The experiments utilized sex- and age-matched inbred C57BL/6 mice from Harlan Laboratories, Germany, which were kept in a pathogen-free environment. The mice were at least 7 weeks old at the beginning of the experiments. Mice were housed under 12:12 light cycle in a relative humidity of 55 ± 10 and a temperature of 22 ± 2°C. Mice were fed ad libitum with control diet (Ssniff, E15510-04, 196 mg/kg of iron) and iron-deficient chow (Ssniff, E15510-24, <10 mg iron/kg) for four weeks followed by FV infection.

The FV complex, comprising of B-tropic Friend murine leukemia helper virus and polycythemia-inducing spleen focus-forming virus, was administered. The FV stock was created as a 15% spleen cell homogenate from susceptible BALB/c mice infected for 14 days with 4,000 spleen focus-forming units (SFFU) of FV. Mice received intravenous injections of 0.1 ml phosphate-buffered saline (PBS) containing 40,000 SFFU of FV and mice were analyzed 7 days post infection (dpi). Importantly, the virus stock did not contain lactate dehydrogenase-elevating virus. Immortalized mouse embryonic fibroblasts (CIM) were infected with Bac-derived wild type Murine Cytomegalovirus (MCMV) Smith strain for virus propagation. At 3-4 days post infection the cells were then collected and the virus was purified from the infected cells by ultracentrifugation at 27,000 g for 1 hour as previously described (31). The virus titer was determined by performing standard plaque assay on Mouse newborn fibroblasts (MNC) (32). Female C57BL/6 mice, aged 6-12 weeks, were infected intraperitoneally with 2x10⁵ PFU in 100 μl PBS for 8 weeks. Herpes simplex virus 1 (HSV-1) F strain was propagated in CIM cells. At 4-5 days post-infection, the cells and supernatant were harvested, and subjected to 3-4 freeze-thaw cycles to enhance virus yield. The mixture was then centrifuged for 30 minutes, and the resulting supernatant was collected for final purification via ultracentrifugation at 23.000 g for 2 hours. Virus titer was determined using an end-point infection assay (TCID50) on CIM cells. Female C57BL/6 mice, aged 6-12 weeks, were infected intravenously with 5x10⁶ TCID50 in 100 μl PBS for 8 weeks. All experiments were approved and conducted in accordance with the guidelines of the State Agency for Nature, Environment, and Consumer Protection (LANUV).

Mus dunni fibroblast and mouse lymphoma (YAC-1) cell lines were cultured in RPMI 1640 supplemented with 10% fetal bovine serum and 100 U/ml Penicillin/100 μg/ml Streptomycin. Cells were cultured in a humidified atmosphere with 5% CO₂ at 37°C.

Single-cell suspensions obtained from femur and tibia of hind legs were prepared and serially diluted in 10-fold increments. These dilutions were then co-cultured with Mus dunni cells for 72 hours. After incubation, cells were fixed with ethanol. To visualize the infectious centers, cells were immunostained using a F-MuLV envelope-specific monoclonal antibody (clone 720). The F-MuLV envelope-specific antibody was generated in a hybridoma cell line (H720). This primary antibody was then detected using a peroxidase-conjugated goat anti-mouse secondary antibody (Dianova). The staining was developed by incubating the cells with aminoethylcarbazol (Sigma, 0.1 mg/ml). Following development, the cells were washed with water to reveal distinct foci, each representing an infectious center.

Splenocytes were isolated from naive C57BL/6 mice and cultured for 6 days in 12.5 ng/ml IL-15 (BioLegend). On day 6, IL-2 (20 ng/ml), IL-12 (10 ng/ml), murine IFNα11 (500 U/ml) (18), murine IFNβ (500 U/ml) or murine IFNλ2 (100 ng/ml, Peprotech) were added for 18 hours, if indicated. Cells were incubated in a humidified atmosphere with 5% CO₂ at 37°C.

Splenocytes were treated overnight with IFNα11 (500 U/ml), IFNβ (500 U/ml), IFNlambda2 (100 ng/ml) and IL-2 (20 ng/ml)/IL-12 (10 ng/ml) or left untreated. Cells were washed twice with serum-free RPMI supplemented with 0.5% BSA. Spleen cells were then incubated with 5 μg/ml of Alexa Fluor 647-conjugated Transferrin (T23366, Invitrogen) for 10 min at 37°C. Uptake was stopped by washing twice with ice-cold acid wash buffer (PBS supplemented with 150 mM NaCl and 20 mM citric acid, pH 5) followed by ice-cold RPMI + 0.5% BSA. Cells were stained for surface markers and analysed by flow cytometry.

Bone marrow cells were flushed from murine tibia and femur into a single-cell suspension and washed. Cells were incubated with fluorochrome-conjugated antibodies for 15 minutes at room temperature, protected from light. After staining, the cells were washed and either analyzed immediately using a multi-parameter flow cytometer or fixed for subsequent intracellular staining using a Fixation/Permeabilization Solution Kit from BD Biosciences.

For cytokine detection, splenocytes were restimulated prior to staining. The cells were incubated for 3 hours at 37°C in RPMI buffer containing ionomycin at a concentration of 500 ng/ml, phorbol myristate acetate (PMA) at 25 ng/ml, monensin at 1X (BioLegend), and brefeldin A at 2 μg/ml. For staining of mitochondria, MitoSpy Green (BioLegend, 1 mM), Tetramethyl-rhodamine, Methyl Ester, Perchlorate (TMRM, Invitrogen; 100 nM) and CellROX (Invitrogen, 5 µM) were added in RPMI supplemented with 1% Pen/Strep (Gibco) and 10% FBS (Sigma) and incubated in a humidified atmosphere with 5% CO₂ for 20 min at 37°C.

Flow cytometry analysis was performed using several antibodies, including CD3 (clone 145-2C11 from Thermo Fisher), CD69 (clone H1.2F3 from BioLegend), CD71 (clone R17217 (RI7 217.1.4) from Thermo Fisher), CD107a (clone 1D4B from Becton Dickinson), CD11b (M1/70 from Becton Dickinson), CD11c (N418 from Becton Dickinson), F4/80 (BM8 from BioLegend), Gr1 (RB6-8C5 from Thermo Fisher), Granzyme A (GzmA) (GzA-3G8.5 from Thermo Fisher), Granzyme B (GzmB) (clone GB11 from Becton Dickinson), IFN-γ (clone XMG1.2 from Thermo Fisher), KI-67 (clone B56, Becton Dickinson), NK1.1 (clone PK136 from Becton Dickinson), Ter119 (clone TER-119 from BioLegend), TNFα (clone MP6-XT22 from BioLegend), TRAIL (clone N2B6 from BioLegend). To exclude dead cells, cells were stained with Zombie Aqua (BioLegend) or fixable viability dye (Invitrogen).

Murine serum was obtained by spinning the clotted blood sample at 8000×g for 5 min and stored at −80°C. Serum iron was quantified using the Atellica CH 930 Analyzer (Siemens Healthineers) and the Iron_2 and Iron_3 methods at University Hospital Essen, Germany. The Iron_2 and Iron_3 assays are methods for the quantitative determination of iron in serum and plasma used on the Atellica CH 930 systems of Siemens Healthineers in routine operation of the Department of Clinical Chemistry at Essen University Hospital. In both methods, Fe3+ is reduced to Fe2+ by transferrin under acidic conditions and simultaneously with ascorbic acid. In the Iron_2 method, a complex formation by ferrozin with Fe2+ occurs. Quantification is carried out by spectrometric measurement of absorption bichromatically at 571 and 658 nm. In the Iron_3 method, on the other hand, a complex of Ferene and Fe2+ is formed, the absorption of which is measured spectrometrically at 596 and 694 nm bichromatically. In the course of the study, the central laboratory had to switch from the Iron_2 to the Iron_3 assay because the Iron_2 method was no longer available. Internal verifications at the Department of Clinical Chemistry showed an excellent agreement between the two methods with almost the same sensitivity. Both methods can be traced back to the reference material SRM 937 of the National Institute of Standards and Technology (NIST).

Single cells were cultured for three days with IL-15 (12.5 ng/ml, BioLegend) and were treated with DFO (Sigma, 100 µM) for 72 hours. FeSO₄ (Sigma, 200 µM) was added for the final 18 hours.

If cells were treated with FeSO₄ only, single cells were plated and freshly prepared FeSO₄ (Sigma, 200 µM) was added for a total of 18 hours. Cells were incubated in a humidified atmosphere with 5% CO₂ at 37°C, washed twice and measured at BD Canto II.

Bone marrow cells of FV-infected animals (7 dpi) were treated with FeSO₄ (Sigma, final concentration 200 µM) or left untreated for 18 hours (37°C) followed by NK cells isolation using magnetic beads. 0.5 x 10⁶ isolated NK cells were transferred into C57BL/6, which were subsequently infected with 40,000 SFFU of FV. Viral loads were determined 3 days post infection by an infectious center assay.

Bone marrow cells were isolated and stained with Tag-it Violet (BioLegend, 5 µM). Cells were plated into 96 well U-bottom plates. IL-15 (50 ng/ml, BioLegend) and IL-18 (50 ng/ml, BioLegend) were added to cells as well as FeSO₄ (50 µM, Sigma). Cells were incubated in a humidified atmosphere with 5% CO₂ at 37°C. After 3 days, proliferation of NK cells was assessed by flow cytometry (BD Canto II) as measured by loss of the Tag It Violet dye.

RPMI 1640 medium, supplemented with 10% Fetal Bovine Serum and 100 U/ml Penicillin/100 µg/ml Streptomycin, was prewarmed. Single cells from bone marrow were treated with FeSO₄ for 18 hours or left untreated, washed, and then resuspended to 96 well plates. Inhibitors were added at final concentrations of 100 mM 2-Deoxy-D-glucose and 1 µM oligomycin, followed by a 30 minutes incubation at 37°C. Subsequently, puromycin was added at a final concentration of 10 µg/ml (Sigma) for 20 minutes. Cells were washed with cold PBS, and FC blockage (BD Biosciences) was performed at 4°C for 5 minutes. Cell surface staining was conducted at 4°C for 30 minutes in the dark, after which the cells were washed. Cells were then fixed using the Fixation/Permeabilization Solution Kit (BD Biosciences). Anti-puromycin staining (Sigma, clone 12D10) was performed in perm-buffer at 4°C for 1 hour. Finally, the cells were analyzed using BD Canto II. Mitochondrial dependency was calculated using the formula: 100 x (no inhibitor – oligomycin)/(no inhibitor – 2-DG Oligo).

Bone marrow cells were treated with or without FeSO₄ for 4 or 18 hours at 37°C. YAC-1 cells were stained for carboxyfluorescein succinimidyl ester (CFSE, BioLegend, 2.5 µM) and seeded in a 200:1 ratio with single cells from bone marrow. After 4 hours, fixable viability dye was used to determine dead cells. YAC-1 killing was calculated by (dead YAC-1 tumor cells)/(all YAC-1 tumor cells) x 100.

Data visualization and statistical analyses were performed using GraphPad Prism version 8 or BioRender.com. For comparisons between two groups, the Mann-Whitney test was applied for nonparametric data, while the unpaired t-test was used for parametric data. When analyzing multiple groups ordinary one-way ANOVA with Tukey’s multiple-comparison test was employed for parametric data. Additionally, correlations between variables were assessed, and linear regression analysis was conducted to determine the relationship between variables and evaluate the goodness of fit. Spider plots were generated with Excel.

Until now, the involvement of IFNs in iron homeostasis, which is vital for the proliferation and function of immune cells, is not known. To understand the role of IFN stimulation on NK cell activation (CD69) and transferrin receptor (CD71) expression, expanded NK cells were treated with murine IFNα11, IFNβ or IFNλ2 and analyzed via flow cytometry ( Figures 1A–F ) (33). IL-2/12 stimulation was used as positive control (34). Stimulation of NK cells with IFNα significantly increased the frequency of NK cell activation ( Figure 1A ) as well as the percentage of CD71⁺ NK cells and CD71 expression ( Figures 1C, D ) similar to IL-2/12-treated NK cells. IFNβ slightly increased the activation and transferrin receptor expression in NK cells, although not significant, whereas IFNλ stimulation had no effect on NK cell activation and CD71 expression. Linking the CD71 expression with the activation of NK cells revealed a positive correlation of both parameters ( Figures 1E, F ). NK cells showed an increased uptake of iron, measured by transferrin uptake, after IFNα11, IFNβ, and IL-2/12 stimulation compared to unstimulated and IFNλ-treated cells ( Figure 1G ). Transferrin uptake of NK cells positively correlated with NK cell activation ( Figure 1H ). To investigate whether the elevated CD71 expression on NK cells following IFNα and IFNβ stimulation was a consequence of increased proliferation and subsequent higher iron demand, we assessed NK cell proliferation using the marker KI-67 ( Figure 1I ). Our analysis revealed no significant differences in the proportion of proliferating NK cells between the experimental groups. These findings suggest that the observed increase in CD71 expression is more likely a direct effect on iron homeostasis rather than a result of enhanced proliferation and demonstrate that type I IFNs, which are rapidly induced upon viral infections, influence the potential of NK cells to take up iron, thus contributing to proper iron supply in NK cell.

In our previous analysis we found a correlation between activation and transferrin receptor expression as well as iron uptake of NK cells after IFNα stimulation ( Figures 1E, F, H ). Building on insights from our in vitro studies, the second part of our investigation examines the impact of dietary iron deficiency on NK cell functionality in vivo. Specifically, we aimed to determine how a low-iron diet affects the antiviral response of NK cells in mice infected with FV, revealing critical implications for immune performance under conditions of iron scarcity. We hypothesized that iron is essential for NK cell functions and the control of viral loads in virus infection. While viruses require iron for their replication and survival, NK cells also depend on iron for their activation and optimal function, creating a complex interplay where iron availability can simultaneously benefit both the pathogen and the host’s immune defense. To address the question whether diet-induced iron levels reduce or enhance viral replication, we fed mice for four weeks with low iron or control chow ( Figure 2A ). Next, we infected these mice with FV for 7 days. The analysis of serum iron showed a significant decrease in mice fed with low iron chow (low Fe) compared to mice fed with control (Ctr) diet ( Figure 2B ). FV mainly infects erythroid precursor cells (Ter119⁺), granulocytes (Gr1⁺), macrophages (F4/80⁺) as well as dendritic cells (CD11c⁺) (35). To elucidate the role of iron on FV target cells, we analyzed cell numbers of these cells in the bone marrow. We did not detect differences in the numbers of FV target cells in low iron fed animals compared to control diet fed mice ( Figure 2C ).

The bone marrow is the most active site of iron utilization in the body (17) and it has been shown that NK cells located in the bone marrow are very sensitive to alterations in iron availability (8). Thus, we analyzed FV replication in the bone marrow of infected animals, which were either fed with control diet (Ctr) or low iron containing diet (low Fe) ( Figure 2D ). Viral replication was significantly increased in animals fed with low iron diet compared to control diet-fed mice. Next, we analyzed the percentages of bone marrow NK cells in control or low iron diet fed animals ( Figure 2E ), which was comparable between both groups. By analyzing GzmB expression in NK cells from the bone marrow we found no differences in the expression of GzmB on NK cells, but a significantly decreased proportion of GzmB⁺ NK cells in animals fed with low iron diet ( Figure 2F ). This suggests that variations in viral loads are attributed to altered elimination of FV-infected cells, rather than solely to differences in the proliferation patterns of FV target cells. Collectively, iron deficiency increased viral replication and decreased GzmB in NK cells implicating a primary role of iron for immune activation rather than in FV replication.

As shown above, insufficient iron levels decreased NK cell’s GzmB expression and increased viral loads ( Figure 2 ). To demonstrate that iron deficiency affects NK cells similarly in vitro as observed in our in vivo experiments with a low iron diet ( Figure 2 ), we employed deferoxamine (DFO), an iron chelator, to induce iron deficiency in cell culture conditions and on the other hand used iron sulfate (FeSO₄) to increase the iron in vitro ( Figure 3A ). Similar to the analyses of the low dietary iron treatment of mice ( Figures 2E, F ), we analyzed frequencies of NK cells ( Figure 3B ) and GzmB⁺ NK cells ( Figure 3C ), which were obtained from FV-infected animals (7 dpi) after in vitro stimulation with DFO or FeSO₄. While we did not find differences in percentages of NK cells between untreated, DFO- or FeSO₄-treated groups ( Figure 3B ), DFO significantly decreased the frequencies of GzmB⁺ NK cells, while FeSO₄ treatment increased the percentage of GzmB⁺ NK cells ( Figure 3C ). The MFI of GzmB⁺ NK cells was not significantly altered by the DFO treatment, consistent with data from in vivo experiments ( Figure 3C ). These findings demonstrate that the low dietary iron experiments in vivo and the DFO/FeSO₄ condition in vitro were well comparable. We further elucidated if iron supplementation can increase the NK cell activation as well as their metabolism, which could be important to improve the treatment of viral infections and cancers. Therefore, NK cells from FV-infected mice (7 dpi) were stimulated with FeSO₄ and analyzed for their effector phenotype (18 hours; Figure 3D ). NK cell proliferation was assessed by tracking cell divisions through the progressive dilution of a fluorescent dye (72 hours; Figure 3D ). NK cell numbers were significantly increased after iron treatment, without any observable differences in NK cell viability ( Figure 3E ). Instead, an increased frequency of proliferation was detectable in iron-treated NK cells. Increased expression and frequencies of the activation marker CD69 and degranulation marker CD107a in iron-supplemented NK cells were identified ( Figure 3E ) and shown as representative histogram in Supplementary Figure 1 or as bar graphs in Supplementary Figure 2 . Interestingly, iron supplementation decreased the expression and frequency of CD71 in NK cells ( Figure 3E ). Upon infection, NK cells rearrange their mitochondria to fuel their increased demands for energy (8) and iron is involved in multiple mitochondrial processes (18). Hence after iron stimulation, NK cells enhanced the expression of MitoSpy, an indicator for mitochondrial mass ( Figure 3F ). Additionally, we detected an increased membrane polarization measured by accumulation of TMRM dye into mitochondrial membrane ( Figure 3F ). Interestingly, iron supplementation also increased the cellular stress in NK cells, as demonstrated by increased CellROX levels ( Figure 3F ). Notably, our investigation revealed that NK cell responses following murine cytomegalovirus (MCMV) or herpes simplex virus type 1 (HSV-1) infection exhibited consistent alterations in response to iron supplementation ( Supplementary Figure 3 , Supplementary Figure 4 ). We then investigated whether extracellular iron could enhance NK cell function by boosting their metabolic activity (7 dpi, FV). For this, we employed SCENITH, a single-cell technique that enables immunometabolic characterization through the analysis of cellular protein translation (36). Our findings demonstrate that NK cells exhibited increased protein translation rates ( Figure 3F ), as evidenced by puromycin incorporation, following iron treatment. Additionally, upon iron stimulation, NK cells showed an increased dependence on oxidative phosphorylation, also measured by SCENITH ( Figure 3F ). These data collectively show that NK cell functions and their mitochondrial machinery can be improved by iron supplementation.

Because iron supplementation can potentially enhance NK cell responses, particularly in the context of iron deficiency or increased iron demand during infections, parameters determining the NK cells cytotoxicity were analyzed in Figure 4A . We examined the cytotoxicity-associated markers GzmA, GzmB and TRAIL in iron-treated NK cells compared to untreated cells ( Figure 4A ). Percentages and median fluorescence intensities (MFI) of GzmA⁺, GzmB⁺, as well as TRAIL⁺ NK cells were significantly increased after iron stimulation. Iron stimulation of NK cells from MCMV or HSV-1 showed similar results ( Supplementary Figure 3 , Supplementary Figure 4 ). Surprisingly, cytokines expressed by NK cells, such as IFNγ or TNF, were rather decreased (TNF) or unchanged (IFNγ) upon iron supplementation (FV, Figure 4A ). Next, we examined the cytotoxicity of NK cells in an in vitro kill assay. NK cell killing was analyzed after 4 or 18 hours of FeSO₄ treatment. To measure the killing capacity of iron-pretreated NK cells, bone marrow cells were cocultured with CFSE-stained NK target cells (YAC-1) for 4 hours. After the coincubation, target cells were quantified by flow cytometry ( Figure 4B ) and NK cell-mediated killing was calculated ( Figure 4C ). The killing capacity of NK cells did not differ between the untreated and treated group after 4 hours of FeSO₄ stimulation. However, after 18 hours of FeSO₄ treatment, killing was 3.5 times higher in the iron-stimulated group compared to non-supplemented group. Next, we analyzed the NK cell killing in vivo after FeSO₄ stimulation of NK cells in vitro. Therefore, we extracted bone marrow cells of FV-infected mice (7 dpi) and stimulated them with FeSO₄ for 18 hours. Untreated bone marrow cells (-FeSO₄) served as control. NK cells were then isolated from treated and untreated bone marrow cells using magnetic beads and 0.5 x 10⁶ NK cells were transferred into naïve mice, which were subsequently infected with FV ( Figure 4D ). Viral loads were determined in the bone marrow at 3 dpi ( Figure 4E ). Mice that received iron-treated NK cells exhibited significantly reduced viral loads compared to mice that received untreated NK cells. These data demonstrate a beneficial and stimulatory effect of iron on the NK cell cytotoxicity in vitro and in vivo resulting in improved FV control.