*3.5. LMP2Avax Delays Tumor Growth*

In order to study the possible impact on an EBV+ tumor expressing a model LMP2A antigen, we next generated a murine epithelial tumor cell line using TC-1 cells that were constructed to express LMP2A using a retroviral transduction system. This may cause high expression of LMP2A relative to EBV-associated tumor cells, but LMP2A protein is expressed in the cancer cells of patients and its epitopes are recognized by T cells [54]. This cell line serves as a vaccine target for our LMP2A immunogens. We generated and selected the LMP2A line as shown in Figure 4A. Briefly, retroviral vectors produced by transfecting Phoenix cells with pBMN plasmids containing GFP and LMP2A were used to stably transduce TC-1 cells. These cells underwent selection via fluorescence-activated cell sorting (FACS) and single-cell cloning to produce a homogenous population expressing LMP2A, and this population was used to introduce tumors into mice.

**Figure 4.** LMP2Avax inhibits tumor growth in mice. (**A**) Workflow to produce tumor cell lines expressing target antigen, in this case TC-1-LMP2A. (**B**) Immunofluorescence assay demonstrating LMP2A expression in TC-1-LMP2A cell line. DAPI is shown in blue, with LMP2A labeled in green. Anti-LMP2A antibodies were used as primary Abs (top), with secondary Abs conjugated to AF647. Anti-EBNA1 primary antibodies were used as a negative control (bottom). Scale bars are 50 μm. (**C**) Vaccination schedule prior to tumor introduction. Two groups of five C57BL/6 mice received three biweekly vaccinations followed by the subcutaneous axillary injection of 2 million TC-1 cells stably expressing LMP2A. The vaccines used 20 μg of DNA in 30 μL of water delivered by electroporation, with the vaccine group receiving plasmid encoding LMP2Avax and the control receiving the empty vector pVAX. Tumor sizes were monitored daily afterwards. (**D**) TC-1-LMP2A tumor volume over time in mice vaccinated with LMP2A or empty vector. Bars show scanning electron microscopy (SEM).

The expression of LMP2A in the derived TC-1-LMP2A tumor line was confirmed by antibody reactivity as demonstrated by immunofluorescence in Figure 4B. We next studied the use of this cell line as a tumor challenge antigen. C57BL/6 mice received either the LMP2Avax DNA vaccine or empty vector three times at biweekly intervals, followed by an axillary injection of 2 million tumor cells after the final vaccination (Figure 4C). LMP2Avax vaccinated mice showed a smaller tumor volume and more rapid tumor shrinkage than those vaccinated with the empty vector, demonstrating the anti-tumor immunogenic potential of the LMP2Avax vaccine (Figure 4D).

#### **4. Discussion**

EBV, formally known as human gammaherpesvirus 4, is responsible for infectious mononucleosis, multiple premalignant conditions, and various EBV-driven cancers. These cancers include Burkitt Lymphoma, Hodgkin's lymphoma, gastric cancer, nasopharyngeal carcinoma, HIV-associated oral hairy leukoplakia, and numerous other lymphoproliferative disorders. Additionally, EBV infection is associated with nonmalignant diseases and significant autoimmune disorders [55]. The worldwide burden of EBV-associated cancer is approximately 150,000 deaths per year, which represents almost 2% of all deaths from cancers. This burden continues to grow. EBV-associated gastric and nasopharyngeal carcinomas are each responsible for over 60,000 cancer deaths per year, and the incidence of the latter is increasing [56]. In light of this burden, additional approaches to EBV immunotherapy are important.

Here, we engineered synthetic consensus DNA vaccines of modified EBV latent proteins to generate immune responses which could impact tumor regression. Latent proteins are present in both lymphomas and carcinomas associated with EBV, and these have been studied as potential targets in various immunotherapeutic strategies. Currently there is no licensed approach for EBV immunotherapy. Cellular therapies have been studied in small trials and have shown some important effects [57,58]. However, these were early studies and additional approaches would be highly beneficial.

Along these lines, work in the HPV setting with SynCon DNA vaccines delivered by adaptive EP has evolved to be a robust approach for induction of antiviral cellular immunity, which can impact tumors and precancers in vivo [26,27]. We tested this approach here for a three-antigen synthetic DNA vaccine approach targeting the major EBV latent oncoproteins. We chose these antigen targets because they are present in EBV-associated cancers. Small trials of cellular therapies targeting EBNA1 [59] and LMPs [25,60,61] have shown improved outcomes against EBV-associated diseases. The high frequency of nasopharyngeal carcinoma concentrated in east Asia makes for a unique environment to test prophylactic and therapeutic approaches targeting the virus [62]. The frequency of Hodgkin's lymphoma in Europe and its temporal association with infectious mononucleosis offers another opportunity [7]. The growing burden of EBV in the US suggests immunotherapy for nasopharyngeal and gastric cancer as well as association of EBV with more common autoimmune disorders may also be important to consider as amenable to robust immunotherapy approaches [61].

Synthetic DNA vaccines can drive in vivo immune responses via MHC class I and II presentation through their delivery of and intracellular production of genetically encoded antigens. Newer delivery approaches have resulted in the generation of more consistent and robust immunity that can target cancer in the clinic [26,27]. Here we show that these designed latent antigen vaccines elicit significant cytotoxic T lymphocyte responses against the encoded vaccine targets EBNA1vax and LMP2Avax, which showed dominant CD8 T cell responses in vivo. These cellular responses are important in protecting mouse models from EBV antigen-expressing tumors in murine vaccine models, as recently shown in a novel heterologous prime-boost approach that impacted an EBNA1 tumor challenge [63]. Importantly, LMP2Avax-induced immunity protected against tumor growth in a TC-1 challenge model where LMP2A was targeted by the immunization. The immune responses produced by EBNA1vax and LMP2Avax merit further study. In addition, continued engineering may be interesting in this regard, as DNA delivery of LMP1 as an immunogen can clearly impact tumor growth as a standalone antigen in some models [28]. Combination development for this group of immunogens appears worthy of additional attention.

Recent developments in the DNA platform in formulation, engineering and delivery by adaptive EP have led to improved immune potency and improved consistency in clinical studies [26,27]. In these studies, we noted that the vaccines were biased towards driving highly desired CD8 immunity against the vaccine targets over CD4 immunity. This CD8 bias may be particularly relevant for clearing virally infected cells by cytotoxic T lymphocyte induction that would ultimately kill tumor cells. These latent antigen vaccines could be studied in the context of epithelial tumors, such as gastric and nasopharyngeal carcinomas, among others. The addition of checkpoint inhibitors in the context of these immunizations, as we have reported for HPV, might also be of interest for impacting EBV-related tumor progression [26,27,64].
