E7 RNA-LPX vaccination and LRT synergize to control established HPV16
In order to evaluate the combination of E7 RNA-LPX and LRT, we designed a schedule employing a subtherapeutic dose of E7 RNA-LPX (late single intravenous immunization) followed by different doses of LRT (two weekly doses of 1.8 Gy, 7 Gy or 12 Gy) in mice bearing well-established (therapy start at 75 mm3) HPV16 E6/E7+ TC-1 tumors. In line with our previous findings  single and late-administered E7 RNA-LPX vaccination significantly promoted tumor rejection (Fig. 1a) and survival in TC-1 tumor-bearing mice compared to control-vaccinated mice (Fig. 1b); however, tumor rejection was followed by cases of relapse (10/14 mice). Double LRT treatment with control RNA-LPX only had a marginal effect on tumor growth compared to non-irradiated mice (control RNA-LPX) but, when combined with E7 RNA-LPX, displayed superior tumor rejection, independently of the tested radiation doses (Fig. 1a, b). Greatest anti-tumor efficacy was achieved when E7 RNA-LPX was combined with double treatment of high-dose LRT (12 Gy), rendering 100% of mice tumor-free up to 100 days after tumor inoculation (Fig. 1b) – a schedule that was chosen for subsequent experiments, however, dispensing the second irradiation to allow the collection of samples for characterization of the tumor immunemicroenvironment by flow cytometry. Seventeen days after E7 RNA-LPX vaccination, 2% of circulating CD8+ T cells were E7-specific whether mice were irradiated or not, indicating the efficient priming of antigen-specific T cell responses after E7 RNA-LPX with or without cytotoxic LRT (Fig. 1c).
Recent reports from studies in mice have shown that radiation dose-fractionation to an intermediate high-dose has superior T cell priming capacity when administered together with an anti-cytotoxic T-lymphocyte-associated protein 4 monoclonal antibody . Therefore, we combined E7 RNA-LPX with 12 Gy fractionated to a similar BED of 3 × 6 Gy (single administration). The rate of survival was at 25% whether the E7 RNA-LPX vaccine was combined with 12 Gy or 3 × 6 Gy LRT, indicating the relevance of total dose rather than LRT dose-fractionation to reach therapeutic synergism with E7 RNA-LPX (Supplementary Fig. 1a, b). The frequency of E7-specific CD8+ T cells in E7 RNA-LPX- and E7 RNA-LPX/LRT-treated mice was monitored over time and displayed a peak around 12 days after vaccination (Supplementary Fig. 1c). Interestingly, E7 RNA-LPX/LRT-treated mice showed the highest persistence of E7-specific CD8+ T cells in the circulation, potentially accounting for prolonged and still ongoing tumor rejection in this treatment group.
The anti-tumor efficacy of combined E7 RNA-LPX/LRT was also evaluated in a second HPV16+ mouse tumor model, C3. In line with findings made in the TC-1 tumor model, the combination of E7 RNA-LPX vaccination and high-dose LRT (reduced to a single treatment) enhanced the survival benefit (Fig. 1d) and the rate of complete responses (CR 6/10 mice; Fig. 1e) of C3 tumor-bearing mice over monotherapies and control-treated mice.
E7 RNA-LPX vaccination alone or in combination with LRT induces high levels of effector immune cell infiltration
The wide clinical use of radiotherapy is based on its cytotoxic and growth inhibitory properties; however, evidence in the last two decades suggests that radiotherapy may also activate the immune system, especially when given at a high-dose and combined with immunotherapy .
To characterize the underlying cellular drivers of tumor rejection after combined E7 RNA-LPX/LRT treatment, we analyzed TC-1 tumor immune infiltrates by flow cytometry (Fig. 2). TC-1 tumors were excised twelve days after a single E7 RNA-LPX vaccination and five days after single dose 12 Gy LRT, which was a point at which tumor growth curves of single therapies already diverged (Fig. 2a). E7 RNA-LPX vaccination significantly increased the infiltration of CD45+ leukocytes from 10% at baseline (control RNA-LPX) to 70%, which was also observed for E7 RNA-LPX/LRT-treated mice (Fig. 2b). LRT with control RNA-LPX also mediated CD45+ leukocyte infiltration compared to non-irradiated mice (control RNA-LPX-treated), although this was to a lesser extent than the E7 RNA-LPX vaccine. Furthermore, E7 RNA-LPX vaccinated mice displayed a higher fraction of intratumoral CD4+, CD8+, NK cells (Fig. 2c) and E7-specific CD8+ T cells in the tumor as well as in spleens and lymph nodes of treated mice (Fig. 2d) when compared to LRT with control RNA-LPX, which only mildly modulated these cell populations. In sum, total immune infiltrates of E7 RNA-LPX/LRT-treated mice largely recapitulated those of E7 RNA-LPX-vaccinated mice (Fig. 2b-d). Within the CD8+ T cell population, 15% of CD8+ T cells from E7 RNA-LPX and E7 RNA-LPX/LRT-treated mice expressed IFNγ and TNFα after ex vivo peptide restimulation (Fig. 2e), indicating comparable effector function of E7-specific T cells in both treatment groups. Interestingly, we detected a higher frequency of the transcription factor TOX1 (Fig. 2f) in intratumoral CD8+ T cells by histology, which is required for a sustained T cell effector function in cancer and chronic viral infection [29, 30].
Tumor-infiltrating myeloid cells were also analyzed (Supplementary Fig. 2). E7 RNA-LPX/LRT-treated TC-1 tumor-bearing mice display an enrichment of total CD11b+ myeloid cells, such as type 2 DC (DC2), M1-polarized inflammatory tumor associated macrophages (TAM) (expressing MHC class II, while negative for CD206), myeloid derived suppressor cells (MDSC) (Supplementary Fig. 2c) and also type 1 dendritic cells (DC1). The expression of the activation markers CD86 and MHC class II (Supplementary Fig. 2d, e) was similar across different myeloid cell subsets independent of the treatment performed, whereas PD-L1 expression was enhanced on myeloid cells of vaccinated mice (Supplementary Fig. 2f).
Combining LRT with E7 RNA-LPX vaccination enhances tumor cell death, reduces hypoxia and promotes CD8+
T cell proliferation
As there was no evidence of cytokine modulation in intratumoral E7-specific CD8+ T cells that explained the superior therapeutic efficacy of E7 RNA-LPX/LRT over E7 RNA-LPX monotherapy, we investigated the impact of treatment on tumor cells, which are the vital cellular subset directly targeted by antigen-specific T cells.
As a cytotoxic therapy, LRT potently induced a threefold reduction of TC-1 tumor cell counts (tumor cells/mg tumor, Fig. 3a), and increased the fraction of apoptotic tumor cells, characterized by the expression of CC3 (Fig. 3b) and the death receptor Fas (Fig. 3c), when compared to control or E7 RNA-LPX vaccinated mice alone. However, the expression of MHC class I, T cell inhibitory ligand PD-L1, and the inhibitory immune receptor Qa-1b (Fig. 3c) were only mildly modulated by LRT. Conversely, E7 RNA-LPX vaccination strongly increased the expression of MHC class I molecules, PD-L1 and, to a lesser extent, Qa-1b (Fig. 3c) on tumor cells, despite not changing the total tumor cell count (Fig. 3a) and only slightly increasing CC3 and Fas expression on tumor cells (Fig. 3b, c) when compared to control RNA-LPX. TC-1 tumors of combination therapy-treated mice shared features of both monotherapies in which LRT-mediated cell death (Fig. 3a, b) was paired with E7 RNA-LPX-mediated induction of MHC class I and PD-L1 expression (Fig. 3c); however, expression of the cell death receptor Fas and T cell inhibitory ligand PD-L1 exceeded those of either monotherapy (Fig. 3c).
Previous in vitro studies using human tumor cell lines have shown that sublethal irradiation can render tumor cells more susceptible to antigen-specific CD8+ T cells, upregulating the expression of cell surface proteins involved in T cell recognition such as Fas, intercellular adhesion molecule 1 and MHC class I molecules . To identify if LRT sensitizes TC-1 tumors to E7-specific CD8+ T cell-induced death, we co-cultured differentially irradiated TC-1 tumor cells (0 to 12 Gy) with CD8+ T cells isolated from spleens of E7 RNA-LPX-vaccinated mice in a controlled manner in vitro and measured the expression of MHC class I, PD-L1 and CC3 on tumor cells and the secretion of effector cytokines by E7-specific CD8+ T cells (Supplementary Fig. 3). Reflecting the previous in vivo observation, co-culture with E7-specific CD8+ T cells strongly increased the expression of MHC class I and PD-L1 on tumor cells (Supplementary Fig. 3b, c), likely a feedback mechanism in response to IFNγ secretion , whereas expression levels remained the same whether tumor cells were irradiated or not. In line with previous reports [31, 33, 34], radiation appeared to sensitize TC-1 tumor cells to E7-specific CD8+ T cell killing with a higher fraction of cells being CC3+ (Supplementary Fig. 3d). In addition, E7-specific CD8+ T cells secreted more IFNγ when co-cultured with irradiated tumor cells (Supplementary Fig. 3e), suggesting that they may have enhanced recognition of irradiated versus non-irradiated TC-1 tumor cells. These in vitro data support the hypothesis that radiation renders tumor cells more susceptible to antigen-specific CD8+ T cell-mediated killing. The molecular mechanisms that drive radiation susceptibility are likely multifactorial and were not further evaluated.
In addition to radiation-induced reduction of tumor cell count and increase of tumor cell killing as compared to control RNA-LPX and E7 RNA-LPX alone (Fig. 3a, b), we characterized radiation-mediated effects on the local tumor microenvironment (TME) that could impact intratumoral E7-specific CD8+ T cells induced by the E7 RNA-LPX vaccine (Fig. 2d). Hypoxia is a hallmark of most solid tumors and commonly promotes immunosuppression . The hypoxia probe pimonidazole  was intravenously injected into control, E7 RNA-LPX-, LRT- and combination therapy-treated TC-1 tumor-bearing mice and the hypoxic tumor areas were analyzed by histology (Fig. 3d). In E7 RNA-LPX/LRT-treated mice, TC-1 tumor hypoxia was significantly reduced compared to all other treatment groups (Fig. 3d). The level of tumor hypoxia thereby correlated with the tumor size at the time point of excision (Supplementary Fig. 4) and tumors of E7 RNA-LPX/LRT-treated mice shared the same hypoxic area than untreated TC-1 tumors at matched tumor sizes (explanted at an earlier time point, day 16). Despite similar levels of tumor oxygenation, vascularization was slightly reduced in combination therapy-treated mice (reveled by CD31 staining of endothelial cells), whereas vasculature architecture was similar (using FITC-dextran vessel leakiness assay) between rejecting E7 RNA-LPX/LRT (day 26) and control tumors (day 16) at matched tumor sizes (data not shown), indicating that neither tumor size nor normalization of the vessel phenotype/morphology seem to play the exclusive role in tumor oxygenation observed after combined E7 RNA-LPX/LRT.
The reduction of tumor hypoxia in E7 RNA-LPX/LRT-treated mice was furthermore was associated with markedly increased proliferation of CD8+ tumor infiltrated lymphocytes (TIL) five days after LRT (Fig. 3e), as shown by the higher incorporation of the base analog BrdU in CD8+ TIL but not in CD4+ TIL.
LRT prolongs the duration of E7-specific CD8
T cell immune responses
Although we did not observe differences in the E7-specific CD8+ T cell response in E7 RNA-LPX and E7 RNA-LPX/LRT-treated mice five days after LRT (Fig. 2d, e), the reduction of tumor cell count, tumor hypoxia and increased CD8+ TIL proliferation led us to investigate the antigen-specific CD8+ T cell response when tumor growth curves diverged more strongly, namely day 29 after tumor injection (Fig. 4a) and hypotheized that a more oxygenated environment can have subsequent effects on immune cell types, such as T cells. Therefore, we characterized functional parameters of E7-specific CD8+ T cell responses such as total tumor infiltration, the secretion of cytokines involved in T cell effector function (IFNγ, TNFα) and proliferation (IL-2), as well as the expression of negative immune checkpoints (TIM-3, PD-1 and NKG2AB [a Qa-1b ligand ) (Fig. 4). Flow cytometry analysis showed that E7 RNA-LPX/LRT treatment induces cell death of tumor cells and hence increases the tumor infiltration of total CD45+ cells and CD8+ T cells, but does not change the fraction of vaccine-induced E7-specific CD8+ T cells among CD8+ T cells, when compared to E7 RNA-LPX-treated mice (Fig. 4b). E7-specific CD8+ TILs from combination treated mice produced significantly more of the effector cytokines IFNγ and TNFα, as well as IL-2, upon in vitro antigen-specific restimulation (Fig. 4c), indicating that combined LRT treatment drives a higher effector function and activation status of vaccine-induced E7-specific CD8+ T cells. E7-specific CD8+ T cells of E7 RNA-LPX/LRT-treated tumors further displayed comparable expression of the immune checkpoint inhibitory receptors TIM-3, a lower expression of PD-1 (Fig. 4d) and a higher expression of the T cell inhibitory receptor NKG2AB (Fig. 4e) than E7-RNA-LPX treated mice. The latter likely correlates with the enhanced IFNγ secretion observed in the combination treated group (Fig. 4c) as NKG2AB is known to be expressed after continuous IFNγ secretion in the TME .
The higher magnitude and effector function of E7-specific CD8+ TILs in E7 RNA-LPX/LRT-treated mice is in agreement with more potent anti-tumor effects observed in vivo.
Together, our data indicate that LRT reduces tumor cell count and tumor hypoxia, thereby amplifying vaccine-induced E7-specific CD8+ TIL effector function and promoting rejection of HPV16+ tumors.