Following mice strains were used throughout the study: C57BL/6 (WT - wild type), Ifnar1 ^-/-, Catchup^IVM-red x C57BL/6, Catchup^IVM-red x Ifnar1^-/- . In Catchup^IVM-red mouse model neutrophil-specific locus Ly6G is modified with a knock-in allele expressing Cre-recombinase and the fluorescent protein tdTomato. Additionally, these mice were crossed with a ROSA26 tdTomato-reporter strain to enable in vivo tracking of fluorescently labeled neutrophils (28). For in vivo luminescence imaging (IVIS^®) a transgenic luciferase-reporter mouse model for IFN-β-production was used (IFN-β^+/Δβ-luc mice) (29).

For the experiments, male littermates between 7-12 weeks were used. All mice used in this study were on the C57BL/6 background. Mice were housed and bred under specific pathogen-free conditions at the animal facility of the University Hospital Essen (Essen, Germany). Animal experiments were performed according to German laws and guidelines of the Federation of European Laboratory Animal Science Associations (FELASA) and were approved by the animal ethics committee.

The murine oropharyngeal carcinoma cell line MOPC (C57BL/6-derived, HPV16 E6/E7⁻) was obtained as a gift from Dr. William Chad Spanos and John H. Lee (30, 31) (Sanford Research/University of South Dakota, Sioux Falls, SD, US). During cultivation, the cell line was regularly tested for mycoplasma contamination with negative results. The cells were cultivated in a special medium (67% DMEM, 22% Hams F12 (both from Gibco by Thermo Fisher Scientific, Waltham, US), 10% Fetal Bovine Serum (FBS, PAN-Biotech, Aidenbach, Germany), 1% penicillin-streptomycin (Gibco), 0.5 µg/ml Hydrocortisone, 8.4 ng/ml Cholera Toxin, 5µg/ml Transferrin, 5 µg/ml Insulin, 1.36 ng/ml Tri-Iodo-Thyronine, 5 µg/ml E.G.F. (all from Sigma by Merck, Darmstadt, Germany) and harvested by 90% confluency. Cells were grown in a monolayer at 37°C in a humidified incubator with 5% CO₂.

The MOPC cells were injected subcutaneously (s.c. 1 x 10⁶ in 100 µl PBS) into the flank of C57BL/6 and Ifnar1 ^-/-mice, tumor growth was measured by caliper every 2 to 3 days. Tumor-free animals from the same strains were used as control animals. Tumor volume was calculated as follows:

Latest on day 14 (unless stated otherwise) after tumor implantation mice were sacrificed and tissues were harvested under aseptic conditions. Blood was taken by heart puncture with a heparinized syringe and stored on ice. Tumors and inguinal TDLNs were harvested and stored in DMEMc (DMEM, 10% FBS, 1 mM sodium pyruvate and 1% penicillin-streptomycin, all from Gibco, by Thermo Fisher Scientific, Waltham, US) on ice for a maximum 30 minutes before processing.

Tissue samples (tumors and TDLNs) were digested in an enzyme solution (dispase 0.2 µg/ml, collagenase A 0.2 µg/ml, DNase I 100 µg/ml, all Sigma-Aldrich/Merck, Darmstadt, Germany, in DMEMc), 1 ml per sample for 45 min at 500 rpm and 37°C. Cells were filtrated through 100 µm sterile strainers (Cell Trics, Partec, Sysmex Europe GmbH, Goerlitz, Germany), centrifuged at 300g +4°C for 5 min and the supernatant was discarded. The pellet was resuspended in 100 µl PBS containing Fc block (anti-mouse CD16/CD32 (BD, Franklin Lakes, US) and incubated for 15 min at +4°C.

For in vivo imaging, MOPC cells were injected subcutaneously (s.c. 1 x 106 in PBS) into the flank of IFN-β+/Δβ-luc mice. On day 14 post tumor implantation mice were i.p. injected with 150 mg/kg d-luciferin (XenoLight D-luciferin, Bioluminescent Substrate, PerkinElmer, Waltham, US) in PBS, anesthetized using Isofluoran (CP-Pharma, Burgsdorf, Germany) and monitored using an IVIS^® 200 imaging system (Caliper LS, PerkinElmer). For IFN-β expression, photon flux was quantified at the projection of the tumor-draining inguinal lymph nodes (same side versus contralateral side) using the Living Image 4.4 software (Caliper LS, PerkinElmer) expressed in photons/s/cm2/steradian.

On day 14 post tumor cell injection, mice were sacrificed, LNs were extracted and snap-frozen in liquid nitrogen in Tissue-Tek O.C.T. Compound (Sakura Finetek, Japan) containing 5% paraformaldehyde. Afterwards, 7-μm cryosections were prepared, fixed with -20°C cold acetone and stained with a hematoxylin-eosin, based on standard protocol. Microscopy was performed using Olympus BX51 manual upright epifluorescence microscope (Shinjuku, Japan) and tissues were analyzed manually for the presence of metastasis based on the morphology.

Murine LNs were dissected and snap-frozen in liquid nitrogen. About 7-μm cryosections were prepared on slides and fixed with -20°C cold acetone. The sections were rehydrated in PBS, blocked with 3% goat serum in PBS and stained with anti-CD3 (clone 17A2, 3 µg/ml), anti-Ly6G (clone 1A8, 3 µg/ml), anti-CD197 (CCR7, clone 4B12, 3 µg/ml), anti-Ki-67 (clone 11F6, 3 µg/ml) and DAPI [(0,7 µg/ml), all from Biolegend, San Diego, US]. After 1.5h incubation at RT, slides were washed with PBS and ddH₂O, dried and mounted with Neo-Mount (Merck, Burlington, US). Microscopy was performed using Zeiss AxioObserver Z1 Inverted Microscope with ApoTome Optical Sectioning equipped with filters for DAPI, FITC, Alexa Fluor 488, GFP, DsRed, and Cy3. Images were processed with ZEN Blue 2012 software and analyzed with ImageJ. Quantification of CCR7 expressing neutrophils in LNs was performed per 1 mm² field of view and calculated per LN.

Single-cell suspensions were stained with antibodies anti-CD11b (clone M1/70, 0,20 µg/million cells) with isotype Rat IgG2b, κ (clone RTK4530, 0,20 µg/million cells), anti-Ly6G (clone 1A8, 0,20 µg/million cells), anti-CD197 (CCR7, clone 4B12, 1,0 µg/million cells) with isotype Rat IgG2a, κ (clone RTK2758, 0,20 µg/million cells), anti-CD3 (clone 17A2, 0,25 µg/million cells), anti-CD4 (clone GK1.5, 0,25 µg/million cells), anti-CD8 (clone YTS156.7.7, 0,25 µg/million cells), anti-H-2K^d (MHC class I, clone SF1-1.1, 0,25 µg/million cells) with isotype Mouse IgG2a, κ (clone MOPC-173, 0,25 µg/million cells), anti-CXCR4 (clone L276F12, 0,25 µg/million cells) with isotype Rat IgG2b, κ (clone RTK4530, 0,25 µg/million cells), CD80 (clone 16-10A1, 0,7 µg/million cells) with isotype Armenian Hamster IgG (clone HTK888, 0,7 µg/million cells), CD86 (clone GL-1) with isotype Rat IgG2a, κ (clone RTK2758, 0,7 µg/million cells), anti-Ki-67 (clone 11F6, 0,25 µg/million cells) with isotype (clone Rat IgG2b, κ, 0,25 µg/million cells) anti-IFN-γ (clone XMG1.2, 0,25 µg/million cells) and Flash Phalloidin ((20 units/million cells) all from Biolegend, San Diego, US)), ICAM-1 (clone KAT1, 0,25 µg/million cells) with isotype Rat IgG2b, κ (clone eBR2a, 0,25 µg/million cells), MHC-II (I-A/I-E, clone M5/114.15.2, 0,2 µg/million cells) with isotype control Rat IgG2b, κ (clone eB149/10H5, 0,20 µg/million cells), both from eBiosciences by Thermo Fisher Scientific, Waltham, US) and viability dye eFluor 780 (0,10 µg/million cells, Invitrogen by Thermo Fisher Scientific). Flow cytometry was performed using BD FACSCanto™ system and data were analyzed using BD FACSDiva™ software (BD, Franklin Lakes, US). The specific marker expression levels are shown as mean fluorescence intensity (MFI) and calculated by substruction of corresponding isotype control MFI from specific marker MFI.

Neutrophils were isolated from LN single-cell suspensions using flow sorting (FACS Aria III Cell Sorter (BD, Franklin Lakes, US), based on the phenotype (single alive Ly6G⁺, CD11b⁺ cells) with the purity >95%. T-cells were isolated from single-cell suspension from spleen of naïve WT mice using flow sorting based on the phenotype (single alive CD3⁺ cells) with the purity >95%. Sorted neutrophils were co-incubated in DMEMc with recombinant murine IFN-λ2 (Novus Biologicals, Littleton, US) in the concentration 10 ng per 50 000 cells, during 4h at 37°C in a humidified incubator with 5% CO₂. After sorting, T-cells were loaded with CFSE (Biolegend, San Diego, US) according to manufacturer recommendation and co-incubated with anti-CD3/28-beads (STEMCELL, Vancouver, Canada) during 72h in the absence or presence of neutrophils (WT and Ifnar1 ^-/-), isolated from TDLNs (50.000 neutrophils and 50.000 T-cells were used for each condition). On day 3 the cell suspension was incubated for 4h at 37°C with Brefeldin A and Monensin (for IFN-γ accumulation) followed by staining of surface and viability markers (30min at 4°C). The intracellular markers [IFN-γ and F-actin (Phalloidin)] were stained in next step, after fixation and permeabilization with the Fix/Perm Buffer Set according to the manufacturer’s instructions (all reagents from Biolegend, San Diego, US).

Anesthesia and surgeries were performed in accordance with FELASA guidelines. The Catchup^IVM-red x C57BL/6 and Catchup^IVM-red x Ifnar1^-/- mice were placed under ketamine/xylazine (100 µg/g/10 µg/g in 0.9% NaCl i.p.) anesthesia before surgery. During surgery and following imaging, mice were anesthetized with isoflurane (1,5% mixed with oxygen, 250 ml/min). The body temperature was kept at ∼36°C using a feedback-controlled heating pad. 30 min prior to the MP-microscopy, T lymphocytes were fluorescein-labeled by tail-vein injection of 100 μl PBS containing 10 μg FITC coupled CD3 antibody (clone 17A2, Biolegend, San Diego, US). Two-photon microscopy of the previously surgically exposed inguinal lymph node was performed using the Leica DM6000 CFS-based MP microscope with simultaneous detection via photomultipliers (PMTs) and hybrid reflected-light detectors (HyDs) and an HCX IRAPO L 25×/0.95-NA (numerical aperture) water-immersion objective (Leica, Wetzlar, Germany). Illumination was performed at 960 nm using a Coherent Chameleon Vision II Ti: sapphire laser. Td-Tomato-expressing neutrophils were detected with a 585/50 filter, and T-cells were detected by FITC signal with a 525/50 filter. The lymph node structures were detected by second harmonic generation (SHG) with a 460/50 filter cube. Rendering three-dimensional (3D) stacks and image processing was performed using IMARIS image visualization and analysis software (OXFORD instruments, Belfast, UK). Cells counts were performed in three different regions of interest (ROIs) of 150x150 µm per lymph node. To account cell movement over time, cell counts were performed at the beginning and at the end of the imaging time frame (15 minutes) and then averaged per ROI. Cell interactions were manually counted in ROIs over the duration of 15 minutes. Immediate cell proximity between fluorescently labeled neutrophils and T-cells in a three-dimensional plane was counted as a cell-cell interaction.

The MOPC cells were injected subcutaneously (1 x 10⁶ s.c. in 100 µl PBS) into the flank of C57BL/6 and Ifnar1 ^-/-mice as described above. For therapy approach, mice were intraperitoneally injected with 8 µg of recombinant murine IFN-λ2 (Novus Biologicals, Littleton, US) in 100 µl of 0,9% NaCl solution at days 0, 2, 4, and 7 after tumor setting. Animals injected only with 100 µl of 0,9% NaCl were used as controls. Tumor growth was measured as described above. On the day 9 mice were sacrificed, the right inguinal TDLNs were removed, single-cell suspensions were prepared and flow cytometry was performed as described above.

The MOPC cells were injected subcutaneously (1 x 10⁶ s.c.) into the flank of Catchup^IVM-red x C57BL/6 and Catchup^IVM-red x Ifnar1^-/- mice. The in vivo multiphoton imaging was performed on day 14 after tumor setting. Treatment with recombinant murine IFN-λ2 (20 µg in 100 µg PBS i.p., Novus Biologicals, Littleton, US) was performed twice: 24 hours and 1 hour before the imaging.

Single cells from TDLNs were collected and stored in RNAlater (Invitrogen) RNA stabilization solution at –20C. The RNA was isolated using Qia Shredder and RNeasy Mini Kit (Qiagen, Hilden, Germany) and the cDNA was produced using the Superscript II Reverse Transcriptase Kit (Invitrogen, Thermo Fisher Scientific, Waltham, MA, US). qRT-PCR was performed at 60°C annealing temperature. As housekeeping gene, Rps9 was used. The mRNA expression was measured using the Luna Universal qPCR Master Mix (New England BioLabs, Ipswich, MA, US). Relative gene expressions were calculated by 2^-ΔCt formulations. Primer sequences (Rps9, Ifnb1, Tnf, Ccl19, ccl21, Cxcl1, Cxcl2) are available upon request.

For analysis of cytokines profiles were used Proteome Profiler Array (Mouse XL Cytokine Array Kit, Cat.N ARY028, R&D Systems, Minneapolis, USA), based on manufacturer guidelines. In short, LN tissue (0.06 g of tissue in 0.2 ml DMEMc) was incubated for 4h at 37°C. Supernatant was collected and incubated overnight at 4°C with antibody-coated spotted membranes from the kit. After incubation, membranes were washed three times, biotin-conjugated antibody mix was added with the following incubation for one hour at room temperature. Next, the membrane was washed three times, and an HRP-conjugated antibody was added. Then, membranes were incubated in a chemiluminescence reagent that produced a light signal proportional to the amount of antibody-target analyte complexes formed. Analysis and quantification of signal intensity were done using FIJI (ImageJ) software. Signal intensities were normalized to the control spots intensities on the corresponding membranes.

Statistical analyses were performed using Kruskal-Wallis ANOVA with the Bonferroni correction for multiple comparisons and Mann-Whitney U test for two independent samples. Data are shown as median with interquartile range or as an individual values with median. P<0.05 was considered significant.

Neutrophils have been shown to control the pre-metastatic niche formation, therefore we aimed to investigate the role of TDLN-infiltrating neutrophils in this process with the focus on type I IFNs as regulators of neutrophil activity. First, we evaluated the growth of transplantable MOPC in WT and type I IFN receptor (IFNAR1) knockout (Ifnar1 ^-/-) mice and observed significantly elevated primary tumor growth in Ifnar1 ^-/- animals ( Figures 1A, B, D, E ) compared to WT controls. Moreover, we observed significantly induced IFN-β expression in TDLNs at day 14 post tumor injection ( Figure S1 ), which could be responsible for the initiation of anti-tumor adaptive immune responses in WT animals. At this time point, we observed the increased size of TDLNs in Ifnar1 ^-/- animals ( Figures 1C, F ), but no visible metastases were present in TDLNs in both mouse strains ( Figure 1G–J ). In parallel, elevated neutrophil and T-cell infiltration was observed in Ifnar1 ^-/- animals ( Figures 1K, L and S2 ).

To underpin the mechanism responsible for elevated neutrophil infiltration of TDLNs in type I IFN deficiency, we analyzed the expression of CCR7 on LN neutrophils, as this receptor is suggested to be involved in the recruitment of neutrophils to T-cell zones (32). We demonstrated a significant increase of CCR7 expression on LN neutrophils in Ifnar1^-/- vs WT mice in comparison to tumor-free levels ( Figure 2A , Figure S3 ) and the elevated fraction of CCR7⁺ neutrophils in tumor-bearing Ifnar1^-/- mice in comparison to tumor-free animals estimated by flow cytometry ( Figure S3 ) (Neutrophils were indicated as single alive CD11b⁺Ly6G⁺ cells, gating strategy is presented on Figure S4 ; representative dot plots are presented on Figure S3 ). These findings are in line with immunofluorescent histological staining, which shows increased amounts of CCR7⁺ LN neutrophils in tumor situation, especially in Ifnar1^-/- mice ( Figures 2B–F , S5 ). In parallel with increased CCR7 expression on neutrophils, expression of CCR7 on lymphocytes in TDLN was also increased in Ifnar1^-/- animals ( Figure S6 ). At the same time, the expression of other molecules possibly involved in neutrophil homing into lymph nodes, such as CXCR4, CD11b ( Figures S7A, B ) and chemotaxis such as CXCL2, TNF-α, CCL21, ( Figures S7F–H ) was not significantly changed on a gene ( Figures S7C, E, G ) and protein levels ( Figures S7D, F, H ). The expression of CCL19 was changed on the gene level ( Figure S7I ), but not on protein level ( Figure S7J ).

Next, we aimed to evaluate changes in the phenotype and functions of TDLN-associated neutrophils related to type I IFN deficiency. We assessed the expression of molecules involved in antigen presentation and T-cell activation on neutrophils and observed a downregulated expression of MHC-I and ICAM-1 on neutrophils derived from Ifnar1^-/- animals ( Figures 3A, B, D ). At the same time, no differences in the expression of other markers associated with antigen-presenting properties of neutrophils, i.e. MHC-II, CD80 and CD86, were found ( Figures S8A–C ).

It was shown that filamentous actin (F-actin) polarization at the immune synapse of antigen-presenting cells maintains synapse stability necessary to induce T-cell activation (33). We performed the phalloidin staining of isolated LN neutrophils to check cytoskeleton organization with flow cytometry, as microscopical evaluation of F-actin in neutrophils is impeded due to the size and complex shape of the nucleus (34). We detected a significantly lower F-actin levels in Ifnar1^-/- as compared to WT ( Figures 3C, D ), suggesting decreased F-actin polarization and therefore impaired immunological synapse stability in IFN I deficiency as a possible reason for impaired interaction between neutrophils and T-cells in TDLNs.

Since MHC-I and ICAM-1 are known to be involved in the formation of immunological synapse between APCs and T-cells, we aimed to visualize the interactions between neutrophils and T-cells in regional LNs of tumor-bearing animals.

The Catchup^IVM-red mice were used for imaging of neutrophils and their cell/cell interactions in tumor-draining inguinal LNs via two-photon intravital microscopy. To assess the interaction of neutrophils with T Lymphocytes in such LNs in vivo, T Lymphocytes were labeled with FITC-coupled CD3 antibody. We observed that tumor injection led to a significant downregulation of neutrophil/T-cell interactions in TDLNs of Ifnar1^-/- mice, but not in WT mice ( Figures 4A–C ) (Supplementary Movies 1, 2). Naïve non-tumor-bearing WT and Ifnar1^-/- mice were used as a control. We did not observe any differences in interaction numbers between control groups ( Figure 4C ). Of note, to avoid artifacts due to different neutrophil numbers in LN, the number of neutrophil/T-cell interactions was recalculated per one neutrophil.

The observed decrease of neutrophil interactions with T-cells in Ifnar1^-/- mice could lead to the impaired activation and proliferation of effector T Lymphocytes. Indeed, we observed significantly reduced activation of T-cells responses in TDLN tissue and isolated cells of Ifnar1^-/- mice, estimated by Ki-67 expression ( Figures 4D–J and S9 ). In line with that, lower IFN-γ expression in CD4⁺ and CD8⁺ T-cells in TDLNs was observed ( Figures 4K, L )

To prove the role of neutrophils in activation of T-cells in vitro, we co-incubated isolated WT vs. Ifnar1^-/- TDLN neutrophils with WT naïve T-cells. Indeed, we observed significantly suppressed T- cell activation after co-incubation with Ifnar1^-/- neutrophils, as visualized by reduced IFN-γ production by both CD4⁺ and CD8⁺ T Lymphocytes ( Figures 4M, N ).

Altogether, these results suggest that deficiency in type I IFNs might be responsible for impaired neutrophil/T-cell interactions, resulting in impaired activation of anti-tumor immune responses. This in turn could be responsible for the observed accelerated tumor growth in IFNAR1-deficient mice.

Therapeutic strategy aiming to restore type I IFN signaling in neutrophils could enhance their anti-tumoral properties (35). Among others, type III interferons (interferon lambda, IFN-λ) are the most functionally similar to type I IFNs and share downstream signaling pathway, although signaling through distinct receptors (36). Therefore, we aimed to evaluate whether IFNAR1 deficiency could be rescued by the activation of IFN signaling downstream of the inactive receptor, using IFN-λ, hereby potentially activating immunostimulatory properties of Ifnar1^-/- TDLN neutrophils.

We first isolated Ifnar1^-/- TDLN neutrophils and treated them in vitro with IFN-λ to evaluate their capacity to stimulate T-cells. Indeed, we observed elevated T-cell activity indicated by their IFN-γ production after co-incubation with such treated neutrophils, as compared to untreated Ifnar1^-/- cells ( Figures 5A–D ). Moreover, treatment with IFN-λ increased MHC-I expression by TDLN neutrophils ex vivo ( Figure 5E ).

Next, we assessed the capacity of IFN-λ to rescue cell stimulatory properties of IFNAR1-deficient neutrophils in vivo. To this end, we treated Ifnar1^-/- tumor-bearing mice with IFN-λ and monitored neutrophil/T-cell interactions using the in vivo multiphoton imaging, as described above. In agreement with our in vitro data, we observed elevated counts of interactions between neutrophils and T-cells in TDLNs of IFN-λ treated mice, compared to untreated animals ( Figure 5F ).

Finally, to prove the therapeutical effect of IFN-λ on anti-tumor immune responses and tumor growth in IFNAR1-deficient animals in vivo, we set tumors into mice and treated animals with IFN-λ as described in methods. Tumor growth was monitored and mice sacrificed at day 9. Indeed, Ifnar1^-/- mice treated with IFN-λ show significantly decreased tumor growth, as compared to untreated Ifnar1^-/- mice ( Figure 5G ).

Collectively, these data suggest that inactive IFNAR1 on neutrophils leads to the prominent defect in the neutrophil-mediated activation of anti-tumor T-cell responses in TDLNs. IFNAR1-independent rescue of type I IFN signaling, using IFN-λ, improves the immunostimulatory capacity of neutrophils in TDLNs and contributes to the suppression of tumor growth.