More than 125 years after its discovery by Robert Koch, Mtb remains one of the most devastating pathogens worldwide, causing over 10 million new infections and 1.5 million deaths annually (World Health Organization, 2023). Despite the availability of effective antibiotics, the standard treatment requires a prolonged combination therapy of at least 6 months with two to four drugs. This regimen is often associated with severe side effects and poor patient compliance. The alarming rise of multidrug-resistant and extensively drug-resistant strains further hampers therapeutic success (Khawbung et al., 2021). Given these challenges, there is an urgent need for novel, safe, and effective therapeutic strategies that can complement or enhance current anti-tuberculosis regimens. Antimicrobial peptides (AMPs) have emerged as promising candidates in this context. AMPs are short, mostly cationic peptides that typically act by disrupting microbial membranes, thereby exhibiting broad-spectrum activity against bacteria, fungi, and viruses (Yeaman and Yount, 2003). They are evolutionarily conserved components of innate immunity and have been isolated from diverse sources including bacteria, plants, invertebrates, and mammals.

In humans, a rich diversity of bioactive peptides—collectively termed the peptidome—can be found in tissues and body fluids. The peptidome encompasses not only classical proteolytic fragments of larger precursor proteins but also peptides derived from short open reading frames, non-canonical translation events, alternative splice variants, and small secreted proteins. Despite its complexity and biological relevance, the human peptidome remains largely unexplored as a reservoir of endogenous antimicrobial compounds. Endogenous AMPs offer several advantages over conventional antibiotics. As naturally occurring host molecules, they are less likely to induce adverse side effects or trigger immune reactions. Their mechanism of action, typically based on electrostatic interactions with bacterial membranes, makes the development of resistance less likely. Furthermore, their relatively small size facilitates large-scale chemical synthesis at reasonable cost (Hancock and Sahl, 2006; Yeaman and Yount, 2003). To systematically explore the human peptidome for novel AMPs, we previously established a platform to generate peptide libraries from various human tissues and fluids and to screen these libraries for antimicrobial activity (Münch et al., 2014; Bosso et al., 2018). Using this approach, we previously identified the antimycobacterial peptides Angie1 and Lys-H1 from serum (Noschka et al., 2021; Maier et al., 2024) and NapFab from bronchoalveolar lavage fluid (Beitzinger et al., 2021). Since tuberculosis is primarily a pulmonary disease, and more than 95% of infected individuals are able to contain the pathogen without ever developing active disease (Migliori et al., 2021), we hypothesized that the human lung harbors innate defense molecules capable of controlling Mtb infection. To test this, we generated a peptide library from healthy human lung tissue, screened it for activity against virulent Mtb and functionally characterized the most potent bioactive molecule (HBB(112–147, RapTB).

To identify antimicrobial peptides from lung lysate, 60 fractions were screened for antimetabolic activity against Mtb (Figure 1). Fraction 42 exhibited activity exceeding 30%. Repeated sub-fractionation led to the identification of sub-subfraction 40, which showed antimetabolic activity of over 90%. Mass spectrometry analysis of subfraction 40 revealed a signal at 7798.09 Da (deconvoluted spectrum), which turned into a signal at 3957.07 Da after carbamidomethylation, indicating the presence of a dimer formed by one intermolecular disulfide bridge (Figure 2). The MS/MS sequencing of the carbamidomethylated sample revealed the presence of HBB 112–147, VCVLAHHFGKEFTPPVQAAYQKVVAGVANALAHKYH (Supplementary Figure 1). This C-terminal peptide has previously been detected in human placenta and menstrual blood and showed antimicrobial activity, extracellular bacteria and viruses (Liepke et al., 2003; Mak et al., 2004; Mak et al., 2007; Groß et al., 2020).

To validate the antitubercular activity of RapTB, the synthetic peptide was tested against extracellular Mtb in an antimetabolic assay. RapTB exhibited a maximum of 53% inhibition at 50 μM (Figure 3A). Higher concentrations did not further enhance activity. To investigate structural effects on the pathogen, Mtb was incubated with RapTB (50 μM) and analyzed by transmission electron microscopy. Compared to untreated bacilli, which appeared as intact rods, RapTB-treated Mtb displayed profound morphological alterations, including vesicle formation, membrane swelling, and loss of membrane integrity (Figure 3B), suggesting membrane disruption as a likely mode of action.

Since Mtb primarily resides in macrophages, the cytotoxicity of RapTB was assessed against primary human macrophages after 24 h of exposure. At 50 μM, RapTB showed minimal cytotoxicity (13%). At lower concentrations (0.0005–5 μM), cell viability remained above 79% (Figure 4), indicating that RapTB is well tolerated by host cells.

To determine whether RapTB could reach intracellular Mtb, we investigated its uptake by macrophages using confocal laser scanning microscopy. Since a RapTB -specific antibody was required for detection in confocal microscopy and RapTB is a fragment of hemoglobin ß, a panel of commercially available anti-hemoglobin ß antibodies were tested for cross-reactivity, one of which recognized RapTB (Supplementary Figure 2). After 24 h, RapTB localized predominantly to vesicular structures near the nucleus (Figure 5). However, not all macrophages internalized the peptide, and no direct colocalization with intracellular bacilli was observed or it could not be detected at the sensitivity level of the microscope.

To investigate the RapTB localization on a molecular level in particular with respect to the intra-cellular Mtb, we used single molecule localization using dSTORM (Figure 6), resulting in super-resolution optical microscopy images. In the dSTORM-reconstructions of the macrophages treated with RapTB (0.5 μM), individual RapTB molecules could be localized throughout the cytosol of macrophages. The molecular density of RapTB inside the diffraction-limited region covered by the Mtb normalized by the cytosolic molecular density yields a value of 0.92 ± 0.07 when averaged over all donors (n = 3) and reconstructions (n = 12). This demonstrates that RapTB does not preferentially colocalize with Mtb intracellularly. The fact that RapTB is underrepresented in the areas occupied by Mtb indicates, that Mtb reside in compartments with low peptide uptake.

To assess the antimycobacterial of RapTB, infected macrophages were treated with RapTB (50 μM) for 4 days, followed by lysis and quantification of viable bacteria by determining the number of colony forming units on agar plates. No reduction in intracellular Mtb was observed (Figure 7A). This result was corroborated using the mycobacterial growth inhibition assay (MGIA) as an independent objective approach (Figure 7B). Together, these findings demonstrate that while RapTB is highly active against extracellular Mtb, it lacks efficacy against intracellular bacilli, likely due to inefficient colocalization with the pathogen inside host cells.

Here, we report the identification and characterization of HBB(112–147), a 39-amino-acid C-terminal fragment of β-hemoglobin, now termed RapTB. RapTB exhibited potent antimycobacterial activity against extracellular Mtb, likely via disruption of the bacterial cell wall. Although RapTB was efficiently internalized by human macrophages, it did not co-localize with intracellular bacilli and failed to inhibit intracellular Mtb.

RapTB belongs to a family of endogenous antimicrobial peptides, known as hemocidins, which are generated by proteolytic cleavage of hemoglobin. We and others described this and related fragments in placenta and menstrual blood and demonstrated broad-spectrum antimicrobial activity against Gram-positive and Gram-negative bacteria, fungi, and HSV-2 (Liepke et al., 2003; Mak et al., 2004; Mak et al., 2007; Groß et al., 2020). These peptides were shown to be active under acidic conditions, with their efficacy modulated by ionic strength and divalent cations such as Mg²⁺ and Ca²⁺, and to act synergistically with other endogenous AMPs including HBD-1, LL-37, and HNP-1 (Mak et al., 2007). In those studies, HBB(112–147) was identified through independent biochemical approaches and shown to be processed by physiological proteases such as Napsin A and Cathepsin D (Liepke et al., 2003; Beitzinger et al., 2021). However, antimycobacterial activity had not been previously reported. Our current study extends the known functional profile of this peptide by demonstrating that RapTB exhibits potent activity against Mycobacterium tuberculosis, the causative agent of tuberculosis. This new activity was discovered through the screening of a human lung-derived peptide library, emphasizing the potential role of hemoglobin-derived AMPs in pulmonary host defense. The identification of RapTB in lung tissue- the primary site of Mtb infection- highlights its likely physiological relevance and introduces a new layer of function to this well-characterized peptide. While RapTB effectively killed extracellular Mtb, it failed to co-localize with intracellular bacilli and showed no activity in infected macrophages. This emphasizes the importance of cellular trafficking and intracellular access for antimicrobial function, especially in the context of intracellular pathogens such as Mtb.

The systematic screening of peptide libraries derived from human tissues has proven to be an effective strategy for identifying endogenous AMPs. Using this approach, we previously discovered angiogenin from hemofiltrate, which showed both extracellular and intracellular anti-Mtb activity (Noschka et al., 2021). In contrast to angiogenin, RapTB lacks intracellular activity but is highly effective against extracellular Mtb. Nevertheless, RapTB could contribute to protection against tuberculosis during the early stages of infection when the pathogens logarithmically replicate in the alveolar space before adaptive immunity is initiated (Ryndak and Laal, 2019). RapTB shares important biochemical features with angiogenin (122 aa) including its cationic charge that promotes interaction with negatively charged mycobacterial membranes (Hancock and Sahl, 2006). With a length of 36 amino acids, RapTB is structurally intermediate yet functionally distinct. Interestingly, RapTB is generated by proteolytic cleavage of hemoglobin by Napsin A (Beitzinger et al., 2021), indicating potential crosstalk between host proteases and AMP production.

Our findings also support a broader concept: the proteolytic processing of abundant precursor proteins such as hemoglobin can generate multiple antimicrobial peptides with complementary or synergistic activity. Proteases like Napsin A and Cathepsin D, which are active in lung and placenta, cleave hemoglobin into bioactive fragments including RapTB and related hemocidins (Liepke et al., 2003; Beitzinger et al., 2021; The UniProt Consortium et al., 2023; Walther et al., 2010). These peptides have been shown to act synergistically with other AMPs such as HBD-1 and LL-37 (Mak et al., 2007), suggesting that localized proteolysis may shape a combinatorial AMP response. This mechanism may enhance antimicrobial efficacy, broaden pathogen coverage, and reduce the likelihood of resistance development.

The presence of hemocidins in lung tissue may reflect erythrocyte leakage due to mechanical disruption or inflammation. This is particularly relevant in tuberculosis, where capillary erosion and hemoptysis are common features (Wang et al., 2024). Such events could release hemoglobin into the lung parenchyma, allowing local proteases to generate bioactive peptides that contribute to the innate immune response.

Despite its lack of intracellular activity in its native form, RapTB may still contribute to host defense by limiting early extracellular Mtb replication. Since one of the primary goals of TB therapy is to control this initial bacterial expansion, AMPs like RapTB may serve as valuable adjuncts, especially in cases of multidrug-resistant tuberculosis. In addition to direct antimicrobial activity, hemoglobin-derived peptides have been shown to enhance cytokine release and macrophage activation, suggesting potential immunomodulatory functions as well (Roth et al., 1994). Furthermore, as RapTB is efficiently internalized by macrophages, it could potentially complement the activity of conventional antimycobacterial drugs. For instance, antimycobacterial agents may alter phagosomal integrity or cytosolic trafficking, thereby enabling RapTB to access intracellular Mtb and act synergistically.

To overcome the current limitations in intracellular activity, several strategies may be pursued. One approach is to improve intracellular delivery using nanocarrier systems. Dendritic mesoporous silica nanoparticles (DMSNs) have shown promise for enhancing peptide uptake, providing sustained release, and maintaining low cytotoxicity (Xu et al., 2019; Croissant et al., 2018; Amin et al., 2021; Du and Qiao, 2015). Similarly, liposomal encapsulation can improve intracellular delivery and efficacy of AMPs (Alzahrani et al., 2022). Targeted delivery via conjugation to toxins like C3bot from Clostridium botulinum, which selectively enters monocyte-derived cells, offer the possibility to optimize targeting to infected macrophages (Fellermann et al., 2022). Nanodiamonds are an emerging platform for AMP delivery, enabling enhanced intracellular accumulation, real-time imaging, and co-delivery of multiple therapeutic agents (Chow et al., 2011; Chen et al., 2009; Wu and Weil, 2022). Another strategy to improve RapTB function involves chemical modification of the peptide backbone. Substitution of L-amino acids with D-enantiomers can increase serum stability and reduce immunogenicity, as shown in VEGF and other peptides (Kreil, 1997; Garton et al., 2018; Uppalapati et al., 2016). Such modifications could prolong the half-life of RapTB and protect it from proteolytic degradation.

In summary, we identified RapTB as a hemoglobin-derived AMP with potent activity against extracellular Mtb. Our findings extend the known antimicrobial spectrum of hemocidins and support a role for protease-generated AMPs in pulmonary host defense. Ongoing studies aim to optimize the stability and intracellular trafficking of RapTB using rational peptide design and advanced delivery systems, with the goal of evaluating its potential for adjunctive therapy in severe or drug-resistant tuberculosis.