PDS0101 (ImmunoMAPK-RDOTAP/HPV-16 E6 and E7 Peptides) vaccine was obtained from PDS Biotechnology (Princeton, New Jersey, USA) as part of a Collaborative Research and Development Agreement (CRADA) with the National Cancer Institute (NCI), National Institutes of Health (NIH). PDS0101 is a lipid-based vaccine containing six HLA-unrestricted epitopes against HPV16 E6 and E7. The dose used was the murine equivalent of the clinical dose, 300 µg R-DOTAP and 40 µg HPV-peptide mix. Bintrafusp alfa (M7824) consists of two TGFßRII fused to a human IgG1 monoclonal antibody blocking PDL1 and showed safety and de novo generation of antigen-specific responses in a phase I trial in HPV-positive malignancies.18 Bintrafusp alfa was used at a dose of 250 µg. NHS-IL12 is an immunocytokine composed of two IL-12 heterodimers, each fused to the NHS76 antibody, which targets histones in necrotic areas of tumor. For murine studies, the surrogate NHS-muIL12, which is the NHS76 antibody fused to two murine IL-12 heterodimers, was used; this is necessary because human IL-12 lacks bioactivity in the mouse.19 NHS-IL12 was used at a dose of 50 µg. NHS-IL12 has shown safety in a phase I clinical study.20 The bintrafusp alfa and NHS-IL12 agents were obtained from EMD Serono (Rockland, Massachusetts, USA) as part of a CRADA with the NCI, NIH.

The TC-1 cell line (murine lung carcinoma cell line transfected with HPV16 E6 and E7 oncoproteins) was a generous gift from Dr. T.C. Wu (Johns Hopkins University, Baltimore, Maryland, USA), and was tested for mycoplasma and viral contamination according to NIH procedures. TC-1 cells were cultured according to previous studies.13 The mEER cell line (murine oropharyngeal cell line transfected with HPV16 E6 and E7 oncoproteins22) was a generous gift from Dr. Clint Allen (NCI, Bethesda, Maryland, USA), and was tested for mycoplasma and viral contamination according to NIH procedures. mEER cells were cultured according to previous studies.22

Mice were housed in microisolator cages under pathogen-free conditions, in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care guidelines. All animal studies were conducted under approval of the NIH Intramural Animal Care and Use Committee (Protocol #057). C57BL/6J or C57BL/6J-COH mice, from in-house breeding, were shaved on the right flank and instilled with 2×10⁴ TC-1 cells or 5×10⁵ mEER cells in a 1:1 mixture with Matrigel basement matrix (Corning, Corning, New York, USA). TC-1 tumors were allowed to grow for 1 week, or until the tumor was ~5–6 mm on one side. mEER tumors were allowed to grow for 4 days, or until the tumor was ~3–4 mm on one side. PDS0101 was thawed and mixed according to the manufacturer’s instructions (PDS Biotechnology), and 100 µL (300 µg R-DOTAP plus 40 µg HPV peptide mixture) was injected subcutaneously (s.c). Bintrafusp alfa was diluted in sterile 1 x phosphate buffered saline (PBS) to a concentration of 250 µg/100 µL and injected intraperitoneally (i.p.). NHS-IL12 was diluted in sterile 1 x PBS to a concentration of 50 µg/100 µL and injected s.c. Tumor volume was measured biweekly, and mice were euthanized when control (PBS treated) mice reached ethical endpoint.

IFNγ ELIspot was performed overnight according to the manufacturer’s (BD Biosciences, San Jose, California, USA) instructions using 2×10⁵ splenocytes per well. HIV-gag (0.625 µg/mL; American Peptide Company, Sunnyvale, California, USA) was used as a negative control, and PMA/ionomycin (1 µg/mL) cocktail was used as a positive control. Overlapping HPV16 E6 and E7 15-mer peptides (JPT, Berlin, Germany) containing HLA-A2 agonist epitopes from HPV1623 were used as the test antigens (0.625 µg/mL). Results are presented as the number of spots per 5×10⁵ splenocytes after subtracting any spots in the background controls.

Immunohistochemistry for CD8 and CD4 was performed using the Perkin Elmer Opal IHC kit, according to the manufacturer’s instructions (Perkin Elmer, Waltham, Massachusetts, USA). Slides were imaged on an AxioScan (Zeiss, Oberkochen, Germany).

Tumors were excised and homogenized via mechanical dissociation and single cell suspensions were prepared by filtering through a 40 µm nylon cell strainer. Cell suspensions were stained on ice with fluorescently conjugated antibodies diluted in FACS buffer. Dead cells were identified via live/dead fixable stain (ThermoFisher, Waltham, Massachusetts, USA). Antibodies used for flow cytometry were purchased from Biolegend (San Diego, California, USA). Mouse antibodies used were: CD3 (Clone # 500A2), CD4 (Clone # RM4-5), CD8 (Clone # 53–6.7), CD45 (Clone # 30-F11), Ki67 (Clone # B56), CD44 (Clone # IM7), CD62L (Clone # MEL-14), CD25 (Clone # PC61), F4/80 (Clone # BM8), GR1 (Clone # RB6-8C5), FoxP3 (Clone # FJk-16s), CD38 (Clone # 90), PDL1 (Clone # 10F.9G2). Intracellular staining was performed using FoxP3/transcription factor kit (eBioscience, San Diego, California, USA), according to the manufacturer’s instructions. Cells were enumerated using AccuCheck Counting beads (ThermoFisher). Cytometric data were obtained via a four laser Attune Flow Cytometer (ThermoFisher). Data were analyzed via FlowJo (FlowJo, LLC, Ashland, Oregon, USA).

TC-1 tumors were dissociated using Mouse Tumor Dissociation kit (Miltenyi Biotec, Gaithersburg, MD), according to the manufacturer’s directions. Briefly, tumors were cut into 2–4 mm pieces, added to Enzyme D, R and A, and transferred to a gentleMACS C tube. C tubes were attached to the gentleMACS dissociator and run on m_tumor1. Samples were incubated for 30 min at 37°C with continuous rotation. Samples were processed again on the gentleMACS dissociator, filtered, spun down at 300xg for 7 min, then counted and resuspended in isolation buffer (PBS, pH 7.2, 0.5% bovine serum albumin and 2 mM ethylenediaminetetraacetic acid (EDTA)) for tumor infiltrating lymphocyte (TIL) isolation. TILs were isolated using mouse TIL Isolation kit (Miltenyi Biotec). DNA was isolated from TILs using QiaAmp DNA mini kit (Qiagen, Germantown, Maryland, USA), according to the manufacturer’s directions. T-cell receptor (TCR) Vβ CDR3 sequencing (TCRseq) was performed by Adaptive Biotechnologies (Seattle, Washington, USA) using the survey resolution ImmunoSeq platform (Adaptive Biotechnologies); analysis was performed using the ImmunoSeq ANALYZER 3.0 (Adaptive Biotechnologies). Repertoire size, a measure of TCR diversity, was determined by calculating the number of individual clonotypes represented in the top 25th percentile by ranked molecule count after sorting by abundance; this measure is relatively stable to differences in sequencing depth, and not strongly influenced by rare clonotypes.

GraphPad Prism V.7 (GraphPad Prism Software, La Jolla, California, USA) was used to perform statistical analyses. Details of the appropriate analysis are found within each figure legend.

We employed the TC-1 syngeneic mouse model to evaluate the antitumor activity of therapeutic vaccination with PDS0101. Flow cytometry analyzes of CD45^NEG cells from transplanted tumors (n=3) showed between 0.1% and 0.4% of cells expressing PDL1 (online supplementary figure 1). R-DOTAP has been reported to enhance dendritic cell uptake and antigen cross-presentation.12 R-DOTAP has also been demonstrated to specifically upregulate type I IFN and associated chemokines primarily within the lymph nodes, thus promoting the recruitment and priming of antigen-specific CD4+ and CD8+ T cells.12

TC-1 is a murine lung carcinoma transformed with HPV16 E6 and E7, and grows aggressively when implanted s.c. in C57BL/6J mice. Treatment was started 7 days postimplantation. First, PBS control vaccination was compared with three weekly injections of the liposomal adjuvant R-DOTAP, to evaluate if the R-DOTAP formulation would mediate any antitumor activity (figure 1A). Tumor volume in the PBS control group compared with R-DOTAP treatment was not significantly different, indicating that in the absence of HPV peptides, s.c. injection of R-DOTAP provided no antitumor efficacy. Next, the PDS0101 vaccine was tested for antitumor activity compared with PBS control or R-DOTAP control. PDS0101 was also injected s.c. once weekly starting on day 7, ipsilateral to the tumor for 3 consecutive weeks. PDS0101 monotherapy resulted in the slowing of tumor growth compared with PBS or R-DOTAP control (figure 1A), and in significantly lower tumor weights compared with PBS control (figure 1F, p<0.01). Tumor weights at the end of study were used as an additional indicator of antitumor activity.

Studies were then conducted to determine if the antitumor efficacy of the weekly PDS0101 vaccine could be increased with the addition of bintrafusp alfa, an anti-PDL1/TGFBRII fusion antibody14–18 (figure 1B). Bintrafusp alfa was injected i.p. starting day 7, every 2 days for 6 days (three times total) to avoid mouse−antihuman immune responses to the human IgG present in the antibody.17 Bintrafusp alfa monotherapy treatment was not significantly different from PBS control (figure 1A), for tumor growth rates as well as end of study tumor weight (figure 1F), most likely due to the lack of PDL1 expression on TC-1 tumor cells (online supplementary figure 1). The combination of PDS0101 plus bintrafusp alfa-treated mice showed enhanced antitumor responses compared with mice treated with PDS0101 alone. One possible explanation for this could be the observation that PDS0101 treatment of mice resulted in a slight increase in PDL1 expression (approximately 2%–4.5% of tumor cells) (online supplementary figure 1). The rapid rate of growth in the TC-1 model and no cures in these experiments reflect an aggressive tumor model, and mice with a tumor volume below 300 mm³ appeared to have controlled tumor growth more effectively than mice above that threshold. The combination of PDS0101 and bintrafusp alfa resulted in 5/16 mice with tumor volumes <300 mm³, compared with 0/16 mice in the PBS-treated group (figure 1G).

Studies were also conducted employing treatment with NHS-IL12, an immunocytokine fusion protein consisting of IL-12 and the NHS76 antibody, which targets the immunocytokine to the TME.19 20 24 25 NHS-IL12 was injected s.c. once at the start of treatment on day 7, contralateral to the tumor. NHS-IL12 monotherapy resulted in a slower tumor growth rate compared with PBS or R-DOTAP control treatment (figure 1C), as well as significantly lower end of study tumor weight (figure 1F, p<0.01). The combination of PDS0101 and NHS-IL12 further reduced tumor growth as compared with PBS control (p<0.0001). It also resulted in better tumor control, with 10/16 mice displaying a tumor volume <300 mm³, compared with 0/16 mice in the PBS control group (figure 1G).

The combination of treatment with NHS-IL12 and bintrafusp alfa was also investigated. This resulted in decreased tumor growth rate (figure 1D), as well as significantly decreased tumor weight at the end of study (figure 1F, p<0.001). Additionally, this combination resulted in a greater number of mice with tumor volumes below 300 mm³ than either treatment alone, with 8/16 displaying lower tumor volume from the combination treated, vs 0/16 for bintrafusp alfa and 6/16 for NHS-IL12 (figure 1G).

Finally, the combination of all three agents, PDS0101, bintrafusp alfa and NHS-IL12, was evaluated, employing the same schedules described above. The triple combination most effectively reduced the rate of tumor growth (figure 1E), and compared with all the other treatment groups resulted in the lowest average tumor weight at the end of study (figure 1F, p<0.001 vs PBS control). The combination of the three agents also yielded more mice (13/16) with tumors smaller than 300 mm³ at the end (figure 1G).

Flow cytometry was performed on both tumors and spleens from mice bearing TC-1 tumors and treated with single agents or the combinations of agents described above. Spleens were used to evaluate the effects of the agents on the peripheral immune cell compartment. Figure 2A shows a representative example of CD8+ and CD4+ T cells in the spleens (upper panels) and tumors (lower panels) of two mice treated with PBS control (left) and the combination of PDS0101, NHS-IL12 and bintrafusp alfa (right). The overall frequencies of T cells did not change in the spleens after treatment, but the combination treatment led to substantial infiltration of T cells into these tumors (lower panels). In agreement with the antitumor data, there were significant increases in CD8+ and CD4+ T cell infiltration per milligram of tumor in the group treated with PDS0101 plus NHS-IL12 (p<0.05) and the group treated with PDS0101, bintrafusp alfa plus NHS-IL12 (p<0.05) (figure 2B). An increase in activated CD8+ T cells with proliferative potential (Ki67+CD44+) was also seen in the tumor of PDS0101 plus NHS-IL12-treated mice, with a trending increase seen in the other treatment groups (figure 2D), as well as in spleens (figure 2E). CD8+ activated effector T cells, defined as CD44+CD62L+CD8+ increased significantly in the tumors of mice treated with NHS-IL12, PDS0101 plus bintrafusp alfa, NHS-IL12 plus bintrafusp alfa, and PDS0101 plus NHS-IL12 compared with controls (figure 2F). In contrast, there were decreases in naïve CD8 T cells (CD44^NEGCD62L+) in the spleens of mice treated with NHS-IL12 plus bintrafusp alfa and in the triple combination group (figure 2G).

Other immune cell populations, such as natural killer (NK) cells, myeloid derived suppressor cells, and regulatory T cells (Tregs) were not significantly changed in tumor after treatment, whereas the CD38+ macrophages (M1 macrophages) showed a significant increase in the PDS0101 plus NHS-IL12 combination group (online supplementary figure 2).

To investigate whether the numbers of immune cells in the tumor were related to the antitumor effects seen in the different groups, that is, tumor size, correlation analyses were performed. CD4+ T cells, CD8+ T cells, Tregs and macrophage infiltration per milligram tumor from all treatment groups inversely correlated to tumor weight, demonstrating that increased infiltration of immune cells was correlated with smaller tumors (figure 3A) with p values of <0.0001 and r values of between −0.701 and −0.824 for the four cell types. Increases in CD4+ (green) and CD8+ T cells (red) can also be seen in the combination groups compared with monotherapy groups in representative immunohistochemistry images (figure 3B). These changes also correspond with the flow cytometry data (figure 2B).

To evaluate T-cell antigen specificity generated by vaccination with PDS0101, splenocytes were isolated from animals from all groups for analysis using IFNγ ELIspot as described in Methods. Overlapping 15-mer peptides from the human HPV16 E7 protein in the PDS0101 vaccine were used as the target antigen. Only mice in the different groups vaccinated with PDS0101 developed significant antigen-specific responses against the HPV16 E7 peptide (figure 4A).

Because the PDS0101 vaccine also contains a single mouse epitope against E7 (RF9 peptide, part of the human KF18 peptide), assays to evaluate the responses to RF9 were conducted as well. Analyses of splenocytes from PDS0101-vaccinated mice in an IFNγ ELIspot assay showed similar responses using the RF9 peptide compared with HPV16 E7 overlapping 15-mers (figure 4B).

To test the efficacy of the PDS0101 vaccine in another syngeneic HPV-positive tumor model, C57Bl/6J mice bearing mEER tumors were vaccinated with PDS0101. mEER is a syngeneic, oropharyngeal cancer cell line transformed with HPV16 E6 and E7 oncoproteins.22 Three weekly treatments with PDS0101 reduced the tumor burden in individual mice compared with PBS-treated controls (figure 4C, p<0.05). ELIspot analysis using HPV16 E7 overlapping 15-mer peptides for stimulation of splenocytes showed significantly higher IFNγ production in PDS0101-treated versus control-treated mice (figure 4D, p<0.05), confirming that PDS0101 treatment increased antigen-specific T cells in vivo in a second model.

To investigate TCR clonality in the tumor, DNA from TILs was isolated from tumors from each treatment group. The PDS0101 monotherapy group displayed a lower number of TCR clones per 25% of the T-cell repertoire, indicating that the vaccine promoted clonality over diversity (figure 5A). The NHS-IL12 monotherapy group showed the highest number of T-cell clones per 25% of the repertoire, which was expected since IL-12 promotes the proliferation of T cells.26 The NHS-IL12 plus bintrafusp alfa group displayed much higher diversity than other combination groups, which is likely due to the absence of the vaccine (figure 5A). The PDS0101 plus NHS-IL12 group and the PDS0101 plus bintrafusp alfa group each displayed a dramatic decrease in diversity (or increase in clonality), potentially due to the antigen-specific response induced by the vaccine and the proliferation. Finally, the triple combination group displayed the largest decrease in diversity/increase in clonality, most likely showing the effects of the T-cell induction of the vaccine, the tumor targeting of bintrafusp alfa, and the proliferation enhancing effects of NHS-IL12 (figure 5A).