Both proteins contain a FecR domain, a module of roughly 100 amino acids that is usually associated with regulation of gene expression of iron transport system proteins in other bacteria. Earlier work by the same research group had shown that two other FecR-containing proteins, LIC_11122 (Lsa19) and LIC_11834 (Lsa33), interact with human extracellular matrix and plasma components, while the abundant outer membrane protein LipL45, which also carries this domain, had been characterized as a candidate adhesin. Because iron limitation acts as a potent environmental signal that reprograms leptospiral metabolism and alters the expression profile of its outer membrane proteins, the team set out to functionally characterize LIC_10902 and LIC_20185 and to determine whether the FecR domain itself mediates host interactions. To address this question, LIC_10902 was divided into two fragments, F1, which contains the FecR domain, and F2, which lacks it.

In silico analysis predicted a molecular mass of approximately 42.5 kDa and an isoelectric point of 9.14 for rLIC_10902, and 28.2 kDa with a pI of 6.87 for rLIC_20185. PFAM analysis revealed a single FecR domain in each protein, predicted to be surface exposed, and CELLO algorithms indicated a high probability of outer membrane anchoring. TMHMM predicted LIC_10902 as a transmembrane protein with an intracellular N-terminus, residues 12 to 29 embedded in the membrane, and the remainder surface exposed, while LIC_20185 was identified as a putative lipoprotein by LipoP and SignalP, both programs detecting a signal peptide encompassing amino acids 1 to 19. Conservation analysis showed that LIC_10902 is highly conserved among the P1 subclade of Leptospira species capable of causing leptospirosis, whereas LIC_20185 displayed high coverage across both P1 and P2 subclades, with identity between pathogenic and intermediate species exceeding 50 percent, features that suggest these proteins could be important for bacterial virulence.

Experimentally, the target genes were amplified from the L. interrogans sv. Copenhageni genome, cloned into the pAE vector, and recombinant proteins were produced in Escherichia coli BL21 Star (DE3) pLysS. Purification was achieved by metal affinity chromatography, and the purified proteins were structurally characterized by circular dichroism spectroscopy on a Jasco J-810 spectropolarimeter. The spectra exhibited a positive peak around 185 nm and a corresponding negative peak in the 200 to 220 nm range, indicative of a predominantly beta-sheet secondary structure. BeStSel analysis predicted that the FecR domains are composed mainly of beta-sheets, with rLIC_10902 showing 47.3 percent beta-sheets and only 5.5 percent alpha-helices, and rLIC_20185 showing 45.3 percent beta-sheets and 9.3 percent alpha-helices, findings corroborated by PSIPRED predictions and AlphaFold tertiary structure modeling. Interestingly, rLIC_20185 appeared to undergo fragmentation into two smaller products ranging between 10 and 15 kDa, a behavior reminiscent of L. interrogans LipL45 and Pseudomonas aeruginosa FoxR, both of which possess a conserved GT motif located in a loop region between two beta-sheets of the FecR domain, suggesting a shared self-cleavage mechanism usually associated with cell-surface signal transduction.

Subcellular localization was evaluated by fluorescence-activated cell sorting using low-passage L. interrogans sv. Copenhageni strain Fiocruz L1-130. When the antibody against the cytoplasmic control VapB2 was used, approximately 11.4 percent of bacteria were labeled, while the antibody against the outer membrane protein LipL46 labeled 37.8 percent. Strikingly, antibodies raised against rLIC_10902 and rLIC_20185 labeled approximately 52.9 percent and 77.7 percent of the bacteria, respectively, exceeding the labeling achieved with the established outer membrane control and providing strong evidence that both proteins are located, at least partially, on the leptospiral outer membrane. The recombinant proteins were also recognized by leptospirosis serum samples: rLIC_10902 reacted with 48 percent of 100 convalescent-stage samples, while rLIC_20185 reacted with 30 percent. Further evaluation against 20 paired samples revealed that rLIC_10902 was recognized by only 6.25 percent of MAT-negative sera and 50 percent of MAT-positive sera, suggesting that antibodies against these proteins are produced later, most probably during the convalescent phase of the disease, and that neither protein exhibits relevant diagnostic potential.

The functional assays demonstrated a broad interaction repertoire. Among extracellular matrix and cell-exposed molecules, rLIC_10902 did not bind any of the tested components, but rLIC_20185 interacted with elastin, E-cadherin, and laminin, with only the interactions with elastin and E-cadherin being dose-dependent and saturable, and therefore considered specific. The affinity of rLIC_20185 for E-cadherin, estimated at 881.70 nM, was similar to that reported for the rLIC_10831 protein dimer, and its affinity for elastin was 358.30 nM, comparable to other leptospiral elastin ligands such as LipL21 and OmpL37. Both proteins also bound the complement components C4b, C7, C8, and C9 in a dose-responsive manner, and both interacted with the plasma proteins plasminogen and vitronectin in dose-dependent and saturable patterns, with dissociation constants for plasminogen of 8.061 nM for rLIC_10902 and 92.28 nM for rLIC_20185, and for vitronectin of 725.1 nM and 480.7 nM, respectively. Notably, the affinity of both proteins for the terminal component C9, at 94.11 nM and 126.5 nM, was the highest among leptospiral outer membrane proteins known to interact with the complement system, indicating that this interaction may occur under physiological conditions.

Fragment experiments pinpointed the FecR domain as the mediator of these interactions. The F1 fragment, which contains the FecR domain, bound C8, C9, and plasminogen, whereas the F2 fragment, lacking the domain, could not interact significantly with any of the molecules analyzed. Because plasminogen is characterized by five kringle domains that typically engage lysine residues in bacterial ligands, the team tested whether the interaction could be disrupted by 6-aminocaproic acid, a lysine analogue. The binding of both proteins to plasminogen was indeed reduced in the presence of varying concentrations of ACA, confirming that the interaction occurs through kringle domains via lysine residues. Furthermore, both proteins captured plasminogen directly from normal human serum, and when bound to either purified plasminogen or plasminogen derived from serum, the captured zymogen was converted into its active form, plasmin, in the presence of urokinase-type plasminogen activator. Because plasmin is a broad-spectrum serine protease capable of degrading extracellular matrix components and cleaving antibodies and C3b, these data suggest that the FecR proteins could facilitate leptospiral penetration through host barriers and contribute to immune evasion and dissemination to target organs through sequestration of the host plasminogen/plasmin system, a strategy shared by pathogens such as Candida albicans, Streptococcus species, Mycobacterium tuberculosis, Borrelia burgdorferi, and Treponema denticola.

Given the importance of complement resistance for survival in the host bloodstream, the ability of the FecR proteins to inhibit zinc-induced polymerization of C9, the terminal component required for membrane attack complex formation, was assessed. Both proteins inhibited the polymerization of 3 micrograms of C9 in a dose-dependent manner: 1.25 micrograms had no visible effect, 2.5 micrograms partially reduced the presence of C9 polymers, and 5 micrograms completely inhibited polymerization, as evidenced by the total absence of C9 polymers and a corresponding increase in the monomeric C9 band. Inhibiting the polymerization of any complement component reduces the probability of membrane attack complex formation, characterizing a mechanism of immune evasion. To test whether these functions translate into serum resistance in a native context, the coding sequences of both proteins were cloned into the pMaGroES vector with a C-terminal 6x-histidine tag and transferred by conjugation into the non-pathogenic saprophytic Leptospira biflexa serovar Patoc. Western blotting confirmed that all recombinant clones produced the target proteins, although expression levels appeared low. However, survival assays with 20 percent active normal human serum showed that production of the FecR proteins did not increase resistance to complement attack; on the contrary, production of LIC_20185 negatively affected the survival of the recombinant leptospires at the 3-hour time point, while all three LbP derivatives showed similar survival when exposed to heat-inactivated serum, supporting the hypothesis that complement factors were responsible for the observed reduction.

The authors propose several explanations for this discrepancy. The wild-type human C9 present in serum may differ structurally from the recombinant human C9 used in the interaction assays, possibly presenting a higher affinity for the FecR proteins. The amount of FecR protein produced by recombinant LbP under the control of the GroES promoter is unknown and may not be sufficient to reproduce the results obtained with purified proteins. Differences in tertiary structure resulting from production in different heterologous systems, along with potential post-translational modifications, may also alter protein function. These findings reinforce the limitations of using simplified systems and heterologous hosts as surrogate organisms to study complex host-pathogen interactions, and highlight the importance of integrating the activity of recombinant proteins within more complex biological and environmental contexts to achieve a deeper understanding of leptospirosis pathogenesis.

Collectively, the study suggests that LIC_10902 and LIC_20185 participate at the interface between L. interrogans and its host. In vitro assays revealed that rLIC_20185 binds the host tissue molecules elastin and E-cadherin, while both proteins could contribute to immune evasion by trapping C8, C9, and plasminogen, probably through the FecR domain, thereby preventing polymerization of the C9 component. Their roles in pathogenesis are reinforced by their amino acid sequence conservation in virulent strains. However, the fact that production in the saprophytic surrogate did not confer protection against human serum highlights the multifactorial nature of leptospiral virulence, suggesting that the full functionality of these proteins may depend on a native context involving precise membrane topology and synergistic cooperation with other virulence factors that simplified in vitro models cannot fully replicate.

,

Subject of Research: Host interaction functions of LIC_10902 and LIC_20185 highlight the multifactorial nature of Leptospira interrogans pathogenesis

Article Title: Host interaction functions of LIC_10902 and LIC_20185 highlight the multifactorial nature of Leptospira interrogans pathogenesis

Article References: Gaspar, J. P., Paixao, J. E. L., Surdel, M. C., Coburn, J., Teixeira, A. F., & Nascimento, A. L. T. O. (2026). Host interaction functions of LIC_10902 and LIC_20185 highlight the multifactorial nature of Leptospira interrogans pathogenesis. Frontiers in Cellular and Infection Microbiology, 16, Article 1988027. https://doi.org/10.3389/fcimb.2026.1988027

Image Credits: AI Generated

DOI: 10.3389/fcimb.2026.1988027

Keywords: Host, interaction, functions, highlight, multifactorial, nature, Leptospira, interrogans, pathogenesis, scientific research