Introduction

Immune checkpoint inhibitors (ICIs) boost anti-tumor immune response by mitigating the self-tolerance mechanism of the immune cells, which is hijacked by tumor cells. The clinical application of ICI has profoundly improved prognosis and life expectancy in metastatic cancer patients suffering from melanoma, non-small-cell lung, and kidney cancer, representing a paradigm shift in cancer therapy1,2,3. Immunotherapy is based on the role of cell-surface receptors and ligands accessory to the T-cell receptor in inhibiting cell-mediated immune response. The first monoclonal antibodies targeting this inhibitory axis were developed against the immune checkpoint PD-1 (programmed death-1), its ligand PD-L1 (programmed death ligand-1) and CTLA-4 (cytotoxic T lymphocyte antigen-4). They were first tested in the treatment of melanoma and then applied to the treatment of other cancers characterized by poor prognosis4,5,6,7,8,9,10,11,12. Although ICIs therapy has improved the survival of many cancer patients, the percentage of patients responding remains low. In order to improve efficacy and patient response rates, new therapeutic strategies combining ICIs with adjuvants that augment immune-dependent attack of cancer cells are needed. For example, targeting alternative pathways such as the co-stimulatory molecules OX40, 4-1BB, glucocorticoid-induced TNFR-related protein (GITR) has proven to enhance T-cell mediated immunity in preclinical model13,14,15,16,17, although no clinical studies have confirmed the efficacy of such treatments in humans.

However, ICIs also causes side effects which are uniquely associated with an increase in autoimmunity due to alteration of self-tolerance. A recent study found that 3.5% of patients initiating ICI experience adverse events requiring hospitalization and immunosuppression18,19. Immune-related adverse events (IRAEs) can affect colon, lungs, liver, skin, pituitary, thyroid, and heart20. Although cardiotoxicity accounts for <1% of IRAEs, the onset of such complications, such as myocarditis, arrhythmia, pericarditis and vasculitis results in death in 50% of cases21.

Cardiovascular immune-related adverse events include myocarditis, pericardial disease, vasculitis, Takotsubo syndrome, destabilization of atherosclerotic lesions, venous thromboembolism, and conduction abnormalities22. ICI-associated myocarditis results from inflammation of the conduction system due to infiltration of T cells and macrophages23. Under physiological conditions, the cardiac lymphocyte infiltrate is limited and the macrophages and dendritic cells resident in the heart control the expression of the immune checkpoint proteins in order to maintain homeostasis. Thus, inhibiting the immune checkpoint pathway could lead to adverse outcomes, by promoting the recruitment of lymphocytes and macrophages and triggering an inflammatory response24.

Indeed, in genetically modified mice, deletion of CTLA4 leads to massive lymphoproliferative disease and diffuse lymphocyte infiltration in almost all organs, including the heart. In contrast the deletion of Pdcd1 (encoding PD-1) in Balb/c mice causes cardiomyopathy due to the development of autoantibodies against troponin I25. On the other hand, activation of OX40 with agonistic antibodies stimulates the release of proinflammatory cytokines (IL6, TNFα, IFNγ) by activated T lymphocytes and antigen-presenting cells (APCs)26,27 thus causing a systemic inflammatory response syndrome26,64,65, this condition is associated with systemic inflammation and cytokine storm66 that determines interstitial cardiac infiltration of macrophages or multifocal lymphocytes67. A key factor for the cytokine storm is Interleukin 6 (IL-6) which causes the activation of the T lymphocyte and the release of inflammatory cytokines, generating a positive feedback of immune activation that leads to myocardial damage68.

This study sets the stage for clinical trials aimed at assessing the ability of FMD to increase the efficacy of immunotherapy while reducing its side effects. These results also indicate that the anti-inflammatory and protective effects of FMD cycles in combination with ICI could affect other organs and systems.

Methods

Our studies were carried out in compliance with all relevant ethical regulations. The in vivo experiments on animals were performed according to the protocols approved by the IFOM Institutional Animal Care and Use Committee (IACUC) and the Ministry of Health.

Tumor cell lines

The B16-F10 melanoma tumor cell and LLC1 cell were obtained from ATCC. Cells were grown in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, penicillin (100 U ml−1) and streptomycin (100 μg ml−1) at 37 °C with 5% CO2 and maintained at a confluence of 70–80%.

Tumor implantation, Immune checkpoint blockade (ICB) treatment and tumor volume measurement

C57BL/6J female mice, 6–8 weeks old, were purchased from Charles River and housed under pathogen-free conditions at 22 ± 2 °C with 55 ± 10% relative humidity and with 12 h day/light cycles in Cogentech animal facility and with food and water ad libitum. All procedures were carried out in accordance with the guidelines established in the Principles of Laboratory Animal Care (directive 86/609/EEC), were approved by the Italian Ministry of Health, and were performed under the supervision of Institutional Animal Care and Use Committee (OPBA) at IFOM- The AIRC Institute of Molecular Oncology.

In the vivo experiments, tumors were implanted in C57BL/6J mice by injecting subcutaneously (s.c.) 2 × 105 B16-F10 or 5 × 105 LLC1 cells per mouse into the right flank at day 0. Three days after tumor injection, mice from the appropriate groups (at least 5 mice per group) were treated intraperitoneally (i.p.) with anti-PD-L1 (at the dose of 100μg per mouse), anti-OX40 (at dose of 100 μg per mouse), anti-PD1 (at the dose of 100μg per mouse), anti-CTLA4 (at dose of 100 μg per mouse). The ICB therapy was administered every other day for three treatments. The combined anti-OX40/anti-PD-L1 treatment was administered sequentially. The mice were treated with anti-OX40 the first week, while the second week with anti-PD-L1. Anti-PD-1/anti-CTLA4 were administered concurrently on 4, 6 and 8 post-injection day. The mice underwent one or two cycles of FMD (4 days each week) starting the third day after tumor implantation and sacrificed on 21 post-injection day.

One FMD cycle consists of alternating four consecutive days of fasting mimicking diet and three days of refeeding with standard diet. FMD components are described in Brandhorst et al. and Di Biase et al.38,40. Briefly the day 1 diet provides 1.88 kcal/g (50% of normal daily intake) and is made by a mix of various low-calorie broth powders, a vegetable medley powder, extra virgin olive oil, and essential fatty acids mixed with hydrogel; day 2-4 diet contains 0.36 kcal/g (10% of normal day intake) and consist of low-calorie broth powders and glycerol mixed with hydrogel.

Tumors were measured every 3–4 days using a digital caliper; tumor volume was calculated using the formula V = (L × W × H)/2, where V is tumor volume, L is the length of the tumor

(longer diameter), W is the width of the tumor (shorter diameter) and H is the height (diameter of tumor perpendicular to length and width). Mice were monitored for tumor growth and survival. Mice were killed when tumor volume reached 1.5 cm3.

Flow cytometry analysis of tumor-infiltrating lymphocytes and apoptosis

For the flow cytometry analysis of tumor-infiltrating lymphocytes, tumors were minced, B16 were digested for 1 h with Collagenase D (10 mg/ml) and DNAseI (10 mg/ml). Processed tumors were load on Lympholyte gradient and centrifuged at 1500 × g for 30 min. The interphase ring, which contains most live leukocytes cells, was collected and used for FACS staining.

1–2 × 106 cells per sample were stained with the LIVE/DEAD stain (Invitrogen), and then with membrane protein marker (CD45, CD3, CD8, CD4, CD44) followed by fixation with formaldehyde. For intranuclear (Foxp3, Tbet, Klrg1, Eomes, Tcf1/7) and cytoplasmic marker (GzmB) staining, cells were permeabilized and fixed with Foxp3/transcription factor staining kit (Invitrogen eBioscence) or BD cytofix/cyto perm kit (BD bioscences). Data acquisition was performed on Attune NxT Flow Cytometer. Results were analyzed with the FlowJo software. Gating strategies for myeloid and lymphoid cells are illustrated in Supplementary Fig 8.

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Antibodies FACS

Anti-mouse/human CD44, PB (IM7)

Biolegend

103020

Anti- muse FoxP3, eFluor 506 (FJK-16s)

eBioscience

69-5773-82

Anti-mouse CD274 (PD-L1, B7-H1), Biotin (1-111 A)

eBioscience

13-9971-81

Anti-mouse Eomes, PE (Dan11mag)

eBioscience

12-4875-82

Anti-mouse CD127 (IL-7Rα), PE (A7R34)

Biolegend

135010

Anti-mouse CD8a, PerCP-Vio700 (53-6.7)

Miltenyi Biotec

130-120-756

anti-mouse CD185 (CXCR5), PE/Cy7 (L138D7)

Biolegend

145516

Anti-mouse CD335 (NKp46), PE-eFluor 610 (29A1.4)

eBioscience

61-3351-82

Anti-mouse CD366 (Tim-3), PE/Dazzle 594 (B8.2C12)

Biolegend

134013

Anti-mouse T-bet, AF 647 (4B10)

Biolegend

644804

Anti-mouse/human KLRG1, APC (2F1/KLRG1)

Biolegend

138412

Anti-mouse Ly108, APC (330-AJ)

Biolegend

134610

Anti-mouse CD3, AF 700 (17A2)

eBioscience

56-0032-82

Anti-mouse MRC1, AF 700 (MR6F3)

eBioscience

56-2061-82

Anti-mouse CD25, APC/Cy7 (3C7)

Biolegend

101918

Anti-mouse PD1, Biotin (RMP1-30)

Biolegend

109106

Anti-mouse CD11c, APC-Vio770 (N418)

Miltenyi Biotec

130-107-461

Anti- mouse TCF1/TCF7, AF 488 (C63D9)

Cell Signaling Technology

BK6444S

Anti-mouse CD11b, VioBlue (REA592)

Miltenyi Biotec

130-113-810

Anti-mouse CD4, VioBright FITC (REA604)

Miltenyi Biotec

130-118-692

Anti-mouse Granzyme B, FITC (REA226)

Miltenyi Biotec

130-118-341

Anti-mouse Ly-6C, FITC (REA796)

Miltenyi Biotec

130-111-915

Anti-mouse MHC Class II, PE (REA813)

Miltenyi Biotec

130-112-231

Anti-mouse CD62L, PE (MEL-14)

Biolegend

104407

Anti-mouse CTLA-4, BV421 (UC10-4B9)

Biolegend

106311

Anti-mouse Ly-6G, PerCP-Vio700 (REA526)

Miltenyi Biotec

130-117-500

Anti-mouse CD45, PE-Vio770 (REA737)

Miltenyi Biotec

130-110-661

Anti-mouse F4/80, APC (REA126)

Miltenyi Biotec

130-116-525

Antibodies in vivo

anti-mouse OX40 (CD134)

BioXcell

BE0031

anti-mouse PD-L1 (B7-H1)

BioXcell

BE0101

anti-mouse CTLA4 (CD152)

BioXcell

BP0032

anti-mouse PD-1

BioXcell

BE0146

Antibodies IHC

Anti-CD3

Abcam

Ab16669

Anti- CD8α

Abcam

Ab217344

Anti-Myeloperoxidase

Abcam

Ab208670

Anti-B220

Bd bioscience

553928

Collagen quantification in myocardial tissues

For ex vivo analyses, hearts were excised and fixed in 10% neutral buffered solution. The myocardial tissues were formalin-fixed and paraffin-embedded for morphometry and immunohistochemistry. General morphology was studied using hematoxylin-eosin staining. Sections were stained with hematoxylin/eosin to evaluate the collagen content in tissue. In order to quantify the total collagen content in the heart tissues, measurement of pro-collagen 1α1 (an established biomarker of cardiac fibrosis) was performed using the Mouse Pro-Collagen I alpha 1 CatchPoint SimpleStep ELISA Kit from Abcam (ab229425). Tissues were homogenized, after protein quantification (Bradford assay) 100 µg of proteins were assayed according to manufacturer’s instruction. Fluorescence was measured at 530/590 using a 96-well plate reader Tecan Infinite M200 plate-reader (Tecan Life Sciences Home, Männedorf, Switzerland). Metalloproteases type 9 (MMP-9) were associated with collagen maturation in heart failure, demonstrating the important role of these enzymes in fibrosis through collagen configuration, activation, and deposition. Therefore, we quantified MMP-9 in heart tissues through Mouse MMP9 ELISA Kit (ab253227, Abcam, Milan, Italy) following the manufacturer’s instructions; results are expressed as pg of MMP-9/mg of protein (determined through Bradford assay).

Immunohistochemistry

Mouse tissues were fixed in 10% buffered formalin and paraffin embedded with Diapath automatic processor. To assess histological features Hematoxylin/Eosin (Diapath) staining was performed according to standard protocol and samples were mounted in Eukitt (Bio-Optica).

For immunohistochemical analysis, paraffin was removed with xylene and the sections were rehydrated in graded alcohol. Antigen retrieval was carried out using preheated target retrieval solution (pH 9.0) for 30 minutes. Tissue sections were blocked with FBS serum in PBS for 90 min and incubated overnight with primary antibodies.

For melanoma tumor antibody binding was detected using the conjugated goat anti-rabbit polymer alkaline phosphatase (AP) (Biocare) followed by a Vulcan red chromogen reaction (Peroxidase substrate kit, DAB, SK-4100; Vector Lab).

For lung tumor antibody binding was detected using a polymer detection kit (GAR-HRP, Microtech) followed by a diaminobenzidine chromogen reaction (Peroxidase substrate kit, DAB, SK-4100; Vector Lab). All sections were counterstained with Mayer’s hematoxylin and visualized using a bright-field microscope (LEICA DM750)

For B220 antibody binding was detected using a ABC kit Vectastain (vector, Pk4100) followed by a diaminobenzidine or Magenta chromogen (ENvision flex hrp magenta substrate chromogen system,dako) reaction.

Quantification of CD3+ and CD8+ cells in myocardial tissues

To evaluate for tissue abundance of CD3+ and CD8+ T-cells and diagnosing any myocarditis induced by immunotherapy, formalin-fixed paraffin-embedded (FFPE) sections of heart (4 μm thickness) were stained with rabbit anti mouse CD3 and CD8 antibodies, diluted 1:150 and 1:300 respectively in Dako AR9352 diluent buffer, upon antigen retrieval with 10 mM citrate buffer pH 6 (CD3) or with EDTA antigen retrieval Buffer pH 8 (CD8). Goat anti-rabbit antibodies were used as secondary antibodies for CD3 and CD8.

Visualization of the antibody–antigen reaction was visualized by peroxidase complex kit reagents (SignalStain® DAB Substrate Kit) as the chromogenic substrate. Finally, sections were weakly counterstained with hematoxylin and mounted. Lymphocyte densities (cells/mm2) were quantified by an experienced pathologist (FT).

Quantification of plasma and myocardial cytokines

Cytokines involved in inflammation (IL-1α, IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12, IL17-α, IFN-γ, TNF-α, G-CSF, and GM-CSF) were quantified in heart tissue extracts by using the 12 mouse cytokine Multiplex Assay Kit (Qiagen, Germantown, USA) following the manufacturer’s instructions and summarized in other recent work63. Results were expressed as pg of cytokine/mg of heart tissue.

Quantification of pro-inflammatory biomarkers

After treatments, mice were sacrificed after the proper anesthesia as described before. Hearts were weighed and snap-frozen in dry ice; after, heart tissues were homogenized in a solution 0.1 M PBS (pH 7.4) containing 1% Triton X-100 and protease inhibitor cocktail. Tissues were well homogenized through a step in a high intensity ultrasonic liquid processor. Obtained homogenates were than centrifuged at 4 °C and supernatants were treated for quantification of several inflammation markers. Specifically, leukotriene B4 expression (pg/mL of tissue lysate) was quantified through the LTB4 ELISA Kit (Enzo, Life Technology). Nuclear factor NF-kappa-B p65 subunit (p65-NF-Κb) expression (ng/mL of tissue extract) was quantified through a mouse, rat RelA/NF-kB p65 ELISA Kit (My BioSource, Seattle, WA, USA). NLRP3 inflammasome expression (ng/mL of tissue extract) was quantified by the NLRP3 ELISA Kit (Mouse) (OKEH05486) (Aviva Systems Biology); quantification of systemic and myocardial hydrogen peroxide, a reactive oxygen species (ROS) marker, was performed by using a Fluorimetric Hydrogen Peroxide Assay Kit (Sigma Aldrich, Milan, Italy), in line with other recent work49,69. This kit utilizes a peroxidase substrate that generates a red fluorescent product (λex = 540/λem = 590 nm) after reaction with hydrogen peroxide that was analyzed by a fluorescent microplate reader.

Statistical analysis

GraphPad Prism 9 was used for statistical analysis and graphing. Differences between three or more groups were analyzed using one-way ANOVA followed by post hoc analysis through Tukey’s multiple comparison test. Comparison of CD3 and CD8 immune cell count in heart sections were performed using the Kruskal–Wallis test with post hoc Mann–Whitney U-test. Statistical tests are indicated in the respective sections and figure captions. Differences were considered statistically significant at p value (p) < 0.05. All data are represented as mean ± SEM of three independent experiments.

For in vivo experiments, sample size estimation was performed with G.Power software using a multifactorial variance analysis (ANOVA) repeated measurement, within-between interaction.

A power analysis with significance level α = 0.05, (assuming a large effect size f = 0.4) indicated that n = 8 mice per group are needed to achieve a power (1 – β)=0.9 which is considered adequate to detect difference between means at least of two groups.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.