THP-1 human monocytic cells (DMSZ ACC 16) were propagated in RPMI (PAN-Biotech) containing 10% FBS (PAN-Biotech), 1% Penicillin/Streptomycin (Lonza), 1% NEAAs (Gibco) and 2% Na-Pyruvat (Gibco). Before infection experiments, 5 × 10⁵ cells were seeded onto 12-well plates and differentiated with 20 ng/mL Phorbol-12-myristate-13-acetate (PMA) for 48 h. After differentiation, macrophage-like cells were washed with PBS and the medium was changed to RPMI containing 1% FBS and no antibiotics.

Klebsiella pneumoniae (ATCC 43816) was grown from overnight cultures under sterile conditions in LB Broth (Sigma-Aldrich) to mid logarithmic growth phase [optical density 600 nm (OD₆₀₀) 0.4–0.6]. Bacterial counts were determined before each experiment using a cell counter and analyzer (CASY, 45 μm capillary, OLS OMNI Life Science).

Bacteria constitutively expressing the fluorescent protein Ypet were previously described (Grubwieser et al., 2022). For the KP iron sensing strain, bacteria were made electro-competent using glycerol/mannitol density step centrifugation, as described in an established protocol (Warren, 2011). The plasmid pAH05 with genes encoding for blue fluorescent protein (BFP) downstream of the iron sensitive RhyB2 promotor, green fluorescent protein (GFP) downstream of the SodB promotor, and mCherry downstream of the constitutively active PybaJ promoter was electroporated into KP (NCBI: OQ979407).

For experiments with heat-inactivated KP, bacteria were incubated at 70°C for 20 min. For inactivation with gentamicin, KP was incubated in 50 μg/mL gentamicin (Life Technologies) for 30 min. After inactivation, bacteria were plated on LB plates to confirm the absence of viable bacteria.

For the bacterial growth assay, bacteria were diluted to OD₆₀₀ of 0.005 in 50 μL cell culture medium in sterile 96-well plates (Greiner bio-one), and directly afterwards incubated in an automated microplate reader (Spark, TECAN) at 37°C, 5% CO₂. OD₆₀₀ and fluorescence intensity were measured every 15 min after a double orbital shaking interval for a total of 14 h.

To quantify intracellular bacteria, we applied a gentamicin-protection assay. Cells were infected with KP at a multiplicity of infection (MOI) of 10. After 1 h of active infection, cells were washed thrice with PBS containing gentamicin (Life Technologies, 50 μg/mL) and incubated in a fresh medium containing 1% FBS and gentamicin (25 μg/mL). This treatment prevents bacterial overgrowth, and as gentamicin exclusively eliminates extracellular bacteria, allows further incubation of macrophages containing intracellular viable bacteria. Uninfected controls were treated with identical washing and incubation steps. To quantify the number of viable intracellular pathogens, cells were washed thrice again in PBS and afterwards lysed in 0.5% sodium deoxycholic acid (Sigma-Aldrich) at indicated time points. Cell lysates were then serially diluted and plated onto LB-plates. After overnight incubation at 37°C, colony-forming units (CFUs) were quantified.

When cells were treated with inactivated bacteria, an MOI of 50 was used to account for the lack of bacterial growth compared to the viable bacteria during the 1 h active infection phase. After 1 h of stimulation, cells were washed and incubated in a fresh medium containing 1% FBS and gentamicin, equal to cells infected with viable bacteria.

Where indicated, cells were stimulated during the gentamicin-protected infection phase (directly after washing and addition of gentamicin-containing medium) with either 50 μg/mL iron-loaded transferrin (TF, Sigma-Aldrich), 1 μM STAT-6 inhibitor AS1517499 (MedChem Express), 40 μg/mL anti-human IL-10 blocking antibody (Biolegend) or 40 ng/mL recombinant human IL-10 (Peprotech).

For immune fluorescence imaging, cells were seeded and differentiated in 8-well chamber slides (Falcon) and infected with fluorescent (Ypet) bacteria. Afterwards, cells in chamber slides were washed with PBS and fixed with 4% paraformaldehyde for 10 min. Cells were then permeabilized with 0.5% saponin (Sigma-Aldrich) for 30 min, stained with Alexa-flour-647-labeled CD71 (TFR1) antibody (Biolegend) for 2 h, and 4′,6-diamidino-2-phenylindole (DAPI, BioLegend) for 30 min. After staining, samples were mounted with Faramount Mounting Medium (Dako). Fluorescence microscopy was performed immediately after sample preparation using a VS120-S6 fluorescence microscope (Olympus). Images were captured with a 40-x objective using 387/440 nm (DAPI), 485/525 nm (Ypet), and 650/684 nm (Alexa-flour-647) lasers and filters, under identical exposure times for every sample. Fluorescence intensity of Alexa-flour-647 was determined with the CellSense (Olympus) program for at least 100 cells per sample.

The quantitative real-time PCR was carried out as described elsewhere (Hoffmann et al., 2021). In short, total sample RNA was isolated using acid guanidinium thiocyanate-phenol-chloroform extraction with peqGOLD Tri-Fast™ (Peqlab). For reverse transcription, 2 μg RNA, random hexamer primers (200 ng/μL) (Roche), dNTPs (10 mM) (GE Healthcare LifeSciences) 20 U RNasin (Promega) and 200 U M-MLV reverse transcriptase (Invitrogen) in first-strand buffer (Invitrogen) were used. Ssofast Probes Supermix and Ssofast EvaGreen Supermix (Bio-Rad Laboratories GmbH) were used according to the manufacturer’s instructions. Real-time PCR reactions were performed on QuantStudio 3 and 5 real-time PCR systems (Thermo Fisher Scientific). Gene expression was normalized using the ΔΔct method. Tubulin (TUB) and ornithine decarboxylase antizyme 1 (OAZ1) were used as reference transcripts. TaqMan PCR primers used in this study are listed in Table 1.

Protein extraction and Western blotting were performed as described previously (Hoffmann et al., 2021). The following antibodies were used: a mouse TFR1 antibody (1,1,000; Sigma Cat# SAB4300398), a rabbit FT antibody (1,500; Sigma, F5012), and a rabbit actin antibody (1,500, Sigma Cat# A2066). Appropriate HRP-conjugated secondary antibodies (1,2000, anti-rabbit, Dako Cat# P0399; 1:4000, anti-mouse; Dako Cat# P0447) were used. For quantification, densitometry data were acquired on a ChemiDoc Touch Imaging System (Bio-Rad) and analyzed with Image Lab 5.2.1. (Bio-Rad Laboratories GmbH).

A human IL-10 ELISA Set (BD Biosciences) was used to measure secreted IL-10 levels in cell culture supernatants, according to the manufacturer’s protocol.

For evaluation of free cellular iron, cells were washed with PBS and stained with 250 μM FerroOrange (FO, Dojindo) for 30 min. FO is a fluorescent probe that detects intracellular Fe²⁺ ions with high specificity and does not react with chelated iron or other bivalent metals (Hirayama et al., 2020). Subsequently, fluorescence intensity (540 nm/585 nm) was determined using a multimode microplate reader (Spark, TECAN) at 16 different localizations in each well.

To assess the fluorescence of intracellular bacteria, flow cytometry was applied. Cells were infected with KP containing the pAH05 plasmid and underwent the gentamicin protection assay as described in 2.2. After 24 h, cells were harvested in FACS-buffer containing 0.5% FBS and 2 mM ethylenediaminetetraacetic acid (EDTA) and immediately subject to flow cytometry analysis. Data were acquired on a CytoFLEX S (Beckman Coulter) and subsequently analyzed with FlowJo 10.7.0 software (Beckton Dickinson). After gating for cells and single cells, infected cells were identified as mCherry+. In infected cells, BFP MFI was acquired to evaluate the status of bacterial iron starvation. Gating of infected cells is depicted in Figure 4.

Statistical analysis was carried out using GraphPad Prism version 9.1 for Windows and Mac (GraphPad Software). Data are presented as mean with 95% CI, SD or SEM as dispersion characteristic. Significant differences between groups were determined using unpaired t test or ANOVA with post-hoc analysis. Multiple comparisons were adjusted using Tukey’s or Holm-Šídák’s methods. For non-normal distributed data, as evaluated by Shapiro–Wilk-test, a Kruskal-Wallis test with Dunn’s multiple comparisons test was performed. p < 0.05 was used as the significance threshold.

To study the KP-macrophage interaction, we infected PMA differentiated THP1 cells with viable bacteria for 1 h. We then applied a gentamicin-protection assay, which eliminates extracellular bacteria but preserved viable bacteria in the intracellular space of macrophages over the course of 24 h (Figure 1A). Although bacterial numbers declined over time, macrophages were unable to fully eliminate the pathogen during the 24 h infection phase. This is indicative of bacterial resistance in this host-pathogen interaction, in which numerous host factors influence the outcome, including cellular iron metabolism. To gain insight into alterations of macrophage iron homeostasis, we performed western blots of critical iron proteins in uninfected controls and KP infected cells after 24 h. Interestingly, protein levels of the main cellular iron importer, the transferrin-receptor-1 (TFR1), were strongly induced in KP infected cells (Figure 1B; Supplementary Figure S1). A time course analysis of TFR1 protein levels revealed that induction was already present after 12 h of infection, but was most pronounced after 24 h of infection (Supplementary Figure S2). Accordingly, the iron storage protein ferritin (FT) was upregulated in KP infected cells. To shed light on this host-pathogen interaction, we applied immunofluorescence using bacteria that were transformed to constitutively express the Ypet fluorescence protein (Figure 1C). Intracellular bacteria were visible in infected macrophages (green, rod shaped) and the induction of TFR1 (red) was again confirmed. TFR1 expression was enhanced in both, cells containing bacteria and bystander cells, indicative of an autocrine and paracrine mechanism driving this upregulation.

Together, these results implicate that once inside the host cell, KP establishes an intracellular niche to facilitate infection, and this process is associated with higher expression of the main cellular iron uptake protein, TFR1.

Next, we aimed to uncover whether TFR1 induction is solely a consequence of the host-response to bacterial invasion, or aids the pathogen to establish infection in macrophages. For this, we stimulated infected cells with 50 μg/mL iron-loaded transferrin (TF), to facilitate increased iron availability for KP via the TFR1-mediated iron uptake. Interestingly, KP infected cells treated with TF showed increased intracellular pathogen numbers compared to not TF supplemented cells after 24 h of infection (Figure 2A). This raises the possibility, that TFR1 induction is mediated by the intracellular pathogen itself, as increased cellular iron availability benefits the intracellular pathogen. Following this lead, we treated cells with either heat-inactivated, gentamicin-killed, or viable KP bacteria and subsequently analyzed TFR1 and FT protein expression (Figure 2B; Supplementary Figure S1). Strikingly, only viable bacteria were able to induce the main iron importer TFR1. Of note, FT induction was solely increased in cells treated with inactivated bacteria. This suggested that a specific factor originating from viable intracellular bacteria is responsible for TFR1 induction. In contrast, sterile pathogen associated molecular patterns from inactivated bacteria, like LPS, seem to be responsible for FT induction. Subsequently, we quantified TFR1 mRNA levels, to shed light on the specific induction mechanism (Figure 2C). Consistent with our protein data, mRNA expression of TFR1 was significantly increased in KP infected cells compared to uninfected controls over the course of infection. We then studied cellular iron accumulation employing the fluorescent probe FerroOrange (FO, Figure 2D). In cells treated with TF, significantly more iron was measured compared to untreated controls. Cells infected with KP showed significantly more FO fluorescence compared to uninfected cells. Interestingly, infected cells treated with TF showed the highest fluorescence level, being indicative for more cellular iron incorporation. Collectively, these results indicate that viable KP induces TFR1-mediated iron uptake in macrophages.

Next, we investigated possible mechanisms underlying increased TFR1-mediated iron uptake in KP infected macrophages. As recent work proposed the STAT6-IL-10-axis to be involved in KP-driven metabolic reprogramming, we investigated the effects of specific pharmacological inhibitors of this pathway (Dumigan et al., 2022). First, we stimulated macrophages with the STAT-6 inhibitor (AS1517499). Interestingly, the KP dependent TFR1 induction could be reduced, as analyzed by Western Blot (Figure 3A; Supplementary Figure S1). Protein levels of FT concomitantly were decreased. Next, we analyzed the cellular expression of the anti-inflammatory cytokine IL-10 over the course of infection (Figure 3B). IL-10 expression was significantly increased in cells infected with KP compared to uninfected cells, and treatment with the STAT-6 inhibitor during infection significantly diminished IL-10 expression in KP infected cells. To follow this up, we performed IL-10 ELISA of cell culture supernatants. In line with gene expression data, IL-10 was detected in high amounts in supernatants of infected cells, whereas treatment with the STAT-6 inhibitor substantially decreased IL-10 concentrations in supernatants (Figure 3C). As IL-10 emerged as a crucial mediator of KP induced metabolic changes, but is also known to regulate TFR1 expression in inflammatory macrophages (Ludwiczek et al., 2003), we applied a specific IL-10 blocking antibody during infection experiments. In agreement with our previous observations, IL-10 neutralization antagonized KP induced TFR1 induction, whereas the effect on FT induction was less pronounced (Figure 3D; Supplementary Figure S1). Interestingly, treatment of uninfected cells with recombinant human IL-10 did not lead to increased TFR1 levels (Figure 3E). In contrast, if cells were treated with inactivated bacteria together with recombinant IL-10, TFR1 was markedly induced, which is in line with previous observations on TFR1 regulation in macrophages (Tilg et al., 2002; Ludwiczek et al., 2003).

To quantify the effects of both specific inhibitors (STAT-6-inhibition and anti IL-10) on intracellular pathogen survival, we quantified intracellular bacteria after 24 h of infection. STAT-6 inhibition (Figure 3F), as well as IL-10 blockade (Figure 3G) led to significantly decreased numbers of intracellular CFUs. To exclude a direct bactericidal effect of the STAT-6 inhibitor AS1517499, a bacterial growth assay was performed (Supplementary Figure S3). This indicated that the STAT-6 inhibitor did not negatively affect the growth of KP in-vitro. We then applied immunofluorescence staining of cells to corroborate our findings. For this, cells were infected with KP carrying the Ypet plasmid (green) and stained for TFR1 (red). Expression of TFR1 was again increased in infected cells compared to uninfected controls, as measured in at least 100 cells per condition using automated CellSense software (Figure 3H). Treatment with a neutralizing IL-10 antibody significantly decreased TFR1 levels. Intracellular bacteria, as well as TFR1 expression patterns are also visible in provided immunofluorescence images (Figure 3I).

In sum, our data indicates that KP induces TFR1 expression in macrophages via the STAT-6-IL-10 axis to ensure a sufficient supply of iron for intracellular multiplication. Of note, inhibition of the STAT-6-IL-10 pathway results in reduced TFR1 expression and leads to impaired survival of KP within macrophages.

To explore, whether inhibition of the STAT-6-IL-10 axis leads to growth limitation because of bacterial iron starvation, we employed flow cytometry of cells infected with KP containing the pAH05 plasmid, which encodes for iron sensitive promoters coupled to the expression of fluorescent proteins (NCBI OQ979407). As such, bacteria constitutively express mCherry (PybaJ promotor), GFP under high iron (SodB promotor) and BFP under iron starved (RhyB2 promotor) conditions (Figure 4A shows illustration, Figure 4B shows BFP fluorescence of bacterial cultures, Supplementary Figure S4 shows mCherry and GFP fluorescence of bacterial cultures). Using this approach, we determined the fluorescence of BFP, indicative of iron starvation, in infected (mCherry+) cells (Figure 4C) via flow cytometry. As the intracellular environment does not expose bacteria to excessive amounts of iron, GFP fluorescence of infected cells was not changed in any of the analyzed conditions (data not shown). Nevertheless, BFP mediated fluorescence was increased when infected cells were treated with anti IL-10, indicating iron limitation for bacteria. More prominently, treatment of infected cells with iron loaded TF reduced BFP expression, and this effect was largely reversed upon IL-10 neutralization (Figure 4D). Finally, we performed measurement of cellular iron content via FO under the same conditions (Figure 4E). In agreement with BFP-mediated indication of iron availability for bacteria, cellular iron levels were increased in infected and TF treated cells, whereas cellular iron content was markedly reduced upon anti-IL-10 treatment of KP infected macrophages.

Taken together, our results demonstrate that activation of the STAT-6-IL-10 axis during infection leads to increased iron supply to intracellular bacteria, and inhibition of this pathway reduces TFR1 mediated iron uptake and elicits iron starvation of intramacrophage bacteria, thereby diminishing their survival.