Article Text

Original research
Type III interferon inhibits bladder cancer progression by reprogramming macrophage-mediated phagocytosis and orchestrating effective immune responses
  1. Bo Wang1,2,
  2. Bingkun Zhou1,2,
  3. Junyu Chen1,2,
  4. Xi Sun1,
  5. Wenjuan Yang2,3,
  6. Tenghao Yang1,2,
  7. Hao Yu1,
  8. Peng Chen1,
  9. Ke Chen1,
  10. Xiaodong Huang1,
  11. Xinxiang Fan1,
  12. Wang He1,2,
  13. Jian Huang1,2 and
  14. Tianxin Lin1,2
  1. 1 Department of Urology, Sun Yat-sen Memorial Hospital, Sun Yat-sen (Zhongshan) University, Guangzhou, China
  2. 2 Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Guangdong-Hong Kong Joint Laboratory for RNA Medicine, Medical Research Center, Sun Yat-Sen Memorial Hospital, Sun Yat-sen (Zhongshan) University, Guangzhou, China
  3. 3 Department of Hematology, Sun Yat-sen Memorial Hospital, Sun Yat-sen (Zhongshan) University, Guangzhou, China
  1. Correspondence to Mr Tianxin Lin; lintx{at}mail.sysu.edu.cn; Dr Jian Huang; huangj8{at}mail.sysu.edu.cn

Abstract

Background Interferons (IFNs) are essential for activating an effective immune response and play a central role in immunotherapy-mediated immune cell reactivation for tumor regression. Type III IFN (λ), related to type I IFN (α), plays a crucial role in infections, autoimmunity, and cancer. However, the direct effects of IFN-λ on the tumor immune microenvironment have not been thoroughly investigated.

Methods We used mouse MB49 bladder tumor models, constructed a retroviral vector expressing mouse IFN-λ3, and transduced tumor cells to evaluate the antitumor action of IFN-λ3 in immune-proficient tumors and T cell-deficient tumors. Furthermore, human bladder cancer samples (cohort 1, n=15) were used for immunohistochemistry and multiplex immunoflurescence analysis to assess the expression pattern of IFN-λ3 in human bladder cancer and correlate it with immune cells’ infiltration. Immunohistochemistry analysis was performed in neoadjuvant immunotherapy cohort (cohort 2, n=20) to assess the correlation between IFN-λ3 expression and the pathological complete response rate.

Results In immune-proficient tumors, ectopic Ifnl3 expression in tumor cells significantly increased the infiltration of cytotoxic CD8+ T cells, Th1 cells, natural killer cells, proinflammatory macrophages, and dendritic cells, but reduced neutrophil infiltration. Transcriptomic analyses revealed significant upregulation of many genes associated with effective immune response, including lymphocyte recruitment, activation, and phagocytosis, consistent with increased antitumor immune infiltrates and tumor inhibition. Furthermore, IFN-λ3 activity sensitized immune-proficient tumors to anti-PD-1/PD-L1 blockade. In T cell-deficient tumors, increased Ly6GLy6C+I-A/I-E+ macrophages still enhanced tumor cell phagocytosis in Ifnl3 overexpressing tumors. IFN-λ3 is expressed by tumor and stromal cells in human bladder cancer, and high IFN-λ3 expression was positively associated with effector immune infiltrates and the efficacy of immune checkpoint blockade therapy.

Conclusions Our study indicated that IFN-λ3 enables macrophage-mediated phagocytosis and antitumor immune responses and suggests a rationale for using Type III IFN as a predictive biomarker and potential immunotherapeutic candidate for bladder cancer.

  • Phagocytosis
  • Urinary Bladder Neoplasms
  • Macrophages
  • Immune Checkpoint Inhibitors

Data availability statement

Data are available on reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Effective immune responses necessitate activation by interferons (IFNs), with type III (λ) being more recently associated with type I IFNs (α). The antitumor efficacy of IFN-λ has been documented in several mouse models. Nonetheless, a more comprehensive understanding of the mechanistic impacts of IFN-λ on the bladder tumor microenvironment and how IFN-λ contributes to immune checkpoint therapy response is warranted.

WHAT THIS STUDY ADDS

  • In this study, we demonstrated that ectopic Ifnλ3 expression in mouse bladder tumor cell enhances T cell and macrophage tumor immunity, which further improves the efficacy of immune checkpoint inhibition. Furthermore, elevated IFN-λ3 expression correlates positively with the presence of effector immune cell infiltrates and the efficacy of immune checkpoint blockade therapy in human urothelial carcinoma of the bladder.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • These findings suggest that IFN-λ facilitates both T cell-mediated and non-T cell-mediated antitumor immune responses, making it a potential candidate for bladder cancer immunotherapy targeting.

Background

Urothelial cell carcinoma of the bladder (UCB) is one of a few solid malignancies routinely treated with immunotherapy.1 2 The most exciting approach has been the immune checkpoint inhibitors, particularly programmed cell death protein 1/programmed cell death ligand 1 (PD-1/PD-L1) inhibitors, which represent a major treatment breakthrough for some patients with locally advanced or metastatic UCB.2 3 However, many patients still experience unsatisfactory clinical outcomes.4 A comprehensive understanding of the fundamental cellular and molecular mechanisms responsible for antitumor immunity could lead to optimized treatment strategies.

Interferons (IFNs) are required to activate effective immune responses during tumor progression.5 They play a central role in reactivating immune cells for immunotherapy-mediated tumor regression.6 7 Currently, IFNs are subdivided into three major classes: type I (α1-13, β, ε, κ, ω), type II (γ), and the recently described type III (λ).5 8 Pre-existing IFN signatures, particularly type I IFNs, are highly associated with immune checkpoint inhibitor treatment efficacy in various human cancer types, including UCB.9–11 However, the limited efficacy and constitutional symptom-related adverse events associated with IFN-α treatment have restricted enthusiasm for its clinical application.12

Type III IFN, which is related to IFN-α, consists of four IFNL genes (IFNL1, IFNL2, IFNL3, and IFLN4) in humans and only two functional Ifnl genes (Ifnl2 and Ifnl3) in mice.8 13–15 IFN-λ plays a crucial role in infections, autoimmunity, and cancer.16–19 The antitumor activity of IFN-λ has been reported in several mouse models, including melanoma, fibrosarcoma, and lung carcinoma.13 20 21 In humans, IFN-λ has been demonstrated to directly induce apoptosis in various tumor cells13 22 or stimulate monocyte-derived macrophage (Mφ) differentiation,23–25 cytotoxicity, phagocytosis, and proinflammatory cytokine secretion23 25 to mediate immune response in vitro. Although IFN-λ induces a weaker STAT1 activation than IFN-α, it can still trigger an effective immune response in a cell type-dependent manner, including in epithelial and myeloid cells.16 24 26 Therefore, IFN-λ-based therapies, alone or with other immunotherapies, could potentially overcome the adverse effects associated with IFN-α-based treatments.

Here, we aimed to elucidate IFN-λ function and mechanisms in bladder cancer progression and determine its clinical significance in human UCB tissues. We developed a retroviral vector encoding mouse IFN-λ3 and transduced tumor cells to evaluate the antitumor effect of IFN-λ3. IFN-λ3 secreted by transduced tumor cells markedly inhibited in vivo subcutaneous tumor growth in the MB49 mouse bladder tumor model. We further confirmed that the effect of IFN-λ3 on tumor progression depended on repolarizing tumor-associated Mφs toward antitumoral phenotypes and functions, including tumor cell phagocytosis and improved adaptive immune responses. Furthermore, we found that PD-L1 blockade was more effective in leading to tumor regression in IFN-λ3high than IFN-λ3low tumors. Finally, we demonstrated that IFN-λ3 is expressed by human UCB tissues and stromal cells and that high IFN-λ3 expression was positively associated with high densities of infiltrating effector T, natural killer (NK), and antigen-presenting cells, and the efficacy of immune checkpoint blockade therapy. Thus, our findings indicate that IFN-λ uniquely activated innate and adaptive immune responses against tumor progression.

Methods

Mice

Female C57BL/6 and BALB/c nude mice aged 6–8 weeks were purchased from a certified vendor (SJA Laboratory Animal, China). The mice were housed under specific pathogen-free conditions with a 12-hour light-dark cycle in the South China University of Technology animal facility.

Cell line

The mouse bladder cancer MB49 cell line was used to investigate the in vitro and in vivo functions of IFN-λ3.27 Cells were tested yearly for mycoplasma contamination using a mycoplasma detection kit (Yeasen, China). The mouse Ifnl3 plasmid was purchased from IGEbio (China). For stable transfection, HEK-293T-derived lentivirus was used in conjunction with polybrene (Beyotime, China). Specifically, HEK-293T packaging cells were seeded at a density of 1×106 cells per medium-sized dish. The HEK-293T cells were transfected with 3 µg pCDH- Ifnl3 vector, 2.25 µg psPAX2, and 0.75 µg pMD2G using 400 µl Opti-MEM and 9 µl x-tremeGENE (Roche, Switzerland). After collecting the supernatant with the viral particles over 48 hours, it was filtered through a 0.45 µm syringe filter and concentrated using 4 M NaCl and PEG 8000. The viral particles were resuspended in phosphate buffered saline (PBS) and stored at –80°C following centrifugation at 3200 g for 20 min. The cells were cultured at 37℃ under 5% CO2 in DMEM (Gibco, USA) supplemented with 10% fetal bovine serum (FBS, ExCell Bio, China), 100 units/mL penicillin, and 100 µg/mL streptomycin (Gibco, USA). In some experiments, MB49 cell line was stimulated with 100 ng/mL recombinant mouse IFN-λ3 (rmIFN-λ3, Novus biologicals, USA) for 16 hours to evaluate its inflammatory response.

In vivo tumor cell injection and treatment

Tumors were established by subcutaneous injection of 3×105 empty vector overexpressed MB49 (MB49-EV) cells or Ifnl3 overexpressed MB49 (MB49-Ifnl3 OE) cells into the flank of C57BL/6 and 1×105 MB49-EV or transfected MB49-Ifnl3 OE cells into the flank of BALB/c nude mice, respectively. Subcutaneous MB49-EV and MB49-Ifnl3 OE bladder cancer models were established in C57BL/6 mice aged 6–8 weeks old to evaluate the effect of anti-PD-L1 treatment. From the seventh day after implantation, the mice were administered intraperitoneally injections, once every 3 days until the end of the experiment, of 100 µg anti-mouse PD-L1 antibody (BioXcell, USA) or Rat isotype IgG (BioXcell, USA) as control. The tumors were measured every 2–3 days, and the volumes were calculated in mm3 using the following formula: volume=[length (mm)×width (mm)2/2. The mice were killed at indicated time or if their tumor volume exceeded 2 cm3 or they exhibited any signs of ulcers.

Human UCB samples

We enrolled 15 patients who underwent transurethral bladder tumor resection or radical cystectomy during 2020–2021 (cohort 1) and 20 who underwent radical cystectomy for resectable UCB during 2021–2022 (cohort 2) at the Sun Yat-sen Memorial Hospital. Cohort 1 UCB tissue samples were used for immunohistochemistry (IHC) and multiplex immunoflurescence (mIF) analysis to assess the expression pattern of IFN-λ3 in UCB tissues and correlate it with immune cells’ infiltration. Cohort 2 received at least three neoadjuvant therapy cycles (anti-PD-1, gemcitabine, and cisplatin) before cystectomy. Pathological complete response was defined as pT0N0. IHC analysis was performed in cohort 2 to assess the correlation between IFN-λ3 expression and the pathological complete response rate. Tumors were graded according to the WHO 2016 classification and staged using the TNM classification (eighth edition, 2017).28 Follow-up was performed every three months in the first year, every 6 months in the second year, and annually thereafter. The patients’ clinicopathological characteristics are described in online supplemental table 1.

Supplemental material

Histology, IHC, and mIF

Tumor samples fixed in formalin and embedded in paraffin were cut into 5 µm sections and stained by H&E, IHC, and mIF as previously described.29 H&E-stained MB49-EV and MB49-Ifnl3 OE tumor sections were analyzed for immune cell infiltration using the Nuclear Phenotype feature on the HALO Image Analysis Platform (Indica Labs, USA).

For IHC, antigen retrieval was done with Tris-EDTA (pH=9.0) in a high-pressure cooker for 15 min. Unspecific protein binding sites were blocked with 5% BSA (Thermo Fisher Scientific, USA) for 1 hour at room temperature. Slides were incubated with primary antibodies (listed in online supplemental table 2) at 4°C overnight. The sections were developed with peroxidase-conjugated secondary antibodies and stained with peroxidase and 3,3′-diaminobenzidine tetrahydrochloride in an Envision System (Agilent Dako, USA). Hematoxylin was used for counterstaining. After dehydration and mounting, animal slides were captured at 200×magnified high-power fields by an ECLIPSE Ni-E/Ni-U microscope (Nikon, Japan) and analyzed by Image J software (http://imagej.nih.gov/ij/). Human slides were scanned by Vectra Polaris Automated Quantitative Pathology Imaging System (PerkinElmer, USA). Subsequently, InForm V.2.5.0 software (PerkinElmer, USA) randomly defined and quantified 16 zones within the bladder cancer tissues, each measuring 0.64 mm2.

For mIF, sections were incubated with anti-mouse or anti-human primary antibodies (online supplemental table 2), followed by incubation with appropriate peroxidase-conjugated secondary antibodies (Agilent Dako, USA). Detection was performed using a PANO 5-plex IHC kit (PANOVUE, China), which includes PPD520-labeled, PPD570-labeled, and PPD690-labeled Tyramide. The slides were counterstained with DAPI. Stained slides were scanned into images. Image analysis was performed using InForm V.2.5.0 software (PerkinElmer, USA). The analysis procedures included tissue and cell segmentation and scoring. Tumor and blank areas were defied in tissue segmentation to accurately calculate the tumor areas. For cell segmentation, the nucleus was defined by a relative DAPI intensity of 0.1, a splitting sensitivity of 0.4, a minimum size of 20 nm, and a cytoplasm thickness of 1.0 nm. Positivity scoring thresholds for the various markers were determined by automatic calculations and manual adjustments.

RNA and protein isolation

The tumor RNA and proteins were extracted, purified, and quantified following previously established methods.30 For RNA extraction, bladder tumor samples were homogenized in RNAiso Plus (Takara, Japan) using a Tissuelyser (Shanghai Jingxin, China) at 60 Hz, work 30 s, rest 30 s, five cycles for 5 min. After centrifugation, the supernatant was stored at –80°C until use. RNA was isolated using the RNA Quick Purification kit (ESscience, China), quantified using Nanodrop One (Thermo Fisher Scientific, USA), and stored at –80°C until use. For protein extraction, bladder tumor samples were homogenized in PBS containing 0.1% Tween 20 (Servicebio, China) and 4% protease inhibitor (CWBio, China) using a Tissuelyser at 60 Hz, work 30 s, rest 30 s, five cycles for 5 min. The homogenate was transferred to a new tube and snap-frozen in liquid nitrogen for 1 min, followed by thawing in a 37°C water bath for 3 min. Subsequently, samples were sonicated for 1 min and centrifuged at 13 000 g for 10 min at 4°C. The supernatant was transferred to a new tube and stored at –80°C until use. Protein concentration was determined by a bicinchoninic acid assay kit (CWBio, China). Specific protein levels in the tumor microenvironment (TME) were quantified as the rate of their concentration to the total protein concentration (ng/mg).

Reverse transcription quantitative real-time PCR

Total RNA was extracted from culture cells or bladder tumor tissues using the RNA Quick Purification kit (Esscience, China). A 1 µg of total RNA was reverse transcribed to cDNA templates using a cDNA synthesis kit (Yeasen, China) and amplified following the manufacturer’s instructions. Reverse transcription quantitative real-time PCR (RT-qPCR) was performed in triplicate using SYBR Green (Accurate Biology, China) and LightCycler 480 II (Roche, Switzerland) or ABI QuantStudio Dx (Thermo Fisher Scientific, USA) in a reaction volume of 10 µL. Gene expression was calculated using the 2−ΔΔCt method. The target genes’ expression was normalized to GAPDH and displayed as fold change relative to the MB49-EV control group for cells and tumor tissues. Primers were purchased from Tianyi Huiyuan Gene Technology (China; online supplemental table 3).

RNA sequencing

Mouse bladder tumor total RNA samples were sent to HaploX Genomics Center (China) for RNA sequencing (RNA-Seq). Each group comprised three samples. The RNA samples were qualified and quantified using Agilent 4200 TapeStation, Qubit 3.0, and Nanodrop. Transcriptional libraries were prepared and sequenced using Illumina PE150 (USA). RNA-Seq data were aligned to the mouse reference genome (mm10). Raw sequencing data were demultiplexed and examined for quality. Differentially expressed genes (DEGs) were screened using the R package limma (https://bioconductor.org/) with a threshold of log2 fold change ≥1 and adjusted P<0.05, and further analyzed by R packages clusterProfiler, DOSE, GO.db, and topGO (https://bioconductor.org/).

Data retrieval and preprocessing

RNA-seq expression data from human bladder cancer and normal samples were obtained from The Cancer Genome Atlas Urothelial Bladder Carcinoma (TCGA-BLCA) public database (https://tcga-data.nci.nih.gov/tcga/). Based on the Creative Commons 3.0 License, the complete expression data and detailed clinical information of the UCB immunotherapy-related IMvigor210 cohort were obtained from http://research-pub.gene.com/IMvigor210CoreBiologies/packageVersions/.

Wilcoxon tests assessed whether the IFNL1, IFNL2, and IFNL3 genes were differentially expressed between tumor and adjacent normal samples in TCGA-BLCA dataset.

The microenvironment cell populations (MCP)-counter method was used to define IFNL signature score and estimate the relative abundance of several immune cell populations.31 We ran the MCP-counter method using an IFNL gene set that included IFNL1, IFNL2, and IFNL3 to construct IFNL signature score. We also identified the effector genes of tumor infiltrating immune cells (TIICs) using previous studies (online supplemental table 4).32 High (top 50%) and low (bottom 50%) IFNL signature scores defined by the MCP-counter method were used for gene set enrichment analysis and correlation analysis with TIICs in human bladder cancers from TCGA-BLCA dataset. Furthermore, the correlation between IFNL signature score and the response rate to immunotherapy was investigated using the IMvigor210 dataset.

Flow cytometry (FACS)

Fresh tumor tissues, lymph nodes, and spleens were harvested from tumor-bearing mice and used for flow cytometric analysis.33 Bladder tumors and tumor-draining lymph nodes were incubated with collagenase type II (Worthington, USA) and filtered through a 70 µm cell strainer to obtain single-cell suspensions. Spleen suspensions were also filtered through a 70 µm cell strainer and treated with an RBC lysis buffer (CWBio, China) to remove the red blood cells. The lymph node, spleen, and tumor cells were stained with live/dead fixable viability dye and subsequently with the following monoclonal surface antibodies: anti-CD45.2, anti-CD3ε, anti-NK1.1, anti-CD4, and anti-CD8α (online supplemental table 5). For intracellular and nuclear staining, cells were fixed, permeabilized with the Foxp3 Transcription Factor Staining Buffer Set (eBiosciences, USA), and stained with antibodies against T-bet, Foxp3, Ki-67, and granzyme B (online supplemental table 5). Stained cells were assayed using a CytoFLEX flow cytometer (Beckman Coulter, USA). The single-cell populations were gated on the forward/side scatter plots for analysis. The live cell population was gated for the fixable viability dye-negative population. Gates for intracellular markers were determined using IgG isotypes. Data were analyzed using FlowJo software (TreeStar, USA).

ELISA

The cytokines and chemokines were quantified by ELISA following the manufacturer’s instructions (R&D Systems, USA). The Mouse IFN-λ2/3 DuoSet ELISA Kits (R&D Systems, USA) were used to measure the expression of IFN-λ3 in mouse bladder tumor tissues or cell line supernatants (online supplemental table 6). In the case of human samples, the expression of CXCL9 and CXCL10 in Mφs, supernatants stimulated with human IFN-λ1 (10 ng/mL) or IFN-α (10 ng/mL) were quantified using the human CXCL9 and CXCL10 DuoSet ELISA Kits (R&D Systems, USA; online supplemental table 6).

In vitro proliferation assays

Cell counting kit-8 (CCK-8) and colony formation assays were conducted to compare the proliferation capacity of MB49-Ifnl3 OE and MB49-EV. In the CCK-8 assay, the transfected tumor cells were seeded in 96-well plates at a density of 1000 cells per well for 4 days. Viable cells were incubated with 10% CCK-8 reagent (ApeXbio, USA) for 2 hours, followed by measurement of absorbance at 450 nm using a microplate reader (TECAN Spark, Switzerland). In the colony formation assay, the transfected tumor cells were seeded in 6-well plates at a density of 1000 cells per well and maintained in DMEM complete medium containing 10% FBS for 5–7 days. The cells were fixed with methanol, stained with 0.01% crystal violet for 15 mins, and imaged and analyzed using vSpot Spectrum (AID, Germany) to quantify colony formation.

Generation of human polarized Mφs

Peripheral blood mononuclear cells were isolated from buffy coats obtained from healthy donors using a Ficoll density gradient as previously described.34 Monocytes were separated from the peripheral blood mononuclear cells using anti-CD14 magnetic beads (Miltenyi Biotechnology, UK) following the manufacturer’s instructions. Resting Mφs were obtained by culturing 106 /mL CD14+ monocytes for six days in RPMI 1640 containing 10% FBS with 25 ng/mL recombinant human M-CSF (PeproTech, USA). Mφs were polarized by stimulating them with 10 ng/mL recombinant human IFN-α (rhIFN-α, R&D Systems, USA) or 100 ng/mL recombinant human IFN-λ1 (rhIFN-λ1, R&D Systems, USA; online supplemental table 6) for either 6 or 12 hours.

Generations of bone marrow-derived Mφs

To generate bone marrow-derived Mφs (BMDMs), bone marrow cells were collected from femurs and tibias of C57BL/6J mice aged 6–8 weeks old.27 35 After lyzing the erythrocytes with RBC Lysis Buffer (CWBio, China), 1×106 cells/mL were seeded in 6-well plates. BMDM culture medium comprised DMEM with 10% FBS, 1% penicillin/streptomycin (Gibco, USA), and 20 ng/mL murine M-CSF or 10 ng/mL murine granulocyte-macrophage colony-stimulating factor (GM-CSF, PeproTech, USA; online supplemental table 6). The culture medium was replaced every 3 days. On days 6–7, the BMDMs were confirmed to be >80% F4/80+ by FACS, performed as described above, and used for Mφ polarization and in vitro phagocytosis assays. Cells were dissociated using a cell scraper and vigorous pipetting. In some experiments, GM-CSF stimulated BMDMs were polarized by stimulating them with 100 ng/mL rmIFN-λ3s (Sino Biological, China; Novus Biologicals, USA; online supplemental table 6) for indicated time.

Western blotting

GM-CSF stimulated BMDMs were treated without (M0) or with 100 ng/mL LPS (M(LPS)) or 10 ng/mL IL-4 (M(IL-4)) for evaluating type III interferon receptor (IFNLR1) expression. The proteins were extracted following protocols as previously described.34 They were separated by 10% SDS-PAGE prior to immunoblotting with antibodies against IFNLR1 (Abclonal, China) or GAPDH (Proteintech, China) and then visualized using an ECL kit (Merck Millipore, USA).

Ex vivo phagocytosis assay

BMDMs were cultured with live MB49 bladder tumor cells as a phagocytosis assay.25 35 Phagocytosis of live tumor cells was measured by FACS. Specifically, BMDMs were cultured with GFP-expressing MB49 cells at a 1:2 ratio for 18 hours. Subsequently, the cells were washed and labeled with PE/Cyanine7 anti-mouse CD11b antibody (Biolegend, USA) or BV510 rat anti-mouse F4/80 antibody (Biolegend, USA) as outlined in online supplemental table 5. The resulting samples were analyzed using a CytoFLEX flow cytometer (Beckman Coulter, USA).

Statistical analysis

Graphs were generated, and statistical analysis was performed using Prism V.8.0 (Graphpad software, USA). Two independent groups were compared using the Student’s t-test when the F test outcome was insignificant; otherwise, the Mann-Whitney U test was used. Means of three or more groups were compared by one-way ANOVA with Tukey’s multiple comparison test. Repeated measures two-way ANOVA compared the tumor growth and Cck-8 curves. Spearman’s rank correlation assessed protein correlations in the TME. The individual data points shown in the graphs represent independent biological replicates. RT-qPCR results were normalized to the zero time point, set to a relative quantity of 1.0. IFN-λ3-positivity in human bladder cancer tissues was defined as ≥1% IFN-λ3+ cells. Two-tailed P values of <0.05 were considered statistically significant. It should be noted that the TCGA and IMvigor210 datasets were analyzed separately from the other experiments described above.

Results

Ectopic Ifnl3 expression significantly inhibits MB49 tumor progression

We used the subcutaneous MB49 bladder cancer mouse model to evaluate the in vivo antitumor effect of IFN-λ3. Given the low endogenous expression of IFN-λ3 in the MB49 cell line, we ectopically overexpressed Ifnl3 in MB49 cells and confirmed the altered expression by RT-qPCR and ELISA (figure 1A,B). We also confirmed high IFN-λ3 expression in MB49-Ifnl3 OE tumor tissues by ELISA assay and IHC staining (and online supplemental figure 1A,B). Although MB49-Ifnl3 OE cells did not change tumor cell proliferation when compared with MB49-EV cells, as indicated by CCK-8 and colony formation in vitro assays (figure 1C and online supplemental figure 1C,D), ectopic Ifnl3 expression significantly reduced MB49 tumor growth and weight in immune-competent C57BL/6 mice (figure 1D–F). H&E staining of sections, analyzed by Nuclear Phenotype on the HALO software platform (Indica Labs, USA), revealed increased immune cell infiltration in MB49-Ifnl3 OE tumors (figure 1G). Moreover, there is a decrease in cell proliferation (Ki-67 staining) and an increase in cell apoptosis (cleaved caspase-3 staining) in MB49-Ifnl3 OE tumors (figure 1H,I). These results suggested that IFN-λ3 induced a potent antitumor response against bladder cancer.

Supplemental material

Figure 1

Ectopic Ifnl3 expression significantly inhibits MB49 tumor progression. IFN-λ3 is highly overexpressed at the mRNA (A) And protein (B) Levels in MB49-Ifnl3 OE compared with MB49-EV cells. (C) The CCK-8 assay indicates that MB49-Ifnl3 OE and MB49-EV cells show similar proliferation ability in vitro (two-way ANOVA). Representative images (D; scale bar: 1 cm), growth kinetics (E; two-way ANOVA), and weight at harvest (F) of MB49-Ifnl3 OE and MB49-EV tumors 20 days after implantation (n=8 each). (G) H&E staining for calculating the immune cell population in MB49-Ifnl3 OE and MB49-EV tumors by the Nuclear Phenotype on the HALO software platform. (H) Representative images of Ki-67 staining and quantification of Ki-67+ cells (percentage of the total number of cells within the tumor sections). (I) Representative images of cleaved caspase-3 staining and quantification of cleaved caspase-3+ cells (percentage of the total number of cells within the tumor sections). The bar graph presents means±SEM. P values were determined by Student’s t-test when the F test outcome was insignificant; otherwise, the Mann-Whitney U test was used. (E, F) Shown are representative of three independent experiments. (G) Data are representative of three biologically independent mice. (H, I) Shown are representative of two independent experiments. ANOVA, analysis of variance.

Ectopic Ifnl3 expression promotes immune activation in MB49-Ifnl3OE tumors

We compared RNA-Seq transcriptome analysis of MB49-Ifnl3 OE and MB49-EV tumors 23 days after subcutaneous tumor cells injection to investigate the potential mechanisms underlying the in vivo antitumor effect of IFN-λ3. The DEGs ranked by fold difference between two groups were listed in online supplemental table 7. As expected, most upregulated genes were associated with types I and III interferon responses, JAK-STAT signaling and lymphocyte activation in MB49-Ifnl3 OE tumors (figure 2 and online supplemental figure 2C). Further analysis revealed significant upregulation of many genes associated with cytokine-cytokine receptor interactions, chemokine signaling pathways, and phagocytosis in MB49-Ifnl3 OE tumors (online supplemental figure 2A,B,D,E). Lymphocyte exhaustion genes such as Cd274 (PD-L1) and Pdcd1 (PD-1) were also significantly upregulated at the detection time (figure 2A). Moreover, gene expressions of selected markers were confirmed by RT-qPCR in a larger group of specimens, including those used for RNA-Seq (online supplemental figure 3A). The expressions of T-cell maturation and activation makers, including Ifng, Il15, and Il18, were increased in MB49-Ifnl3 OE tumors. Furthermore, the expressions of Mφ-associated chemokines (Ccl3, Ccl4 and Ccl5) and proinflammatory genes (Cd86, Tnfsf10 and Nos2), T cell-associated chemokines (Cxcl9 and Cxcl10) were also increased in MB49-Ifnl3 OE tumors. However, the anti-inflammatory gene Arg1 expression was comparable in MB49-Ifnl3 OE and MB49-EV tumors (online supplemental figure 3B). These data indicated the effector T cell and proinflammatory Mφ phenotypes were positively associated with antitumor effects of IFN-λ3.

Supplemental material

Figure 2

Transcriptome analysis of MB49-Ifnl3 OE and MB49-EV tumors. Subcutaneous MB49-Ifnl3 OE and MB49-EV tumors were established in C57BL/6 mice. The tumors were harvested on day 23, and total RNA was extracted for RNA-Seq. (A) Volcano plot of differentially expressed genes (DEGs; threshold: log2 fold change ≥1, adjusted P<0.05) between MB49-Ifnl3 OE and MB49-EV tumors. (B) Gene set enrichment analysis (GSEA) showed the enriched pathways in MB49-Ifnl3 OE tumors. Heatmaps of DEGs involved in types I and III Interferon response (C) and lymphocyte activation (D) genes in MB49-Ifnl3 OE and MB49-EV tumors. Data were generated from one experiment.

Ectopic Ifnl3 expression led to tumor cytotoxic T and myeloid cell infiltration

We used two complementary approaches, multicolor FACS of cells harvested from the tumors, draining lymph nodes, and spleens, and IHC and mIF of cells within the tumor, to assess the tumor-infiltrating cell phenotypes in the TME. The flow cytometry gating strategy is illustrated in online supplemental figure 4. Compared with the MB49-EV group, ectopic Ifnl3 expression induced a significant increase in the proportion of tumor-infiltrating CD45+ leukocytes and NK1.1+ CD3 NK, CD3+ NK1.1 T, total CD4+ T, CD4+ T-bet+ Th1, CD4+ Foxp3+ regulatory T, and CD8+ effector T cells (Teff). However, the Teff to regulatory T cell ratio and cytotoxic or proliferative NK cells were comparable in MB49-Ifnl3 OE and MB49-EV tumors (figure 3A; online supplemental figure 5C). Among T cells, proliferation (Ki-67+) and activation (GZMB+) markers were significantly enhanced in CD4+ and CD8+ T cells in MB49-Ifnl3 OE tumors (figure 3A). Detailed analysis of the immune cells in the spleen (online supplemental figure 5A) and draining lymph nodes on the tumor side (online supplemental figure 5B) showed markedly higher CD3+ T cells, total CD4+ T cells, CD4+ T-bet+ Th1 cells, and CD8+ Teffs in mice with MB49-Ifnl3 OE tumors than in the MB49-EV mice. Intriguingly, only CD8+ T cells in the draining lymph node, but not in the spleen, of mice with MB49-Ifnl3 OE tumors showed increased Ki-67 expression and Teff to regulatory T cell ratio (online supplemental figure 5B–D), indicating that IFN-λ3 induced a systemic antitumor immune response.

Figure 3

FACS, IHC and mIF analysis of MB49-Ifnl3 OE and MB49-EV tumors. Subcutaneous MB49-Ifnl3 OE and MB49-EV tumors were established in C57BL/6 mice. The tumors were harvested on day 23, and multicolor FACS, IHC, and mIF were performed. (A) Percentages of live CD45+, NK1.1+CD3 natural killer, CD3+NK1.1 T, CD4+ T, CD8+ T, CD4+Foxp3+ Tregs, CD4+T-bet+ Th1, CD4+GZMB+ T, CD4+Ki-67+ T, CD8+GZMB+ T, CD8+Ki-67+ T cells and the ratio of tumor-infiltrating CD8+ T cells to CD4+Foxp3+ Tregs that were analyzed by multicolor FACS in MB49-Ifnl3 OE (n=8) and MB49-EV (n=7) tumors. (B, C) Infiltration of T cell subsets, NK cells, F4/80+ Mφs, and CD11c+F4/80 DCs in MB49-Ifnl3 OE and MB49-EV tumor tissue sections (n=8 each). Representative images with the positive cells stained green (CD4+, NCR1+), red (CD8+, F4/80+), and blue (Foxp3+, CD11C+), and their colocalization (examples indicated with arrows) are presented (B) and quantified (C). (D, E) Phenotype transformation detected by mIF. Proinflammatory (iNOS+) and anti-inflammatory (Arg1+) tumor-infiltrating F4/80+ Mφs evaluated in MB49-Ifnl3 OE and MB49-EV tumors (n=6 each). Representative images with the positive cells stained green (F4/80+), red (Arg1+), and blue (iNOS+) and their colocalization (examples indicated with arrows) (D) and percentages of iNOS+ and Arg1+ in total F4/80+ Mφs (E) In MB49-Ifnl3 OE and MB49-EV tumors are presented. Representative images (F) and bar graphs (G) of CXCL9+ and CXCL10+ cell densities in MB49-Ifnl3 OE and MB49-EV tumors (n=8 each) stained with IHC and assessed in 200×high-power fields. Scale bar, 100 µm. Bar graphs show means±SEMs; P values were determined by Student’s t-test when the F test outcome was insignificant; otherwise, the Mann-Whitney U test was used. Data are representative of two independent experiments. FACS, flow cytometry; IHC, immunohistochemistry; mIF, multiplex immunofluorescence.

We further assessed various immune cell markers expressed in the tumors using IHC, mIF, or the ELISA assay. We observed a 3–5 fold increase in NCR1+ NK, CD4+ T, and CD8+ T cells in MB49-Ifnl3 OE tumors compared with the MB49-EV group (figure 3B). Moreover, in myeloid cells, ectopic Ifnl3 expression resulted in a pronounced increase in F4/80+ Mφs infiltrate, and a smaller increase in CD11c+ F4/80 dendritic cell (DC) infiltrate (figure 3C), but a decrease in Ly6G+ neutrophil infiltrate (online supplemental figure 6). We questioned whether a Mφ phenotype transformation occurred in MB49-Ifnl3 OE tumors. To address this, we assessed the expression of known pro-inflammatory and anti-inflammatory Mφ-associated genes in tumor tissues from both groups. Our data revealed increased expression of pro-inflammatory genes in MB49-Ifnl3 OE tumors, including Nos2, Tnfsf10, Tnfa, Cd80, Cd86, Ccl3, Ccl4, Ccl5, Cxcl9, and Cxcl10 (online supplemental figure 3A). Surprisingly, the anti-inflammatory gene Il10 was also upregulated, while Arg1 gene remained unchanged in MB49-Ifnl3 OE compared with MB49-EV tumors (online supplemental figure 3B). Additionally, we found that the number of iNOS+F4/80+ Mφs increased while that of Arg1+F4/80+ Mφs decreased (figure 3D,E), following the increase in CXCL9+ and CXCL10+ cells in the tumor region (figure 3F,G), indicating an effective reprogramming of Mφs toward antitumor functions.

IFN-λ stimulation potentiated the proinflammatory effects of Mφ in vitro

As Mφs are heavily increased and polarized in MB49-Ifnl3 OE compared with MB49-EV tumors, we explored whether Ifnl3 coordinated pro-inflammatory responses by directly signaling to the tumor cells or Mφs in vitro. By using RT-qPCR, we observed that expressions of Viperin and Tnfsf10 were comparable in MB49-Ifnl3 OE and MB49-EV tumor cells. Although the expression of Isg15 and Cxcl10 were smaller increased in MB49-Ifnl3 OE compared with MB49-EV tumor cells (fold change<1.5; P<0.05), expressions of Cxcl9, Cxcl11 and Ccl5 were significantly downregulated in MB49-Ifnl3 OE compared with MB49-EV tumor cells (online supplemental figure 7A). Furthermore, stimulation of MB49 tumor cells with rmIFN-λ3 did not alter the expression of most above interferon-stimulated genes (ISG) or chemokines involved in T cell and myeloid cell recruitment, except for Cxcl11, which showed a slight increase following rmIFN-λ3 stimulation (online supplemental figure 7B). Notably, expression of Cd274 was downregulated by ectopic Ifnl3 expression and remained unchanged following rmIFN-λ3 stimulation in MB49 tumor cells (online supplemental figure 7). These results suggested that IFN-λ3 has limited influence on inflammatory response in MB49 tumors cells.

We next investigated the function of IFN-λ3 in Mφs using mouse BMDMs, which were isolated and stimulated with rmIFN-λ3 cytokines in vitro (online supplemental figure 8). We first found that IFNLR1 was expressed by BMDMs and their subtypes as previous described (online supplemental figure 8A).23 Different from in tumor cells, rmIFN-λ3s stimulation significantly upregulated representative ISG (Isg15, Viperin, Il12p35 and Tnfsf10), T cell recruiting chemokines (Cxcl9, Cxcl10 and Cxcl11) and myeloid cell recruiting chemokines (Ccl2, Ccl3, Ccl4 and Ccl5), but donwregulated anti-inflammatory gene (Arg1) in BMDMs (online supplemental figure 8B,C). Furthermore, CD14+ monocytes isolated from human peripheral blood mononuclear cells were differentiated into resting Mφs in vitro (online supplemental figure 9A). Using IFN-α as a positive control, we confirmed that IFN-λ1 could stimulate Mφs to acquire pro-inflammatory phenotypes (CD169, CXCLl9 and CXCL10 (online supplemental figure 9B–D). Taken together, these results indicated that ectopic Ifnl3 expression in tumors promoted T-cell infiltration and activation, an influx of myeloid cells, and antitumoral Mφ polarization.

The antitumor effect of IFN-λ3 in MB49 bladder tumors relies on T cells and Mφs

As CD8+ T cells within the TME are strongly associated with the T cell-based antitumor immune response following immunotherapy,11 we hypothesized that tumor-infiltrating CD8+ T cells play a crucial role in mediating Ifnl3-driven antitumor immunity. To test this hypothesis, we compared MB49-Ifnl3 OE and MB49-EV tumor inhibition rates in immune-deficient nude mice (lacking T and B cells) and immune-competent C57BL/6 mice. Although MB49-Ifnl3 OE tumors in C57BL/6 mice exhibited greater suppression than in nude mice (figure 4A), a significant delay in tumor growth was observed in MB49-Ifnl3 OE compared with MB49-EV tumors in nude mice (figure 4B,C), indicating a lymphocyte-independent antitumor component and function in MB49-Ifnl3 OE tumors. Moreover, there is a smaller decrease in cell proliferation (Ki-67 staining), whereas no difference in cell apoptosis (cleaved caspase-3 staining) in MB49-Ifnl3 OE tumors compared with MB49-EV tumors in nude mice (online supplemental figure 10).

Figure 4

The antitumor effect of IFN-λ3 in MB49 bladder tumors relies on T cells and Mφs. Tumors were established by subcutaneous injection of MB49-EV or transfected MB49-Ifnl3 OE cells into the flanks of C57BL6 or BALB/c nude mice, respectively. (A) Inhibition rates of MB49-Ifnl3 OE and MB49-EV tumors were analyzed in three independent experiments. Inhibition rates were calculated as follows: inhibition rate (%)=(average MB49-EV tumor volume - average MB49-Ifnl3 OE tumor volume)/average MB49-EV tumor volume×100. The rates were used to compare the inhibition efficacy of ectopic Ifnl3 expression in the C57BL/6 and BALB/c nude mice tumor models. (B) Growth kinetics of MB49-Ifnl3 OE and MB49-EV tumors (n=6 each) in BALB/c nude mice (two-way ANOVA). (C) MB49-Ifnl3 OE and MB49-EV tumor weights per BALB/c nude mouse at the endpoint. (D, E) Densities of antitumoral and protumoral Mφs in MB49-Ifnl3 OE and MB49-EV tumor sections (n=6 each) by multiplex immunofluorescence. Representative images with positive cells stained green (F4/80+), red (Arg1+), and blue (iNOS+) and their colocalization (examples indicated with arrows) are presented (D; scale bar: 100 µm). Arg1+F4/80+ and iNOS+F4/80+ Mφs in MB49-Ifnl3 OE and MB49-EV tumors were quantified and compared (E). (F–J) Four Mφ subpopulations were evaluated by flow cytometry in MB49-Ifnl3 OE and MB49-EV tumors (n=5 each). Representative images of the four Mφ subpopulations (F) and their percentages in CD11b+Ly6G myeloid cells (G–J). Bar graphs show means±SEMs; P values were determined by Student’s t-test when the F test outcome was insignificant; otherwise, the Mann-Whitney U test was used. Data are representative of at least two independent experiments. ANOVA, analysis of variance.

Consistent with immune-competent mice, we observed an increase in total F4/80+ and iNOS+F4/80+ Mφs and a decrease in Arg1+F4/80+ Mφs in MB49-Ifnl3 OE tumors (figure 4D,E). In a mouse pancreatic tumor model, Roehle et al found that the cellular inhibitor of apoptosis proteins 1 and 2 antagonism induces a phagocytic macrophage subset, identified by the coexpression of the surface markers Ly6C and the major histocompatibility complex (MHC) class II molecule IA/IE, which have the ability to phagocytosis of live tumor cells.35 Notably, the phagocytic Ly6C+Ly6G-IA/IE+ Mφ subset, which was highly enriched in MB49-Ifnl3 OE tumors (figure 4F–J). However, we observed relatively balanced infiltrate distributions of NCR1+ NK cells and CD11c+ DC cells in MB49-Ifnl3 OE and MB49-EV tumors in nude mice, while Ly6G+ neutrophils were significantly reduced at the detecting point in MB49-Ifnl3 OE tumors (online supplemental figure 11). These data suggested that Mφs play a role in the antitumor effect of IFN-λ3 in an immune-deficient MB49 bladder tumor model.

Mφs display enhanced phagocytosis of tumor cells with ectopic Ifnl3 expression

As Mφs can act as phagocytes and directly kill tumor cells, and RNA-Seq data indicated phagocytosis pathway augmentation in MB49-Ifnl3OE tumors (online supplemental figure 2E), we tested their phagocytic capacity in in vivo and in vitro experiments. We implanted nude mice with MB49-Ifnl3OE or MB49-EV tumors expressing green fluorescence protein (GFP) to measure in vivo phagocytosis. After the tumors were harvested, infiltrating Mφs were analyzed by FACS (online supplemental figure 12A). The results showed increased phagocytosis in Ly6C+Ly6GIA/IE+ phagocytic Mφs rather than CD11b+ myeloid cells in MB49-Ifnl3OE tumors compared with MB49-EV tumors (figure 5A–D). Phagocytosis was directly visualized in excised MB49-Ifnl3OE tumor tissue in situ by mIF staining using antibodies to F4/80 and GFP, which the MB49 tumor cells expressed (figure 5E,F). The fraction of tumor cell-containing Mφs was significantly higher in MB49-Ifnl3OE than in MB49-EV tumors. To better model the effects of ectopic Ifnl3 expression in MB49 tumor cells on Mφ phagocytosis, a coculture of BMDMs and viable tumor cells was developed to conduct the in vitro phagocytosis assay. FACS detection showed more MB49-Ifnl3OE than MB49-EV cells were phagocytized by the BMDMs (online supplemental fifure 12B; figure 5G,H). These findings demonstrated that ectopic Ifnl3 expression in MB49 tumors enhanced tumor cell uptake by mononuclear phagocytes.

Figure 5

Mφs display enhanced phagocytosis of tumor cells with ectopic Ifnl3 expression. (A, B) Green fluorescence protein (GFP)-positive MB49-EV or transfected MB49-Ifnl3 OE cells were injected subcutaneously into the flanks of BALB/c nude mice. Mφ phagocytosis of tumor cells (GFP+) was evaluated by flow cytometry on day 23. Representative plots showing tumor+ CD11b+ myeloid cells (A) And proportions of GFP+ tumor cells in CD11b+ myeloid cells in MB49-Ifnl3 OE and MB49-EV tumors (B; n=5 each). (C, D) Phagocytosis by phagocytic Ly6C+Ly6GIA/IE+ Mφs. Representative plots showing tumor+ Ly6C+Ly6GIA/IE+ Mφs (C) And proportions of GFP+ tumor cells in Ly6C+Ly6GIA/IE+ Mφ populations in MB49-Ifnl3 OE and MB49-EV tumors (D; n=5 each). (E,F) Phagocytosis assessment by multiplex immunofluorescence in MB49-Ifnl3 OE and MB49-EV tumors (n=5 each). Representative images of GFP+ F4/80+ Mφs (E) and proportions of GFP+ tumor cells in F4/80+ Mφ populations (F). Data are representative of two independent experiments. (G,H) In vitro phagocytosis assays were assessed by flow cytometry after bone marrow-derived Mφs (BMDMs) were cocultured with live MB49-Ifnl3 OE and MB49-EV tumors (n=8 each). Representative images of GFP+F4/80+ Mφs (G) and a bar graph (H) are presented. Each dot represents a biological replicate. Bar graphs show means ± SEMs; P values were determined by Student’s t-test when the F test outcome was insignificant; otherwise, the Mann-Whitney U test was used.

PD-1/PD-L1 axis blockade leads to better antitumor effects in MB49-Ifnl3OE tumors

As IFNs are known to contribute to the quality of antitumor immunity and response to immunotherapy,5 we evaluated the relationship between ectopic Ifnl3 expression and the PD-1/PD-L1 axis in a mouse model. Using IHC staining, we observed that the MB49-Ifnl3 OE tumors had a higher concentration of infiltrating PD-1+ cells than MB49-EV tumors (online supplemental figure 13). We next investigated whether blockade of the PD-1/PD-L1 axis could improve IFN- λ antitumor efficacy in MB49 tumors. As shown in figure 6A, MB49-Ifnl3 OE and MB49-EV tumors were generated in C57BL/6 mice. PD-L1-blocking and IgG control antibodies were injected intraperitoneally, once every three days, at a dose of 100 µg per mouse. As expected, we detected a significant delay in tumor in mice with MB49-Ifnl3 OE tumors. While no benefit was detected in MB49-EV tumors treated with anti-PD-L1 antibodies, a further reduction in tumor growth and prolonged survival was observed in MB49-Ifnl3 OE tumors treated with anti-PD-L1 antibodies (figure 6B–E). These findings indicated that blocking the PD-1/PD-L1 axis led to a higher IFN-λ3 antitumor efficacy in MB49 tumors.

Figure 6

The antitumor effects in MB49-Ifnl3 OE tumors are better than in MB49-EV tumors after PD-1/PD-L1 axis blockade. Subcutaneous MB49-Ifnl3 OE and MB49-EV tumors were established in C57BL/6 mice. For the PD-1/PD-L1 pathway blockade experiment, the mice were injected intraperitoneally with 100 µg/dose of anti-PD-L1 antibodies or IgG every 3 days from day 7 after tumor cell injection. At day 7 after tumor injection, mice with similar tumor volumes were randomly divided into four groups: MB49-EV+IgG, MB49-EV+anti-PD-L1, MB49-Ifnl3 OE+IgG or MB49-Ifnl3 OE+anti-PD-L1 (n=5–7 per group). (A) Schematic diagram of the PD-L1 blockade experiment. Graphs depicting tumor growth in each mouse (B) and the combined survival curves (C) of the mice. P values were calculated using the log-rank test. Data are representative of at least two independent repeats. PD-1, programmed cell death protein 1; PD-L1, programmed cell death ligand 1.

IFN-λ is associated with innate and adaptive immune responses in bladder cancer

The antitumor effect of ectopic Ifnl3 expression in MB49 tumors, which depends on T and myeloid cells, led us to investigate the relationship between IFN-λ and the innate and adaptive immune response in human bladder cancer. Using data from TCGA-BLCA, we demonstrated that expression of the IFN-λ genes (IFNL1, IFNL2, and IFNL3) were upregulated in tumor tissues compared with adjacent normal tissues (online supplemental figure 14A). Using the MCP counter algorithm, we defined a human IFNL signature comprised all of IFN-λ1, IFN-λ2, and IFN-λ3. Based on the median level of the IFNL signature, we divided the 412 patients into high and low IFNL signature groups. We found that the expression level of IFNL signature was positively with IFNG and slightly positively associated with IFNB, but not with IFNA on RNA levels in TCGA-BLCA dataset (online supplemental figure 14B). Moreover, we identified the TIICs from previous study and calculated their infiltration level using the MCP-counter algorithm in the TCGA-BLCA dataset.36 Further analysis revealed that the IFNL signature was positively associated with the infiltration levels of CD8+ T cells, Th1 cells, NK cells, Mφs, and DCs (figure 7A). Gene set enrichment analysis showed that response to type I/III interferon, positive regulation of leukocyte activation and phagocytosis pathways was enriched in the high-IFNL signature group (figure 7B). These results were validated by IHC and mIF staining in a human bladder cancer cohort (n=15). IHC staining showed IFN-λ3 is expressed on tumor and stromal cells (figure 7C). mIF staining of serial sections was used to evaluate TIIC infiltration, including CD8+ T, CD4+ T, NCR1+ NK, and HLA-DR+ antigen-presenting cells. Consistently, tumors in which IFN-λ3 was detected appeared to have higher TIIC infiltration rates than tumors in which IFN-λ3 was not detected (figure 7D,E). These results were confirmed in a cohort of patients with bladder cancer receiving anti-PD-1 neoadjuvant immunotherapy (n=20). IHC staining data showed that IFN-λ3 detection was higher in responders (80%, 8/10) than in non-responders (40%, 4/10) in this cohort (p<0.001; figure 7F). Furthermore, we evaluated the IFNL signature and anti-PD-L1 immunotherapy efficacy in the IMvigor210 cohort.37 The IFNL signature was positively associated with the immunotherapy efficacy (P=0.0342; online supplemental figure 15). Taken together, these results indicated that IFN-λ is associated with innate and adaptive immune responses and the efficacy of immunotherapy in human bladder cancer.

Figure 7

IFN-λ is associated with innate and adaptive immune responses in bladder cancer. High (top 50%) and low (bottom 50%) IFNL signature scores, determined using microenvironment cell populations (MCP)-counter method, were used for correlation analysis with effector genes of tumor infiltrating immune cells and gene set enrichment analysis (GSEA) in the TCGA-BLCA dataset. (A) Heatmap of effector gene sets of the various immune cells determined by MCP-counter method in TCGA dataset. (B) GSEA analysis of positive regulation of response to type I/III interferon (top), leukocyte activation (middle) and phagocytosis (down) in a bladder cancer cohort from TCGA. (C–E) Clinical bladder cancer samples were divided into IFN-λ3high (n=7) and IFN-λ3low (n=8) groups. IFN-λ3high was defined as ≥5% IFN-λ3+ cells in human bladder cancer tissues using IHC staining. Representative images of IFN-λ3high and IFN-λ3low cells with IHC staining (C). Positive cells stained cyan (NCR1+), green (CD4+), red (CD8+), and blue (HLA-DR+) by multiplex immunofluorescence are presented (D) and quantified (E). (F) IFN-λ3 expression and anti-PD-1 immunotherapy efficacy in the neoadjuvant immunochemotherapy cohort (IFN-λ3high: n=10 and IFN-λ3low: n=10). Bar graphs show means±SEMs; P values were determined by Student’s t-test when the F test outcome was insignificant; otherwise, the Mann-Whitney U test was used. IHC, immunohistochemistry PD-1, programmed cell death protein 1 TCGA, The Cancer Genome Atlas.

Discussion

Our study highlighted the potential of type III interferon to induce Mφ-mediated immune responses against MB49 bladder tumor progression. In immune-proficient tumors, ectopic Ifnl3 expression in tumor cells significantly increased the infiltration of cytotoxic CD8+ T cells, Th1 cells, NK cells, Mφs, and DCs. Transcriptome analysis revealed significant upregulation of many genes associated with an effective immune response, including lymphocyte recruitment, activation, and phagocytosis. These findings were consistent with the increased antitumor immune infiltrates and tumor inhibition. Increased total Mφs and phagocytic Mφ subsets in T cell-deficient MB49-Ifnl3 OE tumors enhanced tumor cell phagocytosis. As immunosuppressive regulatory T cells and immune checkpoint molecules also increase in the TME, PD-L1 blockade therapy proved more efficient in MB49-Ifnl3 OE than in MB49-EV tumors. Furthermore, IFN-λ expression was positively associated with effector immune infiltrates and the efficacy of immune checkpoint blockade (ICB) therapy in patients with UCB. These results indicated that IFN-λ enables T cell-mediated and non-T cell-mediated antitumor immunity and could serve as a candidate target for bladder cancer immunotherapy.

Type III interferon (IFN-λ) can influence tumor progression by directly acting on tumor cells or indirectly regulating the immune microenvironment.19 In a preclinical study for melanoma, IFN-λ showed significant benefits in which the antitumor effect was associated with cell cycle arrest and apoptosis.13 In contrast, IFN-λ induced matrix metalloproteinase 9 expression and promoted tumor migration and invasiveness in a bladder cancer model.38 The differential effects of IFN-λ on tumor cells may be attributed to tumor heterogeneity. Moreover, above studies did not assess the regulation of immune microenvironment components by IFN-λ. During infection with influenza A virus, IFN-λ signaling can modulates DC IL-10 immunoregulatory network, and activates ISG expression to facilitate CD8+ T cell-mediated protective immunity.18 Besides its potential function in combating viruses, there is also a role for IFN-λ in displaying antitumor effects across various murine models.13 20 21 In the current study, ectopic Ifnl3 expression had no effect on tumor cell proliferation and limited effect on inflammatory cytokine production in vitro, but MB49-Ifnl3 OE tumors showed heavy immune cell infiltration, increased cell death and grew significantly slower than MB49-EV tumors. Further investigation indicated that types I and III interferon responses, chemokine signaling pathway and lymphocyte activation were upregulated in MB49-Ifnl3 OE tumors. These results showed that the anti-tumoral efficacy of IFN-λ was at least partly due to enhancing the infiltration, proliferation, and effector function of tumor reactive T cells and NK cells within the TME. Moreover, IFN-λ seems to have immunomodulatory functions, because CD4+Foxp3+ regulatory T cells and anti-inflammatory gene Il10 were also increased by IFN-λ upregulation in MB49-Ifnl3 OE tumors. Mennechet et al also demonstrated that IFN-λ-treated DCs specifically induced proliferation of FOXP3-expressing suppressor T cells in mixed lymphocyte reaction experiments.39 Therefore, further investigation is required to fully understand the impact of IFN-λ on driving T/NK cell-mediated anti-tumor immunity.

IFN-λ activity is limited to specific tissues because of selective expression of IFNLR1.15 Human DCs, neutrophils and Mφs are highly responsive to IFN-λ,18 24 26 whereas mouse Mφs has produced conflicting results.23 24 In a fibrosarcoma cell tumor model, Numasaki et al confirmed that neutrophils were involved in the antitumor role of IFN-λ. In contrast, our study found that while neutrophils decreased, Mφs highly infiltrated the MB49-Ifnl3 OE tumors and repolarized to antitumor phenotypes, even in T cell-deficient nude mice. This discrepancy is probably due to differences between tumor cell lines and different tumor harvest time. In our study, both human and mouse IFN-λs can stimulate Mφs expressed high levels of ISG, Mφ/T cell-associated chemokine and activating molecules in vitro. Consistent with our findings, Cheng et al also found that IFNLR1 was expressed by mouse BMDMs and their subtypes, and IFN-λ inhibited anti-inflammatory Mφ polarization via inhibiton of STAT3 and JNK signaling pathways in vitro.23 Mallampalli et al also found that IFNLR1 expression and IFN-λ signaling was induced by the differentiation of human myeloid cells to Mφ,24 whereas they did not observe ISG induction in differentiated BMDMs by mouse IFN-λ2 and IFN-λ3 stimulation (R&D systems, 4635 ML and 1789 ML, respectively).24 The conflicting results may be due to different commercial IFN-λ cytokines used in different studies. Different from IFN-λ3 cytokines used in Mallampalli et al study (R&D systems), mouse IFN-λ3 was purchased from Sino Biological (China) in our study and the study by Cheng et al.23 Moreover, we used another commercial IFN-λ3 cytokines from Novus Biologicals (USA) to stimulate BMDMs, which showed a high degree of consistency with IFN-λ3 cytokine from Sino Biological (China). As such, analyzing the active structure of IFN-λ is required to unequivocally define its stimulatory effect on both human and mouse Mφs.

Mφs are multifunctional innate immune cells comprizing a continuum of phenotypes in TME, spanning from antitumor ‘defenders’ to protumor ‘remodelers’.40 We discovered that IFN-λ could redirect Mφ ‘programming’ to the antitumor state in the MB49 model. Consistent with our findings, Read et al have shown that IFN-λ stimulates cytotoxicity and phagocytosis in human-derived Mφs and the secretion of proinflammatory cytokines in vitro.25 In a mouse pancreatic tumor model, by using single-cell transcriptional proofing of total CD45+ cells from vehicle and a cellular inhibitor of apoptosis protein treated tumors, Roehle et al identified a putative phagocytic Mφs with both Ly6C and MHC class II positive phenotypes and confirmed this Mφ subset has highly capacity to phagocytosis live tumor cells.35 Specifically, we also found that phagocytic Mφs coexpressing MHC class II and Ly6C significantly increase and augment phagocytosis of tumor cells in MB49-Ifnl3 OE tumors (average 26% vs 9.5%, respectively). Although phagocytosis rates are lower compared with those in vivo experiments, we still showed about two times more MB49-Ifnl3OE than MB49-EV cells were phagocytized by the mouse BMDMs in vitro (average 8.5% vs 4.6%, respectively). The differences between in vitro and in vivo experiments may be due to the simplicity of coculture experiments in vitro, while in vivo models involve other factors such as immune responses mediated by NK cells and/or neutrophils, which can further enhance Mφ phagocytic activity against tumor cells.41 42 Previous studies have shown that antagonism between Toll-like receptor 9 agonist and cellular inhibitor of apoptosis protein can enhance phagocytosis by reprogramming Ly6C+ Mφs toward a tumor-destructive phenotype.35 43 Moreover, our transcriptome analyses revealed that phagocytosis-associated molecules such as SLAMF7 and complement 3 were significantly upregulated in MB49-Ifnl3 OE tumors. SLAMF7 associates with macrophage-1 antigen, composed of integrins CD11b and CD18, and forms a protein complex on the Mφ cell surface.44 This complex interacts with two immunoreceptor tyrosine-based activation motif-containing receptors on the Mφ cell surface, FcRγ and DAP12, eliciting signaling via Src kinase, spleen tyrosine kinase, and Btk kinase to activate the phagocytic machinery.44 Additionally, macrophage-1 antigen is essential in inducing phagocytosis of complement fragment C3bi-opsonized pathogens and apoptotic cells.45 However, the mechanisms underlying IFN-λ-induced phagocytic repolarization of Mφs require further investigation.

Antitumor immune responses are frequently accompanied by immune suppressive mechanisms, which can be exploited by tumors cells to evade immune surveillance.46 We observed that the PD-1/PD-L1 axis was highly activated in MB49-Ifnl3 OE tumors and that suboptimal PD-L1 blockade had a therapeutic effect on MB49-Ifnl3 OE but not MB49-EV tumors. This finding suggests that IFN-λ upregulation can increase sensitivity to ICB therapy. Previous understanding of the IFN-λ system mostly came from mice and human cell line studies.19 25 In human breast cancer, IFN-λ plays a key role in inducing IL-12p70, IFN-γ, CXCR3 ligands, CX3CL1, cytokines, and chemokines involved in NK and Teff cell recruitment and activation.36 It was revealed that IFNL1 or IFNLR1 gene expression was associated with favorable patient outcomes.36 Our study of patients with UCB found that IFN-λ is present in tumor tissues, positively associated with tumor-infiltrating immune cells and can predict ICB treatment efficacy. Similar results were confirmed in the TCGA-BLCA and IMvigor210 datasets. As only approximately 20% of patients with UCB show an effective response to anti-PD-1 or PD-L1 monotherapy, a combination of anti-PD-1 or PD-L1 therapy and other drugs is the key approach to overcoming ICB resistance.47 48 Unfortunately, most combination therapy clinical trials failed for the lack of rational mechanisms.49 This study provides a rationale for combining IFN-λ with anti-PD-L1 drugs for UCB treatment. Pegylated IFN-λ medication has been used in clinical trials to treat patients with hepatitis C virus50 and to prevent clinical events among outpatients with acute symptomatic COVID-19.51 However, as we known, pegylated IFN-λ was not used for the treatment of human cancers. TLR angonist can stimulate the expression of IFN-λ.52 The rapid development and clinical deployment of COVID-19 vaccines worldwide has highlighted the potential of mRNA-based technologies as useful tools for cancer treatment.53 These potential methodologies to induce IFN-λ production could help develop new therapeutic strategies.

Conclusions

To our knowledge, this was the first study to demonstrate the expression of IFN-λ in UCB tissues and evaluate the relationship between its expression and tumor-infiltrating immune cells and immunotherapeutic efficacy. Furthermore, our study indicated that IFN-λ upregulation in MB49 tumors reprograms Mφs to an antitumor state, enhances phagocytosis and inflammatory cytokines secretion, and activates adaptive immunity to inhibit tumor progression. Since IFN-λ can promote phagocyte reprogramming in T-cell-deficient mice, we propose that IFN-λ-mediated antitumor immunity is a potential mechanism for controlling ICB refractory tumors.

Supplemental material

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and all human samples were anonymously coded following the local ethical guidelines as stipulated by the Declaration of Helsinki. Written informed consent was obtained from all patients, and the protocol was approved by the Ethics Committee of Sun Yat-sen Memorial Hospital (SYSKY-2023-716-01). Participants gave informed consent to participate in the study before taking part.

References

Supplementary materials

Footnotes

  • BW, BZ, JC and XS contributed equally.

  • Contributors BW, BZ and JC designed and performed the experiments, analyzed the data, and wrote the manuscript; XS, WY, TY, HY, PC, KC, XH, XF and WH contributed to the experimental design and data analysis; JH and TL supervised the research and provided critical feedback on the experimental design, data analysis, and manuscript writing. TL as the guarantor for the overall content, approved and supervised the project. All authors have read and approved the manuscript and agree with their inclusion as a coauthor.

  • Funding This study was supported by the National Key Research and Development Program of China (grant No. 2018YFA0902803); the National Natural Science Foundation of China (grant No. 82000198, 82002682, 81825016, 81961128027, 81772719 and 81772728); Guangdong Provincial Clinical Research Center for Urological Diseases (2020B1111170006); Elite Young Scholars Development Program of Sun Yat-Sen Memorial Hospital to BW; Guangdong Medical Science and Technology Research Fund Project (A2021107).

  • Competing interests No, there are no competing interests.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.