Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Oct 25.
Published in final edited form as: J Am Chem Soc. 2026 Sep 16;148(36):38641–38653. doi: 10.1021/jacs.6c04584

Native Proteomics Reveals Protease-Mediated Hemophore Release by Bacterial Pathogens

Andrew K Goring 1, Lindsey R Lyman 4, Jordan Ford 1, Jingjing Huang 5, Christopher Mullen 5, Rafael D Melani 5, Michael P Schmitt 4, Robert T Clubb 1,2,3,*, Joseph A Loo 1,2,3,*
PMCID: PMC13694841  NIHMSID: NIHMS2214371  PMID: 42915530

Abstract

Iron is an essential micronutrient for nearly all forms of life, including pathogenic microbes that must acquire it from their host during infection. At the host–pathogen interface, humans restrict microbial access to iron through nutritional immunity, which many pathogens overcome by secreting hemophores that scavenge extracellular heme (iron protoporphyrin IX). However, identifying hemophores and other ligand-binding proteoforms in complex proteomes remains challenging using conventional peptide-based bottom-up mass spectrometry (MS). Here, we introduce ProteoMIX (Proteome Analysis by Mixing), a function-based native top-down proteomics workflow that combines slow-mixing mode native MS with charge reduction to identify ligand-binding proteoforms directly from complex mixtures. Applying ProteoMIX to the Corynebacterium diphtheriae exoproteome identified ChtA30-314, an abundant soluble hemophore generated by proteolytic processing of the surface-exposed ChtA heme receptor. Using native top-down MS sequencing, cell fractionation, and gene deletion, we show that the protease DIP2069 releases ChtA30-314 by removing ChtA’s transmembrane helix, producing a soluble proteoform that delivers heme to support microbial growth. In contrast, the related paralog ChtC is not processed, enabling C. diphtheriae to generate localization-specific heme-binding proteoforms from related gene products. Extending this approach to Staphylococcus aureus, ProteoMIX also revealed soluble IsdA hemophores that coexist with surface-anchored variants, demonstrating the generality of the method and suggesting that protease-mediated hemophore release may operate across gram-positive pathogens.

Keywords: Bacterial pathogenesis, Conserved Region (CR) domain, NEAr iron Transporter (NEAT) domain, native mass spectrometry, hemophore, heme, iron, Proteome Analysis by Mixing (ProteoMIX), Corynebacterium diphtheriae, Staphylococcus aureus

Graphical Abstract

ProteoMIX pairs SLOMO native MS with DIA charge-reduction to find ligand-binding proteoforms (heme from hemoglobin) in unfractionated exoproteomes, revealing protease-mediated hemophore release.

graphic file with name nihms-2214371-f0008.webp

INTRODUCTION

Iron is an essential micronutrient for virtually all forms of life, including pathogenic bacteria that must obtain it from their host during infection.1 At the host–pathogen interface, there is a battle for iron, with humans restricting microbial access to this metal through nutritional immunity mechanisms that pathogens attempt to overcome using a variety of iron-scavenging strategies.2–4 To capture iron in their surroundings, most pathogens secrete small-molecule siderophores that chelate iron and deliver it back to the cell to support growth.5,6 Pathogens also secrete hemophores that bind heme (iron-protoporphyrin IX) primarily derived from hemoglobin (Hb), which is a major reservoir for this nutrient as it contains ~70% of the body’s total iron.2,7–10 While nearly all pathogens secrete siderophores, hemophores have been primarily discovered in gram-negative bacteria and appear to be rare in gram-positive pathogens.5,7,11–19 Despite their functional importance, identifying bacterial hemophores and elucidating the mechanisms through which they are released from the bacterium can be challenging. For example, L. monocytogenes Hbp2 is reported to be largely secreted with only a minor peptidoglycan-associated fraction, consistent with a hemophore-like role that complicates simple ‘cell-associated vs secreted’ classification.20 These challenges arise because bacterial genomes often encode several heme-binding proteins that, based on their sequences, are expected to remain cell associated.21–25 Moreover, their experimental detection can be cumbersome, requiring cell fractionation and the generation of appropriate antibodies to detect them by western blotting.26–29 Bacteria can also secrete a range of other ligand-binding proteins, which can be challenging to detect using conventional methods (e.g., proteins that bind siderophores, metals, and other small molecules).30,31

Corynebacterium diphtheriae is the causative agent of diphtheria, a severe upper respiratory tract infection.32 To thrive in the iron-limited environment of its host, it produces a series of heme-binding proteins (HtaA, ChtA, ChtC, HtaB, and ChtB) that scavenge iron-laden heme and import it into the cell.33–36 Each protein binds heme via a Conserved Region (CR) domain and contains an N-terminal signal peptide for secretion through the Sec translocon and a C-terminal transmembrane (TM) helix that presumably embeds the proteins in the plasma membrane for display on the microbial surface. The microbe also displays a receptor (HbpA) that tethers Hb to the cell surface by contacting its α-globin.22,37,38 Surprisingly, peptide-based bottom-up proteomics and western blot data indicate that the heme-binding ChtA and HtaA proteins are released from the cell into the exoproteome, despite the fact that their full-length forms contain a TM helix for membrane association.29,38

Commonly employed MS methods are generally poorly suited for rapidly identifying specific ligand-binding proteins and their proteoforms within whole proteomes. For example, even though LC-MS/MS proteomics methods that measure proteolytically generated peptide fragments enable facile protein identification and quantification of their relative abundance, proteoform identification can be challenging compared to top-down proteomics.39 Moreover, native top-down MS (nTD-MS) methods can identify both proteoforms and their ligand-bound complexes, such as heme-bound hemophores. In this study, we present an nTD-MS approach that can rapidly identify ligand-binding proteins and their proteoforms within complex proteome mixtures without the need for extensive separation and fractionation. The method we call ProteoMIX (proteome analysis by mixing) obtains large gains in speed and efficiency by combining slow-mixing mode native MS (SLOMO-nMS)40 to monitor ligand binding with data-independent acquisition (DIA)-charge reduction to simplify the mass spectra.41 While SLOMO-nMS and DIA-charge reduction have been described previously, integrating them into a single function-based workflow that identifies ligand-binding proteoforms directly within an unfractionated proteome has not been reported, and this constitutes the central methodological advance of ProteoMIX. We demonstrate the utility of this method by discovering novel soluble, truncated proteoforms of ChtA and IsdA acting as hemophores in pathogenic C. diphtheriae and S. aureus that are released by proteolytic processing of surface-associated heme-binding proteins. Using ProteoMIX in combination with genetic mutations, cellular fractionation, western blot and peptide-based proteomics, we define the mechanism of soluble hemophore release in C. diphtheriae: the DIP2069 protease cleaves ChtA, removing its transmembrane helix and generating a soluble hemophore capable of scavenging heme from human Hb. Together, these results establish ProteoMIX as a function-based native proteomics workflow that identifies ligand-binding proteoforms directly within unfractionated proteomes, and they uncover a previously unrecognized mechanism of hemophore release in gram-positive bacteria.

RESULTS

Native mass spectrometry enables the facile detection of hemophores within proteomes.

Previously, we used bottom-up proteomics to identify 435 unique proteins within the C. diphtheriae exoproteome under iron-limiting conditions.29 Identifying which of these proteins are capable of binding heme from hemoglobin (Hb), however, is not readily achieved by peptide-based methods and requires new approaches. In principle, native TD-MS provides a direct means to detect hemophores and other ligand-binding proteins; however, in complex proteomic samples the resulting spectra are highly congested with overlapping charge state envelopes that obscure proteoform assignments.42 To expedite the identification of functionally active hemophores and their proteoforms, we developed the ProteoMIX method and demonstrated its utility by applying it to the exoproteome of C. diphtheriae. The ProteoMIX workflow is outlined in Figure 1, with a detailed step-by-step schematic and a table reporting the instrument, software, and approximate time for each step provided in the Supporting information (Fig. S8, Table S3). ProteoMIX uses a two-step process to identify ligand-binding proteins, which in this application are hemophores that scavenge heme from human hemoglobin (Hb) (Fig. 1A). In the first step, a plug of native exoproteome is loaded into an electrospray ionization (ESI) emitter tip followed by the addition of exoproteome that has been pre-incubated with Hb (Fig. 1B). This establishes slow-mixing mode native MS (SLOMO-nMS)40 between the exoproteome and analyte. A stable native electrospray is then established to generate a congested mass-to-charge (m/z) spectrum of the exoproteome whose intact masses are uninterpretable due to their convolved mass-to-charge ratios (Fig. 1C, left). Next, narrow m/z windows are isolated and charge-reduced in a data-independent acquisition (DIA) scheme by either proton transfer charge reduction (PTCR43) on an Orbitrap Ascend or electron capture charge reduction (ECCR44) on a Q-Exactive UHMR modified with an e-MSion ExD cell (Fig. 1C, middle). One cycle of DIA-charge reduction, requiring approximately 1 to 5 minutes, produces a snapshot of proteoform masses and relative intensities from the native ESI mass spectrum (Fig. 1C, right). Notably, PTCR on 175 consecutive isolation windows from 2,000 to 6,000 m/z shows that 371 native proteoforms can be confidently detected from a single direct infusion (Fig. S1). DIA-charge reduction cycles are repeated as Hb diffuses from the back of the emitter tip to the front, and heme-binding proteoforms are identified from the reciprocal intensity changes that occur during mixing, as they pick up heme (Fig. 1D). A proteoform that binds heme loses intensity as its heme-bound form, greater in mass by 616 Da (the mass of heme), gains intensity (Fig. 1D). These paired shifts are detected automatically using UniDec and custom Python scripts, and can be inspected manually from the ProteoMIX dataset (see Experimental Methods). Once candidate hemophore masses are identified, their most abundant charge states are isolated and fragmented by higher-energy collisional dissociation (HCD) on the UHMR for native top-down sequencing (Fig. 1E). Fragment ions are matched to candidate heme-binding protein sequences to assign each proteoform, either by manual interpretation or with established top-down search engines (see Experimental Methods). By using DIA-charge reduction coupled with SLOMO-nMS, proteomes containing hundreds of proteoforms (up to 371 in C. diphtheriae’s exoproteome in this example) can be screened for ligand-binding proteoforms in as little as one hour, and their most intense charge states can be prioritized for nTD-MS identification. This type of function-based prioritization of nTD-MS can reduce data collection time, as each proteoform identification requires non-trivial optimization in terms of data collection and analysis. Although we demonstrate the utility of ProteoMIX to identify heme-binding proteoforms, in principle it should be useful to study the binding of any ligand within a proteome that causes a measurable mass shift.

Figure 1.

Figure 1.

ProteoMIX-based approach for discovering ligand-binding proteoforms in unfractionated proteomes. (A) Schematic overview of the ProteoMIX workflow. Functionally active proteoforms in an unfractionated proteome are identified by DIA-charge reduction during SLOMO-nMS mixing, prioritized for native top-down MS (nTD-MS) based on ligand-binding function, and verified by nTD-MS. (B) Buffer-exchanged exoproteome is loaded into the front of the nanospray tip followed by a stack of hemoglobin (Hb)-incubated exoproteome and a stable native ESI-MS spray is established. (C) Proteoforms in the resulting congested mass-to-charge spectrum are then resolved using DIA-charge reduction on narrow isolation windows spanning a majority of the exoproteome (e.g. 2000-5500 m/z, with 50 m/z spacing, for a total of 70 isolation windows). The charge-reduced spectra are deconvolved and summed in UniDec,45 and the resulting masses are evaluated by deconvolution score to produce a snapshot of proteoform masses and their intensities shown on the right in blue. (D) Hemophores are identified from changes in native proteoform intensities as hemoglobin slowly mixes into the exoproteome. Hemophore proteoforms are verified by paired mass shifts of 616 Da (Δheme). (E) Heme-binding proteoforms are then isolated based on their most abundant precursor mass-to-charge peak and fragmented using HCD to identify the hemophore’s proteoform.

A truncated ChtA hemophore facilitates microbial growth by scavenging heme-iron.

ProteoMIX reveals that C. diphtheriae releases from the membrane three soluble proteoforms of the ChtA and HtaA proteins capable of scavenging heme: ChtA30-314, HtaA31-272, and HtaA273-555 (Fig. 2A). These provide an explanation for their previously observed high secretion levels in culture media under iron-limitation,29 which was unexpected due to the presence of a C-terminal transmembrane helix in their gene sequence. Each proteoform contains a heme-binding “conserved region” (CR) module, but lacks the C-terminal transmembrane (TM) helix that tethers the full-length membrane-anchored proteoform to the bacterium. An inspection of the N-termini in ChtA30-314 and HtaA31-272 reveals that they lack signal sequences, suggesting that the proteoforms originate from ChtA and HtaA proteins that have been secreted via the Sec translocon46. Heme capture by the highly abundant ChtA30-314 proteoform is readily observed from the 30,912 Da deconvolved mass gaining 616 Da (heme) during the mixing time-course (Fig. 2B). The full time-course tracking the intensities of apo- and holo-ChtA30-314 during mixing is shown in Figure S2A. nTD-MS reveals that the secreted ChtA30-314 hemophore contains an intramolecular disulfide bond between residues Cys179 and Cys188 (Fig. S2B), as previously observed in the crystal structure of ChtA’s CR domain (ChtACR).47 HtaA273-555 was also identified from a clearly resolved 616 Da mass shift and reciprocal intensity changes for its apo- and holo- heme-bound proteoforms at 28,981 Da and 29,597 Da, respectively (Fig. 2C, Fig. S2A). Additionally, reciprocal intensity changes for the apo- and holo- HtaA31-272 proteoforms were crucial for its identification, as its apo-mass overlaps with the mass for a secreted protease, DIP035030-274, and therefore could have been missed without SLOMO-nMS (Fig. 2D). HtaA proteoforms were also verified by nTD-MS (Fig. S2C, S2D). Collectively, the data illustrate the utility of DIA charge-reduction combined with SLOMO-nMS for detecting ligand-binding proteins in complex mixtures, here revealing truncated ChtA and HtaA hemophores within the exoproteome of C. diphtheriae that are capable of scavenging heme.

Figure 2.

Figure 2.

ProteoMIX identifies heme-binding proteins within C. diphtheriae’s exoproteome. (A) Full-length sequences with domains and sequence motif annotations for ChtA and HtaA (ChtA-FL and HtaA-FL, respectively). Hemophores identified using ProteoMIX (ChtA30-314, HtaA31-272, and HtaA273-555) are shown below the full-length sequences, and non-canonical termini are annotated below their illustration. SPI, signal peptide (cleaved by signal peptidase SpI following secretion through the Sec translocon); TMH, transmembrane helix; CR1 and CR2, first and second conserved region heme-binding domains numbered from the N-terminus; CTD, uncharacterized C-terminal β-propeller domain. (B) Select deconvolved mass spectra from DIA-charge reduction time points acquired using ProteoMIX show ChtA30-314 (blue) acquiring heme (Δ616 Da). The semi-apo Hb dimer (Hsα2-142β2-147 + heme) is also observed in red. (C) Analogous spectra show apo-HtaA273-555 binding heme (yellow; Δ616 Da). Trace amounts of human carbonic anhydrase (HsCA) from the Hb sample are also observed diffusing into the exoproteome. (D) Deconvolved mass spectra (left) and extracted ion intensities from the full ProteoMIX time course (right) show the conversion of apo- to holo-HtaA31-272; the horizontal axes report units in terms of DIA cycle number and total mixing time, while the y-axis reports extracted ion intensity; reciprocal intensity changes were crucial for identification due to overlap with DIP035030-274.

To determine the relative abundance of each ProteoMIX-detected hemophore, we performed bottom-up gel-enhanced LC-MS/MS proteomics (GeLC-MS/MS).48 The exoproteome was separated by SDS-PAGE and then sectioned into 17 consecutive slices that were analyzed by LC-MS/MS (Fig. 3A). A total of 239 proteins were identified by GeLC-MS/MS. Both full-length and truncated proteoforms of ChtA and HtaA are distinguishable and each exhibits a bimodal separation into low- and high-MW gel slices (Fig. 3A). When the molar abundance of all forms of each protein is summed, ChtA and HtaA are the 3rd and 4th most prevalent proteins in the exoproteome (at 8 ± 2% and 6 ± 1% molar abundance, respectively); the most abundant species is DIP2069, a peptidase S1 (clan PA) domain-containing protein (20 ± 8%), followed by the diphtheria toxin (10 ± 2%) (Fig. 3B). These data reveal that >99% of ChtA in the exoproteome exists in its ChtA30-314 truncated proteoform, as nearly all of its intensity is located in the low-MW fraction. In contrast, peptides originating from HtaA are distributed between low- and high-MW fractions, revealing that only ~30% of this protein exists as truncated HtaA31-272 and HtaA273-555 proteoforms, while the remainder is not resolved by nMS, consistent with its presence in high-molecular-weight complexes as previously observed for full-length HtaA.49 Given that ChtA30-314 is the most abundant heme-binding protein in C. diphtheriae’s exoproteome, we wondered whether its heme-bound form could serve as an alternate iron source to support bacterial growth. Indeed, exogenously supplied heme-loaded ChtA30-314 supports C. diphtheriae growth, as recombinant heme-loaded ChtACR domain restores growth in iron-limited media to levels comparable to cells using Hb as an iron source (Fig. 3C). Thus, the abundant truncated ChtA30-314 proteoform is not merely a degradation product, but a functional hemophore capable of supporting iron acquisition.

Figure 3.

Figure 3.

ChtA is a major exoproteome component, predominantly present as a truncated proteoform, and supports iron-limited C. diphtheriae growth through its heme-binding CR domain. (A) GeLC-MS/MS on sequential SDS-PAGE slices shows ChtA is more than 99% present as a truncated proteoform, enriched in low-MW slices consistent with native MS. Only 32% of HtaA signals are present in low-MW slices. Intensities corresponding to full-length proteoforms are indicated with “FL”. (B) LC-MS/MS relative intensities of all proteins within the C. diphtheriae exoproteome show ChtA is highly abundant, following DIP2069 and DT. Relative intensities are in terms of iBAQ values (proxy for molar abundance) and correspond to full-length protein. Combining % intensity in (B) with % proteolysis from (A) provides an estimate of hemophore relative abundance. (C) Growth of C. diphtheriae on Hb and heme-loaded ChtACR at equivalent heme concentrations (0.6 μM) shows ChtACR supports growth better than hemoglobin. An evolutionarily distant heme-binding protein (heme-loaded IsdANEAT domain from Staphylococcus aureus) does not support growth to the same levels.

DIP2069 releases ChtA30-314 from the cell surface, whereas its ChtC paralog remains surface associated.

Reasoning that ChtA and possibly HtaA are released from the cell surface by the action of a bacterial protease(s) that removes their C-terminal TM helices, we searched for other truncated proteins within the exoproteome that may also be processed by these enzymes. A prioritized target list of charge states in the raw spectrum was created for nTD-MS identification that corresponds to unique masses in the DIA-PTCR dataset. First, we collected 175 PTCR mass spectra from consecutively spaced isolation windows encompassing the bulk of ion current from direct infusion of the exoproteome (Fig. S1). Masses detected from the DIA-PTCR spectrum were then associated with their most abundant charge state in m/z space (Fig. 4A) and fragmented using HCD to reveal their primary sequence. Prioritization was critical for low abundance proteoforms. Without it, nTD-MS sequencing is prohibitively slow because each identification requires testing multiple fragmentation settings, each requiring long averaging times. The amino acid sequences from 60 proteoforms within the exoproteome were determined, which originate from 48 distinct proteins (Table S1, Fig. 4A). Interestingly, 7 of these 48 proteins correspond to novel truncated proteoforms resembling ChtA- and HtaA-like processing, as they contain C-terminal residues that are at either phenylalanine (ChtA, DIP2069, FrgD, RpfA) or alanine (HtaA, SprX, CtaC), respectively (Table S1). Specific cleavage sites within ChtA, HtaA and FrgD are positioned to remove C-terminal TM helix or lipid membrane anchoring elements, whereas cleavage of the other proteins presumably does not affect their location as their full-length forms are also expected to reside within the exoproteome (six are shown in Fig. 4B; all seven are listed in Table S1). HtaA and DIP2069 are each cleaved at multiple sites, generating truncated fragments with specific C-terminal alanine and phenylalanine residues, respectively, suggesting defined rather than nonspecific proteolytic processing. Notably, all truncated DIP2069 proteoforms were observed with an additional ~40 Da mass consistent with Ca2+ binding, supporting native cofactor engagement and a canonical chymotrypsin-like serine protease fold for DIP2069 in the exoproteome (Fig. S4). Other proteoforms also appear in the exoproteome, and several of them show N-terminal cleavages corresponding to canonical signal peptidase cleavage sites (SpI and SpII) likely occurring after secretion through the Sec or Tat pathways (Table S1).50 Notably, ChtA30-314 is one of the most abundant truncated proteoforms within the exoproteome when assessed by native MS with DIA-charge reduction, whereas cleaved forms of HtaA (HtaA31-272, and HtaA273-555) are less prevalent because the full-length form of this protein predominates (Fig. 3A).

Figure 4.

Figure 4.

C-terminal phenylalanine cleavage reveals proteolysis of iron-acquisition proteins and exoenzymes. (A) Rank abundance of proteoforms detected by DIA-PTCR charge reduction, if they were identified by native top-down nMS (nTD-verified), heme binding (hemophore), and possess a non-canonical truncation site. The y-axis shows the intensity of each proteoform’s most abundant charge state in the DIA-PTCR dataset. (B) A total of 10 novel truncated proteoforms from 6 different proteins of the exoproteome showed 9 that resulted from a C-terminal phenylalanine or alanine cleavage from their primary sequence. Two proteins involved in iron acquisition are missing membrane anchoring moieties which could explain their presence in the exoproteome (FrgD’s lipidated cysteine; ChtA’s C-terminal transmembrane helix). An exoprotease DIP2069 contains three novel C-terminal truncations at 3 closely spaced phenylalanine residues. RpfA which was found in its full-length and truncated proteoform shows a C-terminal Phe cleavage.

To identify the protease responsible for releasing ChtA30-314 from the cell surface, we examined the effects of deleting the DIP2069 and SprX (DIP0964) extracellular proteases. They were chosen because they are abundant and exist as truncated proteoforms that contain either C-terminal phenylalanine (DIP206923-356, DIP206923-359, DIP206923-360) or alanine residues (SprX27-238). The native exoproteomes of C. diphtheriae strains containing non-polar deletions in either dip2069 (Δdip2069) or sprX (ΔsprX) were analyzed using DIA-charge reduction MS. While all three strains exhibit similar growth behavior in iron-replete media, the Δdip2069 exoproteome differs markedly compared to the WT or ΔsprX deletion strain (Fig. S3). Notably, truncated ChtA30-314, among other truncated proteoforms, is absent (Fig. S3B). To further probe the effect of the dip2069 deletion on ChtA proteolysis, we performed western blotting with an anti-ChtA antibody (α-ChtA) that recognizes epitopes outside of ChtA’s CR domain, as confirmed by strong signal for recombinant full-length ChtA but not the CR-domain-only construct (Fig. 5A). Intact ChtA accumulates in the dip2069 exoproteome, consistent with loss of proteolytic processing. Anti-diphtheria toxin (α-DT) blotting confirmed equivalent secretion across strains (Fig. 5A) and anti-HtaA blotting showed that full-length HtaA is unaffected by the dip2069 deletion (Fig. S5). We then performed label-free quantitative proteomics on exoproteome and surface fractions from WT and Δdip2069 strains. Compared to the Δdip2069 strain, the WT exoproteome showed higher levels of peptides corresponding to the ChtA30-314 proteoform but significantly lower levels of the C-terminal region (Fig. 5B), further supporting that dip2069 is responsible for forming ChtA30-314. Notably, in the absence of dip2069, ChtA is approximately 32-fold more abundant on the surface of the Δdip2069 strain relative to WT (Fig. 5C). Thus, the native MS, western blot analysis, and LC-MS/MS proteomics data indicate that the DIP2069 protease liberates ChtA30-314 from the cell surface through proteolytic processing.

Figure 5.

Figure 5.

DIP2069 generates the secreted ChtA30-314 hemophore. (A) Cultures of iron-limited wild-type (WT) and Δdip2069 C. diphtheriae were grown to stationary phase and the resulting exoproteome (exo.) was subjected to SDS-PAGE with Coomassie staining or prepared for western blotting with anti-ChtA (α-ChtA) or diphtheria toxin (α-DT) antibody. The α-ChtA antibody recognizes epitopes present in full-length ChtA (ChtA-FL) but does not recognize the CR domain alone, as verified by blotting recombinant full-length and CR domain with a C-terminal Strep-tag II (ChtA-CR). Blotting with α-DT serves as a secretion and loading control, confirming equivalent exoproteome protein levels across strains. (B) ChtA peptide intensity differences from exoproteome preparations of WT and Δdip2069 C. diphtheriae cultures. Only peptides with a significantly different abundance between each sample as defined by a P-value of 0.05 from an unpaired Student’s t-test are shown. (C) A volcano plot showing abundance differences at the cell surface between WT and Δdip2069 strains from the same cell cultures that exoproteomes were prepared from for the data in (B). Data from panels (B) and (C) are from bottom-up proteomics preparations as described in methods.

In addition to ChtA, C. diphtheriae also expresses the closely related ChtC protein which shares 47% sequence similarity.21 Both contain related heme-binding CR domains and C-terminal helices, however, ChtA and ChtC are unique members of the heme uptake machinery because they are also predicted to harbor a central beta-propeller domain whose function is not known. Interestingly, the MS data reveal that these proteins localize to distinct sites – ChtA partitions to the exoproteome as ChtA30-314, whereas ChtC remains associated with the microbe and exposed on the cell surface (Fig. 6B).29 An inspection of their primary sequences suggests that this difference may be a result of distinct susceptibilities to DIP2069 cleavage, as the phenylalanine residue that is the cleavage-site in ChtA is conserved across species, whereas in the C. diphtheriae ChtC protein this residue is replaced with a tryptophan (Fig. 6A). Indeed, compatible with it not being a substrate for DIP2069 in both WT and Δdip2069 strains the ChtC protein remains cell-associated (Fig. 6C), whereas ∆dip2069 deletion prevents formation of ChtA30-314 (Fig. S3B) in the exoproteome and leads to accumulation of ChtA on the cell surface (Fig. 5C). Together, the data suggest that DIP2069 is responsible for proteolytically releasing ChtA30-314 from the cell surface, whereas a subtle sequence alteration in its paralog ChtC causes it to remain associated with the cell surface.

Figure 6.

Figure 6.

DIP2069 selectively liberates the ChtA30-314 hemophore from the cell surface, but not its paralog ChtC. (A) Sequence alignment of ChtA-like proteins across prominent corynebacterium pathogens show Phe314 is highly conserved and flanked by charged and/or polar residues. Similarly, polar and charged residues are present in DIP2069, FrgD, and RpfA novel proteoforms that arise from cleavage events C-terminal to phenylalanine. (B) Peptide intensity differences from exoproteome and cell surface preparations of WT and Δdip2069 C. diphtheriae cultures. (C) Western blot analysis of ChtC in the exoproteome (“E”) and cell surface (“C”) fractions from WT and Δdip2069 strains showing full-length ChtC across all fractions.

ProteoMIX reveals S. aureus releases soluble IsdA proteoforms for heme scavenging.

To demonstrate its broad utility, we used ProteoMIX to characterize the exoproteome of pathogenic S. aureus. Previous studies have shown that this bacterium expresses heme-binding Iron-regulated Surface Determinant (Isd) proteins that are covalently attached to the peptidoglycan (IsdA, IsdB, IsdC, and IsdH).23 Application of ProteoMIX revealed soluble IsdA proteoforms in the exoproteome capable of scavenging heme released from Hb (Fig. S6A, B). Similar to C. diphtheriae ChtA30-314, these IsdA proteoforms (IsdA46-224, IsdA46-226, and IsdA46-238) lack C-terminal segments that tether the full-length, cell wall-associated protein to the microbe. Three truncated IsdA proteoforms are generated by cleavage of the segment connecting the heme-binding NEAr iron Transporter (NEAT) domain to the C-terminal LPXTG motif (Fig. S6A), which covalently attaches the surface-associated protein to the peptidoglycan.51 GeLC-MS/MS studies indicate that IsdA partitions roughly equally between cell wall-associated and extracellular forms.52

S. aureus and C. diphtheriae can sometimes inhabit overlapping sites in the upper respiratory tract, and the two bacteria are common co-infectants in subcutaneous wounds,53 suggesting their hemophores may compete for heme iron to sustain growth. Interestingly, growth assays show that heme delivery to C. diphtheriae is specific for ChtA, as supplementation of iron-limited C. diphtheriae cultures with the recombinant heme-loaded form of the structurally distinct IsdANEAT hemophore from S. aureus does not appreciably rescue growth (Fig. S6C). A similar result is observed for S. aureus under iron-deplete conditions. The bacteria grow when supplemented with either Hb or recombinant heme-loaded IsdANEAT, whereas heme-loaded ChtACR from C. diphtheriae is largely ineffective in supporting S. aureus growth in iron-limited media (Fig. S6C). These results indicate that productive heme delivery is largely species-specific, with each hemophore preferentially supporting growth of its native bacterium.

DISCUSSION

Protein-ligand binding events are fundamental to nearly every aspect of cellular function (catalysis, signaling, growth, and metabolism among others), yet identifying these interactions within complex proteomes remains challenging.54 Here we report the development of ProteoMIX, an nMS approach that efficiently identifies specific ligand-binding proteins and their proteoforms within proteomic mixtures. ProteoMIX builds on two recent advancements in native MS: (i) slow mixing mode (SLOMO) nMS that quantitatively tracks gradual changes in protein intensities as a ligand binds,40,55,56 and (ii) DIA-charge reduction that dramatically reduces spectral complexity (Fig. 1).41,43,44 By integrating these methods, ProteoMIX provides an efficient function-based route to discovering ligand-binding proteoforms directly within unfractionated proteomes, which we use to identify heme-binding proteoforms from the exoproteomes of C. diphtheriae and S. aureus and shed new light on how these gram-positive pathogens scavenge heme during infections. SLOMO was used to track heme uptake from Hb by this microbe’s exoproteome with simultaneous DIA-charge reduction to deconvolve the mass spectra. We show that heme-binding proteoforms can be identified with high confidence by paired losses and gains matching the mass of heme, enabling the prioritization and ensuing top-down MS sequencing of their most abundant charge states (Fig. 2). The prioritization step greatly improves efficiency for hemophore identification, as the alternative would be to fragment every peak in the native mass spectrum. Given current throughput, confident MS2 identification of low-to-moderate abundance proteoforms is slow, making comprehensive peak-by-peak sequencing currently impractical (often hours).

A central methodological advantage of ProteoMIX is its use of SLOMO-nMS rather than a differential comparison of separately acquired apo- and holo-proteome samples, enabling one to resolve overlapping proteoforms in complex mixtures. In principle, heme-binding proteoforms could be flagged by acquiring one proteome sample without Hb and a second pre-incubated with Hb, then searching for mass pairs separated by 616 Da whose intensities are anticorrelated between the two acquisitions. While a differential strategy is adequate for purified proteins or relatively simple mixtures, it becomes unreliable at proteome complexity. In a proteome containing several hundred proteoforms, however, intensity differences between independently acquired spectra arise not only from ligand binding but also from emitter-to-emitter geometry, spray stability, and spray voltage, which can change between different native MS nanospray tip loadings. Spurious anticorrelated 616 Da pairs can therefore occur by chance, with no internal criterion to separate a genuine binding event from coincidental fluctuation. SLOMO-nMS removes this ambiguity by replacing two distinct measurements with a continuous mixing trajectory acquired from a single emitter under constant ionization conditions. A true hemophore produces a sustained, reciprocal decrease in apo-proteoform intensity and increase in holo-proteoform intensity across the consecutive DIA-charge reduction cycles of the mixing time course (Fig. 1C, Fig. S2A). Random intensity fluctuation does not reproduce this sustained reciprocal pattern, so the paired reciprocity itself serves as a built-in false positive filter. Time-resolved reciprocity is also essential when an apo-proteoform has the same mass as another protein in the complex mixture (i.e., overlapping masses), as observed for the HtaA31-272 hemophore, which shares its apo-mass with the secreted protease DIP035030-274 (Fig. 2D). A between-sample intensity comparison at that mass would be difficult to interpret with confidence, whereas the coupled rise of holo-HtaA31-272 as the blended apo signal declines (to roughly half its initial intensity over the mixing time course) allows the binding event to be assigned with confidence. Finally, because a single SLOMO experiment samples a continuous Hb gradient, it yields a continuous range of Hb concentrations within a single experiment, without the additional sample preparation, instrument time, and emitter-to-emitter variability inherent to multisample titration.

The hemophore proteoforms characterized in this study were abundant components of their respective exoproteomes, and it is worth considering how ProteoMIX would perform to identify lower abundance species from complex samples. The main practical constraint lies in proteoform sequencing rather than in detecting heme binding. The HtaA proteoforms had intensities approximately 2 orders of magnitude lower than the most intense native proteoform (Fig. 4A) and are predicted to encompass 2% of the molar abundance of the exoproteome when accounting for % cleavage from the intact proteoform (Fig. 3A, 3B). Confident assignment of these proteoforms required between 1,000 and 3,000 microscans, roughly 15 and 45 minutes, respectively). This constraint is shared by native top-down proteomics generally, and motivates the functional prioritization central to ProteoMIX. Across the unfractionated exoproteome, nTD-MS sequencing yielded 60 proteoform identifications (Table S1), which had an intensity of at least 104 from their most abundant charge state, as determined from our DIA-charge reduction dataset (Fig. 4A). Consistent with this, we detected two additional candidate heme-binding species, one of which corresponds to HtaA31-272 without its potassium adduct, from their 616 Da mass shifts during mixing that were at too low an intensity for confident sequencing by native top-down MS (Fig. S7). For more complex proteomes such as microbiome secretomes, the depth and dynamic range of analysis can be extended by tuning the DIA charge-reduction reaction time, by prefractionation prior to analysis, and by continued improvements in instrument sensitivity.

Using ProteoMIX we discovered that C. diphtheriae secretes ChtA30-314 to scavenge heme from Hb and other hemoproteins. This was surprising, since like other proteins in C. diphtheriae’s heme uptake system that are instead cell wall associated, the full-length ChtA protein contains a C-terminal TM helix that is presumably embedded in the bilayer to enable surface display. In follow-up experiments that used a combination of GeLC-MS/MS (Fig. 3), nMS (Fig. S3B), and western blotting applied to wild-type and Δdip2069 strains (Fig. 5, Fig. 6), we provide evidence that a highly abundant extracellular chymotrypsin-like protease DIP2069 is responsible for releasing ChtA30-314 from the cell. ChtA30-314 is liberated when the surface-displayed full-length ChtA protein is cleaved by DIP2069 between its heme-binding CR domain and its C-terminal propeller domain. Once outside the cell, ChtA30-314 can capture heme and deliver it back to the bacterium to enable growth when iron is scarce (Fig. 3C), suggesting it plays an important role in iron uptake during infection. Other pathogenic corynebacterial species presumably use a similar protease-mediated release mechanism to release soluble ChtA hemophores for iron scavenging, as they also contain DIP2069 homologs and ChtA proteins that share a conserved sequence surrounding the site of cleavage. A systematic analysis of the nMS spectra reveals several other truncated proteoforms in the exoproteome that may be generated by DIP2069, as they share conserved C-termini with ChtA30-314 (FrgD, an iron-siderophore/vitamin B12 binding protein; RpfA, a growth resuscitation factor; and DIP2069 itself) (Fig. 4B).

GeLC-MS/MS and western blot analysis revealed that HtaA is secreted as both full-length and truncated proteoforms, with the full-length form predominating (~70%) in the exoproteome (Fig. 3A, Fig. S5). ProteoMIX detected two proteolytically cleaved HtaA proteoforms (HtaA31-272 and HtaA273-555) in the exoproteome, each containing one of HtaA’s two CR domains (CR1 and CR2) but lack the TM helix present in the parent protein. Both proteoforms are capable of scavenging heme from Hb (Fig. 2, Fig. S2). These proteoforms are presumably generated by a yet-to-be-identified protease, as HtaA abundance does not change substantially in the Δdip2069 strain (Fig. S5), and the C-termini of HtaA31-272 and HtaA273-555 lack the C-terminal phenylalanine residue likely present at the DIP2069 cleavage site. Notably, in contrast to ChtA, which is almost completely converted into ChtA30-314 by the DIP2069 protease, only ~30% of HtaA is cleaved. Outside the cell, the longer TM-helix-containing HtaA proteoform predominates, whereas HtaA31-272 and HtaA273-555 are roughly five times less abundant than ChtA30-314. Interestingly, the longer HtaA proteoform is undetectable by nMS, consistent with it being insoluble or sequestered in mycomembrane vesicles too large for detection. This idea aligns with recent findings that the related bacterium Corynebacterium glutamicum produces mycomembrane vesicles containing its heme-binding proteins.57 Interestingly, large HbpA aggregates have previously been observed in the exoproteome,37,38 raising the possibility that it may also be packed into vesicles.

Our findings clarify the roles of individual components within the C. diphtheriae heme-uptake system and reveal distinct functional specializations of its paralogous ChtA and ChtC proteins. Figure 7 summarizes the positioning of these protein components based on our results and previously reported protease shaving and cell fractionation studies.29 HbpA and ChtC are exposed on the mycomembrane, where they are positioned to bind Hb and heme, respectively. Also exposed on the surface are portions of the heme-binding ChtA and HtaA proteins, which are significantly less abundant than HbpA and ChtC; the approximate abundance ratios are 10:5:1:1 for HbpA:ChtC:ChtA:HtaA. Each of these exposed proteins is presumably tethered to the underlying plasma membrane via their C-terminal TM helices, with N-terminal ligand-binding domains positioned outward. HtaA is unique in containing two heme CR domains, CR1 and CR2. Considering the dimensions of the corynebacterial envelope, CR1 is likely exposed on the surface, whereas the C-terminal CR2 domain is buried within the peptidoglycan. The heme-binding HtaB and ChtB proteins are also buried within the peptidoglycan and tethered to the membrane via C-terminal TM helices. Outside the bacterium, our results show that ChtA30-314 scavenges heme from Hb to facilitate cell growth after it is proteolytically released from the cell surface by extensive cleavage of surface-embedded ChtA by the DIP2069 protease (the majority of ChtA is cleaved, such that ChtA30-314 is ~32 times more abundant than cell-surface ChtA) (Fig. 5). Interestingly, even though ChtC and ChtA share highly similar primary sequences,58 ChtC is a poor substrate for DIP2069 and remains intact on the cell surface (Fig. 6). At least three heme-scavenging proteoforms of HtaA are also present extracellularly: a nearly intact proteoform containing its TM helix and both heme-binding domains (Fig. 7), and two less abundant proteolytic fragments, each containing a single heme-binding domain (HtaA31-272 and HtaA273-555) (Fig. 2). Thus, in its quest for iron, C. diphtheriae releases and displays an array of proteins to scavenge heme, presumably delivering it to the microbial surface for transfer across the cell envelope. How heme is transferred among these components is not known, but transfer studies using a fluorogenic reporter suggest it may occur via transient heme-transfer complexes involving their CR domains.59

Figure 7.

Figure 7.

A model of protease-mediated heme acquisition by Corynebacterium diphtheriae. Heme-acquisition begins with red blood cell lysis near the site of infection, releasing high concentrations of tetrameric hemoglobin. Tetrameric hemoglobin releases free heme and dissociates into dimeric hemoglobin, which releases heme approximately 8 times faster. The free heme is scavenged by host hemopexin and Corynebacterium diphtheriae’s secreted ChtA30-314 hemophore, which it ultimately uses as an iron source. The ChtA30-314 hemophore is produced from surface-associated full-length ChtA, which DIP2069 selectively processes, in contrast to ChtA’s surface-associated paralog, ChtC, which remains on the cell surface during iron-limited conditions. Heme bound to ChtA30-314 is then passed to C. diphtheriae to obtain its iron. This presumably occurs when it transfers heme to cell associated HtaA, HtaB and ChtB heme-binding proteins, or directly to the HmuT for import by the HmuTUV ABC transport cassette.

Application of ProteoMIX to evolutionarily distinct gram-positive S. aureus revealed that this pathogen also releases soluble hemophore proteoforms (Fig. S6). We identified truncated IsdA proteoforms in the exoproteome that coexist with longer, cell wall-anchored variants, revealing a modular system in which surface-bound IsdA proteins capture local heme while secreted proteoforms presumably extend acquisition range. The parallel processing of IsdA in S. aureus and ChtA30-314 in C. diphtheriae suggests that proteolytic liberation of hemophores is a conserved strategy among gram-positive bacteria. Growth assays highlight species-specific functionality: while heme-loaded ChtA30-314 delivers heme to C. diphtheriae, delivery by the structurally unrelated IsdANEAT hemophore occurs far less efficiently (Fig. 3C). The converse also is true; exogenously supplemented heme-loaded IsdANEAT is capable of delivering heme and sustaining microbial growth of S. aureus better than ChtACR (Fig. S6C). This suggests that these hemophores have coevolved with their cognate uptake machinery to provide a competitive growth advantage in nutrient-limited niches colonized by diverse genera of bacteria. Notably, C. diphtheriae reaches a higher optical density when supplied with heme pre-loaded onto ChtACR than with an equivalent concentration of Hb, consistent with hemophore-mediated delivery bypassing passive Hb-heme extraction on the cell surface. In contrast, S. aureus grows slightly better on Hb than on IsdANEAT, reflecting the presence of dedicated Hb receptors (e.g., IsdB and IsdH) that efficiently harvest heme from Hb.60–62 These observations suggest that secreted IsdA proteoforms in S. aureus may primarily extend the range of heme scavenging or sequester heme to limit access by competitors, rather than serving as the main mechanism for heme acquisition from Hb. Because relatively few gram-positive hemophores have been identified to date, many more may be uncovered using ProteoMIX. Collectively, these findings establish ProteoMIX as a powerful platform for discovering functional hemophores, revealing previously hidden layers of bacterial nutrient acquisition, and opening new avenues for targeting iron metabolism in pathogens.

CONCLUSION

In conclusion, we have developed a novel native proteomics method, ProteoMIX, that combines DIA-charge reduction with SLOMO-nMS to identify ligand-binding proteoforms from an unfractionated exoproteome. DIA-charge reduction of the native Corynebacterium diphtheriae exoproteome resolved hundreds of proteoforms, and tracking them through SLOMO-nMS with Hb as the mixing species identified novel truncated hemophore proteoforms, of which ChtA30-314 was the most abundant. We showed that this method translated easily to a second organism: applying it to the Staphylococcus aureus exoproteome identified novel truncated hemophore proteoforms from IsdA. C. diphtheriae grew selectively on its own recombinant hemophore, and the reverse was true for S. aureus, indicating these proteoforms contribute to species-specific heme acquisition. The DIA-charge reduction dataset allowed us to build a prioritized fragmentation list for native top-down sequencing, which identified DIP2069 as a candidate protease mediating ChtA cleavage and release in C. diphtheriae. This was corroborated by western blotting and LC-MS/MS proteomics of wild-type and DIP2069 deletion strains. The specific cleavage of ChtA, but not its paralog ChtC, shows how proteoforms connect protein processing to cellular localization and function. Because ProteoMIX relies only on a defined ligand-induced mass shift, it should generalize readily beyond heme to other ligands, and native proteomes, extending native proteomics from cataloguing proteoforms to resolving their function.

Supplementary Material

JACS SI

Additional Experimental Methods, Figures S1 to S8, and Table S1–S3, including details of the ProteoMIX workflow and instrumentation parameters.

The code for DIA-charge reduction processing are openly available at https://github.com/AndrewGoring/DIA_CD_Analysis.

ACKNOWLEDGMENTS

The authors would like to thank Dr. Brendan J. Mahoney and other members of the Clubb and Loo Labs for useful discussions. This work was supported by grants from the NIH (R35GM145286 to J.A.L, R01AI161828 to R.T.C.). A.K.G. acknowledges support from the NIH NIDCR (T90DE030860). Additional support was provided by the NIH National Institute of General Medical Sciences (NIGMS)-funded predoctoral fellowship for J.F. (T32GM145388).

REFERENCES

  • (1).Cassat JE; Skaar EP. Iron in Infection and Immunity. Cell Host & Microbe 2013, 13 (5), 509–519. 10.1016/j.chom.2013.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (2).Murdoch CC; Skaar EP. Nutritional Immunity: The Battle for Nutrient Metals at the Host–Pathogen Interface. Nat Rev Microbiol 2022. 10.1038/s41579-022-00745-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (3).Hood MI; Skaar EP. Nutritional Immunity: Transition Metals at the Pathogen–Host Interface. Nat Rev Microbiol 2012, 10 (8), 525–537. 10.1038/nrmicro2836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (4).Sheldon JR; Laakso HA; Heinrichs DE. Iron Acquisition Strategies of Bacterial Pathogens. Microbiol Spectr 2016, 4 (2), 4.2.05. 10.1128/microbiolspec.VMBF-0010-2015. [DOI] [PubMed] [Google Scholar]
  • (5).Kramer J; Özkaya Ö; Kümmerli R. Bacterial Siderophores in Community and Host Interactions. Nat Rev Microbiol 2020, 18 (3), 152–163. 10.1038/s41579-019-0284-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (6).Perry WJ; Spraggins JM; Sheldon JR; Grunenwald CM; Heinrichs DE; Cassat JE; Skaar EP; Caprioli RM. Staphylococcus Aureus Exhibits Heterogeneous Siderophore Production within the Vertebrate Host. Proc. Natl. Acad. Sci. U.S.A. 2019, 116 (44), 21980–21982. 10.1073/pnas.1913991116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (7).Wandersman C; Delepelaire P. Bacterial Iron Sources: From Siderophores to Hemophores. Annu. Rev. Microbiol. 2004, 58 (1), 611–647. 10.1146/annurev.micro.58.030603.123811. [DOI] [PubMed] [Google Scholar]
  • (8).Antelo GT; Vila AJ; Giedroc DP; Capdevila DA. Molecular Evolution of Transition Metal Bioavailability at the Host–Pathogen Interface. Trends in Microbiology 2021, 29 (5), 441–457. 10.1016/j.tim.2020.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (9).Parrow NL; Fleming RE; Minnick MF. Sequestration and Scavenging of Iron in Infection. Infect Immun 2013, 81 (10), 3503–3514. 10.1128/IAI.00602-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (10).Cescau; Cwerman H; Létoffé S; Delepelaire P; Wandersman C; Biville F. Heme Acquisition by Hemophores. Biometals 2007, 20 (3–4), 603–613. 10.1007/s10534-006-9050-y. [DOI] [PubMed] [Google Scholar]
  • (11).Butaitė E; Baumgartner M; Wyder S; Kümmerli R. Siderophore Cheating and Cheating Resistance Shape Competition for Iron in Soil and Freshwater Pseudomonas Communities. Nat Commun 2017, 8 (1), 414. 10.1038/s41467-017-00509-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (12).Garénaux A; Caza M; Dozois CM. The Ins and Outs of Siderophore Mediated Iron Uptake by Extra-Intestinal Pathogenic Escherichia Coli. Veterinary Microbiology 2011, 153 (1–2), 89–98. 10.1016/j.vetmic.2011.05.023. [DOI] [PubMed] [Google Scholar]
  • (13).Gao H; Bian X. Editorial: Microbial Siderophores: Biosynthesis, Regulation, and Physiological and Ecological Impacts. Front. Microbiol. 2022, 13, 892485. 10.3389/fmicb.2022.892485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (14).Harrington JM; Duckworth OW; Haselwandter. The Fate of Siderophores: Antagonistic Environmental Interactions in Exudate-Mediated Micronutrient Uptake. Biometals 2015, 28 (3), 461–472. 10.1007/s10534-015-9821-4. [DOI] [PubMed] [Google Scholar]
  • (15).Leventhal GE; Ackermann M; Schiessl KT. Why Microbes Secrete Molecules to Modify Their Environment: The Case of Iron-Chelating Siderophores. J. R. Soc. Interface. 2019, 16 (150), 20180674. 10.1098/rsif.2018.0674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (16).Śmiga M; Ślęzak P; Wagner M; Olczak T. Interplay between Porphyromonas Gingivalis Hemophore-Like Protein HmuY and Kgp/RgpA Gingipains Plays a Superior Role in Heme Supply. Microbiol Spectr 2023, 11 (2), e04593–22. 10.1128/spectrum.04593-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (17).Bateman TJ; Shah M; Ho TP; Shin HE; Pan C; Harris G; Fegan JE; Islam EA; Ahn SK; Hooda Y; Gray-Owen SD; Chen W; Moraes TF. A Slam-Dependent Hemophore Contributes to Heme Acquisition in the Bacterial Pathogen Acinetobacter Baumannii. Nat Commun 2021, 12 (1), 6270. 10.1038/s41467-021-26545-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (18).Maresso AW; Garufi G; Schneewind O. Bacillus Anthracis Secretes Proteins That Mediate Heme Acquisition from Hemoglobin. PLoS Pathog 2008, 4 (8), e1000132. 10.1371/journal.ppat.1000132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (19).Barber MF; Elde NC. Buried Treasure: Evolutionary Perspectives on Microbial Iron Piracy. Trends in Genetics 2015, 31 (11), 627–636. 10.1016/j.tig.2015.09.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (20).Malmirchegini GR; Sjodt M; Shnitkind S; Sawaya MR; Rosinski J; Newton SM; Klebba PE; Clubb RT. Novel Mechanism of Hemin Capture by Hbp2, the Hemoglobin-Binding Hemophore from Listeria Monocytogenes. Journal of Biological Chemistry 2014, 289 (50), 34886–34899. 10.1074/jbc.M114.583013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (21).Allen CE; Burgos JM; Schmitt MP. Analysis of Novel Iron-Regulated, Surface-Anchored Hemin-Binding Proteins in Corynebacterium Diphtheriae. Journal of Bacteriology 2013, 195 (12), 2852–2863. 10.1128/JB.00244-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (22).Mahoney BJ; Lyman LR; Ford J; Soule J; Cheung NA; Goring AK; Ellis-Guardiola K; Collazo MJ; Cascio D; Ton-That H; Schmitt MP; Clubb RT. Molecular Basis of Hemoglobin Binding and Heme Removal in Corynebacterium Diphtheriae. Proc. Natl. Acad. Sci. U.S.A. 2025, 122 (1), e2411833122. 10.1073/pnas.2411833122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (23).Mazmanian SK. Passage of Heme-Iron Across the Envelope of Staphylococcus Aureus. Science 2003, 299 (5608), 906–909. 10.1126/science.1081147. [DOI] [PubMed] [Google Scholar]
  • (24).Tarlovsky Y; Fabian M; Solomaha E; Honsa E; Olson JS; Maresso A. A Bacillus Anthracis S-Layer Homology Protein That Binds Heme and Mediates Heme Delivery to IsdC. J Bacteriol 2010, 192 (13), 3503–3511. 10.1128/JB.00054-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (25).Xiao Q; Jiang X; Moore KJ; Shao Y; Pi H; Dubail I; Charbit A; Newton SM; Klebba PE. Sortase Independent and Dependent Systems for Acquisition of Haem and Haemoglobin in Listeria Monocytogenes. Molecular Microbiology 2011, 80 (6), 1581–1597. 10.1111/j.1365-2958.2011.07667.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (26).Maffei B; Francetic O; Subtil A. Tracking Proteins Secreted by Bacteria: What’s in the Toolbox? Front. Cell. Infect. Microbiol. 2017, 7, 221. 10.3389/fcimb.2017.00221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (27).Saka HA; Thompson JW; Chen Y-S; Dubois LG; Haas JT; Moseley A; Valdivia RH. Chlamydia Trachomatis Infection Leads to Defined Alterations to the Lipid Droplet Proteome in Epithelial Cells. PLoS ONE 2015, 10 (4), e0124630. 10.1371/journal.pone.0124630. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (28).Sinclair SHG; Garcia-Garcia JC; Dumler JS. Bioinformatic and Mass Spectrometry Identification of Anaplasma Phagocytophilum Proteins Translocated into Host Cell Nuclei. Front. Microbiol. 2015, 6. 10.3389/fmicb.2015.00055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (29).Goring AK; Hale S; Dasika P; Chen Y; Clubb RT; Loo JA. The Exoproteome and Surfaceome of Toxigenic Corynebacterium Diphtheriae 1737 and Its Response to Iron Restriction and Growth on Human Hemoglobin. J. Proteome Res. 2025, 24 (1), 77–93. 10.1021/acs.jproteome.4c00443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (30).Liu L; Gueguen-Chaignon V; Gonçalves IR; Rascle C; Rigault M; Dellagi A; Loisel E; Poussereau N; Rodrigue A; Terradot L; Condemine G. A Secreted Metal-Binding Protein Protects Necrotrophic Phytopathogens from Reactive Oxygen Species. Nat Commun 2019, 10 (1), 4853. 10.1038/s41467-019-12826-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (31).Pagliuso A; Tham TN; Allemand E; Robertin S; Dupuy B; Bertrand Q; Bécavin C; Koutero M; Najburg V; Nahori M-A; Tangy F; Stavru F; Bessonov S; Dessen A; Muchardt C; Lebreton A; Komarova AV; Cossart P. An RNA-Binding Protein Secreted by a Bacterial Pathogen Modulates RIG-I Signaling. Cell Host & Microbe 2019, 26 (6), 823–835.e11. 10.1016/j.chom.2019.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (32).Hadfield TL; McEvoy P; Polotsky Y; Tzinserling VA; Yakovlev AA. The Pathology of Diphtheria. J INFECT DIS 2000, 181 (Supplement_1), S116–S120. [DOI] [PubMed] [Google Scholar]
  • (33).Allen CE; Schmitt MP. HtaA Is an Iron-Regulated Hemin Binding Protein Involved in the Utilization of Heme Iron in Corynebacterium Diphtheriae. JB 2009, 191 (8), 2638–2648. 10.1128/JB.01784-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (34).Allen CE; Schmitt MP. Novel Hemin Binding Domains in the Corynebacterium Diphtheriae HtaA Protein Interact with Hemoglobin and Are Critical for Heme Iron Utilization by HtaA. Journal of Bacteriology 2011, 193 (19), 5374–5385. 10.1128/JB.05508-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (35).Allen CE; Schmitt MP. Utilization of Host Iron Sources by Corynebacterium Diphtheriae: Multiple Hemoglobin-Binding Proteins Are Essential for the Use of Iron from the Hemoglobin-Haptoglobin Complex. J. Bacteriol. 2015, 197 (3), 553–562. 10.1128/JB.02413-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (36).Muraki N; Kitatsuji C; Okamoto Y; Uchida T; Ishimori K; Aono S. Structural Basis for the Heme Transfer Reaction in Heme Uptake Machinery from Corynebacteria. Chem. Commun. 2019, 55 (92), 13864–13867. 10.1039/C9CC07369H. [DOI] [PubMed] [Google Scholar]
  • (37).Lyman LR; Peng ED; Schmitt MP. Corynebacterium Diphtheriae Iron-Regulated Surface Protein HbpA Is Involved in the Utilization of the Hemoglobin-Haptoglobin Complex as an Iron Source. J Bacteriol 2018, 200 (7), e00676–17, /jb/200/7/e00676-17.atom. 10.1128/JB.00676-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (38).Lyman LR; Schaeffer J; Ruppitsch W; Schmitt MP. Analysis of the HbpA Protein from Corynebacterium Diphtheriae Clinical Isolates and Identification of a Putative Hemoglobin-Binding Site on HbpA. J Bacteriol 2022, 204 (12). 10.1128/jb.00349-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (39).Roberts DS; Loo JA; Tsybin YO; Liu X; Wu S; Chamot-Rooke J; Agar JN; Paša-Tolić L; Smith LM; Ge Y. Top-down Proteomics. Nat Rev Methods Primers 2024, 4 (1), 38. 10.1038/s43586-024-00318-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (40).Bui DT; Li Z; Kitov PI; Han L; Kitova EN; Fortier M; Fuselier C. Quantifying Biomolecular Interactions Using Slow Mixing Mode (SLOMO) Nanoflow ESI-MS. ACS Central Science 2022, 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (41).Schachner LF; Mullen C; Phung W; Hinkle JD; Beardsley MI; Bentley T; Day P; Tsai C; Sukumaran S; Baginski T; DiCara D; Agard NJ; Masureel M; Gober J; ElSohly AM; Melani R; Syka JEP; Huguet R; Marty MT; Sandova W. Exposing the Molecular Heterogeneity of Glycosylated Biotherapeutics. Nat Commun 2024, 15 (1), 3259. 10.1038/s41467-024-47693-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (42).Tamara S; den Boer MA; Heck AJR. High-Resolution Native Mass Spectrometry. Chem. Rev. 2022, 122 (8), 7269–7326. 10.1021/acs.chemrev.1c00212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (43).Huguet R; Mullen C; Srzentić K; Greer JB; Fellers RT; Zabrouskov V; Syka JEP; Kelleher NL; Fornelli L. Proton Transfer Charge Reduction Enables High-Throughput Top-Down Analysis of Large Proteoforms. Anal. Chem. 2019, 91 (24), 15732–15739. 10.1021/acs.analchem.9b03925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (44).Le Huray KIP; Wörner TP; Moreira T; Dembek M; Reinhardt-Szyba M; Devine PWA; Bond NJ; Fort KL; Makarov AA; Sobott F. To 200,000 m / z and Beyond: Native Electron Capture Charge Reduction Mass Spectrometry Deconvolves Heterogeneous Signals in Large Biopharmaceutical Analytes . ACS Cent. Sci. 2024, 10 (8), 1548–1561. 10.1021/acscentsci.4c00462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (45).Marty MT; Baldwin AJ; Marklund EG; Hochberg GKA; Benesch JLP; Robinson CV. Bayesian Deconvolution of Mass and Ion Mobility Spectra: From Binary Interactions to Polydisperse Ensembles. Anal. Chem. 2015, 87 (8), 4370–4376. 10.1021/acs.analchem.5b00140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (46).Teufel F; Almagro Armenteros JJ; Johansen AR; Gíslason MH; Pihl SI; Tsirigos KD; Winther O; Brunak S; Von Heijne G; Nielsen H. SignalP 6.0 Predicts All Five Types of Signal Peptides Using Protein Language Models. Nat Biotechnol 2022, 40 (7), 1023–1025. 10.1038/s41587-021-01156-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (47).Ford J; Goring AK; Lee Y; Chen M; Mahoney BJ; Sawaya MR; Shafaat HS; Loo JA; Clubb RT. Structural Basis of Heme Scavenging by the ChtA and HtaA Hemophores in Corynebacterium Diphtheriae. Journal of Biological Chemistry 2025, 110633. 10.1016/j.jbc.2025.110633. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (48).Dzieciatkowska M; Hill R; Hansen KC. GeLC-MS/MS Analysis of Complex Protein Mixtures. In Shotgun Proteomics: Methods and Protocols; Martins-de-Souza D., Ed.; Springer New York: New York, NY, 2014; pp 53–66. 10.1007/978-1-4939-0685-7_4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (49).Lyman LR; Peng ED; Schmitt MP. The Corynebacterium Diphtheriae HbpA Hemoglobin-Binding Protein Contains a Domain That Is Critical for Hemoprotein-Binding, Cellular Localization and Function. J Bacteriol 2021. 10.1128/JB.00196-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (50).Auclair SM; Bhanu MK; Kendall DA. Signal Peptidase I: Cleaving the Way to Mature Proteins. Protein Science 2012, 21 (1), 13–25. 10.1002/pro.757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (51).Ellis-Guardiola K; Mahoney BJ; Clubb RT. NEAr Transporter (NEAT) Domains: Unique Surface Displayed Heme Chaperones That Enable Gram-Positive Bacteria to Capture Heme-Iron From Hemoglobin. Front. Microbiol. 2021, 11, 607679. 10.3389/fmicb.2020.607679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (52).Byrum SD; Loughran AJ; Beenken KE; Orr LM; Storey AJ; Mackintosh SG; Edmondson RD; Tackett AJ; Smeltzer MS. Label-Free Proteomic Approach to Characterize Protease-Dependent and -Independent Effects of sarA Inactivation on the Staphylococcus Aureus Exoproteome. J. Proteome Res. 2018, 17 (10), 3384–3395. 10.1021/acs.jproteome.8b00288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (53).Lowe CF; Bernard KA; Romney MG. Cutaneous Diphtheria in the Urban Poor Population of Vancouver, British Columbia, Canada: A 10-Year Review. J Clin Microbiol 2011, 49 (7), 2664–2666. 10.1128/JCM.00362-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (54).Hough MA; Prischi F; Worrall JAR. Perspective: Structure Determination of Protein-Ligand Complexes at Room Temperature Using X-Ray Diffraction Approaches. Front. Mol. Biosci. 2023, 10, 1113762. 10.3389/fmolb.2023.1113762. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (55).Bui DT; Favell J; Kitova EN; Li Z; McCord KA; Schmidt EN; Mozaneh F; Elaish M; El-Hawiet A; St-Pierre Y; Hobman TC; Macauley MS; Mahal LK; Flynn MR; Klassen JS. Absolute Affinities from Quantitative Shotgun Glycomics Using Concentration-Independent (COIN) Native Mass Spectrometry. ACS Cent. Sci. 2023, 9 (7), 1374–1387. 10.1021/acscentsci.3c00294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Dhillon SK; Bui DT; Kitova EN; Mahal LK; Klassen JS. Going with the Flow: Mechanistic Insights into Slow Mixing Mode Native Mass Spectrometry. J. Am. Soc. Mass Spectrom. 2025, 36 (11), 2502–2508. 10.1021/jasms.5c00244. [DOI] [PubMed] [Google Scholar]
  • 57.Wang M; Nie Y; Wu X-L. Extracellular Heme Recycling and Sharing across Species by Novel Mycomembrane Vesicles of a Gram-Positive Bacterium. ISME J 2021, 15 (2), 605–617. 10.1038/s41396-020-00800-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Allen CE; Burgos JM; Schmitt MP. Analysis of Novel Iron-Regulated, Surface-Anchored Hemin-Binding Proteins in Corynebacterium Diphtheriae. J Bacteriol 2013, 195 (12), 2852–2863. 10.1128/JB.00244-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Mahoney BJ; Goring AK; Wang Y; Dasika P; Zhou A; Grossbard E; Cascio D; Loo JA; Clubb RT. Development and Atomic Structure of a New Fluorescence-Based Sensor to Probe Heme Transfer in Bacterial Pathogens. Journal of Inorganic Biochemistry 2023, 249, 112368. 10.1016/j.jinorgbio.2023.112368. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Valenciano-Bellido S; Caaveiro JMM; Nakakido M; Kuroda D; Aikawa C; Nakagawa I; Tsumoto K. Targeting Hemoglobin Receptors IsdH and IsdB of Staphylococcus Aureus with a Single VHH Antibody Inhibits Bacterial Growth. Journal of Biological Chemistry 2023, 299 (9), 104927. 10.1016/j.jbc.2023.104927. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Ellis-Guardiola K; Clayton J; Pham C; Mahoney BJ; Wereszczynski J; Clubb RT. The Staphylococcus Aureus IsdH Receptor Forms a Dynamic Complex with Human Hemoglobin That Triggers Heme Release via Two Distinct Hot Spots. Journal of Molecular Biology 2020, 432 (4), 1064–1082. 10.1016/j.jmb.2019.12.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.De Bei O; Marchetti M; Ronda L; Gianquinto E; Lazzarato L; Chirgadze DY; Hardwick SW; Cooper LR; Spyrakis F; Luisi BF; Campanin; Bettati. Cryo-EM Structures of Staphylococcal IsdB Bound to Human Hemoglobin Reveal the Process of Heme Extraction. Proc. Natl. Acad. Sci. U.S.A 2022, 119 (14), e2116708119. 10.1073/pnas.2116708119. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

JACS SI

RESOURCES