METTL9-catalyzed histidine methylation of S100A9 suppresses the anti-Staphylococcus aureus activity of neutrophils

Graphical abstract

Dear Editor,

Protein histidine methylation has been found to be a widespread post-translational modification (PTM) in mammalian cells and represents more than 13% of all protein methylation events in the human methylome (Ning et al., 2016; Kapell and Jakobsson, 2021). Recently, we and other groups independently identified METTL9 as the first broad-specificity histidine N1 methyltransferase in mammals (Davydova et al., 2021; Lv et al., 2021). METTL9 methylates a variety of substrates with a common x-His-x-His motif (His is for histidine; the second x is preferably small amino acids: A/C/G/S) and catalyzes methyl transfer to the last histidine. Although METTL9-mediated histidine N1 methylation is pervasive in mammalian cells, its pathophysiological functions are undescribed. None of the in vivo phenotypes of Mettl9^−/− mice have been reported thus far (Kwiatkowski and Drozak, 2020; Davydova et al., 2021; Jakobsson, 2021).

Alarmin S100A9, which heterodimerizes with S100A8, is constitutively expressed in neutrophils and monocytes. Once released upon infection, the S100A8/A9 complex exhibits broad-spectrum antimicrobial activity against numerous microorganisms (Roth et al., 2001) by sequestering essential trace metals such as Zn²⁺ and Mn²⁺, which are required for bacterial growth (Wang et al., 2018). S100A9 was found to carry histidine methylation at the His107 residue in previous mass spectrum data (Raftery et al., 1996). We and other groups have previously reported that METTL9 methylates S100A9 at the His107 residue (Daitoku et al., 2021; Davydova et al., 2021; Lv et al., 2021). However, whether METTL9-catalyzed His107 methylation of S100A9 has bona fide physiological functions remains unknown. Consistent with previous works, we confirmed by in vitro methylation assays that METTL9 methylates S100A9 at residue His107; no methylation was detected upon mutating S100A9’s His107 residue to Gly (Fig. S1A).

To better detect His107 methylation of S100A9 in vivo, we generated site-specific polyclonal antibodies to detect His107 methylation. We tested the specificity of these antibodies by incubating recombinant GST-tagged S100A9 with or without recombinant METTL9 in the presence of S-adenosylmethionine (SAM), followed by immunoblotting with these antibodies. Methylation of recombinant GST-tagged S100A9 was detected only in the presence of METTL9, suggesting the good quality of our antibodies (Fig. S1B). To confirm that METTL9 methylates S100A9 in vivo, we generated a Mettl9-deficient (Mettl9^−/−) mouse strain [see Fig. S1C–E for the gene knockout (KO) strategy and confirmation of gene knockout expression] and further harvested WT and Mettl9^−/− bone marrow (BM) cells. Immunoblotting showed that in WT cells that expressed METTL9, S100A9 His107 was methylated; in contrast, almost no His107 methylation was detected in Mettl9^−/− cells (Fig. 1A). We next assessed the methylation status of secreted S100A9, as S100A9 is secreted upon neutrophil activation (Wang et al., 2018). Immunoblotting of the culture medium samples from PMA-stimulated mouse primary neutrophils showed that secreted S100A9 can also be methylated (Fig. 1B).

Using site-specific antibodies, we evaluated whether His107 methylation of S100A9 was relatively static or dynamic during bacterial infection. Due to our failure to obtain effective METTL9 antibodies, we engineered a Mettl9 FLAG knock-in (KI) mouse strain wherein a FLAG tag was inserted into the endogenous Mettl9 locus (see Fig. S2A–C for the gene KI strategy and confirmation of gene KI expression). We next isolated peritoneal neutrophils from Mettl9 FLAG KI mice and treated the cells with S. aureus for one to 6 h, followed by immunoblotting with the indicated antibodies. Although the protein levels of FLAG-fused METTL9 remained similar, the His107 methylation of S100A9 was rapidly reduced in neutrophils upon S. aureus infection (Fig. 1C).

To next assess whether His107 methylation of S100A9 is also reduced in vivo during bacterial infection, we utilized a murine model of S. aureus skin infection, in which neutrophils play a primary role in the innate immune response that controls S. aureus infection by forming an abscess to wall off the infection and facilitate bacterial clearance (Kobayashi et al., 2015). Specifically, we induced S. aureus skin infection in WT mice, euthanized the animals at different time points after infection, and monitored the disease states by measuring the lesion area, spleen size, local bacterial CFUs, and recruitment of innate immune cells into infected sites (Fig. 1D). The subcutaneous injection of S. aureus resulted in the formation of a measurable lesion starting on days 1–2, with a maximum size achieved on ~day 2, followed by resolution of infection and a decrease in the lesion area (Fig. 1E). The bacterial burden in the abscess was high at the acute stage (days 0–2) and then gradually decreased to a low level at the resolving stage (days 8–10) (Fig. 1F). Consistently, neutrophils/monocytes were rapidly recruited into infected sites after onset and maintained a level of approximately 95% of total immune cells throughout the infection (Fig. S3A–C).

To examine the protein levels of His107 methylated and total S100A9, we harvested local abscess-infiltrating cells, BM cells and spleen cells at different time points during S. aureus infection. Given that it was not feasible to obtain abscess tissue at day 0, we mainly evaluated His107 methylation of S100A9 at the acute stage (day 2) and the resolving stage (days 8–10) of infection. In abscess-infiltrating cells, His107 methylation was significantly reduced (~5-fold) at day 2 compared with days 8–10 (Fig. 1G and 1H). Consistently, His107 methylation in abscesses was significantly lower than that in BM at day 2, when the bacterial CFU was high, while His107 methylation in abscesses increased by ~4-fold and exhibited no appreciable difference from that in BM at day 10, when most of the bacteria had been cleared (Fig. 1I and 1J). Together, these data suggest that His107 methylation of S100A9 is significantly decreased at local infected sites during the acute stage of S. aureus infection.

We next asked how His107 methylation affects the antibacterial activity of S100A9. Human S100A9 His105 residue, as well as mouse His107 residue, are structurally characterized as a zinc or manganese-binding site (Damo et al., 2013). From structural perspectives, His107 methylation, as well as its H107G mutation, would both lead to decreased zinc or manganese-binding compared to His107 (Fig. 1K). Indeed, His107 methylation of an S100A9 peptide (16 amino acids) showed reduced zinc-binding activity (Daitoku et al., 2021).

To examine whether a recombinant full-length methylated S100A9 (S100A9me) protein had reduced zinc-binding, we co-transformed the S100a9 plasmid with and without the Mettl9 plasmid into Escherichia coli bacteria and purified the S100A9 protein (see Fig. S4A–C and Methods for detailed protocols of protein purification). Immunoblotting and mass spectrometry showed that the purified S100A9 protein was methylated at the His107 residue when the S100a9 plasmid was co-transformed with the Mettl9 plasmid; no His107 methylation was detected when the S100a9 plasmid was transformed alone (Fig. S4B and S4C). Using an ITC-binding assay, we confirmed that full-length methylated S100A9 had significantly reduced zinc-binding activity compared to its unmethylated control (K_ds: 90.9 µmol/L vs. 14.3 µmol/L) (Fig. S4D).

Given that the zinc-binding site is important for the antimicrobial function of S100A8/A9 (Corbin et al., 2008), we next assessed whether His107 methylation affects the anti-S. aureus activity of S100A8/A9 in vitro. Using similar methods as above, we purified S100A8/A9 with and without His107 methylation (Fig. S4E). We also constructed an S100a9 mutant plasmid with its His107 residue mutated to Gly (H107G) and purified the recombinant S100A8/A9 H107G complex. Next, we incubated the recombinant S100A8/A9 protein with S. aureus and monitored the growth of S. aureus by measuring the optical density (OD_{600 nm}) (see Methods for a detailed protocol). As a positive control, unmethylated S100A8/A9 protein exhibited effective anti-S. aureus activities in a dose-dependent manner (Fig. 1L), indicating that our purified S100A8/A9 protein was functional and that the in vitro anti-S. aureus assay was successful. Using this assay, we found that both methylated S100A8/A9 and H107G-mutated S100A8/A9 significantly reduced their anti-S. aureus functions compared to unmethylated S100A8/A9 (Fig. 1M).

We further explored whether His107 methylation reduces the anti-S. aureus function of S100A8/A9 in vivo. Specifically, we subcutaneously administered 30 µg of different forms of recombinant S100A8/A9 proteins, together with 2 × 10⁸ S. aureus, on the backs of WT mice and monitored the skin lesion area daily. We observed that WT mice administered unmethylated S100A8/A9 had significantly reduced local lesion areas and bacterial CFU loads in the abscess tissue compared to those of WT mice administered methylated S100A8/A9 and H107G mutated S100A8/A9 (Fig. 1N–P), indicating that His107 methylation indeed reduces anti-S. aureus function of S100A8/A9 in vivo.

To further investigate the physiological function of METTL9-catalyzed His107 methylation of S100A9 in vivo, we took advantage of Mettl9^−/− mice, in which His107 methylation of S100A9 was completely blocked (Fig. 1A). At steady state, Mettl9^−/− mice were healthy and had no obvious weight abnormalities. Six- to eight-week-old WT and Mettl9^−/− mice had similar numbers and percentages of major cell populations, including neutrophils, monocytes, and macrophages, in their BM (Fig. S1F and S1G).

Given that His107 methylation reduces anti-S. aureus function of S100A8/A9, we postulated that Mettl9 KO mice might have increased antibacterial activity against S. aureus infection. Specifically, we inoculated WT and Mettl9^−/− mice subcutaneously on the back with 3 × 10⁸ S. aureus, monitored the disease state daily by measuring the skin lesion area, and euthanized the mice 5 days after infection. Compared to WT control mice, Mettl9^−/− mice exhibited a significantly decreased lesion area and S. aureus burden in the abscess tissue (Fig. 2A–C). H&E staining of abscess tissue showed that Mettl9^−/− mice had reduced skin thickness and local inflammation (Fig. 2D). These data suggested that Mettl9^−/− mice indeed had an increased anti-S. aureus phenotype. Although the recruitment of neutrophils in abscesses showed no significant differences between WT and Mettl9^−/− mice (Fig. S5A), the antimicrobial activity of neutrophils might increase in Mettl9^−/− mice.

Neutrophils use multiple antimicrobial mechanisms when engaging S. aureus. In addition to S100A8/A9-mediated metal chelation (Corbin et al., 2008), neutrophils can phagocytose microbes and are activated to produce reactive oxygen species (ROS) (DeLeo et al., 1999). Neutrophils can also release neutrophil extracellular traps (NETs) to kill microbes (DeLeo and Allen, 2020). We observed that upon activation with phorbol 12-myristate 13-acetate (PMA), WT and Mettl9^−/− neutrophils induced similar levels of ROS (Fig. S5B) and NETs (Fig. S5C and S5D), suggesting that these two mechanisms might not be the major contributors to the anti-S. aureus phenotype observed in Mettl9^−/− mice.

We next evaluated whether neutrophil lysates from Mettl9^−/− mice have increased anti-S. aureus activity. Previous studies demonstrated that S100A8/A9 does not contribute to neutrophil phagocytic killing, but neutrophil cytoplasmic extracts containing S100A8/A9 inhibit bacterial growth through metal chelation (Roth et al., 2001; Corbin et al., 2008). Briefly, we extracted cell lysates from peritoneal neutrophils, incubated the lysates with S. aureus in TSB medium, and monitored bacterial growth by measuring the optical density (OD_{600 nm}) for 8 h. Since TSB medium is rich in zinc, we serially diluted the zinc chelator N,N,Nʹ,Nʹ-tetrakis (2-pyridylmethyl) ethylenediamine (TPEN) to set up a low-zinc culture condition wherein S. aureus still grew normally (Fig. S6A). Under the optimal TPEN concentration, as the amounts of neutrophil lysates increased, their inhibition of S. aureus growth increased and exhibited a dose-dependent effect (Fig. S6B–I), indicating that the lysate-based anti-S. aureus assay was successful. Using this assay, we determined that the lysates from Mettl9^−/− neutrophils indeed exhibited significantly increased anti-S. aureus ability compared to that from WT neutrophils, while adding excessive zinc into the lysates led to a rescue of anti-S. aureus activity (Fig. 2E), confirming that Mettl9^−/− neutrophil lysates increase the anti-S. aureus activity in a zinc-dependent antibacterial mechanism.

We next sought to determine whether the increased anti-S. aureus phenotype in Mettl9^−/− mice was attributed to the loss of S100A9 His107 methylation. We reasoned that if the His107 residue was functionally important for the increased antibacterial activity of Mettl9^−/− mice, we would expect that mutation of this residue could lead to a rescue of antibacterial activity. We tested this possibility by producing an S100a9^H107G point-mutated mouse strain (see Fig. S7A–C for the CRISPR/Cas9 point-mutation KI strategy and confirmation of point-mutation expression) and further obtained a Mettl9^−/− S100a9^H107G double-deficient mouse strain.

We first assessed the anti-S. aureus activity of neutrophil lysates from WT, Mettl9^−/−, S100a9^H107G and Mettl9^−/− S100a9^H107G mice. Similar to the aforementioned phenotypes, the lysates from Mettl9^−/− neutrophils exhibited significantly increased anti-S. aureus ability compared to that of WT neutrophils. Lysates from Mettl9^−/− S100a9^H107G mice showed no appreciable difference from those from S100a9^H107G mice, suggesting that mutation of the His107 residue led to a significant rescue of the antibacterial activity observed in Mettl9^−/− neutrophil lysates (Fig. 2F).

We further assessed the anti-S. aureus phenotype of WT, Mettl9^−/−, S100a9^H107G and Mettl9^−/− S100a9^H107G mice in vivo. Specifically, we inoculated these mice subcutaneously on the back with 3 × 10⁸ S. aureus, monitored the skin lesions daily and euthanized the mice three days after infection. Similarly, Mettl9^−/− mice exhibited a significantly decreased lesion area and S. aureus burden in the abscess tissue compared to WT controls (Fig. 2G). In contrast, no significant changes were observed when comparing the lesion sizes and CFU load in Mettl9^−/− S100a9^H107G mice to those in S100a9^H107G mice (Fig. 2H–J). Notably, although S100a9^H107G neutrophil lysates had a significantly increased anti-S. aureus activity compared to WT lysates in vitro (Fig. 2F), S100a9^H107G mice did not show an obvious antimicrobial phenotype compared to WT mice (refer to the Discussion). Nevertheless, these data demonstrate that the S100a9^H107G mutation largely abolished the increased anti-S. aureus function observed upon deletion of Mettl9, suggesting that METTL9 controls S. aureus infection via the S100A9 His107 axis in vivo.

In the PTM field, the dynamic and spatiotemporal study of protein PTMs is important to fully understand their complex roles in physiology and disease. As a rarely studied modification, our data reveal that protein histidine methylation could be dynamic (Fig. 2K): (i) The METTL9-S100A9 His107 methylation axis suppresses the antibacterial activity of S100A8/A9 at steady state; (ii) Upon S. aureus infection, the ratio of methylated/total S100A9 is markedly reduced at the acute stage, which endows S100A8/A9 with increased zinc-binding and zinc-dependent anti-S. aureus activity; (iii) The ratio is gradually increased at the resolving stage to act as a brake on the host’s anti-S. aureus activity.

In summary, our work reports the profound physiological significance of the METTL9-S100A9 His107me axis in the regulation of the host’s antibacterial responses and deepens our current understanding of the chemical, enzymological, and physiological aspects of protein histidine methylation beyond protein lysine/arginine methylation. This work will open up new perspectives for future research on the importance of histidine methylation of proteins in eukaryotes.

References

[1] Corbin BD, Seeley EH, Raab A et al. Metal chelation and inhibition of bacterial growth in tissue abscesses. Science 2008;319:962-965.

[2] Daitoku H, Someya M, Kako K et al. siRNA screening identifies METTL9 as a histidine Nπ-methyltransferase that targets the proinflammatory protein S100A9. J Biol Chem 2021;297:101230.

[3] Damo SM, Kehl-Fie TE, Sugitani N et al. Molecular basis for manganese sequestration by calprotectin and roles in the innate immune response to invading bacterial pathogens. Proc Natl Acad Sci USA 2013;110:3841-3846.

[4] Davydova E, Shimazu T, Schuhmacher MK et al. The methyltransferase METTL9 mediates pervasive 1-methylhistidine modification in mammalian proteomes. Nat Commun 2021;12:891.

[5] DeLeo FR, Allen LH. Phagocytosis and neutrophil extracellular traps. Fac Rev 2020;9:25.

[6] DeLeo FR, Allen LA, Apicella M et al. NADPH oxidase activation and assembly during phagocytosis. J Immunol 1999;163:6732-6740.

[7] Jakobsson ME. Enzymology and significance of protein histidine methylation. J Biol Chem 2021;297:101130.

[8] Kapell S, Jakobsson ME. Large-scale identification of protein histidine methylation in human cells. NAR Genom Bioinform 2021;3:lqab045.

[9] Kobayashi SD, Malachowa N, DeLeo FR. Pathogenesis of Staphylococcus aureus abscesses. Am J Pathol 2015;185:1518-1527.

[10] Kwiatkowski S, Drozak J. Protein histidine methylation. Curr Protein Pept Sci 2020;21:675-689.

[11] Lv MY, Cao D, Zhang L et al. METTL9 mediated N1-histidine methylation of zinc transporters is required for tumor growth. Protein Cell 2021;12:965-970.

[12] Ning Z, Star AT, Mierzwa A et al. A charge-suppressing strategy for probing protein methylation. Chem Commun 2016;52:5474-5477.

[13] Raftery MJ, Harrison CA, Alewood P et al. Isolation of the murine S100 protein MRP14 (14 kDa migration-inhibitory-factor-related protein) from activated spleen cells: Characterization of post-translational modifications and zinc binding. Biochem J 1996;316:285-293.

[14] Roth J, Goebeler M, Sorg C. S100A8 and S100A9 in inflammatory diseases. Lancet 2001;357:1041.

[15] Wang S, Song R, Wang Z et al. S100A8/A9 in inflammation. Front Immunol 2018;9:1298.

RIGHTS & PERMISSIONS

©The Author(s) 2023. Published by Oxford University Press on behalf of Higher Education Press.