Chemically synthesized HMOs and HMOs produced by fermentation were prepared in house by Glycom A/S (Hørsholm, Denmark). Their structures were verified by 1H and 13C-NMR. Endotoxin measurements on HMOs produced by fermentation were performed with the Limulus Amebocyte Lysate Kinetic-QCL™ from Lonza (US License No. 1775, catalog No.:50-650U). HMO structures used in this study are depicted in Figure 1A .

The clinically derived methicillin resistant S. aureus (MRSA) WT MRSA (SADR1/A9781) and SNP MRSA (SADR2/A9788) strains were previously described (original names in brackets) (49–51). Culture and UV-inactivation of S. aureus were performed as previously described (49). In brief, overnight S. aureus colonies were inoculated in trypsin soy broth to OD₆₀₀ ~ 0.03 and grown in Erlenmeyer flasks to early stationary phase (OD600 = 5-6). Afterwards, the bacteria were washed and resuspended in PBS to OD600 = 1 (~10⁹ S. aureus/ml). S. aureus solutions were transferred to petri dishes and subjected to pulsed UV-radiation of 10.000μJ/cm² for 120 seconds (monochromatic wavelength of 254nm; CL-1000 crosslinker; UVP, Cambridge, United Kingdom) for UV-inactivation. Bacterial death was verified by plating and incubating on TSA plates at 37°C overnight ( Supplementary Figure 1 ).

Primary monocytes were isolated from buffy coats, as described above. The isolated monocytes were seeded at 3 x 10⁵ cells/ml and cultured in standard RPMI 1640 media, supplemented with 20 mM of HMO and 40 ng/mL M-CSF (Peprotech, 300-25). On day three, the cells received half new medium supplemented with the concentration of HMOs and M-CSF doubled. On day 6 all medium was discarded and replaced with fresh standard RPMI 1640 media containing 20 mM of HMO. The cells were then treated with 12,5 x 10⁶ S. aureus/ml cell suspension and the monocyte-derived macrophages were further activated with S. aureus strains for two days. On day 8 the cells were taken for analysis.

Total RNA purification of primary monocytes was completed on day 8, after cultivation, as described above. Cell media was removed, the cells were lysed in RLT buffer (Qiagen, 79216) and stored at -80°C until purification. Total RNA purification was performed on a QIAcube Connect Instrument using the RNeasy Mini QIAcube Kit (Qiagen, 74116) with on-column DNase treatment, according to the manufacturer’s instructions. Purity and concentration were measured on a Denovix DS11 spectrophotometer. RNA quality was assessed on a bioanalyzer. An A260/A280 ratio above 1.95 and a RIN score above 7 was required for transcriptomic analysis. Hereafter RNA was stored at -20°C until further analysis. Transcriptomic profiling was done using Clariom™ D Assay (ThermoFischer Scientific, 902922) and performed by Center for Genomic Medicine, Copenhagen University Hospital. Data analysis was done using Transcriptome Analysis Console 4.0.2 (ThemoFischer Scientific). Mean probe signal intensity on each chip was used to normalize for loading differences between chips. Fold change values were reported by the program.

Levels of secreted cytokines were measured in harvested cell culture supernatants. The analysis was performed using chemiluminescence-based assays from Meso Scale Discovery (MSD, Gaithersburg, MD, USA). The studies included two different assays from the V-PLEX platform: V-PLEX Human IL-8 Kit (K151RAD) singleplex and V-PLEX Viral Panel 2 Human Kit (K15346D). Assay set up, execution and analysis were done according to the manufacturer’s instructions. The analysis was performed using QuickPlex Q 120 instrument (MSD) and DISCOVRY WORKBENCH^® 4.0 software.

Investigation of the NF-κB-pathway was done using THP1-NF-κB-GFP reporter cells, kindly provided by Dr. Peter Steinberger (48). The reporter cells were seeded at a concentration of 3 x 10⁵ cells/ml in standard RPMI media supplemented with HMO and incubated overnight at 37°C, 5% CO₂. The next day S. aureus bacteria were added to the cells at a concentration of ~50 x 10⁶ bacteria/ml cell suspension, and again incubated overnight at 37°C, 5% CO₂. The GFP-signal was evaluated by flow cytometry the following day. The cells were harvested, transferred to a 96-well plate, and washed twice in PBS + 2% FBS, before analysis by flow cytometry.

To evaluate cell-proliferation, monocytes were stained with Carboxyfluorescein Succinimidyl Ester (CFSE, Molecular Probes, C34554) prior to seeding as previously described (49, 52). Shortly, this was done by dissolving cells in PBS + 5% FBS to a concentration of 5 x 10⁶ cells/ml. The CFSE stock was diluted according to the manufacturer’s manual and added 1:1 to the cell suspension for final concentration of 5 μM. The cells were incubated dark for 5 min, while rotating at room temperature. Afterwards the cells were washed twice in PBS + 5% FBS and resuspended in standard 1640 RPMI media, supplemented with HMOs and M-CSF and set up as described above.

On day 8, prior to analysis, the cells were inspected in the microscope and photographed in the petri dish at 10x optics. Photos were taken using the ECHO Rebel Microscope.

Monocytic derived macrophages were cultured in the presence of 20 mM HMOs and 40 ng/mL M-CSF as earlier described and photographed on day 7. The photos of the cells were taken using Incucyte^®, SX1 Live-Cell Analysis System.

UV inactivated S. aureus strains were labelled with AF647 conjugated succinimidylester (SE-AF647; Molecular Probes, A-20006), as described in (49). Briefly, bacteria were resuspended in the original volume of sodium bicarbonate buffer (pH 8.5), SE-AF647 was added at a concentration of 10 ng/μL and bacteria were incubated at 4°C for 1 h with vigorous agitation. Stained bacteria were washed and resuspended to original volume in PBS and stored at -20°C until use. The staining was confirmed on the flow cytometer ( Supplementary Figure 4 ). Data was acquired with an Accuri C6 instrument using Accuri C6 software and analyzed in Flowlogic 8.6 Software.

U937 cells were seeded at a concentration of 3 x 10⁵ cells/ml in standard 1640 RPMI media supplemented with 20 mM HMO and incubated overnight at 37°C, 5% CO₂. The next day, UV-killed SE-AF647 labelled S. aureus bacteria were added to the cells at a concentration of ~50 x 10⁶ bacteria/ml cell suspension and incubated overnight at 37°C, 5% CO₂. The following day, the uptake of UV-killed SE-AF647-labelled S. aureus bacteria was evaluated. Cells were transferred to a 96-well plate, and washed twice in PBS + 2% FBS, before analyzing them by flow cytometry.

Primary monocytes were cultured as previously stated, supplemented with 20 mM HMOs and 40 ng/mL M-CSF (Peprotech, 300-25) for a total of 8 days. On day 8, 0,1μg S. aureus bioparticles (pHrodo^® Red S. aureus Bioparticles^® for Incucyte^®, 4619) were added in 50µl/well (30.000 cells/well) in a 96-well plate. The plate was incubating for 15 minutes at room temperature to let the particles descend. The phagocytic uptake was analyzed by Incucyte^®, SX1 Live-Cell Analysis System.

The graphical abstract was created in BioRender.com. Publication license, including confirmation of publication and licensing rights are available under agreement number is RM26AK4PC3.

Data preparation and statistical analysis was performed using GraphPad Prism, version 9.3.1 (GraphPad Software). Statistical analysis was performed as stated in the figure legends, and the level of statistical significance was determined by a p-value of less than 0.05 and is presented as follows: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

To investigate the effect of individual HMO structures on macrophages in response to S. aureus, transcriptomic profiling was performed on human blood monocyte-derived macrophages. Monocytes were isolated from healthy donors and treated with three different HMOs; 2’FL, LNnT or 6’SL ( Figure 1A ), as well as M-CSF, followed by treatment with S. aureus. On day 8, transcriptomic analysis was performed on total RNA, which showed that all three HMOs changed gene regulation of a high number of genes ( Figure 2 ). The highest number of genes affected is seen in 6’SL-stimulated macrophages challenged with WT MRSA as depicted in Figures 2A, B . We observed an overlap in gene-regulation between the HMOs, but clearly the different structures also hold very specific regulatory potential ( Figure 2A ).

To further explore the effect of HMOs on macrophages in response to S. aureus, the gene expression of several phenotypic macrophage-related surface molecules was investigated. From transcriptomic profiling, we found that all three HMO structures altered several macrophage-related surface proteins in response to S. aureus. Specifically, macrophages exposed to 6’SL had increased transcripts for the costimulatory activation marker CD137L ( Supplementary Figure 5A ), costimulatory molecule CD80 ( Supplementary Figure 5B ) and adhesion molecule CD18 ( Supplementary Figure 5C ). Concurrently, a decrease in the M2-markers CD200R and CD163 ( Supplementary Figures 5D, F ) was observed. The different responses are linked to the presence of S. aureus and are not notably affected by macrophage exposure to 6’SL alone. This indicates a boosted immune response against S. aureus.

To examine if the altered gene expression correlated with subsequent surface expression, human blood-derived monocytes were exposed to combinations of M-CSF, HMOs and S. aureus by flow cytometry. Data is gated on viable cells ( Supplementary Figure 2 ). Isotype controls were included in all surface marker studies but showed no unspecific binding ( Supplementary Figure 3 ). Moreover, cells were CD14 positive, indicating myeloid origin ( Supplementary Figure 6 ). Surface molecule expression results confirmed that 6’SL is the most potent of the three HMO structures evaluated. Following S. aureus challenge 6’SL increased cell surface expression of the activating molecule CD137L ( Figure 3A ; Supplementary Figure 7A ), the co-stimulatory molecule CD80 ( Figure 3B ; Supplementary Figure 7B ) and adhesion molecule CD18 ( Figure 3C ; Supplementary Figure 7C ), and at the same time decreased the anti-inflammatory and myeloid suppressive surface molecules CD200R, PD-L1 and CD163 in combination with S. aureus ( Figures 3D–F ; Supplementary Figures 7D–F ). Interestingly, CD169 (Siglec-1) also decreased in the presence of HMOs and S. aureus ( Figure 3G ; Supplementary Figure 7G ). Expression of additional classical myeloid cell surface molecules were analyzed but showed no difference ( Supplementary Figure 8 ).

These findings support the results from the gene array analysis and illustrates that most proteins regulated by HMOs on a transcriptional level are subsequently expressed on the cell surface. This is true for most of the displayed surface molecules. As an exception, S. aureus-induced gene expression of PD-L1 was potentiated by 6’SL and 2’FL, but did not translate into a higher surface expression. Contrarily, S. aureus-induced surface expression of PD-L1 was lower upon addition of HMOs ( Figure 3E ; Supplementary Figure 7E ).

Cytokine-genes were also potently regulated by HMOs ( Supplementary Figure 9 ). Again, 6’SL was the most potent of the three tested HMOs, although 2´FL and LNnT also impacted cytokine gene transcription ( Supplementary Figures 9A, B, D, E, G ). The data show that macrophages treated with 6’SL before S. aureus challenge stimulate an increased production of genes encoding the pro-inflammatory cytokines: TNF-α, IL-6, IL-8, IFN-γ and IL-1β ( Supplementary Figures 9A, D, E, F, G ), and simultaneously decreased gene expression of the anti-inflammatory cytokine IL-10 ( Supplementary Figure 9B ). As with the expression of surface proteins, the HMO effect on the expression of pro-inflammatory cytokines is dependent on the presence of S. aureus bacteria for the majority of the tested cytokines.

In order to further investigate HMO-regulated secretion of cytokines, monocyte-derived macrophages were stimulated as described above to collect supernatants for cytokine measurements. The cytokine levels were measured by mesoscale multiplex analysis, and cell numbers were evaluated by flow cytometer-mediated counting. To ensure that the macrophage detachment method did not affect the read-out, two different methods were tested: Trypsination and citric saline (54) combined with cell scraping. As shown in Supplementary Figure 10 , there was no difference between the methods, and the citric saline combined with cell scraping method was applied to enable subsequent flow cytometry analysis of the cells. In parallel with the observation from the transcriptomic profiling, HMO treatment (in particular 6’SL) followed by S. aureus challenge increased the pro-inflammatory cytokines TNF-α, IL-6, IL-8, IFN-γ and IL-1β ( Figures 4A, D–G ). No change in the regulation of the anti-inflammatory cytokine IL-10 was observed ( Figure 4B ), whereas IL-4 increased relative to the untreated control ( Figure 4C ). In summary, these observations are aligned with the results from the gene array and cell surface analysis, indicating that HMOs can boost the inflammatory response against the S. aureus pathogen without affecting unexposed macrophages.

Next, we investigated the potential of HMOs to activate the NF-κB-pathway. The NF-κB-pathway is a common pathway activated during infection and inflammation (55). We used a monocytic THP1-reporter cell line that expresses GFP in response to NF-κB activation, and it is therefore comprehensive to monitor myeloid cell activation by flow cytometry. As shown in Figure 5A , all tested HMOs increased the inflammatory NF-κB-response against S. aureus. Furthermore, the HMOs did not induce NF-κB-activity in macrophages without exposure to bacteria, highlighting that the effect is linked to the presence of pathogens ( Figure 5A ). Since HMO-mediated NF-κB activation was only observed together with S. aureus exposure, and as NF-κB can drive pro-inflammatory responses in macrophages (55), this could likely be a causative factor in the regulation of macrophages. Evaluation of data from microarray assay revealed that multiple genes relevant in the NF-κB-pathway were upregulated in macrophages treated with 6’SL and challenged with WT MRSA ( Figure 5B ). More specifically, there was an upregulation of the genes NFKBIE, NFKBIA, NFKB1, RELA, CHUK and IKBKG with 6’SL compared to untreated control. REL and IKBKB showed no difference in response to 6’SL. These upregulations were not seen in absence of the WT MRSA bacteria nor with the other HMOs (data not shown), pointing to a specific effect of 6’SL against pathogens.

Most manufactured HMOs are produced by a fermentation process (56). To rule out that byproducts from the manufacturing process could impact or stimulate the immune cells, we did a comparison between HMOs produced by fermentation and HMOs produced chemically. Both fermented and chemically produced HMOs caused NF-κB activation, and there was no difference in activation between fermented and chemically produced HMOs ( Figure 6 ). Moreover, an endotoxin analysis was performed on the fermented HMOs to rule out presence of endotoxins. The endotoxin test could not detect endotoxins above the assay threshold level ( Table 1 ), clearly suggesting that the effect of HMOs is not caused by endotoxins. The studies in this article were carried out with the commercially available fermented HMOs.

Microscopic inspection showed that HMOs affected growth activation patterns. Therefore, proliferation of macrophages was evaluated by CFSE-labelling and culturing with combinations of M-CSF, HMOs and S. aureus. On day 8, the macrophages were analyzed by flow cytometry. As seen in Supplementary Figures 11A, B , cells receiving HMOs, most significant 6’SL, proliferated more compared to untreated controls, shown as a decrease in CFSE-MFI value, compared to untreated control ( Supplementary Figure 11B ). Microscopic evaluation further corroborated that 6’SL treatment affected macrophage differentiation and morphology by making them more adherent and elongated ( Supplementary Figure 12 ). Together this indicates an altered activation profile for macrophages exposed to 6’SL compared with the untreated control.

Having established that the HMOs can push macrophages towards an M1-like phenotype, the ability of HMOs to affect bacterial uptake was evaluated in U937 monocytes, using fluorescent-labelled S. aureus ( Supplementary Figure 4 ). Each HMO increased the uptake of S. aureus, suggesting that the phagocytic potential is augmented by HMOs ( Supplementary Figures 13A, B ). We further analyzed the phagocytic uptake by primary macrophages cultured with HMOs ( Figures 7A, B ; Supplementary Figure 13C ). Here 6’SL significantly enhanced the phagocytosis of the S. aureus bioparticles compared to the 2’FL, LNnT and the untreated control. Together, this indicates that HMOs, and in particular 6’SL, have the potential to increase both the inflammatory and phagocytic response against a bacterial infection, in this case S. aureus.

We further investigated the effect of HMOs on macrophages against a more antibiotic-resistant S. aureus strain. The SNP MRSA strain has clpP and rpoB SNP-mutations, related to increased antibiotic-resistance. The HMOs, especially 6’SL, increased the pro-inflammatory phenotype against the more antibiotic-resistant SNP MRSA, but to a lower extent than that observed for WT MRSA. This include increased gene expression of the adhesion molecule CD18 ( Supplementary Figure 5C ), and decreased gene expression of CD200R, CD163 and D169 ( Supplementary Figures 5D, F, G ). HMOs also increased gene expression of pro-inflammatory cytokine genes, including TNF, IL-6, IL-8 and IL-1β ( Supplementary Figures 9A, D, E, G ), and decreased the gene expression of IL-10 ( Supplementary Figure 9B ). When evaluating the effect of HMOs on surface expression of several proteins, 6’SL showed similar effect when co-cultured with SNP MRSA as when co-cultured with WT MRSA. This includes an increase in activating molecule CD137L and adhesion molecule CD18 ( Figures 3A, C ), and a decrease in the anti-inflammatory surface molecules CD200R and PD-L1 ( Figures 3D, E ), as well as CD169 ( Figure 3G ). Regulation of the co-stimulatory molecule CD80 and the M2-marker CD163 was not affected by SNP MRSA exposure, suggesting that the more resistant SNP MRSA has a potential to evade the immune system, as previously shown (43). However, HMOs can indeed activate macrophages against the resistant S. aureus strain. In agreement with this, the response of pro-inflammatory cytokines IL-6, IL-8 and IL-1β ( Figures 4D, E, G ) was enhanced by co-cultivating SNP MRSA with 6’SL. The macrophage proliferation was also higher in response to SNP MRSA, when co-culturing with HMOs ( Supplementary Figure 11 ). In addition, a slight increase in SNP MRSA uptake was observed in U937 monocytes pretreated with 2’FL ( Supplementary Figures 13A, B ). However, the HMOs has no effect on the NF-κB pathway in SNP MRSA-exposed macrophages ( Figure 5 ). These results indicate that HMOs, have the potential to enhance the immune response against S. aureus (WT MRSA), without causing aberrant activation of naïve cells. To a lower extent, a similar effect is observed for S. aureus with SNPs related to antibiotic resistance (SNP MRSA).