C57BL/6 J male mice were purchased from The Jackson Laboratory and housed under specific pathogen-free conditions in standard temperature and lighting conditions with free access to food and water. Tumor inoculation was performed when mice reached 8–10 weeks of age. All experiments were performed with approval of the Institutional Animal Care and Use Committee (IACUC) at Baylor College of Medicine (BCM) and followed established protocols.

MEER tumor cell line expressing HPV16 E6, E7 and hRas was obtained from Dr. Chad Spanos at the Sanford Research center/ University of South Dakota and maintained in E-media as previously described [26]. C57BL/6 J mice were injected subcutaneously (s.c.) with 10⁶ mEER cells in the flank.

MOC2 cell line was a generous gift from Dr. Ravindera Uppaluri at Brigham and Women’s Hospital/ Harvard Medical School and maintained as previously described [27]. C57BL/6 J mice were injected subcutaneously (s.c.) with 15X10⁴ MOC2 cells in the flank.

Mice were monitored twice a week for tumor growth. Measurements were performed by using calipers and tumor area (mm²) is expressed as L x W, where L is Length and W is Width. Growth curve experiments were stopped once tumors reached 225 mm².

All mice were randomized prior to treatment. Once tumors become established (average tumor size of 65 mm² or 19 mm², day 17–18 or 8 after mEER and MOC2 inoculation, respectively) treatment was initiated over two weeks. The chemoradiotherapy treatment combined a fractionated tumor-directed radiation regimen (10 X 3Gy daily) (see Supplementary Material for further details) with weekly cisplatin (83μg/mouse; Selleck Chemicals) intraperitoneal injections. The immunomodulatory treatment combined weekly intraperitoneal injections of cyclophosphamide (2 mg/mouse; TCI Chemicals) and continuous oral administration of L-NIL (0.2%; Enzo Life Sciences) in the drinking water ad libitum.

For CD8 depletion experiments, all mice received the combinatory treatment and were injected with 1 mg depleting InVivoMAb anti-mouse CD8α (clone 53–6.7; BioXCell) or InVivoMAb rat IgG2a isotype control (clone 2A3; BioXCell) 2 days prior the treatment, and further treated with weekly 250 μg antibodies for 3 consecutive weeks.

The 9 patients included in this study were part of a larger (N = 51) published cohort of HPV⁺ OPSCC patients [28, 29]. All 9 patients included in this study received standard-of-care adjuvant (chemo) radiotherapy following surgical resection of their primary tumor. An overview of patient characteristics and treatments received is available in the prior publication [29]. Patients were involved after signing informed consent and studies were conducted in accordance with the Declaration of Helsinki and approved by the local medical ethical committee of the Leiden University Medical Center (LUMC) and in agreement with the Dutch law.

Following tumor resection, high-dimensional single cell mass cytometry (cyTOF) analysis and functional studies were performed to analyze immune cell population tumor infiltration and cytokine of the general tumor infiltrating lymphocyte. The tumor microenvironment was reanalyzed for the relative low and strong presence of immune cell phenotypes and grouped according to the presence (immune response positive (IR+)) or absence (immune response negative- (IR-)) of HPV16-specific T cell tumor infiltration [28, 29].

To characterize tumor immune cell infiltration, mEER tumors were harvested, digested and stained using the method previously described [30]. Briefly, tumors were digested in RPMI 1640 (Sigma-Aldrich) containing DNase I (20 U/ml; Sigma-Aldrich), Collagenase I (1 mg/ml; EMD Millipore) and Collagenase IV (250 U/ml; Worthington Biochemical Corporation) prior to mechanical disaggregation to form single cell suspensions. Following digestion, tumor infiltrating leukocytes were enriched using Lymphoprep™ (STEMCELL Technologies). Single cell suspensions were also prepared from tumor-draining inguinal lymph node and spleen with additional lysis of splenic red blood cells (RBC) using RBC lysis buffer (Invitrogen). Leukocytes were blocked with anti-mouse CD16/CD32 Fc block (BD Biosciences) and stained using one of various antibody panels (Additional file 1: Table S1 and S2). The viability of cells was determined using LIVE/DEAD™ Fixable Blue Dead Cell Stain Kit (Invitrogen). For intracellular staining, cells were fixed and permeabilized with Intracellular Fixation and Permeabilization Buffer Set (eBioscience) and stained using the appropriate antibody panel (Additional file 1: Table S1). Data were acquired on a LSRII and LSR Fortessa (BD Biosciences) flow cytometers, for myeloid and T cell panels respectively, and analyzed using FlowJo v10 software (FlowJo, LLC).

Sectioning and Staining: After harvesting, tumors were immediately fixed overnight in 10% neutral-buffered formalin. Fixed tumors were embedded in paraffin and sections were cut at a thickness of 5 μm. Full-section slides of tumor tissues were stained using Opal multiplex 6-plex kits, according to the manufacturer’s protocol (PerkinElmer), for DAPI, Epcam (polyclonal; Abcam, 1:100 dilution), CD3 (clone SP7; Spring Biosciences; 1:100 dilution), CD8 (clone 4SM15; ThermoFisher; 1:500), CD4 (clone 4SM95; eBioscience, 1:50), Foxp3 (polyclonal; ThermoFisher, 1:500), and Granzyme B (polyclonal; Abcam, 1:200). Single color controls and an unstained slide were also included.

Multispectral imaging and analysis: Multispectral image capture was done at 20X magnification using Vectra (PerkinElmer, Hopkinton, MA). Images were analyzed using inForm software version 2.4.1 (PerkinElmer) as previously described [31]. Further details are presented in the Supplementary Materials.

Tumor samples were harvested and flash frozen in liquid nitrogen. Total RNA was extracted with the RNeasy Mini Kit (Qiagen) as per the manufacturer’s instructions. Gene expression profiling was performed on 100 ng RNA using the nCounter® PanCancer Immune Profiling Panel (NanoString Technologies, Inc) containing 770 genes involved in cancer immune response. Further details are presented in the Additional file 1.

To observe whether changes in the tumor myeloid compartment after treatment influence intratumoral CD8⁺ T cell cytotoxicity, CD8⁺ T cells were purified from the spleen of naïve C57BL/6 J mice with a magnetic bead-based CD8⁺ T cell negative selection kit (Miltenyi Biotec) and CD11b⁺/CD11c⁺ cells were isolated from mEER tumors undergoing treatment using a magnetic bead-based CD11c and CD11b positive selection kit (Miltenyi Biotec). 10⁵ CD8⁺ T cells were co-cultured with 3X10⁴ CD11b⁺/CD11c⁺ cells in enriched DMEM (20% FBS, 2 mM L-glutamine, 1% non-essential amino acid, 1 mM sodium pyruvate, 50 nM 2-mercaptoethanol, 1% penicillin/ streptomycin; as previously described [32]) including 10 ng/ml IL-2 (BioLegend), 2 μg/ml anti-CD3 (clone: 145-2C11; BioLegend) and 5 μg/ml anti-CD28 (clone: 37.51; BioLegend). After 4 days of co-culture, CD8⁺ T cells were collected and co-cultured with 3X10³ mEER tumor cells at a ratio of 4:1 (CD8⁺ T cell: tumor) in enriched DMEM including IL-2 (10 ng/ml). After 24 h, tumor cell apoptosis was observed via Cytation Cell Imaging Reader (Biotek) and quantified via flow cytometry using an Annexin V/Dead Cell Apoptosis kit (Invitrogen).

Data sets were tested for Gaussian distribution using the D’Agostino-Pearson normality test. For parametric data sets, statistical significance was determined by: unpaired t test for two-tailed data and ANOVA test followed by selected comparison by Tukey’s multiple comparison tests with multiple comparison correction. For non-parametric data sets, statistical significance was determined by: Mann-Whitney test for two tailed data and Kruskal-Wallis test followed by selected comparison by Dunn’s multiple comparison tests with multiple comparison correction. Survival was analyzed by the Kaplan– Meier method using Log-rank test. (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, non-significant). Outliers from flow cytometry analysis were determined using ROUT (Q = 1%) method.

The immunosuppressive effects of CRT and attempts to overcome them have been well documented [33, 34]; however, development of immunomodulatory strategies capable of reversing the balance of CRT immune effects towards activation remains a critical need. To address this, we developed an immunomodulatory strategy capable of rendering the tumor immune microenvironment susceptible to CRT through a “cold” to a “hot” immunologic transition (Fig. 1, Additional file 1: Figure S1). In this study, immunologically “cold” tumors are best characterized by high immunosuppressive cell infiltrate (i.e. MDSC, Tregs), low anti-tumor immune cell infiltrate (CD8⁺ T cells, M1 macrophages, dendritic cells (DCs)), and a lack of immune-favorable gene expression. Alternatively, immunologically “hot” tumors present with favorable effector/suppressor immune cell ratios and evidence of anti-tumor immune activation.Fig1Fig. 1

CTX / L-NIL reverses the unfavorable immune microenvironment of CRT treated tumors. a-c Subcutaneous established mEER tumors (day 17–18 post tumor cell injection) were treated with tumor-directed radiation (5 X 3Gy) and/or weekly cisplatin (83 μg/mouse) i.p. injections, according to the schedule in (a); mice were euthanized when tumors reached 225 mm². b Average tumor area (top) and statistical comparison of tumor sizes at time of first euthanasia (bottom; Tukey’s multiple comparison test). c Survival curves (top) and statistical comparison between treatments (bottom; Log-rank test); (b and c; N = 1 representative of 2; n = 6–8/group). d Heatmap depicting the relative abundance of various immune cell populations from CyTOF analysis in HPV16 immune reactive (IR+) and non-reactive (IR-) OPSCC human tumors based on the presence of HPV16-specific tumor infiltrating lymphocytes (TIL). Frequencies of DC (CD14-HLADR⁺CD11c⁺), M1 macrophages (CD163⁻CD14⁺HLA-DR⁺), monocytic MDCS (mMDSC; CD14⁺HLADR⁻), immature B cells (IgM⁺CD38⁺CD20⁺), memory B cells (IgM⁻CD38⁻CD20⁺), total CD4 and CD8⁺ T cells, naïve (Tn; CD45RA⁺CCR7⁺), central memory (Tcm; CD45RA⁻CCR7⁺) and effector memory (Tem; CD45RA⁻CCR7⁻) CD4⁺ and CD8⁺ T cells, CD4⁺CD161⁺ Tem, CD103⁺CD161⁻ and CD103⁺CD161⁺ CD8⁺ T cells. Th-denotation in indicates amount of IFNγ (Th1), IL-5 (Th2) and IL-17A (Th17) produced by the total TIL culture upon phytohemagglutinin stimulation. All data depicted as log-transformed values (base 2) relative to the median for each individual parameter and each column represents an individual patient (n = 9 total patients). e-f Established mEER tumors were treated with CRT and/or CTX/L-NIL immunomodulation (CTX at 2 mg/mouse i.p. and L-NIL at 0.2% in drinking water) and total tumor RNA was extracted and processed for gene expression analysis after 1 week of treatment, according to schedule in (e). f Heatmap of differential immune gene-set pathway enrichment represented as z-scores between treatment groups (See Additional file 1: Table S3 for immune pathway gene list). g Gene-set based immune cell type enrichment comparing CRT and CRT + CTX/L-NIL represented as z-scores (left; See Additional file 1: Table S4 for immune cell type gene list) and statistical comparison for each immune cell type (right; unpaired t test); (e and f; N = 1; n = 9/group). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, not significant

Based on the murine gene expression analysis we next assessed, in a cohort of human HPV16⁺ oropharyngeal squamous cell carcinoma (OPSCC) patients, whether a favorable tumor immune microenvironment could influence standard-of-care treatment benefit (Fig. 1d). Briefly, the 9 patients included in this study received standard-of-care (chemo) radiotherapy following surgical resection of their primary tumor. Patients were grouped based on the presence (immune response positive (IR+)) or absence (immune response negative- (IR-)) of HPV16-specific T cell tumor infiltration observed prior to treatment as previously described [28, 29]. In a prior study, IR+ OPSCC patients treated with surgery followed by adjuvant CRT were shown to have a 3-fold improvement in 5-year survival compared to IR- (ca. 90% vs 30%) [28, 29]. In our study, immune cell type analysis showed that the immune microenvironment of non-responsive (IR-) tumors was “cold” compared to that of immune-responsive (IR+) OPSCCs. IR+ tumors showed increased levels of several immune-stimulating myeloid subsets (i.e. dendritic cells and M1 macrophages) and effector lymphoid subsets (i.e. effector memory CD8⁺ T cells and Th1), thus linking quality of innate and adaptive intratumoral responses to treatment benefit (Fig. 1d). Together, these observations suggest that modulation of the tumor immune microenvironment to an immune responsive phenotype could increase the efficacy of standard-of-care CRT treatment.

To activate the tumor immune microenvironment, we utilized an immunomodulatory treatment previously developed by our group [20] combining weekly CTX injections with continuous delivery of the iNOS inhibitor, L-NIL, in drinking water (Additional file 1: Figure S1A). To first characterize the immunomodulatory effects of the CTX/LNIL regimen in murine tumors, gene expression analysis was used to study the effects of each component of the treatment. Principal component analysis (PCA) revealed aggregate differences in gene expression and clustering based on treatment component (Additional file 1: Figure S1B; PC1 explained 54.13% and PC2 explained 14.25% of the observed variance). Immune gene-set analysis further revealed a “cold to hot” immunologic transition of the tumor following CTX and CTX/LNIL treatments (Additional file 1: Figure S1C). This dataset and existing literature [23, 35] support the notion that CTX treatment induces favorable immunomodulatory effects. Nevertheless, its combination with L-NIL further enhances immune activation and favors a unique immunologic gene-set upregulation including STING, innate response, and Th1. Further immune cell type enrichment analysis revealed gene signatures consistent with increases in favorable myeloid and lymphoid subsets following CTX and CTX/LNIL treatment (Additional file 1: Figure S1D). Interestingly, CTX/LNIL promoted significant enrichment of effector CD8⁺ T cells and Th1 cells compared to CTX alone, consistent with the infiltrates found in immune responding HPV16⁺ OPSCC human tumors (Fig. 1d). Together, these observations showed that the combination of CTX and L-NIL functions as a potent immunomodulatory regimen capable of multifactorial enhancement of the intratumoral immune microenvironment.

Given the potent immunomodulatory effects induced by CTX/LNIL treatment, we hypothesized that this combination could improve the tumor immune microenvironment, rendering HPV-HNSCC tumors more responsive to CRT. Thus, we developed a combination treatment regimen in which mice bearing established mEER tumors were treated with concurrent CRT and CTX/L-NIL immunomodulation and we profiled the tumors for gene expression changes after the first week of treatment (Fig. 1e). PCA analysis revealed unique clustering of the combinatory CRT + CTX/L-NIL from that of CRT alone, but a similar clustering pattern is noted in comparison of CTX/LNIL and CRT + CTX/LNIL (Additional file 1: Figure S2A; PC1 explained 43.87% and PC2 explained 18.22% of the observed variance), revealing that CTX/LNIL further modulates the CRT-induced tumor immune microenvironment. These data are further supported by immune gene-set analysis showing that numerous anti-tumor immune response pathways downregulated by CRT were upregulated by CRT + CTX/L-NIL, including MHC, toll-like receptor (TLR), complement system, antigen processing, DC and T cell functions (Fig. 1f; for individual sample heat map see Additional file 1: Figure S2B). Interestingly, several immunosuppressive pathways (i.e. iNOS, hypoxia-adenosine and immune check points) but also the immunogenic cell death (ICD) pathway that were upregulated by CRT strongly decreased in expression after the combinatory treatment. Additional immune cell type enrichment analyses highlighted a significant increase in several innate immune cell gene subsets following the combinatory CRT + CTX/L-NIL treatment; however, only inflammatory and plasmacytoid DCs were significantly upregulated compared to CRT alone (Fig. 1g). Overall these results show that, at the level of gene expression, CTX/L-NIL enhances favorable and suppresses unfavorable intratumoral immunologic effects of CRT.

To determine the therapeutic benefit of CRT and/or CTX/L-NIL immunomodulation, we assessed tumor growth and survival (Fig. 2). Briefly, mice bearing established mEER tumors were treated with CRT and/or CTX/L-NIL over a period of two weeks and were then continually monitored for long-term survival (Fig. 2a). While CRT and CTX/LNIL treatments alone each induced modest tumor growth delays compared to untreated mice, neither promoted complete tumor clearance. The combination of CRT + CTX/L-NIL significantly delayed tumor growth compared to singlet treatments and induced an average of 21% complete tumor clearance (12.5 to 30% depending on the experiment; Fig. 2b to d). To further validate these findings and determine if the treatment benefit was restricted to HPV-positive tumors, which bear viral antigens and typically have a better response to CRT, we next assessed the effects of combinatory treatments in an aggressive HPV-negative HNSCC cancer model, MOC2 [36] (Additional file 1: Figure S3). MOC2 cells were injected subcutaneously in the flank of mice and treatment was initiated once tumors became established (approximately 19 mm² in size; Additional file 1: Figure S3A). Combinatory treatment significantly delayed tumor growth compared to CRT and CTX/LNIL alone (Additional file 1: Figure S3 B and C) and increased overall survival by 23 days compared to conventional CRT (Additional file 1: Figure S3D). Collectively, this dataset shows that CTX/LNIL immunomodulation enhances the treatment benefit of conventional CRT regardless of HPV-status, suggesting that this therapeutic approach may be applicable to treatment of a variety of solid tumors.Fig2Fig. 2

CTX/L-NIL combines with CRT to induce durable control of established tumors. Mice bearing established mEER tumors were treated with CRT (10 X 3Gy tumor-directed radiation and 83 μg/mouse weekly cisplatin i.p.) and/or CTX/L-NIL immunomodulation (CTX at 2 mg/mouse i.p. and L-NIL at 0.2% in drinking water) according to the schedule in (a); mice were euthanized when tumors reached 225 mm². b Individual tumor growth curves shown by treatment group, with each mouse represented as a single line. c Average tumor area (top) and statistical comparison at time of first euthanasia (bottom; Tukey’s multiple comparison test). d Survival curves (top) and statistical comparison between treatments (bottom; Log-rank test). (b-d; N = 1 representative of 2; n = 8–9/group). **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, not significant

We have previously shown that CTX/L-NIL inhibits the intratumoral infiltration of immunosuppressive MDSCs [20] and our prior gene expression analysis revealed numerous anti-tumor innate gene-sets and cell types enriched following CRT + CTX/LNIL treatment compared to CRT alone (Fig. 1f and g). Overall, these observations suggest that CTX/LNIL immunomodulation may favorably alter the myeloid tumor microenvironment. Thus, using flow cytometry, we assessed the impact of CTX/LNIL and/or CRT treatment on the tumor myeloid composition and function (Fig. 3). To eliminate immunologic bias due to tumor size, we chose to isolate tumors and observe the immune response at an intermediate treatment timepoint (after one week) when tumor sizes were comparable between all groups (similar to Fig. 1e). The analysis of myeloid cell subpopulations (among CD11b⁺/ CD11c⁺ cells) was performed using flow cytometry (for myeloid gating strategy see Additional file 1: Figure S4A, as previously described in [37]).Fig3Fig. 3

CRT + CTX/L-NIL reprograms the myeloid tumor microenvironment. Established mEER tumors were treated with CRT and/or CTX/L-NIL and harvested after the first week of treatment, as shown in to Fig. 1e, and tumor myeloid cell infiltrate was analyzed using flow cytometry (a-e; see Additional file 1: Figure S4 for myeloid flow cytometry gating strategy) and ex vivo co-culture (f-h). a Myeloid-focused tSNE (among CD11b⁺/CD11c⁺ cells) from flow cytometry data for each treatment group. b Corresponding tSNE color-map (left) and radar plot (right) showing myeloid sub-type alterations between each treatment group as z-scores (myeloid sub-type color in radar plot corresponds with their color in tSNE map; N = 1 representative of 3; n = 8–9/group). c-e Percentage of myeloid sub-types among total myeloid cell tumor infiltrate (CD11b⁺/CD11c⁺), including inflammatory monocytes (c), M1-like macrophages (d), and granulocytic MDSCs (e) (N = 3; n = 23–25/group; Tukey’s multiple comparison test for inflammatory monocytes and Dunn’s multiple comparison test for M1-like macrophages and granulocytic MDSCs). f Experimental schematic used to test treatment-induced myeloid influence on CD8⁺ T cell cytotoxicity. Naïve splenic CD8⁺ T cells were stimulated with anti-CD3/CD28 antibodies and IL-2 in the absence or presence of myeloid cells (CD11b⁺ and CD11c⁺ cells) isolated from tumors that received different treatments for 3 days and then co-cocultured with mEER cells in presence of IL-2. After 24 h, apoptotic tumor cells were detected by Annexin V/ PI staining. g Representative images of mEER tumor cells stained for Annexin V (shown in green) following co-culture (scale bars show 200 μm). h mEER tumor cell apoptotic fold change (Annexin V⁺ / PI⁻) normalized to non-primed T cells (T cells not co-cultured with myeloid cells) (N = 4; n = 20–26/group; Dunn’s multiple comparison test). All bar graphs show mean +/− SD and each dot represents an individual mouse. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Since myeloid cells can both promote and suppress T cells according to their subtype and activation status [40], we sought to determine whether the favorable shift in the myeloid microenvironment induced by combinatory treatment influenced T cell anti-tumor cytotoxicity. To assess this, we performed an ex vivo CD8⁺ T cell cytotoxic assay in which naïve splenic CD8⁺ T cells were stimulated (using anti-CD3, anti-CD28, and IL-2) in the presence of myeloid cells (CD11b⁺ and CD11c⁺ cells) isolated from treated tumors (as previously described in [32]). The myeloid-primed and stimulated CD8⁺ T cells were then co-cultured with mEER tumor cells and assessed for T cell induced cytotoxicity using Annexin-V/PI staining (Fig. 3f). Microscopic examination of Annexin V showed an increase in tumor cell apoptosis after CRT + CTX/L-NIL treatment compared to single therapies (Fig. 3g). Quantitative flow cytometry analysis revealed a significant improvement in the CD8⁺ T cell cytotoxicity, compared to non-primed stimulated CD8⁺ T cells, when influenced by the CRT + CTX/L-NIL intratumoral myeloid compartment (Fig. 3h). These data suggest that the favorable myeloid shift induced with the combination of CRT and CTX/LNIL lends to enhanced CD8⁺ T cell cytotoxicity.

Based on the above data and our previous report showing that CTX/LNIL immunomodulation promotes CD8⁺ T cell tumor infiltration [20], we assessed whether CTX/LNIL and/or CRT could favorably alter the intratumoral lymphoid compartment using multi-color immunofluorescence and flow cytometry (Figs. 4 and 5). After one week of treatment (similar to Fig. 1e) we profiled tumors by multiplex immunofluorescence [31]. We observed an increase in CD8⁺ T cell tumor infiltration and a higher level of cytoplasmic Granzyme B after CTX/L-NIL and combinatory treatment (Fig. 4a). Image quantification of T cell subsets further revealed that CTX/LNIL immunomodulation promoted a CD8⁺ T cell dominated tumor and a drastic depletion of Tregs, and this effect was maintained when combined with CRT (Fig. 4b). Additionally, we observed that CTX/LNIL improved CD8⁺ T cell activation compared to CRT, as evidenced by the expression of cytoplasmic Granzyme B. When combined, although non-significantly, CRT + CTX/L-NIL induces a 4-fold increase in the intratumoral density of CD8⁺ T cells expressing Granzyme B compared to CRT alone (Fig. 4c).Fig4Fig. 4

CRT + CTX/L-NIL promotes intratumoral CD8⁺ T cell infiltration and activation. Established mEER tumors were treated with CRT and/or CTX/L-NIL and harvested after the first week of treatment, as shown in Fig. 1e, and tumor lymphocyte infiltrate was analyzed using quantitative multiplex immunofluorescence (qmIF; a-c). a Representative multiplex images of mEER tumors showing DAPI (nuclei, dark blue), EpCAM (tumor, red), CD8α (CD8⁺ T cells, cyan), and Granzyme B (T cell cytotoxic marker, green). b Pie-charts showing T cell subset fractions among total T cells for each treatment (fraction averaged across 5 images per tumor and n = 3/group). c Log₂ fold change of Granzyme B (GzB)⁺ CD8⁺ cells normalized to control group (N = 1; n = 3/group; Tukey’s multiple comparison test). Bar graphs show mean +/− SD. *p < 0.05; **p < 0.01

To further understand the consequences of the improved CD8⁺ T cell infiltration induced by the combinatory treatment, we next analyzed the tumor-infiltrating CD8⁺ T cells for various functional markers. We first tested whether CRT + CTX/L-NIL could improve CD8⁺ T cell tumor specificity by evaluating E7-specific tetramers binding by flow cytometry (Additional file 1: Figure S6B). This revealed that the proportion of intratumoral CD8⁺ T cells specific for the E7 antigen expressed by mEER tumor cells was significantly increased following CRT + CTX/LNIL treatment (more than 22.7% of CD8⁺ T cells on average) (Fig. 5d). Based on a consensus nomenclature defined for murine and human CD8⁺ T cell phenotypes [41], we monitored the expression of several markers associated with T cell effector, memory, and exhaustion functions using flow cytometry (Fig. 5e and Additional file 1: Figure S8). CRT + CTX/L-NIL promoted a significant increase in the expression of killer cell lectin-like receptor G1 (KLRG1), a marker commonly associated with effector CD8⁺ T cell subsets [42]. However, perforin expression on CD8⁺ T cells was significantly decreased after CRT + CTX/L-NIL treatment compared to CRT alone, suggesting that CD8⁺ T cells had entered a later killing phase [43]. Collectively, these observations show that the tumor infiltrating CD8⁺ T cells promoted by the combinatory CRT + CTX/LNIL treatment exhibit improved specificity and effector phenotypes compared to CRT alone.

Since our data show that CTX/L-NIL immunomodulation enhances susceptibility of immune-refractory tumors to CRT by changing the myeloid compartment, leading to enhancements in CD8⁺ T cell specificity and function, we examined whether improved tumor regression and survival was dependent on CD8⁺ T cells (Fig. 6). To accomplish this, depleting anti-mouse CD8 antibodies were injected two days before treatment and then weekly over the course of CRT + CTX/LNIL treatment (Fig. 6a). Flow cytometry analysis validated the depletion of CD8⁺ T cells observed in the blood (Fig. 6b). In the absence of CD8⁺ T cells, the combinatory treatment induced a small tumor growth delay, compared to control tumors (Fig. 6c and d). The stall in tumor growth observed in the absence of CD8⁺ T cells is likely a combination of direct tumor cell killing by the chemotherapeutic and radiation components of the combination treatment and non-CD8-dependent cytotoxic mechanisms (e.g. Natural Killer-mediated killing). Nevertheless, long term survival benefits are entirely lost in the absence of CD8⁺ T cells (Fig. 6e). These results confirm that CD8⁺ T cells are critical component of the therapeutic benefit induced by the combinatory CRT + CTX/LNIL treatment.Fig6Fig. 6

Tumor growth inhibition induced by CRT + CTX/L-NIL requires CD8⁺ T cells. Established mEER tumors were treated with CRT + CTX/L-NIL and anti-CD8α depleting antibody, or isotype control antibody, according to the schedule in (a); mice were euthanized when tumors reached 225 mm². b CD8⁺ T cell percentages among total viable cells in the blood at day 29 of treatment as assessed by flow cytometry (N = 1; n = 10/group, each as an individual dot; data are means +/− SD; Mann-Whitney test). c Individual tumor growth curves shown by treatment group, with each mouse represented as a single line. d Average tumor area (top) and statistical comparison at time of first euthanasia (bottom; Tukey’s multiple comparison test). e Survival curves (top) and statistical comparison between treatments (bottom; Log-rank test). (c-e; N = 1 representative of 2; n = 10/group). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, not significant