The alarming rise of antimicrobial resistance (AMR) in KP necessitates a multi-pronged approach, including judicious antibiotic use, enhanced infection control, and continual exploration of novel therapeutic strategies [40]. Repurposing existing drugs, for example, the rapid FDA approval of Remdesivir for COVID-19 [41], presents a promising avenue for combating KP infections. Our study lays the groundwork for repurposing Rifaximin against KP infections. While its poor gut absorption may pose challenges in in-vivo studies our in-vitro data demonstrate its antibacterial activity against KP. This paves the way for designing novel RfaH inhibitors based on rifaximin scaffold, potentially with improved gut absorption. By synergistically integrating insights from disease pathophysiology, drug pharmacology, and computational analysis, we can prioritize promising drug candidates for KP infections. This interdisciplinary approach holds immense potential for accelerating the development of effective KP therapies, ultimately improving patient outcomes and mitigating the healthcare burden.
Cloning, expression and purification of RfaH
The RfaH gene was cloned into pET-28a(+) vector using gene-specific primers with 6X His tag. The positive colonies were confirmed by using restriction digestion enzymes NdeI and XhoI (Fig. 1A). The plasmid was then transformed into DH5α to increase the copy number and in BL21 (DE3) cells for expression. The positive clones were screened using a specific antibiotic kanamycin, which was further confirmed by sequencing (Fig. 1B). His-tagged recombinant RfaH protein was purified using Ni-NTA affinity chromatography. Eluted fractions were analyzed by SDS-PAGE depicting a single purified band (Fig. 1C).
Fluorescence binding studies
To investigate the antibacterial effect of Rifaximin through RfaH inhibition, we employed fluorescence binding studies. The aromatic residues, particularly tryptophan, tyrosine, and phenylalanine, have unique fluorescence properties that are influenced by factors like solvent polarity, temperature, and interactions with other molecules. When a ligand binds to a protein, it can alter the local environment around these aromatic residues, leading to changes in their fluorescence emission intensity, spectra, or lifetime. This helps to monitor the formation and dynamics of protein-ligand complexes in real-time, providing valuable insights into binding affinity, specificity, and potential mechanisms of action.
We used a quenching approach to assess the interaction between Rifaximin and RfaH, determining binding parameters like the binding constant (Ka) and the number of binding sites (n). Through fluorescence binding studies, we unravelled a strong interaction between Rifaximin and RfaH. Dose-dependent quenching of intrinsic fluorescence of RfaH indicated complex formation with Rifaximin (Fig. 2A). Fitting the quenching data to the modified Stern-Volmer equation yielded Ka of 7.38 x 106 M− 1 and n of 1 (Fig. 2B). This suggests a strong association of Rifaximin with a single binding site on RfaH. This robust interplay, falling within the typical range for protein-ligand complexes, implicates a potential role of Rifaximin in RfaH inhibition and related pathways, paving the way for further investigation into its antibacterial mechanisms and the development of novel therapeutic strategies.
Determination of MIC and MBC for Rifaximin against KP
To quantitatively assess the antibacterial efficacy of Rifaximin, we determined the MIC and MBC values against the K. pneumoniae (ATCC 700603) strain. The INT-colorimetric assay revealed a MIC of 100 µM, highlighting the potential of Rifaximin against KP. While MIC tells us the lowest concentration just prevents bacterial growth, the MBC tells us whether the agent kills the bacteria. This distinction is crucial because even if an inhibitor stops bacteria from growing, they might still be alive and potentially cause harm later. Time-kill curve experiments established an MBC of 200 µM, signifying almost complete eradication at this concentration. Notably, even at the MIC of 100 µM, Rifaximin achieved a remarkable 4-log10 reduction in bacterial population compared to the initial inoculum of 1.5 X 105 CFU/mL (Fig. 3). These findings suggest that Rifaximin is a promising antibacterial agent against KP, warranting further investigation of its therapeutic potential, particularly in the face of multidrug-resistant strains [42].
Capsule production estimation
Capsules play a significant role in bacterial virulence by enhancing drug resistance and hindering the immune system's ability to recognize surface antigens [43]. RfaH homologs are crucial regulators in diverse bacterial species (from E. coli to B. amyloliquefaciens) controlling the expression of operons responsible for producing capsules, LPS core, antibiotics, toxins, and pili [44]. Disrupting RfaH in K. pneumoniae reduces capsule production, mirroring effects seen in E. coli due to the high similarity of their capsule biosynthesis clusters [45, 46]. A recent study has demonstrated Eco RfaH's ability to suppress Rho-dependent termination within capsule operons, further underscoring the importance of RfaH in capsule regulation [47].
A study has investigated if Eco RfaH regions essential for E. coli gene activation are also crucial in K. pneumoniae, employing a lux reporter assay. They used the RfaH gene to delete the K. pneumoniae TOP52 strain, rendering it deficient in endogenous RfaH activity. The lux reporter utilized the Photorhabdus luminescens lux operon placed downstream of an ops element. This element is known to bind both Eco and Kpn RfaH. Notably, a similar reporter was previously used to identify key functional residues in Eco RfaH. The study findings revealed that both Eco RfaH and Kpn RfaH significantly increased lux expression (p < 0.0001 compared to vector control), mirroring their effects on LPS and capsule biosynthesis operon activation. This suggests a high degree of functional conservation in the Eco RfaH regions critical for gene activation across these bacterial species [15].
In another study mutation in the RfaH gene resulted in a severe attenuation (over 10,000-fold decrease) of the mutant strain's growth within the lungs compared to the wild-type strain. This growth defect was significantly restored by complementation with the wild-type RfaH gene, highlighting the critical role of RfaH for bacterial fitness in the lung environment. Furthermore, the RfaH mutant exhibited a smaller colony size compared to the wild type, indicative of a potential impairment in capsule biosynthesis. India ink staining revealed a substantial capsule surrounding the wild-type bacteria, while the RfaH mutant lacked a visible capsule [45].
In both E. coli and KP, the protein RfaH plays a critical role in maintaining cell envelope integrity. Deletion of RfaH in E. coli results in dramatic sensitivity to the detergent sodium dodecyl sulfate (SDS), mirroring the effect of a polar mutation within its target operon, which was responsible for lipopolysaccharide (LPS) biosynthesis [48]. This sensitivity can be alleviated by mutations in rho, a termination factor [49]. KP exhibits a similar reliance on RfaH, with its deletion leading to reduced capsule production [45]. This parallels observations in E. coli, suggesting a conserved role for RfaH in capsule biosynthesis due to the close resemblance of the corresponding gene clusters in both species [46].
Mutations within genes encoding rfaH, in K. pneumoniae, have also been linked to the development of phage resistance in this bacterium. RfaH plays an essential role in regulating and synthesizing both CPS and LPS, which are important components of the bacterial cell wall. Studies have consistently shown that mutations affecting genes involved in CPS and/or LPS synthesis contribute to phage resistance. This resistance mechanism arises from the loss of specific receptors on the bacterial surface due to the altered cell wall composition, hindering phage attachment and subsequent infection [50, 51]. These findings highlight the crucial contribution of RfaH to cell envelope stability and its intricate interplay with other regulatory factors like Rho in both E. coli and KP.
In K. pneumoniae ATCC 700603, treatment with 100 µM Rifaximin led to a dramatic reduction in capsule production, exceeding 50% compared to untreated controls (Fig. 4A). To further affirm the effect of Rifaximin on KP outer capsule, 6 hours cultivated K. pnuemoniae cells were negatively stained and visualized [2]. As observed, Rifaximin treated and untreated K. pnuemoniae cells showed variation in bacterial outer capsule thickness. The untreated cells showed comparatively thicker outer capsule when compared to sub-MIC Rifaximin treated cells that showed smaller cells with relatively thin outer capsule (Fig. 4B). This observation suggests a potential role for Rifaximin in inhibiting RfaH, a critical protein in the KP capsule biosynthesis pathway.
The capsule, a polysaccharide-based extracellular structure, serves as a vital shield against diverse environmental assaults, including phagocytosis by the immune system, phage infection, and desiccation [52]. Consequently, a substantial decrease in capsule content, as induced by Rifaximin, is likely to heighten the susceptibility of K. pneumoniae cells to these external stresses, potentially culminating in cell death. Further investigation is warranted to elucidate the precise mechanism by which Rifaximin interacts with RfaH and to assess the functional impact of this capsule reduction on KP virulence and its overall adaptability within the host and environment [53, 54].
Molecular docking
The molecular docking between RfaH and Rifaximin revealed a binding affinity of -9.3 kcal/mol. An in-depth analysis of the interaction between Rifaximin and the RfaH binding pocket was conducted for all 9 docked conformations of the ligand. While Rifaximin displayed interactions with RfaH at multiple sites, its favored docking position exhibited the highest binding strength when compared to alternative positions. Literature studies have demonstrated that disrupting specific contact points in Eco RfaH, namely with the β’CH (Tyr54), ops DNA (Arg73), βGL (Thr66), and S10 (Ile146), impairs RfaH-dependent gene activation in E. coli and the functional importance of these contact points extends to K. pneumoniae also. As anticipated, disrupting interactions with β'CH, S10, and ops DNA, abolishes the activity of operons regulated under RfaH. This suggests that some RfaH interaction points are universally critical for both E. coli and K. pneumoniae bacterial species [55, 56]. Our docking results indicated that the rifaximin forms bonds with the Tyr54, Phe78, Arg80, and Asp147 residues of the RfaH protein. As Tyr54 which is a critical residue of β′ clamp helices (CH) domain and thus for RfaH activity, hence disruption of this specific contact point by rifaximin will lead to RfaH inhibition. The binding configuration of Rifaximin with RfaH is illustrated in Fig. 5. Rifaximin binds within the binding pocket cavity of RfaH, interacting with residues situated in the β′ clamp helices (CH) domain, as depicted in Figs. 5A and B. Figure 5C indicates a deep penetration of Rifaximin to the pocket of RfaH.
MD simulations
The MD simulation is employed to gain insights into the atomic-level dynamics of protein-ligand complexes [57]. Moreover, it aids in evaluating the flexibility of the docked complex in comparison to the native protein state [58]. In this investigation, MD simulation was applied to appraise the stability of protein-ligand docked complexes, specifically those involving RfaH and Rifaximin within a water model. Employing the CHARMM36 force field, we conducted a 100-nanosecond simulation for both the RfaH and RfaH-Rifaximin complex. Detailed information on MD simulation outcomes is provided in Table 1.
Table 1
Various dynamic and structural parameters were examined and analyzed for a period of 100 ns for the RfaH-Rifaximin complex.
System | RMSD (nm) | RMSF (nm) | Rg (nm) | SASA (nm2) | #H-bonds |
RfaH | 0.39 | 0.17 | 1.81 | 114.26 | 96 |
RfaH-Rifaximin | 0.46 | 0.22 | 1.77 | 114.14 | 96 |
The protein undergoes structural dynamics upon binding to a ligand molecule. The calculation of root mean square deviation (RMSD) is a fundamental method employed to quantify the structural alterations in a protein following ligand binding. RMSD values for both the native state and the protein-ligand complex are charted across a 100 ns timeframe. The RfaH native structure exhibits slight fluctuations during this period. Notably, the time evolution of RMSD for the RfaH-Rifaximin complex displays some fluctuations after 60 ns, indicating that the system undergoes complexity and experiences minor instability during the simulation period (Fig. 6A).
During MD simulation, the residual flexibility of a protein over duration is assessed through a root mean square fluctuation (RMSF) plot. We have generated RMSF plots for both the RfaH native state and the RfaH-Rifaximin complex (Table 1). Throughout the simulation, the RfaH-Rifaximin complex exhibits slightly elevated RMSF values. Despite this, both systems demonstrate nearly synchronized RMSF distributions (Fig. 6B). This implies that the protein-ligand complex maintains stability post-binding. The RMSF analysis reveals minimal differences in RMSF between the two systems, signifying a stable complex.
In MD simulation, the evaluation of the structural folding and conformational dynamics of a protein involves estimating the radius of gyration (Rg). The Rg is computed by determining the average distance of each atom from the center of mass of the protein molecule, employing the square of each atom's distance. Throughout the simulation, variations in the size and shape of the protein molecule are observed through the Rg, providing insights into the stability of the protein. Throughout a 100 ns duration, we computed the Rg values for both the RfaH native structure and the RfaH-Rifaximin complex (Fig. 7A). There is not any significant difference in the Rg values between both systems. By the 100 ns mark, the Rg values for both the free protein and the complex converge. This convergence suggests that RfaH, upon binding to Rifaximin, exhibits stable conformational dynamics and folding.
The solvent-accessible surface area (SASA) indicates the portion of the protein's surface that is accessible to solvent molecules. Assessing the solvent-accessible surface area is a pivotal technique for evaluating interactions between a protein and a ligand. This measure is valuable in evaluating the stability of the protein-ligand complex and identifying potential binding sites. The analysis of SASA is widely used to understand the stability and folding properties of proteins and protein-ligand complexes. We have graphed the SASA values for both the RfaH native structure and the RfaH-Rifaximin complex throughout the entire simulation duration (Fig. 7B). The mean SASA value for both systems remains constant, exhibiting no significant alteration. Throughout the simulation period, no notable distinctions are observed, and the values converge by 100 ns. This convergence implies a stable structural folding and dynamics of the protein upon binding to the ligand.
Dynamics of hydrogen bonds
The stability of the protein-ligand complex relies on the establishment of hydrogen bonds [59]. Intra-molecular hydrogen bonds were calculated for both the RfaH native structure and RfaH after binding with Rifaximin, and the results were plotted over 100 ns duration (Fig. 8A). The plot generated indicates that there were no significant alterations observed in the hydrogen bonding interactions within the RfaH protein upon the formation of a complex with Rifaximin. The plot displays a constant number of hydrogen bonds for both systems. Intermolecular hydrogen bonds were also estimated to infer the stability of interactions of RfaH with Rifaximin. These interactions revealed the formation of up to six hydrogen bonds, with four consistently present bonds throughout the trajectory (Fig. 8B).
Evaluation of secondary structures
Examining the dynamics of the secondary structure content of protein is a means to understand its conformational behavior and folding mechanism [60]. We calculated the alterations in the secondary structure for RfaH when bound to Rifaximin. The structural elements in the unbound RfaH exhibit nearly constant and equilibrated characteristics throughout the 100 ns simulation period (Fig. 9A). However, a small decrease in the β-sheets and a slight increase in α-helix and content of RfaH can be seen upon compound binding (Fig. 9B). The average number of residues engaged in secondary structure formation differs in the case of the RfaH-Rifaximin complex compared to free RfaH (Table 2). Despite this, there is no major change observed in the secondary structure of RfaH upon the binding of Rifaximin, indicating strong stability of the complex.
Table 2
Evaluation of Secondary Structures during MD simulation of RfaH native and RfaH-Rifaximin complex
| RfaH | RfaH-Rifaximin |
Coil | 36 | 38 |
β-sheet | 37 | 34 |
β-bridge | 1 | 2 |
Bend | 15 | 15 |
Turn | 17 | 16 |
α-helix | 53 | 56 |
310-helix | 2 | 0 |
PCA and FEL analysis
The PCA is a crucial technique for assessing the collective motion of atoms in a protein-ligand complex [61]. PCA is employed to analyze the conformational changes in both the native RfaH and the RfaH-Rifaximin complex, utilizing projections of Cα atoms to estimate the conformational dynamics of these systems. The plots illustrating these analyses are presented in Fig. 10A. The subspaces occupied by free RfaH closely align with those of the protein-ligand complexes. The native RfaH occupies one subspace, while the complex occupies two subspaces, signifying a reduction in stability within the complex. The protein-ligand systems show some variability concerning the free state of the protein.
Assessing the protein folding mechanism involves the utilization of Free Energy Landscape (FEL) analysis. This analysis is employed to determine global and local minima points within the energy landscape of a protein. The FEL plots for both the unbound RfaH and the RfaH-Rifaximin complex are depicted in Figs. 10B and C. Regions in deep blue color within the plots represent low-energy states, closely associated with the native states. The free state of the protein showed one large basin. In the case of RfaH-Rifaximin three distinct basins are formed. It suggests that the global minimum of free RfaH is slightly disturbed by the binding of Rifaximin. In summary, the FEL analysis suggests that the interaction of Rifaximin with RfaH does not induce protein unfolding throughout the 100 ns timeframe.