synDNA vaccine against TCR chains and neoantigens for T cell lymphoma therapy

Abstract

T cell lymphomas constitute approximately 10% of all non-Hodgkin lymphomas and are associated with poor prognosis. Patients experiencing early relapse post-treatment exhibit a 5-year overall survival rate of 11%, underscoring the need for improved therapeutic strategies. The clonality of T cell cancers makes the T cell receptor (TCR) an appealing target for immunotherapy. Here, we developed and evaluated a synDNA vaccine against the TCR⍺, β, and γ chains (TCRfullvax) of the EL4 murine T cell lymphoma model. Immunogenicity studies revealed induction of robust T cell responses against all three TCR chains, with identification of immunodominant epitopes for each chain. Notably, we observed no significant differences in the number of live T cells between TCRfullvax-vaccinated group and control groups, indicating the vaccine's ability to selectively break tolerance against vaccinated TCR without broadly depleting T cells. In a minimal residual disease model, TCRfullvax delayed EL4 tumor progression. Tumors from TCRfullvax-treated mice revealed downregulation of TCR expression, suggesting a potential immune escape mechanism. Neoantigens, derived from somatic mutations within tumor genome, present another promising target for anticancer vaccine development, accordingly we developed a second vaccine targeting 15 neoantigens identified through sequencing of EL4 cells (EL4neovax). EL4neovax elicited strong immune responses against 5/15 encoded neoantigens and controlled EL4 tumors. Co-administration of TCRfullvax and EL4neovax demonstrated superior tumor control compared to either vaccine design alone, further supporting that neoantigen targeting can partially mitigate TCR loss. These findings highlight the potential of combining TCR-targeted and neoantigen-based immunotherapies for the treatment of T cell lymphomas. Further investigation of this dual-vaccine approach is warranted to optimize the therapeutic efficacy for this difficult disease.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00262-026-04302-5.

Introduction

T cell lymphomas (TCLs) represent about 10% of all non-Hodgkin’s Lymphomas (NHL) and can be divided into several categories based on maturation stage of the T cell at time of tumorigenesis [1, 2]. TCLs are a relatively rare tumor with fewer than 8000 new diagnoses in the US annually [3]. TCLs are heterogenous and can be subdivided into more than 30 different subcategories according to the World Health Organization [4]. The diversity and heterogeneity of TCLs account for significant differences in patient prognosis and results in very few therapeutic options which are specific for TCLs [3]. Peripheral T cell lymphomas (PTCLs) comprise of about 26% of all TCLs and are the most common form with few therapeutic options, especially in the relapsed/refractory setting [2]. For cutaneous T cell lymphoma (CTCL), which is considered a less aggressive tumor, the median survival for patients diagnosed with stage III/IV disease is 5 years [5]. On the other hand, the median survival for PTCL patients is 5 months for refractory patients and 11 months for relapsed patients [6]. These statistics underscore the need for new approaches to better impact patient outcomes.

Over the last few years, immunotherapies such as monoclonal antibodies, immune checkpoint inhibitors (ICIs), bispecific antibodies and CAR-T cells have impacted the treatment landscape for several cancers. Immunotherapy demonstrated promising results in patients with relapsed and refractory disease in solid tumors as well as B cell cancers [7–10]. Slow growing TCLs such as CTCL are characterized by an infiltration of CD8+ T cells which primarily exhibit an exhausted phenotype and express immune inhibitory markers such as PD1, Tim3 and Lag3 [11]. Exhausted T cells are also found in more aggressive T cell tumors and are associated with disease progression [2]. These results highlight that TCLs could potentially benefit from reinvigoration of exhausted T cells and that immunotherapy could be successful in treatment of TCLs. Given the efficacy of immunotherapies in the clinic, their efficacy for treatment of TCLs should also be explored.

Over the last few years there has been a resurgence of interest focused on next generation approaches for cancer vaccines as a tool to drive T cell immunity. The targets for cancer vaccines are derived from several distinct sources such as tumor associated antigens (TAAs), lineage markers, viral oncogenes and neoantigens (shared or private) [12, 13]. Additionally, for B and T cell cancers, the immune cell receptor (BCR or TCR) can also serve as a unique antigenic target. The hypervariable regions of the receptors are produced via random joining of the V,D,J segments with further mutations and insertions/deletions of amino acids [14]. This suggests that the immune cell receptor can potentially be a lower bar immune target as the T cells which would recognize them would not have been subject to central T cell tolerance. Accordingly, the BCR of a B cell cancer was shown to be a driver of mutation for the disease [15]. This suggests that the downregulation of BCR in response to targeted therapies against the BCR would be costly to the tumor. It has been shown that in patients suffering from B and T cell cancers, peptides derived from cancerous BCR/TCR can specifically bind to the patient’s Class I MHC and be expressed as peptide-MHC complexes for recognition by CD8+ T cells [16–18]. Clinically, an idiotype vaccine targeting the B cell receptor in patients with lymphoplasmacytic lymphoma was safe, produced idiotype specific T cell responses in 87.5% of the patients and reduced tumorigenic B cell (but not plasma cell) burden in 66.7% of the patients in a Phase 1 trial [19]. The safety associated with a vaccination-based immunotherapy approach is another advantage, especially for indolent and slowly progressive cancers. The toxicities associated with autologous or allogeneic cell therapies would most likely be prohibitive and prevent their use in such settings.

Neoantigens are antigenic peptides derived from somatic mutations in tumor cells. Rapid cell division, a feature of tumor cells, leads to several errors in replication. This process can be further enhanced due to either inherited or de novo mutations in enzymes which normally ensure fidelity of DNA replication, whose loss of function can contribute to increasing tumor mutational burden [20]. These mutations can be recognized by the adaptive immune system because they are not subject to central immune tolerance. Further, neoantigen expression is limited to tumor cells, making them ideal targets for therapeutic cancer vaccines [21]. While the majority of neoantigens are derived from nonsynonymous mutations, they can be also be derived from insertions and deletions (INDEL) mutations which modify the open reading frame and gene/RNA fusions [22, 23]. Every tumor evolves uniquely, hence each patient needs a highly personalized vaccine for their tumor therapy.

An ideal vaccine platform for personalized therapeutic cancer vaccines should allow for simple delivery, rapid and cost effective manufacturing, targeting of multiple neoantigens for optimal clinical response and generate strong CD8 + T cell responses [21, 24, 25]. Using a synDNA platform for neoantigen targeting, we have previously demonstrated robust CD4+ and CD8+ T cell responses in animals, which led to significant tumor control of murine lung, ovarian and colon tumors in vivo [25, 26]. We have further demonstrated that the platform can be used to deliver up to 40 different neoantigens in a single plasmid without compromising immune responses and in a metastatic model of mouse lung and ovarian tumor, a 40-mer vaccine controlled growth of lung and ovarian tumors with very different mutation profiles [25]. synDNA vaccines can be rapidly manufactured at low cost and allow for co-delivery of gene encoded cytokine/chemokine adjuvants which can further boost immune responses [27–30]. These results were replicated in a Phase 1/2 clinical trial evaluating synDNA vaccines co-administered with a plasmid-encoded IL12 as a molecular adjuvant in combination with pembrolizumab in heavily pretreated patients with advanced, unresectable/metastatic hepatocellular carcinoma. With this immunization approach the authors described induction of robust CD4+ and CD8+ T cell responses leading to shrinking of primary and metastatic lesions resulting in an overall response rate of 30.6%. This includes 3 patients demonstrating a complete response and a fourth patient achieving secondary resectability while on treatment and also becoming tumor free. The trial also supported the importance of simultaneously targeting multiple neoantigens as the authors observed a positive correlation (p = 0.025) between the number of neoantigens included in the vaccine and the clinical response achieved. In this trial, patients whose vaccines encoded for 30 or more neoantigens had a significantly higher likelihood of achieving an objective clinical response compared to those with fewer than 30 neoantigens [31].

In our present study, we tested, for the first time, whether a synDNA vaccine targeting the TCR can control the growth of EL4 tumors (a mouse syngeneic T cell lymphoma cell line). We observed robust T cell immune responses against the TCRα, TCRβ and TCRγ chains expressed in the EL4 cell line. We identified immunogenic epitopes from each chain and found that the most immunogenic epitopes were part of the CDR regions of the TCR. This highlights the safety of this approach as the CDR regions of each TCR are unique and there are minimal chances of cross reactivity against non-tumorigenic T cells. The TCRfullvax controlled EL4 tumor growth, for a period of time, in vivo resulting in enhanced survival of tumor bearing mice. However, we observed that tumors downregulate TCR expression and regression of tumor control occurs. This suggests a potential mechanism that tumors develop in response to immune pressure to evade the immune system. We hypothesized that a second immunogen targeting neoantigens derived from EL4 tumors could complement the TCRfullvax and further improve tumor control. Vaccination induced both CD4+ and CD8+ T cells and led to control of EL4 tumors in vivo. Finally, we demonstrate that mice treated simultaneously with TCRfull and EL4neo vaccines had improved tumor control compared to those treated with either vaccine alone and this combination therapy significantly improved animal survival compared to those treated with empty vector controls. While additional improvement is important, these studies highlight that synDNA vaccines could be useful in treating T cell cancers to drive impactful T cell responses. Further development and enhancement of synDNA immunogens for this difficult-to-treat cancer is warranted.

Results

TCRfullvax design and characterization

For TCRfullvax, the sequences of the TCRα, TCRβ and TCRγ chains were obtained from direct RNA sequencing data. The sequences were optimized for codon usage and mRNA expression. A codon optimized human IgE leader sequence was added at the beginning of the vaccine insert to improve the expression and induced immune responses as previously reported [32]. Each individual chain was separated by furin cleavage sites to allow for individual expression of single chains and minimize formation of junctional epitopes (Fig. 1A). For immunogenicity assessment, each mouse was immunized with 25ug DNA in a prime/boost schedule (Fig. 1B). Spleens were harvested one week post the final vaccination, and immunogenicity was measured via IFNγ ELISpot and intracellular cytokine staining by flow cytometry.

Fig. 1.

TCRfullvax design and characterization. (A) TCRfullvax design representing order in which each TCR chain was included. (B) Schematic representation of immunization schedule for TCRfullvax immunogenicity characterization. (C) Mean IFNγ spots generated against TCRα, TCRβ and TCRγ chains by the TCRfullvax. (D) IFNγ spots generated against each peptide pool derived from the TCRα chain. (E) IFNγ spots generated against each peptide pool derived from the TCRβ chain. (F) IFNγ spots generated against each peptide pool derived from the TCRγ chain. (G) Schematic of tumor challenge testing efficacy of TCRfullvax with EL4 cells in vivo. (H) Mean EL4 tumor sizes in mice treated with TCRfullvax or empty vector control pVax. (I) Survival of EL4 tumor bearing mice demonstrating statistically significant improvement in survival of mice treated with TCRfullvax

We observed the strongest immune response against the TCRγ chain with a mean of 3900 IFNγ spots per million splenocytes. This was followed by the TCRα chain (mean 633 IFNγ spots), and the TCRβ chain which was the least immunogenic with a mean of 362 IFNγ spots (Fig. 1C). We performed epitope mapping using mixed matrix peptide pools to identify the most immunogenic epitopes for each TCR chain, the results of which are listed in Table 1. For the TCRα chain, the strongest immune response was against pools 2 and 7 (Fig. 1D and Supplementary Fig. 1A) which contain the epitope SIFTNQ which is predicted to be in the top 1.5% of binders to HLA H2-Kb. This epitope is derived from the CDR1 region of the TCRα chain highlighting that the strongest immune response is being made against the variable regions of the TCR chain. For TCRβ chain, the immune responses were more diverse and pools 3,4,6,8 and 9 elicited similar immune responses (Fig. 1E and Supplementary Fig. 1B). There are several epitopes contained in each pool that are predicted to be within the top 2% of binders to either HLA H2-Kb or H2-Db. Of these the epitope ASSTGTETL (derived from the CDR3 region) is predicted to be within the top 1% of binders to HLA H2-Db and is represented in pools 4 and 9 as well as partially in pool 3 which are among the most immunogenic pools. For the TCRγ chain, the strongest immune responses were against pools 4 and 9 and pools 2 and 7 had the second highest immune response (Fig. 1F and Supplementary Fig. 1C). Pools 4 and 9 contain the epitope VWIEYSSGF (derived from CDR3 region), and pools 2 and 7 contain the epitope VSLPCFSNT (derived from CDR1 region) both of which are predicted to be within the top 1% of binders to HLA H2-Kb. These results highlight that a synDNA vaccine against the TCR chains can break tolerance which results in generation of strong immune responses against TCRα, TCRβ and TCRγ chains. This is further supported by the fact that we did not observe any differences in the number of total splenocytes or percentage of live CD3+ T cells among the naïve mice and TCRfullvax mice (Supplementary Fig. 2A and 2B) suggesting that the immune responses generated are specific for the EL4 TCR chains only and that we do not see any off target immune responses.

Table 1.

Most immunogenic epitopes for each TCR chain

TCR Chain	Most immunogenic pool #	Common epitope	CDR	Percentile of predicted BA to Class 1 MHC
TCRα	2,7	ENAELQCSFSIFTNQ	1	1.23% (H2-Db)
TCRβ	3,6	VLIANHTDAGVTQTP	–	1.2% (H2-Db)
	3,9	PKDSAVYLCASSTGT	3	> 2%
	4,8	AFKDRFKAEMLNSSF	–	1.3% (H2-Kb)
	4,9	YLCASSTGTETLYFG	3	0.89% (H2-Db)
	3,8	LMEDGGAFKDRFKAE	–	> 2%
	4,6	TDAGVTQTPRHEVAE	–	> 2%
TCRγ	4,9	TYYCAVWIEYSSGFH	3	0.96% (H2-Kb)
	2,7	ISCIVSLPCFSNTAI	1	0.88% (H2-Kb)

BA: Binding Affinity

To further characterize the immune response via flow cytometry, we cultured splenocytes from vaccinated mice with peptide pools comprised of overlapping peptides derived from TCRα, TCRβ or TCRγ chains for 5 h. For the TCRα and TCRβ chains, we detected IFNγ, TNFα and IL2 expression on CD8 + T cells, and TNFα and IL2 expression on CD4 + T cells (Supplementary Fig. 3A-3C). For the TCRγ chain, we observed robust secretion of all three cytokines by both CD4 + and CD8 + T cells (Supplementary Fig. 3A-3C). This highlights that TCRfullvax can elicit functionally active T cells that can secrete multiple cytokines.

To test whether the immune responses generated are sufficient to control tumor growth, we vaccinated C57BL6 mice three times, and one week post the final dose, we injected 2e5 EL4 tumor cells subcutaneously on the right flank (Fig. 1G). There was a significant delay in tumor growth in the TCRfullvax immunized mice compared to mice injected with empty vector pVax controls (Fig. 1H and Supplementary Fig. 4A). The delay in tumor growth led to a statistically significant improvement in survival of TCRfullvax-treated mice (Median survival 18 days vs 16 days) (Fig. 1I).

TCRfullvax-treated tumors have greater CD8+ T cell infiltration

To further characterize the impact of vaccine generated immunity on the tumors, we collected tumors at the terminal timepoint, dissociated them into single cells and performed flow cytometry. We observed greater CD8+ T cell infiltration in the tumors from TCRfullvax-treated mice compared to the pVax treated mice (Fig. 2A). These CD8+ T cells in tumors isolated from TCRfullvax mice were less exhausted and less anergic as evidenced by decreased expression of PD1 (Fig. 2B and D) and KLRG1 (Fig. 2C and E). Boolean gating for PD1 and KLRG1 revealed that 94.6% of the CD8+ T cells from TCRfullvax mice did not express either PD1 or KLRG1 and only 0.17% were PD1+ KLRG1+ double positive T cells (Supplementary Fig. 4B and 4C). In contrast, only 59.25% of CD8+ T cells from pVax treated tumors were PD1-KLRG1-double negative and 2% of all CD8+ T cells were PD1+ KLRG1+ double positive (Supplementary Fig. 4B and 4C). Intratumoral CD8+ T cells from pVax mice were 11.7 times more KLRG1+ PD1+ , 8.7 times more KLRG1- PD + , 2.5 times more KLRG1+ PD1- and only 0.6 times KLRG1-PD1- compared to those from TCRfullvax-treated mice (Supplementary Fig. 4D). These data indicate that CD8+ T cells from pVax treated mice were more exhausted compared to those from TCRfullvax-treated mice. We extended our analysis to different memory subsets of the CD8+ T cell population and observed lower expression of PD1 and KLRG1 on both central memory (T_cm) and effector memory (T_eff) populations (Supplementary Fig. 4E-4H).

Fig. 2.

TCRfullvax-treated tumors have greater CD8+ T cell infiltration. (A) CD8+ T cell infiltration in tumors of mice immunized with either pVax or TCRfullvax. (B) %PD1 expression on CD8+ T cells in tumors of mice immunized with either pVax or TCRfullvax. (C) %KLRG1 expression on CD8+ T cells in tumors of mice immunized with either pVax or TCRfullvax. (D) PD1 MFI on CD8+ T cells in tumors of mice immunized with either pVax or TCRfullvax. (E) KLRG1 MFI on CD8+ T cells in tumors of mice immunized with either pVax or TCRfullvax. (F) %TCRβ+ TCRVβ12+ (TCR expressing EL4 cells) in tumors of mice immunized with either pVax or TCRfullvax. (G) Flow cytometry plots showing TCRβ+ TCRVβ12+ double positive cells in tumors of mice immunized with either pVax or TCRfullvax

To elucidate mechanisms of resistance to the TCRfullvax, we isolated tumors and digested them into single cell suspensions for flow cytometry analysis (Supplementary Fig. 5). We observed a significant decrease in expression of TCR Vβ12 in the vaccinated mice vs control mice (Fig. 2F, G and Supplementary Fig. 6). This highlights a potential immune escape mechanism as the tumor has downregulated expression of target antigen in response to targeted immune pressure. It would be desirable to have longer term control even in this highly aggressive mouse model.

EL4neovax design and characterization

Neoantigens are derived from mutations occurring in the tumor cells during their growth and serve as important and unique targets for therapeutic cancer vaccines [13, 21, 25]. We hypothesized that an immunogen design targeting neoantigens derived from the EL4 cells would be immunogenic and should also provide T cell control the growth of EL4 tumors. We identified a total of 27 nonsynonymous mutations that could potentially serve as neoantigen vaccine targets. From these we down selected 15 mutations to encode in our vaccine based on the selection criteria described previously [21] (Supplementary Table 1). The designed immunogen contained 8 single nucleotide variants, and 7 mutations derived from frameshift mutations (Fig. 3A). Of the 15, 2 had a predicted binding affinity of < 500nM (high affinity), 7 had a binding affinity 500nM-2000nM (medium affinity) and 6 had a binding affinity lower than 2000nM (low affinity) based on NetMHC4.0 predictions (Fig. 3B and Supplementary Table 1). Each neoantigen was encoded as a 33 amino acid sequence and separated by furin cleavage sites (Fig. 3C). We immunized C57BL6 mice as in Fig. 1b and analyzed the immune responses one week after final dose. Based on ELISPOT data, 5/15 neoantigens elicited significantly higher numbers of IFNγ SFUs compared to naïve controls (Fig. 3D-3F). Of the neoantigens that elicited significant immunity, 4/5 were predicted to be medium affinity binders and 1/5 was predicted to be a high affinity binder to class I MHC (Fig. 3G). We next performed flow cytometry to further characterize immune responses in these animals (Supplementary Fig. 7A-7C). Of the 5 neoantigens that were immunogenic, 2 epitopes generated both CD4+ and CD8+ T cell responses, 2 epitopes generated CD4+ T cell response only and 1 epitope generated CD8+ T cell response only (Fig. 3H).

Fig. 3.

EL4neovax design and characterization. (A) Bar graphs representing number of single nucleotide variants (SNVs) and Frameshift (FS) mutations which were encoded into EL4neovax. (B) Bar graphs representing breakdown of neoantigens based on predicted binding affinity to Class I MHC. (C) Schematic of EL4neovax design. (D) Mean IFNγ spots generated against 5 immunogenic neoantigens by EL4neovax. (E) Breakdown of percentage of neoantigens that were immunogenic vs non-immunogenic. (F) Representative images showing IFNγ spots generated against the 5 immunogenic neoantigens in EL4neovax. (G) Bar graphs indicating binding affinity of the 5 immunogenic neoantigens in EL4neovax. (H) Bar graphs indicating whether each individual epitope elicited CD4+ , CD8+ T cell response or both

EL4neovax controls EL4 tumors in mice

To test whether the immune responses against neoantigens can impact EL4 tumor growth, we vaccinated mice 3 times and one week post final dose, we injected EL4 tumors on the right flank (Fig. 4A). We observed a statistically significant delay in the growth of EL4 tumors in mice vaccinated with EL4neovax compared to those immunized with empty vector controls (Fig. 4B and C). While there was a trend in improvement of survival in vaccinated animals (median survival 25 days vs 22 days), this did not reach statistical significance (p value = 0.08) (Fig. 4D).

Fig. 4.

EL4neovax controls EL4 tumors in mice. (A) Schematic of tumor challenge testing efficacy of EL4neovax with EL4 cells in vivo. (B) Mean EL4 tumor sizes in mice treated with EL4neovax or empty vector control pVax. (C) Tumor size of each individual mouse treated with either EL4neovax or pVax. (D) Survival of EL4 tumor bearing mice demonstrating improvement in survival of mice treated with EL4neovax

Synergistic effect of TCRfullvax and EL4neovax

Given the impact of both individual vaccines in slowing EL4 tumor growth, we next tested whether the combination of both vaccines would be synergistic in controlling this tumor. We vaccinated C57BL6 3 times with both vaccines and injected EL4 tumors on the right flank 7 days post the third dose (Fig. 5A). Mice that received single vaccination had significantly delayed tumor growth compared to pVax controls. Mice that received dual vaccination had smaller tumors than the mice that received either TCRfullvax or EL4neovax (Fig. 5B and C). All groups that received treatment significantly improved animal survival compared to pVax controls (Fig. 5D and Table 2). These data highlight that the loss of antigen expression driven by the TCRfullvax can at least partially be overcome by combining with DNA vaccine targeting neoantigens.

Fig. 5.

Synergistic effect of TCRfullvax and EL4neovax. (A) Schematic of tumor challenge testing synergy of TCRfullvax and EL4neovax for controlling EL4 cells in vivo. (B) Mean EL4 tumor sizes in mice treated with TCRfullvax, EL4neovax, both vaccines co-delivered or empty vector control pVax. (C) Tumor size of each individual mouse treated with either TCRfullvax, EL4neovax, both vaccines co-delivered or pVax. (D) Survival of EL4 tumor bearing mice demonstrating improved survival of mice treated with combination therapy compared to either vaccine alone

Table 2.

Survival data for synergy experiment

Treatment	Median Survival (Days)	Statistical Significance
		pVax	TCRfullvax	EL4neovax	Combo
pVax	15	–	*	*	*
TCRfullvax	19	*	–	NS	NS
EL4neovax	19	*	NS	–	NS
Combo	19	*	NS	NS	–

Discussion

In this manuscript, we demonstrate that a synDNA vaccine against the TCR of a T cell lymphoma is highly immunogenic and can impact the growth of TCR expressing tumors in vivo. TCRfullvax elicited strong CD4+ and CD8+ T cell responses and majority of the T cell responses were against the CDR regions suggesting that healthy T cells would be spared by the vaccine generated immune response. We observed CD8+ IFNγ responses against all three chains (TCRα, TCRβ and TCRγ) included in the vaccine whereas CD4+ IFNγ responses were only observed against the TCRγ chain, highlighting the value of our vaccine in driving stronger CD8+ T cell responses. We observed a decrease in expression of TCR in mice vaccinated with TCRfullvax indicating a potential immune evasion strategy employed by the tumor cells. Several studies have demonstrated that tumor cells downregulate or modulate antigen expression in response to immunotherapies such as CAR-T cells, Bispecific T cell engagers and vaccine induced immune pressure [33–36]. To counteract the loss of antigen expression, we sequenced the EL4 tumors and identified 27 potential neoantigens. We down selected 15 neoantigens based on their predicted binding affinity to Class I MHC and RNA expression data and designed a new EL4neovax construct. The EL4neovax elicited potent CD4+ and CD8+ T cell responses against 5/15 neoantigens, including against frameshift mutations, and controlled growth of EL4 tumors in mice. Historically, neoantigen-based cancer vaccine development has focused predominantly on the targeting of neoantigens predicted to have high binding affinity to class I MHC (kd < 500 nm). TCLs and other hematological malignancies are often categorized as tumors with a low tumor mutational burden (TMB), with small numbers of targetable neoantigens and therefore unlikely to benefit from targeting of small numbers of private neoantigens [37]. Our data demonstrate that the immune system has the ability to drive strong T cell responses to more neoantigens including against those that are not predicted to be the highest affinity candidates. Indeed, we observed T cell responses to approximately similar proportions of predicted high and medium affinity neoantigens—i.e., 1 of 2 (50%) high affinity antigens and 4 of 7 (57.1%); cumulatively 5/9 (55.5%) medium and high affinity antigens—thereby expanding the tumor directed T cell repertoire via vaccination with a broader range of targetable neoantigens [21, 26, 31]. Finally, we show that animals treated with combination of TCRfullvax and EL4neovax had better tumor control compared to mice treated with either vaccine alone highlighting that downregulation of TCR expression to evade the immune response can at least partially be overcome by vaccinating against tumor derived neoantigens.

Immunotherapies such as monoclonal and bispecific antibodies, CAR-T cells and ICI are playing an essential and growing role in cancer care over the last few years which has improved clinical outcomes for solid tumors and B cell cancers [7–10]. The role of immunotherapy in treatment of TCLs has been limited so far. Three monoclonal antibodies targeting CD52, CD30 and CCR4 have been approved for different subtypes of TCLs depending on antigen expression and clinical subtype [2]. Of these alemtuzumab, an anti-CD52 antibody had an overall response rate of ~ 35% for the treatment of peripheral or cutaneous T cell lymphoma (PTCL or CTCL). Suppression of T cell activity is one of the side effects of anti-CD52 antibodies and treatment with alemtuzumab led to a high rate of opportunistic infections including some that were fatal in about 30% of the patients highlighting the risks associated with this therapy [38, 39]. CART cells are another type of immunotherapy that are being tried for TCL therapy. Most targets for CART cell therapy are also expressed on healthy T cells, which creates the risk of fratricide which leads to manufacturing issues. It can also lead to T cell aplasia which can significantly increase the risk of infections and compromise the quality of life [2]. CART cells against CD30 have been tried in 2 patients with Anaplastic Large Cell Lymphoma (ALCL), one of which experienced a complete response without compromising anti-viral immunity [40]. This highlights the potential utility of CART cell therapy against TCLs and several clinical trials against different subtypes targeting different antigens are ongoing, the results of which will be exciting to see [2]. Among immune checkpoint inhibitors, in early-stage trials with small numbers of patients, anti-PD1 inhibitors have demonstrated an ORR of 15–100% in different TCL subtypes [41–45]. The highest response rates were observed in patients with NKTCLs which is linked with EBV infection which most likely provides additional antigens for T cell targeting and results in higher response rates. These results need to be confirmed in larger trials with more patients. Additionally, in cases of TCL, PD1 can also act as a tumor suppressor and PD1 inhibition can lead to accelerated tumor growth. Ratner et al. reported this phenomenon in a clinical trial of adult T cell lymphoma–leukemia which was terminated after the first 3 patients experienced rapid progression of tumors after anti-PD1 therapy [46]. This serves as a cautionary tale for the use of anti-PD1 therapy for TCLs.

Anticancer vaccine therapy for TCL has been limited to using dendritic cells pulsed with whole tumor lysates and intratumoral injection of TLR9 agonist to stimulate CD8+ T cell activity in CTCL patients with limited efficacy [47, 48]. In recent times, personalized therapeutic vaccines targeting neoantigens have demonstrated impressive results across different tumor types using different vaccine platforms such as DNA, RNA and adenoviral vectors [31, 49–51]. This includes shrinking of large, established HCC tumors which is traditionally considered an immunologically cold tumor [31]. The neoantigen approach led to induction of CD8+ T cell responses in 50% of the patients (although with limited numbers of T cell clones driven) in PDAC patients, another type of cancer that is considered a low TMB tumor and is immunologically cold [50]. CTCL which is a relatively less aggressive form of TCL is perhaps an optimal candidate for personalized anticancer vaccines targeting TCR and/or neoantigens. Recently, Song et al. described CTCL as having a high tumor mutation burden [52] which suggests that CTCL tumors could harbor a high number of neoantigens and it would be critical to target as many as possible to get optimal clinical response. The synDNA platform is optimally placed to address this need as preclinical and clinical studies have shown the feasibility of targeting up to 40 neoantigens simultaneously without compromising immune response or tumor control [25, 31]. In preclinical studies, a 40-mer vaccine was able to control growth of murine lung and ovarian tumors in the same mouse highlighting the value of including as many neoantigens as possible in controlling multifocal tumors with different mutational profiles [25]. This suggests that a synDNA vaccine targeting the TCR and neoantigens derived from CTCL could be effective in the potential treatment of CTCL [52].

Targeting the TCR region of T cell lymphoma has previously been reported to provide some minimal protection in different mouse models [36, 53]. Gonthier et al. used a peptide-based vaccine against the TCR Vβ12 variable region and observed protection in 40–60% of the mice challenged with L12R4 (a mouse TCL cell line) [53]. The study did not specifically analyze vaccine induced immune responses or reasons for why the vaccine was only effective in 40–60% of the mice. This strategy can potentially lead to immune responses against other TCRVβ12 T cells which can also deplete healthy T cells. Tusup et al. describe a mRNA-based vaccine encoding just the CDR3 region of α and β chains of the EL4 cell line which demonstrated some control of the EL4 tumor growth in mice. The authors also observed a decrease in TCR expression indicative of a potential escape mechanism. The authors do not report vaccine induced immune responses and while they did observe a delay in tumor growth, they did not observe an improvement in animal survival [36]. With our vaccine TCRfullvax designs, we did not see any changes in numbers of splenocytes or live T cells. Additionally, majority of the T cell response was against the CDR regions highlighting the safety of the vaccine construct and suggesting that healthy T cells would not be affected by the anti-vaccine immune responses.

Tumors from TCRfullvax-treated mice exhibited greater CD8+ T cells compared to those from control mice. These CD8+ T cells also expressed lower levels of PD1 and KLRG1 compared to intratumoral CD8+ T cells from the control mice. High PD1 expression on T cells is associated with a more exhausted phenotype leading to decreased cytokine secretion, proliferation and cytotoxic potential [54]. KLRG1+ T cells are also poor at proliferation and cytokine productions suggesting diminished ability to kill the tumors [55]. Co-expression of PD1 and KLRG1 on T cells is further associated with an inhibitory phenotype and presence of PD1 KLRG1 double positive CD8+ T cells in the tumors would indicate a reduced ability of these T cells to fight the tumors [56]. Detailed flow cytometry analysis of the tumor infiltrating lymphocytes revealed that tumors from TCRfullvax-treated mice had almost no CD8+ PD1+ KLRG1+ T cells, whereas those from pVax treated mice had almost 2% of their CD8+ T cells co-express PD1 and KLRG1 (11.7-fold higher). Thus, TCRfullvax not only drives greater CD8+ T cells to the tumor, but these tumors are better equipped to control the tumor compared to controls suggesting a mechanistic insight into vaccine’s efficacy in slowing down tumor growth.

In conclusion, the combination of TCR and neoantigen targeting vaccine can be effective in controlling TCLs. This approach would most likely have synergistic effects with other immunotherapy approaches such as ICI which are also being developed for TCLs and IFNα which is used for some TCL therapy. Clinically, the combination of gene delivered pIL12 as a cytokine adjuvant has demonstrated improved T cell responses to vaccine antigens in the cancer and infectious disease settings [31, 57–61]. Such an approach has not been tested in settings of hematological cancers, and it would be interesting to study if cytokine adjuvants can further boost immune responses for enhanced treatment of TCLs. Further development of this approach for treatment of T cell lymphomas is warranted.

Materials and methods

Animals and cell lines

6- to 8-week-old female C57BL/6 mice were purchased from The Jackson Laboratory. All animal experiments were approved by the Institutional Animal Care and Use Committee at The Wistar Institute. The EL4 cell line was purchased from ATCC. EL4 tumors were generated by injecting 200,000 EL4 cells into the right flank. All cell lines were maintained at low passage (< 10 passages) and thawed directly from a master stock generated upon receipt of the cells for all experiments. The cell lines were not genetically authenticated but were examined for morphologic authenticity in cell culture.

Mice were vaccinated by injecting indicated amounts of DNA resuspended in 50 μL of water into the tibialis anterior muscle followed by electroporation with the CELLECTRA-3P device (Inovio Pharmaceuticals). For each vaccination, mice were delivered two 0.1-amp electric constant current square-wave pulses.

DNA and RNA sequencing

We sequenced EL4 cell lines from in vitro cultures. As a control, we used tails from C57Bl/6 mice. The mouse exome and RNA sequencing were performed on the Illumina HiSeq-2500 platform. The SureSelect Mouse All Exon Kit (Agilent Technologies; cat #5190–4642) was used. All samples generated greater than 13 Gb of data, with greater than 98% of the exomes covered at ≥ 150 × . Overall, 99% of the reads aligned to the mouse reference genome. Mapping quality for 80% of the aligned reads was ≥ Q60. Duplicate % was low: 4%–6%. Somatic variant calling was performed using Strelka program v1.0.14 (Illumina Inc.). The identified somatic variants were further filtered (using Strelka parameters such as read filtering, indel calling, SNV calling and other parameters described in https://github.com/Illumina/strelka.), and only passed and on-target variants were considered for further analysis.

The RNA sequencing was done using TrueSeq RNA library prep kit v2 (Illumina, cat. # G9641B). All samples generated > 100 million reads. Reads mapping to the ribosomal and mitochondrial genome were removed before performing alignment. The reads were aligned using STAR (2.4.1) aligner (open source software distributed under GPLv3). Overall 96% to 98% of the total preprocessed reads were mapped to the reference gene model/genome (Mus musculus GRCm38 DNA). The gene expression was estimated using Cufflinks v2.2.1 (Trapnell and colleagues, Broad Institute of MIT and Harvard).

Vaccine design

For TCRfullvax, the sequences for the TCRα, TCRβ and TCRγ chains were obtained from the RNA sequencing data. Each chain was separated by furin cleavage sites to allow for individual expression of each chain and minimize formation of junctional epitopes.

For EL4neovax, neoepitopes were prioritized from either nonsynonymous coding missense mutants or frameshift mutations, where the mutant allele expression was ≥ 1 FPKM. MHC class I binding analysis was performed for all coding mutations. The 9-mer epitopes were analyzed using NetMHCons v4.0 [62] (http://www.cbs.dtu.dk/services/NetMHCcons/) on the C57Bl/6 MHC alleles (H-2-Kb and H-2-Db). We included a total of 15 neoantigens with a wide range of predicted binding affinities for class 1 MHC. For all mutations, we kept the 9-mer epitope including the mutation in the central position and included 12 wildtype amino acids on each side to create a 33-mer epitope. Each epitope was separated by furin cleavage sites. A list of all neoantigens included in the vaccine is provided in supplementary Table 1.

The human IgE leader sequence was added to each construct to improve expression. All constructs were codon and RNA optimized for expression and encoded into pVax1 expression vector.

Flow cytometry

Antibodies conjugated to murine CD4 (RM4-5), CD8b (YTS156.7.7), IFN-γ (XMG1.2), TNF-α (MP6-XT22) and IL-2 (JES6-5H4) were acquired from BioLegend. A Live/Dead Violet viability kit from Invitrogen was utilized to exclude dead cells from the analysis. Intracellular cytokine production was analyzed by culturing splenocytes (1 million) from vaccinated mice with peptides (5 μg/mL) derived from corresponding mutated neoantigens (or pooled peptides derived from the TCRα, TCRβ and TCRγ chains) along with a protein transport inhibitor (eBioscience) for 5 h. Following surface staining, we permeabilized the splenocytes using the eBioscience FoxP3 staining kit. Data were acquired on a BD FACSymphony (BD Biosciences) and analyzed with FlowJo. The neoantigen peptides were designed as 15-mer sequences overlapping by 9 amino acids, covering the entire 33-mer used in vaccination. We vaccinated mice three times at 2-week intervals euthanized mice one week after the final vaccination. We collected spleens, and splenocyte suspensions were prepared using a Stomacher 80 Biomaster (Thomas Scientific), followed by red blood cell lysis (Thermo Fisher Scientific).

ELISPOT

Mice were vaccinated three times at 2-week intervals. One week following the final vaccination, splenocytes were harvested and co-incubated with each neoantigen-derived peptide pool comprising 15-mer overlapping by 9 aa (5 μg/mL). To determine immunogenic epitopes in TCR chains, peptides (15-mer overlapping by 9 aa) derived from TCRα and TCRγ chains were pooled into 10 pools and TCRβ was pooled into 9 distinct pools. Splenocytes were cultured at 37°C. 24 h later, we performed the murine IFN-γ ELISPOT according to the manufacturer’s instructions (Mabtech, no. 3321-4APT-10). Spots were quantified using MabTech IRIS Fluorospot/ELISpot reader, normalized to an unstimulated control (R10). Concanavalin A was used as a positive control. The threshold for immunogenicity was set at 30 SFU per million splenocytes.

Tumor challenge

Mice were vaccinated with indicated construct 3x at 2-week intervals. One week after the final dose, 200,000 EL4 cells (in PBS) were injected on the right flank subcutaneously. Tumor size was monitored via caliper measurements. Mice were euthanized when the length of tumor reached 20 mm or tumor volume reached greater than 2,000 mm [3]. Tumor volume was calculated using the formula V = [(length × width2) × 3.14]/2, where width is considered the side with smaller measurement. Graph represents data from two independent experiments (n = 5 mice/ group). For TCR expression analysis, the tumors from a separate experiment were isolated at day 10 and broken down into a single cell suspension using the mouse tumor dissociation kit (Miltenyi Biotec). The cells were stained for anti-mouse TCRβ and anti-mouse TCR Vβ12 for FACS analysis. Tumors were stained for CD8b (YTS156.7.7), PD1 (RMP1-30), KLRG1(2F1), CD62L (Mel-14) and CD44 (IM7). A Live/Dead Violet viability kit from Invitrogen was utilized to exclude dead cells from the analysis.

Statistical analysis

The difference between the means of experimental groups was calculated using a Tukey’s multiple comparison test. Differences in CD8+ T cell infiltration and expression of PD1 and KLRG1 on CD8+ T cells were calculated using unpaired t tests. Comparisons between two or more groups with multiple subjects were done using ordinary one-way ANOVA. Comparisons between tumor size at each time point were done using two-way ANOVA with Fisher’s least significant difference (LSD) test. Error bars represent standard error of the mean. For mouse survival analysis, significance was determined using a log-rank (Mantel–Cox) test. All statistical analyses were done using GraphPad Prism 10.0. p < 0.05 was considered statistically significant. Error bars represent SEM unless otherwise stated.

Supplementary Information

Below is the link to the electronic supplementary material.

Author’s contribution

PSB, APP and DBW conceptualized the study. PSB, APP and JY designed the vaccine constructs. PSB, DB and JJ performed all experiments. PSB and DBW analyzed the data. All authors contributed to writing and revision of the manuscript.

Data availability

All data supporting the findings of this study are contained within the article and its supplementary content. The raw data can be obtained from the corresponding author upon reasonable request.

Declarations

Conflict of interest

APP, JY and NYS are employees of Geneos Therapeutics (APP is currently at JnJ). DBW has received grant funding, participates in industry collaborations, has received speaking honoraria, and has received fees for consulting, including serving on scientific review committees. Remunerations received by DBW include direct payments and equity/options. DBW also discloses the following associations with commercial partners: Geneos (consultant/advisory board), AstraZeneca (advisory board, speaker), Inovio (board of directors, consultant), Sanofi (advisory board), Pfizer (advisory board), and Advaccine (consultant). The other authors declare that they have no competing interests.