express LMP2 protein as determined by immunoblot or reverse-transcriptase polymerase chain reaction (PCR).

Interest in killing NK/T lymphoma stems from the inadequacies of present therapeutic approaches. Extranodal NK/T-cell lymphomas (formerly lethal midline granulomas), often present in the nasal region, are usually associated with EBV, and are often poorly responsive to anthracycline-based combination chemotherapies.²  Although progress has been made with alternative chemotherapy regimens and concurrent radiation therapy, these tumors are still most often fatal when in advanced stage.^3,4  Other EBV-associated tumors, most notably posttransplant lymphoproliferative disease, are responsive to adoptive immunotherapy approaches involving transfer of EBV-targeted T cells.⁵  Thus, there was an interest in possible immunotherapeutic approaches to this family of tumors.

Why focus on this particular protein? Although the EBV genome encodes 80 or more proteins, only a few are made in tumors. Posttransplant lymphoproliferative disease, which has proven to be most responsive to adoptive T-cell approaches, typically expresses 8 or more viral proteins, among these the immunodominant Epstein-Barr nuclear antigen 3A (EBNA3A), EBNA3B, and EBNA3C proteins that are targeted by CD8⁺ T cells with high frequency. Other tumors, notably EBV-associated Burkitt lymphoma and gastric carcinoma, may express only a single EBV protein, EBNA1. This protein is problematic as a target because the protein prevents its own presentation in class I major histocompatibility complex, so tumor cells expressing only EBNA1 may be invisible in CD8⁺ immunosurveillance.⁶  LMP1 and LMP2 are different EBV-encoded membrane proteins that are both targeted by CD8⁺ T cells. Both are expressed in EBV-associated Hodgkin lymphoma and variably in other EBV-associated peripheral T-cell lymphomas.⁷  Although not immunodominant, CD8⁺ T-cell responses to LMP2 are especially well characterized.⁸  In addition to serving as a target for adoptive T-cell immunotherapy, several therapeutic vaccine studies using vaccines designed to enhance the CD8⁺ T-cell response to LMP2 are under way or have recently concluded.⁹ 

Characterization of the sensitivity of tumor cell lines to CD8⁺ T-cell lines of varying specificities led to the curious observation of target-specific killing (LMP2 targeted CD8⁺ T cells killing EBV(+) NK-cell lines) in the apparent absence of the target (immunoblot and RNA detection showed no LMP2 expression). In the EBV virion, the genome is linear double-stranded DNA. Upon infection, that DNA circularizes inside cells. The process involves fusion of the terminal repeats at each end of the linear genome leading to creation of the LMP2A gene.¹⁰  LMP2B is transcribed from a downstream start site and so lacks the first coding exon of LMP2A. The first exon of LMP2B is noncoding. Subsequent exons are identical in LMP2A and LMP2B and encode the membrane-spanning domains. As it happens, the available monoclonal antibodies detecting LMP2 protein are directed against epitopes present in the first exon of LMP2A, so there are, as yet, no antibody reagents to detect LMP2B. Nonetheless, LMP2A and LMP2B RNA transcripts are readily detected and distinguished by reverse-transcriptase PCR. In some of the NK-cell line studies, there was no evidence of LMP2 by immunoblot, no LMP2A transcription, and minimal transcription ofLMP2B. The explanation for LMP2-specific killing observed is the existence of a third isoform of the protein, with a transcript likely beginning in the terminal repeats themselves. As such, the encoded protein lacks epitopes recognized by extant monoclonal antibodies and is missed by many of the PCR primer pairs used to characterize EBV gene expression. The authors term the new protein LMP2-TR.

Success with any or all of the strategies targeting LMP2-TR in extranodal NK/T lymphoma may mean that the discovery of the new protein isoform, rather than merely a footnote to the gene expression map of EBV, will begin an interesting new chapter in the book defining immunotherapeutic targets in EBV-associated malignancies.