Retro-Inversion Imparts Antimycobacterial Specificity to Host Defense Peptides

Retro-Inversion Imparts Antimycobacterial Specificity to Host Defense Peptides | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Retro-Inversion Imparts Antimycobacterial Specificity to Host Defense Peptides Scott Medina, Hugh Glossop, Gebremichal Gebretsadik, Sabiha Sultana, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6497899/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Antimicrobial host defense peptides are promising alternatives to resistance prone small molecule antibiotics. To overcome the poor physiologic stability of these therapeutic candidates it is common to prepare proteolytically resistant retro-inverso analogues, where sequence backbone direction and amino acid chirality are reversed. However, in many cases, gains in stability are offset by altered assembly propensities and reduced biologic potency. Here, we show that, contrary to the dogma for non-mycobacterial pathogens, retro-inversion of antimycobacterial host defense peptides improves their potency, specificity and host safety by an order of magnitude. Biophysical assays suggest that altered mycomembrane thermodynamics, instead of improved proteolytic stability, plays a causative role in retro-inverso mediated potency gains. Additional bacteriologic assays using a lead retro-inversed candidate, MAD1-RI, demonstrates this analogue rapidly sterilizes both replicating and dormant cultures of Mycobacterium tuberculosis , is effective towards drug-resistant clinical isolates of the pathogen, and synergistically enhances the activity of co-incubated antibiotics. Transcriptomic studies uncover complementary membrane destabilizing and metabolic mechanisms of antitubercular action for MAD1-RI, and in doing so identify sequence retro-inversion as a simple, but powerful, modality in the de novo design of non-natural antimycobacterial peptides. Biological sciences/Biochemistry/Peptides Biological sciences/Biotechnology/Molecular engineering/Antimicrobials/Antibiotics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Main Mycobacterial lung and disseminated infections arise from several species of environmental and pathogenic mycobacteria. While M. tuberculosis ( Mtb ), the causative agent of tuberculosis, has been studied for centuries, non-tuberculous mycobacteria are emerging pathogens that also cause pulmonary and soft tissue infections, particularly in cystic fibrosis and immunocompromised patients. Increasing global incidence of both tuberculous and non-tuberculous mycobacterial infections, in addition to their acquisition of pan-drug resistance, 1 has urged the development of novel therapeutic strategies. Peptide antibiotics, particularly host-defense peptides (HDPs), are promising alternatives to traditional small molecule antibiotics in treating drug-resistant infections. 2 For example, colistin (polymyxin E) is now a drug of last resort against pan-resistant, gram-negative bacteria in clinical settings. 3 Due to the diversity of HDP modes of action, which are mechanistically distinct from classical small molecule antibiotics, phenotypic resistance towards HDPs is expected to be rare. In fact, despite millennia of exposure to HDPs as part of innate immune responses, widespread acquisition of resistance has not yet been documented. Indeed, antimicrobial peptides have co-evolved with bacterial pathogens, leading to a diverse array of membrane-lytic HDPs produced across species, 4,5 habitats, 6 and time. 7,8 Yet, natural HDPs are limited to the twenty canonical, L -chiral, amino acid building blocks, thereby making them proteolytically unstable. Sequence retro-inversion, where both backbone direction (retro) and amino acid chirality (inversion) are reversed, represents an attractive strategy to impart innate protease resistance to HDPs, while maintaining side chain topology. 9,10 However, nearly four decades of research have shown that gains in peptide stability from retro-inversion often come with a penalty of decreased biologic activity due to altered secondary structure and self-assembly propensity. 11-13 Here, we show that, contrary to the prevailing paradigm observed in non-mycobacterial pathogens, retro-inversion of HDPs improves their potency by nearly an order of magnitude towards mycobacteria. Bacteriologic assays suggest this enhanced potency is only realized for mycobacterial targets, and that retro-inversion does not appreciably change HDP toxicity towards non-mycobacterial species. Consequently, the lead HDP, identified from a library of screened retro-inverted sequences, demonstrates a 4-fold increase in selectivity towards Mtb over non-mycobacteria, and a greater than 30-fold reduction in host toxicity, compared to the native peptide. Additional bacteriologic testing of this lead candidate, referred to as MAD1-RI, shows the sequence is capable of sterilizing replicating and dormant Mtb cultures, synergistically enhances the activity of clinical TB antibiotics, and is effective towards drug-resistant clinical Mtb isolates. Finally, using a battery of biochemical, biophysical and transcriptomic analyses we elucidate the mechanistic basis of this enhanced activity, and demonstrate that gains in potency operate through altered mycomembrane thermodynamics of retro-inverted sequences. Thus, we identify retro-inversion as a potentially unique driver of anti-mycobacterial specificity in the design of HDPs that utilize mechanisms distinct from the traditional dogma of increased proteolytic stability. Results Bacteriologic screening of retro-inverted sequences To explore the bacteriologic effects of retro-inversion we began by chemically modifying the model antimycobaterial peptide, MAD1 ( m ycomembrane- a ssociated d isruption sequence 1). 14 MAD1 exerts antitubercular activity through disruption of the mycolic acid rich outer mycomembrane, putatively due to its structural mimicry of the mycobacteria specific porin MspA. Inhibition assays utilizing a series of Mtb complex strains ( M. bovis, mc 2 6230, mc 2 7901, H37Ra, and H37Rv strains) and non-tuberculous mycobacteria ( M. smegmatis and M. abscessus ) cultures showed that retrosynthesis (backbone reversal, MAD1-R) or chiral inversion (enantiomerization, MAD1-I) of MAD1 did not, individually, alter minimum inhibitory concentration (MIC) of the analogues relative to the parent sequence ( Table 1 ; see Supplementary Table S1 for peptide mass and purity). Conversely, the combination of these features, producing MAD1-RI, resulted in an approximate order of magnitude improvement in MIC against several of the tested mycobacteria relative to native MAD1. For example, single digit micromolar MICs were recorded for MAD1-RI against M. smegmatis , M. bovis and Mtb (H37Ra). Against the virulent Mtb strain H37Rv, MAD1-RI demonstrated an MIC of 40µM, while the native MAD1 was weakly active (MIC > 40 µM). MAD1-RI was similarly effective against a panel of drug-resistant clinical Mtb isolates (PMID: 30908506), with MICs ≤ 40µM (Supplementary Table S2). All four MAD1 sequences were subsequently counter screened against the non-mycobacterial pathogens, methicillin-resistant S. aureus (MRSA), P. aeruginosa , and A. baumanii to assess mycobacterial selectivity. A selectivity index (SI, see Table 1) was then calculated to quantitively compare microbial specificity. Remarkably, retro-inversion yielded a > 16-fold increase in mycobacterial selectivity (MAD1-RI SI = 32) relative to the MAD1, MAD1-R, and MAD1-I peptides (SI = 1–2). Finally, safety of the MAD1 and MAD1-RI sequences were compared by measuring their half-maximal inhibitory concentration (IC 50 ) against A549 human lung epithelial cells, and subsequently calculating a toxicity index (TI, see Table 1). Gratifyingly, retro-inversion decreased the peptide’s toxicity, increasing the TI from 5.3 for the native sequence to 368.4 for MAD1-RI; nearly two orders of magnitude improvement. To evaluate the impact of proteolytic stability on MAD1-RI activity we performed degradation assays in the presence of proteinase K, a particularly promiscuous and aggressive protease ( Fig. 1a ). Chromatographic results show that, as expected, backbone reversal alone (retrosynthesis, MAD1-R) did not enhance the proteolytic stability of the parent sequence, with both showing near complete digestion after 1 hour. Conversely, amino acid enantiomerization (inversion, MAD1-I) yielded a sequence that was highly resistant to proteolysis under the tested conditions, which was conserved when combined with retrosynthesis to produce MAD1-RI. Additional time-dependent stability studies in the presence of Mtb -secreted proteases indicate that native MAD1 was completely proteolyzed after 14 days ( Fig. 1b ), while MAD1-RI generally remained stable up to two weeks in culture ( Fig. 1c ). Taken in context with our MIC results, these findings indicate that enhanced proteolytic stability, alone, does not explain the significantly greater antimycobacterial potency of the retro-inversed MAD1-RI peptide, as typical dogma suggests. Instead, later studies indicate altered thermodynamic stability of retro-inverted sequences in the unique mycolate membrane of Mtb may play a causative role. We next set out to assess whether these mycobacterial-specific effects represent a conserved phenomenon of retro-inversion across membrane-active HDPs. To test this, we prepared five pairs of native peptides and their retro-inverted counterparts (see Table 1), which included two analogues of MAD1 identified through artificial intelligence guided screening (MAD2.2, MAD5.1) and three antimicrobial sequences isolated from insect venoms (Latarcin 3a, 15 Euminitin F, 16 Pinipesin 17 ). Three of the retro-inverted peptides, MAD2.2-RI, Latarcin 3a-RI and Euminitin F-RI, showed a significant enhancement in activity relative to the native sequence across the majority of mycobacterial cultures studied; in some cases improving activity by 16 times. MAD5.1, on the other hand, showed no improvement following sequence retro-inversion. This may be a consequence of its high formal charge (+ 8), where cationic membrane depolarization, and subsequent detergent-like destabilization, by the peptide may be insensitive to retro-inversion mediated conformational changes. Additionally, we observed no significant antibacterial activity of the centipede derived Pinipesin peptide across all tested species. Performing additional experiments under nutrient poor, and reduced temperature conditions (30°C), as were employed in the original publication, 17 did not yield improved potency (data not shown). Nevertheless, collectively our studies suggest that retro-inversion may represent a broadly applicable design strategy to enhance the antimycobacterial activity of HDPs, which operates through mechanism(s) independent of enhanced proteolytic stability. We further explore this assertion through a series of conformational assays performed in both model systems and mycobacterial membranes. Conformational consequences of peptide retro-inversion Circular dichroism was employed to establish the secondary structure of the four conformational MAD1 variants in pH 7.4 phosphate buffer ( Fig. 2a,b ). As previously reported, MAD1 displays mixed α-helical (minima at 206nm and 219nm) and β-sheet (minima at 212nm) conformations under physiologic conditions, 14 with an unusual exciton band at 228 nm indicative of a tryptophan zipper. 18 , 19 As expected, MAD1 and MAD1-I (Fig. 2a), as well as MAD1-R and MAD1-RI (Fig. 2b), congeners display mirrored CD spectra as a result of residue enantiomerization. Unexpectedly, although the spectra of MAD1 and MAD1-I show the putative tryptophan zipper signal at 228nm, the retro-synthetic MAD1-R and MAD1-RI analogues have lost this feature. This suggests backbone reversal may alter residue side chain orientation and disrupt tryptophan-mediated peptide assembly. Additional in silico conformational analyses utilizing the BeStSel algorithm 20 – 22 show that MAD1 and MAD1-RI possess similar anti-parallel β-sheet content, with accompanying turn-like arrangements, in low ionic strength conditions ( Fig. 2c ). However, nearly 50% of the spectra for both sequences are ascribed to disordered conformations (other), suggesting significant conformational flexibility that may potentiate varied and environmentally sensitive self-assembly phenotypes. Based on this observation, we next compared the structural dynamics of MAD1 and MAD1-RI in the presence of increasing concentrations of sodium dodecyl sulfate (SDS) micelles, a common model for anionic lipid membranes ( Fig. 2d,e ). In the absence of SDS, MAD1 displays intense exciton-coupled bands at 212 and 228 nm, indicative of long-range assembly. This is significantly disrupted in the presence of SDS, with plots of the concentration-dependent change in ellipticity showing that MAD1’s β-sheet (θ 212nm ) and tryptophan-zipper (θ 228nm ) structures are nearly completely abolished in the presence of ≥ 1mM SDS (Fig. 2d, right). MAD1-RI, on the other hand, was significantly more conformationally rigid, showing only minor changes in structural states up to 90mM SDS. Parallel studies performed in the presence of the Mtb cell envelope glycolipid, trehalose dimycolate, showed a comparable trend, with MAD1 becoming more disordered and MAD1-RI conversely adopting greater structural order (Supplementary Fig. S1 ). Finally, CD experiments conducted under varying ionic strength ( Fig. 2f,g ) and temperature ( Fig. 2h,i ) conditions displayed a similarly contrasting conformational behavior. Remarkably, MAD1-RI maintains its secondary structure even at temperatures as high as 95°C (Supplementary Fig. S2), suggesting significant secondary structural stability. Taken together, our ex cellulo results suggest that retro-inversion leads to greater conformational rigidity of the MAD1 peptide in physiologic solutions. This, in turn, may lower the entropic barrier to interpolation and assembly of the sequence within mycomembranes, thereby enhancing its antimycobacterial potency. Using a series of cell based and biophysical assays, we next test this assertion in the context of living mycobacterial cells. Fluorescent confocal and electron microscopy were initially utilized to probe MAD1-RI localization to the Mtb cell surface, as well as investigate peptide-mediated morphologic changes in myco-envelope architecture. Fluorescence micrographs shown in Fig. 3a demonstrate that the peptide rapidly engages the Mtb cell wall and co-localizes with outer membrane mycolic acids ( Fig. 3b ). When employed at sub-lethal concentrations (0.5x MIC), these outer membrane interactions cause the typically smooth Mtb surface to adopt a ruffled morphology ( Fig. 3c ). Increasing concentration to 5x MIC led to a ruffling of the pathogen outer membrane, which was complemented by the presence of extracellular aggregates; presumably released intracellular contents. Close inspection of the Mtb cell surface revealed the formation of MAD1-RI nano-assemblies ( Fig. 3d ), indicating the pore-formation capabilities of MAD1 have been maintained after retro-inversion. To further interpret the nature of these assemblies, and better understand how conformational plasticity of the peptides may contribute to their membrane-dependent assembly phenomena, small angle x-ray scattering (SAXS) and analytical ultracentrifugation studies were designed. SAXS curves shown in Fig. 3e demonstrate that MAD1-RI adopts a more extended and enlarged conformation relative to MAD1, with a radius of gyration (R g ) and maximum dimension of 8.2 and 22.2 Å for MAD1 and 10.2 and 30.6 Å for MAD1-RI, respectively. This further supports our assertion that MAD1-RI is more conformationally rigid than the parent sequence. A double maximum observed in the P(r) SAXS curves ( Fig. 3f ) suggests that both peptides adopt an ensemble of homodimeric conformers. Area under the curve (AUC) analyses confirmed this observation and demonstrated a dimeric species for both peptides in physiologic solution ( Fig. 3g ). Curiously, the predicted molecular weight of the MAD1-RI dimer was ~ 300 Da larger than the MAD1 dimer, despite their identical true molecular weight. It is possible this difference is due to changes in the hydration shell and bound ions for the MAD1-RI dimer assembly. Further SAXS experiments performed in the presence of SDS yielded a model of peptide assembly that suggests MAD1 adopts a head-to-head dimer, while MAD1-RI associates head-to-tail ( Fig. 3h ). This MAD1-RI dimer conformation would span approximately 3 nm in its longest dimension, 1 nm larger than the prediction for MAD1 dimers. Importantly, the head-to-tail alignment of MAD1-RI dimers may allow propagation of these conformers to completely span the ~ 8 nm thickness of the outer Mtb mycomembrane. Thus, compared to MAD1, MAD1-RI may more readily associate into a pore that spans the Mtb outer membrane and potentiate the leakage of intracellular contents; further corroborating the intracellular aggregates observed during SEM microscopy (Fig. 3d). Antitubercular activity of MAD1-RI We next evaluated time dependent killing of Mtb by MAD1-RI, in both proliferating and dormant TB cultures. Initial studies confirmed that, at equivalent concentrations, MAD1-RI showed a ~ 4-log greater anti-TB efficacy relative to MAD1 ( Fig. 4a ). Additional kinetic growth assays demonstrated that the Minimum Bactericidal Concentration (MBC, > 99.9% reduction in colony forming units) of MAD1-RI towards Mtb was twice its MIC at ~ 80 µM ( Fig. 4b ) and complete culture sterilization occurred at 4x MIC (160 µM) between 7 and 14 days of exposure (Fig. 4a,b). Next, combinatorial studies were performed to identify potential synergistic pairings of MAD1-RI and clinical TB antibiotics ( Fig. 4c , Supplementary Fig. S3). Fractional inhibitory concentration (FIC) scores show that MAD1-RI did not antagonize the activity of any of the tested antibiotics, and for the majority of candidates acted independently or in an additive fashion. Two antibiotics, Moxifloxacin (MOX) and Meropenem, displayed synergistic activity with MAD1-RI (FIC score ≤ 0.5). MOX is a second-line fluoroquinolone commonly used to treat drug-resistant TB, and therefore this drug was prioritized for follow up combinatorial kill kinetic studies. In proliferating Mtb ( Fig. 4d ), the co-incubation of MAD1-RI (40 µM) and MOX (0.25 µg/mL) lead to complete sterilization of the culture after two weeks of incubation, which represents a ~ 6-log enhancement in activity relative to the monotherapy controls. This activity is also superior to the front-line drugs Isoniazid and Rifampicin, which, despite a ~ 2-log reduction in the first week of exposure, permitted regrowth of treated Mtb cultures after 3 and 10 days, respectively (Supplementary Fig. S4). Finally, we performed similar Mtb kill kinetics studies under nutrient starvation ( Fig. 4e ) and hypoxic ( Fig. 4f ) conditions. This was done as Mtb adopts a quiescent state in the phagosomes of infected macrophages within granulomatous lesions. This metabolically inactive state promotes persistence of the pathogen by rendering it phenotypically resistant to most antibiotics. This was confirmed in our experimental models, where under nutrient starvation and hypoxic conditions Isoniazid was inactive (Supplementary Fig. S5), and the efficacy of MOX significantly blunted (Fig. 4e,f). Conversely, we hypothesized that MAD1 would remain active under these conditions due to the insensitivity of its mycomembrane destabilizing mechanism of action to the pathogen’s metabolic state. Results shown in Supplementary Fig. S6 confirm this assertion, and again show enhanced activity of MAD1-RI relative to MAD1 toward persister Mtb . As expected, the combination of MAD1-RI and MOX showed more effective killing than the individual therapies in nutrient starved Mtb cultures (Fig. 4e). However, under hypoxic conditions, the combinatorial formulation was equal in its efficacy to MOX alone, and was significantly less active than the anaerobic positive control drug Metronidazole (Supplementary Fig. S5b). Notably, although effective in vitro , metronidazole lacks efficacy in in vivo TB models due, at least in part, to insufficient anaerobic conditions to allow reductive activation of the drug. 23 In sum, our results show that MAD1-RI is effective towards dormant Mtb persisters, and suggest that modulation of particular metabolic pathways during dormancy may blunt the antimycobacterial activity of the peptide; an assertion further explored through genomic studies. MAD1-RI antimycobacterial mechanisms To characterize the transcriptional response to MAD1-RI exposure, H37Rv Mtb cultures were treated for 24 hours with 25 µM of the peptide before performing RNA sequencing. Results in Fig. 5a show that 16 genes were induced and 7 genes repressed (> 1-fold, q < 0.05, principal components analysis shown in Supplementary Figure S7). Within the upregulated data set, the five-gene operon espACD-Rv3613c-Rv3612c is particularly notable given its required role in ESX-1 secretion, which is essential for Mtb virulence and host cell survival. 24 Additionally, induction of the PE20 gene encodes for the PPE20 protein, which forms a complex with PE15 to facilitate molecular transport across the Mtb cell envelope. 25 Finally, MAD1-RI also induced Rv1057 , which is proposed to be upregulated in response to envelope stress induced by sodium dodecyl sulfate (SDS) exposure and stabilizes MmpL3 complexes that shuttle lipid components to the cell wall. 26 Thus, it was hypothesized that MAD1-RI-treated Mtb would exhibit increased sensitivity to membrane-targeting agents, such as detergents like SDS. Indeed, after 24 hours of drug treatment, survival differences were observed between treated and untreated cultures upon SDS exposure. Here, the viability of cultures pre-treated with MAD1-RI and lysocin E, a peptide that causes membrane disruption, 27 declined more rapidly and to a greater extent than untreated cultures, indicating heightened sensitivity to SDS ( Fig. 5b ). Together, these results suggest that MAD1-RI compromises the integrity of the Mtb cell envelope by disrupting cell wall secretion, transport and biosynthetic processes. To empirically test this assertion, we utilized the lipophilic dye DiOC2(3) to monitor MAD1-RI mediated changes in Mtb membrane potential. This fluorophore intercalates into the mycobacterial outer membrane and displays decreased red fluorescence as membrane potential is reduced. As expected, increasing MAD1-RI concentration from 1.25 to 37.5 µM led to a systematic reduction in DiOC2(3) fluorescence, producing similar membrane depolarization at the highest tested peptide concentration to the carbonyl cyanide 3-chlorophenylhydrazone (CCCP) protonophore positive control ( Fig. 5c ). Finally, it is interesting to note that MAD1-RI exposure also repressed prpD and Rv1129c ( prpR ) transcription in Mtb . Both genes encode for key enzymes in the methylcitrate cycle utilized by Mtb to process fatty acids as a carbon source during survival within macrophages. 28 Additionally, bfrB encodes the iron storage protein ferritin that regulates iron homeostasis in Mtb during survival and proliferation within infected host cells. 29 Thus, repression of these genes suggests that, in addition to its membrane destabilizing effects, MAD1-RI can disrupt fatty acid metabolism and lipid biosynthesis to compromise Mtb fitness in vivo . This may explain the reduction in MAD1-RI activity observed in our in vitro dormant models relative to the proliferating cultures (Fig. 4). Here, although lipid metabolism may be downregulated by the peptide, the pathogen is rescued by other carbon sources available in the rich media. However, evidence from animal studies 30 indicating that Mtb virulence in vivo is dependent on a shift to lipid catabolism suggests that, under nutrient depravation conditions within an infected host, MAD1-RI’s anti-TB activity may be significantly enhanced. Conclusions While peptide retro-inversion is an established approach to improve the proteolytic stability of therapeutic peptides, we show here that it also enhances the antimycobacterial potency and specificity of HDPs through mechanisms distinct from enzymatic digestion. MAD1-RI exhibits a particularly strong effect compared to other tested sequences, likely due to a gramicidin-like ion-leaking antimycobacterial mechanism. Biophysical studies suggest that this mode of action differs from general membrane destabilization, potentially involving a head-to-tail arrangement forming intercalating pores or a β-helical pore with two copies of the retro-inversed MAD1-RI sequence in the mycomembrane. We show this activity not only leads to rapid killing of replicating Mtb , but potently synergizes with clinically relevant TB antibiotics. These findings, in combination with its regulation of genes important to Mtb virulence and cell wall biosynthesis, suggest that MAD1-RI may be a potent addition to the anti-TB therapeutic arsenal. In addition to its clinical potential, the rapid binding of MAD1-RI to mycobacterial cells may also enable its use as a diagnostic probe to monitor pathogens in wastewater and agricultural settings. More broadly, our findings suggest that retro-inversion may represent a unique chemical approach to design narrow-spectrum bactericidal peptides with anti-mycobacterial specificity. Materials and Methods Peptide synthesis, purification and characterization All peptides were synthesized following conventional Fmoc/tBu solid phase peptide synthesis (SPPS) using an automated Liberty Blue 2.0 microwave synthesizer (CEM Corp., North Carolina) following a high-efficiency SPPS protocol. 31 Crude peptides were purified by HPLC and purity and mass spectra were confirmed by LCMS on a LCMS-2020 instrument (Shimadzu Corp.). Fluorescence images were collected with a Biotek Cytation 3 plate reader and Zeiss LSM 880 confocal laser scanning microscope and scanning electron micrographs were obtained with a Zeiss Sigma scanning electron microscope. Circular Dichroism (CD) spectroscopy CD measurements were conducted using a JASCO J-1500 spectrometer equipped with a Peltier-controlled PTC-517 thermostat cell holder. Spectra were recorded from 260 nm to 185 nm at a scan speed of 50 nm/min, with a bandwidth of 1 nm at 25°C. A 1 mm pathlength quartz cuvette was used, and the peptide concentration was 100 µM. A buffer blank was measured before the sample for baseline subtraction, and the data were converted to molar ellipticity. Analysis was performed using the Jasco Multivariate SSE program along with BeStSel algorithm 20 – 22 for single-spectrum analysis and secondary structure prediction. Temperature-dependent CD experiments were carried out from 20°C to 95°C in 5°C increments, with a heating rate of 1°C/min. Spectra were recorded at 50 nm/min from 260 nm to 185 nm, with a data pitch of 1 nm, a DIT of 4 seconds, and a bandwidth of 1 nm. Wavelength-dependent changes at 208 nm and 220 nm were plotted against temperature. All CD experiments utilized phosphate buffer except for data reported in Fig. 2 f and Fig. 2 g in which SDS in pure water was used to determine the effect of SDS concentration on secondary structure. AUC Methods Peptides, dissolved in phosphate buffer (pH 7.4) at a concentration of 33 µM, were loaded into 12 mm Epon-charcoal centerpieces within AUC cells featuring sapphire windows. These cells were placed into an An50 titanium rotor, which had been pre-equilibrated to 37°C, the experimental temperature. The rotor was then inserted into the chamber of a Beckman-Coulter Optima multiwavelength AUC equipped with absorbance optics. A full vacuum was applied in the chamber, and the rotor was allowed to re-equilibrate for 2 hours. A method scan was created using the UltraScan III software and transferred to the instrument, where the experiment began after temperature equilibration. The rotor was spun at 40,000 RPM for 11 hours, capturing radial scans every three minutes at 280 nm. Once the AUC experiment was complete, the data were imported into UltraScan III. Reference scans were automatically chosen to convert the raw radial intensity data into pseudo-absorbance. The air-liquid meniscus was manually selected for each sector. The dataset was also manually cropped, typically between 6.1 cm and 7.1 cm, and the first 5–10 scans, as well as those following complete sedimentation, were excluded from analysis. The edited data were then processed using the LIMS supercomputer at Penn State, with an S-value range set from 1 to 10, a resolution of 100, and a frictional ratio range from 1 to 4, with a resolution of 64. Time-invariant noise was also accounted for during the initial analysis. When the residuals fell below 0.003, the data were refitted, this time incorporating both time and radially invariant noise, along with 11 meniscus fits to ensure precise determination of the meniscus. Once the correct meniscus was identified, a final time and radial invariant noise fit was performed using an iterative method. Lastly, the data were analyzed using a genetic algorithm with Monte Carlo simulations (selecting 1–2 species per sample). 32 Monte Carlo simulations were run using 16 processors, and the resulting pseudo-3D plots were analyzed for final calculations of the s-value, frictional ratio, and molecular weight. BioSAXS Small angle X-ray scattering (BioSAXS) data were collected on peptides at a concentration of 4.5 mg/ml using X-rays with a wavelength of 1.54 Å from an in-house Rigaku MM007 rotating anode X-ray source. This system was coupled with the BioSAXS2000nano Kratky camera, which features OptiSAXS confocal max-flux optics specifically designed for SAXS experiments and a HyPix-3000 Hybrid Photon Counting detector for high sensitivity. The sample was positioned in a capillary with a detector-to-sample distance of 495.5 mm, calibrated using silver behenate powder from The Gem Dugout (State College, PA). The scattering vector q-space (q = 4πsin(θ)/λ, where 2θ is the scattering angle) ranged from q_min = 0.008 Å⁻¹ to q_max = 0.6 Å⁻¹. The X-ray beam energy was 1.2 keV, with a Kratky block attenuation of 22% and a beam diameter of approximately 100 µm. Peptide samples were automatically loaded onto a quartz capillary flow cell via a Rigaku autosampler, which was cooled to 4°C and aligned with the X-ray beam. The sample cell and entire X-ray flight path, including the beam stop, were kept in a vacuum (below 1 × 10⁻³ torr) to eliminate air scattering. The Rigaku SAXSLAB software controlled automated data collection for each peptide, incorporating thorough cleaning cycles between samples. Data reduction, including image integration, normalization, and background subtraction, was also handled by SAXSLAB software. Six ten-minute images, along with three replicates of both protein and buffer samples, were collected, averaged, and inspected to confirm that no radiation damage occurred. Overlays of the SAXS data confirmed no radiation decay or sample loss during the 60-minute collection period. Following this, buffer subtraction was performed to isolate the raw SAXS scattering curve of the peptide. The forward scattering intensity (I(0)) and the radius of gyration (Rg) were calculated using the Guinier approximation, which assumes that at very small angles (q < 1.3/Rg), the intensity follows I(q) = I(0)exp[− 1/3(qRg)²]. The results were consistent with the expected size of peptide dimers. Further analysis of the data, including radius of gyration, Dmax, Guinier fits, Kratky plots, and pair distance distribution functions, was carried out using the ATSAS software. Solvent envelopes were calculated with DAMMIF, an algorithm for deriving ab initio bead models directly from solution scattering data. Fluorescence and electron microscopy For fluorescence microscopy, Mtb mc 2 6230 was grown to an OD of 0.02 in supplemented Middlebrook 7H9 broth. The bacteria was isolated by centrifugation and re-suspended in 7H9 broth containing cyanine-5 tagged MAD1-RI (10 mM) for 30 min prior to drying and heat fixation on a glass microscopy slide. If necessary, Hoechst 33343 stain (1 mg/mL was also added to the bacteria prior to fixation). A glass cover slip was then adhered to the top of the sample with a dab of ProLong™ Diamond Antifade Mountant (Invitrogen) overnight in the dark. For SEM samples, bacteria were grown identically to the fluorescence microscopy protocol and incubated in the presence of 0.5 x MIC, 1 x MIC, and 5 x MIC of MAD1-RI for 30 min before being passed through a 0.2 micron pore filter disc. The bacteria trapped on the filter disc were then fixed with glutaraldehyde (2.5% v/v) in phosphate buffer for 30 min followed by step-wise dehydration with 10, 25, 50, 75, 85, 95, and 2 x 100% solutions of ethanol. The discs were then dehydrated with a Leica EM CPD300 Critical Point Dryer (Leica) prior to being mounted on titanium stubs and sputter coated with a 4.5 nm layer of iridium prior to viewing. Bacteriologic testing Antimicrobial activity was determined by standard broth microdilution minimal inhibitory assays wherein peptide was serially diluted across a 96-welled plate and bacteria adjusted to an OD 600 of 0.002 was added to each concentration of peptide. 27 All bacteria were grown and assayed in specialized broth at 37°C. M. smegmatis (mc 2 155) and Mtb complex lab strains ( M. bovis , mc 2 6230, H37Rv and H37Ra) were grown in Middlebrook 7H9 and supplemented 7H9 broth, respectively. Mtb clinical isolates were grown in Middlebrook supplemented 7H9 containing 40 mM pyruvate. All virulent strains of Mtb were handled under BSL-3 conditions following institutionally approved protocols. S. aureus (USA300), P. aeruginosa (PAO1), and A. baumanii (ATCC19606) were grown in Mueller-Hinton broth (MHB). 7H9 broth for the auxotrophic mc 2 6230 strain was prepared with glycerol (0.2% v/v), oleic acid-albumin-dextrose catalase (OADC; 10% v/v), Tween-80 (0.02% v/v), and pantothenic acid (50 mg/L). For colony growth, the mycobacteria species were grown on Middlebrook 7H10 agar with the same supplementation except for Tween-80. The other bacteria were all grown on MHB agar. Minimum inhibitory concentration (MIC) assays were performed by dissolving peptides to 320 µM in sterile nuclease-free water and 2-fold serially diluting the peptide stock in 7H9 broth (50 µL volume after dilution) in a sterile 96-well round bottom plate in triplicate. Bacterial growth was measured via OD 600 , and the culture was diluted to an OD 600 value of 0.002 for the mCherry expressing Mtb H37Rv and other Mtb clinical isolates. 50 µl of bacterial suspension was added to the treated wells in a 1:1 v/v ratio (100 µL final volume). 32 H37Rv and H37Ra growth were assessed by measuring fluorescence at 570/610 nm using a plate reader after incubating at 37°C for 96 hours. MTT (3- (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays were employed to evaluate the MIC against Mtb clinical isolates. The growth of clinical isolates was assessed by measuring fluorescence at 570 nm using a plate reader after 7 days of incubation. The MIC for each drug was defined as the lowest concentration that reduced fluorescence by 90%. Assays were repeated with at least 2 independent replicates for each strain (n ≥ 6). 33 Minimum bactericidal concentration (MBC) for replicating cultures was determined by adjusting exponentially growing Mtb to OD 600 of 0.1 and treating bacilli with 40, 80, 160 µM of MAD1-RI. Culture plating was done on 7H10 agar at day 0, 1, 3 and 7. The MBC was defined as the lowest concentration which reduced CFU count by 99% relative to the time-zero inoculum on day 7. Twenty-one-day kill kinetics were assessed to evaluate the bactericidal activity of MAD1-RI in combination with other drugs by CFU count. For this purpose, a mid-log-phase Mtb culture was diluted in fresh medium (OD 600 of 0.1). After aliquoting the culture to 30 ml square bottles, the indicated concentration of anti-TB drugs was added to each sample, and cultures were incubated for 21 days. Each culture dilution was plated at selected time intervals on 7H10 agar plates, and the Mtb CFU on each plate were enumerated after incubating at 37°C in a 5% CO 2 enriched atmosphere for 3–4 weeks. 27 Drug synergy Checkerboard assays were used to test for interactions between MAD1-RI and drugs with known anti-TB activities by the broth microdilution method in 96-well microtiter plates. After preparing serially diluted drugs on two different plates, 25 µl from each well of the MAD1-RI plate (plate 1) and another drug plate (plate 2) were transferred to the corresponding wells of a new plate (plate 3) and mixed (the final drug volume of each well was 50 µl). Then, like the MIC assay, 50 µl of mCherry expressing Mtb H37Rv suspension (OD 600 of 0.002) was added to the wells of each plate considered for testing. 32 The MIC for each drug or combination was defined as the lowest concentration that reduced fluorescence by 90% after incubating the plates at 37°C for 4 days. To evaluate the effect of each combination, the obtained MIC values were used to calculate the fractional inhibitory coefficient index (FICI) as follows: FICI = (MIC of drug A in combination/MIC of drug A alone) + (MIC of drug B in combination/MIC of drug B alone). The FIC index (FICI) calculated for each drug combination was categorized based on FICI: 0.5 to ≤ 1 as additive; and > 2 as antagonistic. 34 Kill curves with non-replicating Mtb To study the bactericidal activity of MAD1-RI against non-replicating mycobacteria (NRP), nutrient-starved and hypoxia-induced nonreplicating Mtb were used. To obtain the hypoxia-induced, NRP Mtb phenotype, H37Ra cultures in the early log phase were transferred to an anaerobic chamber (atmosphere of 5% CO 2 , 10% H 2 , 85% N 2 ) and grown for 3 weeks at 100 rpm, 37°C. Then, to evaluate the bactericidal activities of MAD1-RI against NRP Mtb , H37Ra cultures were transferred to 15 ml test tubes and treated with different concentrations of the indicated drugs. The cultures were then incubated for two weeks and plated on 7H10 agar at selected time intervals. 27 To obtain nutrient-starved NRP Mtb , H37Rv bacilli were grown to the exponential phase. Bacteria were then harvested by centrifugation (3,200 rpm; 4°C; 5 min) and washed twice with PBS containing 0.025% Tween 80. The bacteria were diluted to a final OD 600 of 0.2 in PBS and incubated for 3 weeks at 37°C. Clumps were removed by low-speed centrifugation (2000 x g ; 3 min) just before treatment. To evaluate the bactericidal activities of the peptides against nutrient-starved NRP, Mtb H37Rv was transferred to 30 ml square bottles, and treated with MAD1-RI, isoniazid, or moxifloxacin at the indicated concentration. Each culture dilution was plated at selected time intervals on 7H10 agar plates, and the Mtb CFU on each plate was enumerated after incubating at 37°C in a 5% CO 2 enriched atmosphere. 35 Transcriptional profiling and analysis For RNA isolation, Mtb H37Rv was cultured in 7H9 media to an OD 600 of 0.1 before being treated with 25 µM MAD1-RI for 24 h. 36 After treatment, mycobacterial RNA was extracted using a Zymogen RNA Miniprep Kit (Sigma-Aldrich, USA) with minor modifications to the manufacturers protocol. Briefly, cells were resuspended in trizol reagent and 1% Polyacyl carrier and disrupted using Zirconia beads (Biospec, Bartlesville, OK, USA) by beating two times for 1 minute and cooling on ice for two minutes in between. The samples were spun down at 14,000 RPM for 10 min at 4°C and the supernatant was transferred to clean tubes. Each sample was vortexed after 50 µl of BCP (bromo 3-chloro propane) was added. The samples were then harvested into a fresh RNase-free E-tube and one volume of ethanol (95–100%) was added directly to one volume sample (1:1). Samples were mixed well by vortexing. The sample mixtures were then loaded into a Zymo-Spin Column in a collection tube, and centrifuged at 14,000rpm for 1 minute. To clean up the RNA, the columns were washed with 400 µl RNA wash buffer by centrifuging for 30 seconds. To decontaminate DNA, DNase I Reaction Mix was added directly to the column matrix which was then incubated at RT (20–30℃) for 15 minutes, followed by centrifugation for 30 seconds. The samples were then washed with 400 µl Direct-zol RNA PreWash twice by centrifuging for 1 minute. Samples were washed once again by adding 700 µl RNA Wash Buffer to the column and centrifuging for 1 minute. Finally, 50 µl of Dnase/Rnase-Free Water was directly added to the column matrix and centrifuged for 1 minute. The eluted RNA was stored at -80℃ and sent for processing and sequencing. cDNA libraries were prepared by SeqCenter (Pittsburg, PA) with a Stranded Total RNA Prep using the Ribo-Zero Plus 563 Microbiome kit (Illumina Inc). cDNA libraries were sequenced using the Illumina Novaseq platform optimized for 150 bp paired-end reads and producing approximately 12 million reads. RNA sequencing data was analyzed by preprocessing the raw .fastq output files using the pipeline available at https://github.com/MDHowe4/RNAseq-Pipeline . FastQC was utilized to measure the quality control of the reads. Read lengths and t-overhangs were trimmed via Cutadapt, the minimum read length for which was set to 30 bp. Alignment to the Mtb H37Rv genome (NC_000962.3) was performed by the STAR aligner with spliced alignment detection disabled (--alignIntronMax 1). The read counts for given genes were procured via featureCounts, with exclusion criteria defined as genes with fewer than 10 reads across all samples. A negative binomial generalized linear model was created with DESeq2 to quantify the differential expression of genes. Significance was reserved for genes which met both a log2-fold change ≥ 1 or ≤ -1 and an adjusted p-value of < 0.05 criteria. The ggplot2 package was used to generate volcano plots for these data. SDS sensitivity assay and assessment of membrane potential Mtb H37Rv was grown in 7H9 medium supplemented with OADC and 0.05% Tween 80 to an OD600 of 0.5. Cultures were then treated with 160 µM MAD1-RI and 6 µg/ml Lysocin E for 24 hours. Following treatment, bacterial pellets were washed twice with sterile PBS and resuspended to an OD600 of 0.1 in PBS. SDS was added to a final concentration of 0.005%, and cultures were plated on 7H10 agar at 0 and 3 hours post-SDS exposure. After incubation at 37°C for three to four weeks, CFUs were enumerated, and percent survival at 3 hours was calculated relative to the starting CFUs at 0 hours. 36 Membrane potential was assessed as previously described, 37 with modifications. Briefly, mid-exponential- phase Mtb H37Ra cells grown in supplemented 7H9 medium were centrifuged (3,000 rpm for 10 min) and resuspended to a final OD 600 of 0.1 in 3 ml supplemented 7H9 medium at pH 7.0 in 30-ml square Nalgene bottles. Peptides and ionophores were added to the final concentrations indicated in the graph and incubated at 37°C. At 30 min, 180 µl was removed, 20 µl of 150 µM 3,3-diethyloxicarbocianide chloride [DiOC2(3)] was added, and the mixture was incubated at room temperature for 30 min. Cells were then washed, resuspended in supplemented 7H9 medium, transferred to a black-walled 96-well plate, and analyzed in a Molecular Devices SpectraMax M2 plate reader, and fluorescence was assessed by excitation of samples at 488 nm and recording the emissions at 530 nm and 610 nm. Ratios of the emission at 610 nm/emission at 530 nm were calculated. Declarations Data Availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Primary RNA-seq data are publicly available through the National Center for Biotechnology Information via SRA link https://www.ncbi.nlm.nih.gov/sra/PRJNA1157802 . Acknowledgements Research reported here was supported by SIG S10 of the National Institutes of Health under award number # S10-OD028589 for the small angle X-ray scattering, NIH grant S10 OD032215-01 for the Optima AUC and S10 OD030490 for the Wyatt SEC-MALS-DLS system to Dr. Neela Yennawar. We thank William R. Jacobs, Jr and Michael Berney of Albert Einstein College of Medicine for gifting the various mycobacterial strains used in this study. We also wish to thank the assistance of Ms. Julia Fecko at the X-ray Crystallography core at the Penn State Huck Institutes of the Life Sciences and the UMN BioSafety Level 3 Program for facility management. Funding for this work was provided by NIH R01-AI165996 to S.H.M. References Zhang, Y. & Yew, W. Mechanisms of drug resistance in Mycobacterium tuberculosis: update 2015. Int. J. Tuberc. Lung Dis. 19, 1276–1289 (2015). Rima, M., Rima, M., Fajloun, Z., Sabatier, J. M., Bechinger, B. & Naas, T. Antimicrobial Peptides: A Potent Alternative to Antibiotics. Antibiotics 10 (2021). Landman, D., Georgescu, C., Martin, D. A. & Quale, J. Polymyxins revisited. Clin. Microbiol. Rev. 21, 449–465 (2008). Rezaei Javan, R., Van Tonder, A. J., King, J. P., Harrold, C. L. & Brueggemann, A. B. Genome sequencing reveals a large and diverse repertoire of antimicrobial peptides. Front. Microbiol. 9, 2012 (2018). Lazzaro, B. P., Zasloff, M. & Rolff, J. Antimicrobial peptides: Application informed by evolution. Science 368, eaau5480 (2020). Santos-Júnior, C. D. et al. Discovery of antimicrobial peptides in the global microbiome with machine learning. Cell (2024). Maasch, J. R. M. A., Torres, M. D. T., Melo, M. C. R. & de la Fuente-Nunez, C. Molecular de-extinction of ancient antimicrobial peptides enabled by machine learning. Cell Host Microbe 31, 1260–1274.e1266 (2023). Wan, F., Torres, M. D. T., Peng, J. & de la Fuente-Nunez, C. Deep-learning-enabled antibiotic discovery through molecular de-extinction. Nat. Biomed. Eng. (2024). Chorev, M. & Goodman, M. A dozen years of retro-inverso peptidomimetics. Acc. Chem. Res. 26, 266–273 (1993). Goodman, M. & Chorev, M. On the concept of linear modified retro-peptide structures. Acc. Chem. Res. 12, 1–7 (1979). Doti, N., Mardirossian, M., Sandomenico, A., Ruvo, M. & Caporale, A. Recent Applications of Retro-Inverso Peptides. Int. J. Mol. Sci. 22 (2021). Li, C. et al. Limitations of peptide retro-inverso isomerization in molecular mimicry. J. Biol. Chem. 285, 19572–19581 (2010). Preston, G. W. Different directions for retro-inverso peptides. J. Pept. Sci. 28, e3384 (2022). Simonson, A. W. et al. Pathogen-specific antimicrobials engineered de novo through membrane-protein biomimicry. Nat. Biomed. Eng. 5, 467–480 (2021). de Moraes, L. et al. First generation of multifunctional peptides derived from latarcin-3a from Lachesana tarabaevi spider toxin. Front. Microbiol. 13, 965621 (2022). Rangel, M. et al. Chemical and biological characterization of four new linear cationic α-helical peptides from the venoms of two solitary eumenine wasps. Toxicon 57, 1081–1092 (2011). Chaparro-Aguirre, E., Segura-Ramírez, P. J., Alves, F. L., Riske, K. A., Miranda, A. & Silva Júnior, P. I. Antimicrobial activity and mechanism of action of a novel peptide present in the ecdysis process of centipede Scolopendra subspinipes subspinipes. Sci. Rep. 9, 13631 (2019). Cochran, A. G., Skelton, N. J. & Starovasnik, M. A. Tryptophan zippers: Stable, monomeric β-hairpins. Proc. Natl. Acad. Sci. 98, 5578–5583 (2001). Liu, J., Yong, W., Deng, Y., Kallenbach, N. R. & Lu, M. Atomic structure of a tryptophan-zipper pentamer. Proc. Natl. Acad. Sci. 101, 16156–16161 (2004). Micsonai, A. et al. Accurate secondary structure prediction and fold recognition for circular dichroism spectroscopy. Proc. Natl. Acad. Sci. 112, E3095-E3103 (2015). Micsonai, A. et al. BeStSel: webserver for secondary structure and fold prediction for protein CD spectroscopy. Nucleic Acids Res. 50, W90-W98 (2022). Micsonai, A. et al. Disordered–Ordered Protein Binary Classification by Circular Dichroism Spectroscopy. Front. Mol. Biosci. 9 (2022). Klinkenberg, L. G., Sutherland, L. A., Bishai, W. R. & Karakousis, P. C. Metronidazole Lacks Activity against Mycobacterium tuberculosis in an In Vivo Hypoxic Granuloma Model of Latency. J. Infect. Dis. 198, 275–283 (2008). Hunt, D. M. et al. Long-Range Transcriptional Control of an Operon Necessary for Virulence-Critical ESX-1 Secretion in Mycobacterium tuberculosis. J. Bacteriol. 194, 2307–2320 (2012). Boradia, V., Frando, A. & Grundner, C. The Mycobacterium tuberculosis PE15/PPE20 complex transports calcium across the outer membrane. PLoS Biol. 20, e3001906 (2022). Pang, X., Cao, G., Neuenschwander, P. F., Haydel, S. E., Hou, G. & Howard, S. T. The β-propeller gene Rv1057 of Mycobacterium tuberculosis has a complex promoter directly regulated by both the MprAB and TrcRS two-component systems. Tuberculosis 91 Suppl 1, S142–149 (2011). Wiegand, I., Hilpert, K. & Hancock, R. E. Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat. Protoc. 3, 163–175 (2008). Tang, S., Hicks, N. D., Cheng, Y. S., Silva, A., Fortune, S. M. & Sacchettini, J. C. Structural and functional insight into the Mycobacterium tuberculosis protein PrpR reveals a novel type of transcription factor. Nucleic Acids Res. 47, 9934–9949 (2019). Kurthkoti, K. et al. The mycobacterial iron-dependent regulator IdeR induces ferritin (bfrB) by alleviating Lsr2 repression. Mol. Microbiol. 98, 864–877 (2015). Savvi, S., Warner, D. F., Kana, B. D., McKinney, J. D., Mizrahi, V. & Dawes, S. S. Functional characterization of a vitamin B12-dependent methylmalonyl pathway in Mycobacterium tuberculosis: implications for propionate metabolism during growth on fatty acids. J. Bacteriol. 190, 3886–3895 (2008). Collins, J. M., Porter, K. A., Singh, S. K. & Vanier, G. S. High-Efficiency Solid Phase Peptide Synthesis (HE-SPPS). Org. Lett. 16, 940–943 (2014). Geberetsadik, G. et al. Lysocin E Targeting Menaquinone in the Membrane of Mycobacterium tuberculosis Is a Promising Lead Compound for Antituberculosis Drugs. Antimicrob. Agents Chemother. 66, e0017122 (2022). Martin, A. et al. Multicenter study of MTT and resazurin assays for testing susceptibility to first-line anti-tuberculosis drugs. Int. J. Tuberc. Lung Dis. 9, 901–906 (2005). Odds, F. C. Synergy, antagonism, and what the chequerboard puts between them. J. Antimicrob. Chemother. 52, 1 (2003). Gengenbacher, M., Rao, S. P. S., Pethe, K. & Dick, T. Nutrient-starved, non-replicating Mycobacterium tuberculosis requires respiration, ATP synthase and isocitrate lyase for maintenance of ATP homeostasis and viability. Microbiology 156, 81–87 (2010). Jeon, A. B. et al. 2-aminoimidazoles potentiate ß-lactam antimicrobial activity against Mycobacterium tuberculosis by reducing ß-lactamase secretion and increasing cell envelope permeability. PLoS One 12, e0180925 (2017). Peterson, N. D., Rosen, B. C., Dillon, N. A. & Baughn, A. D. Uncoupling Environmental pH and Intrabacterial Acidification from Pyrazinamide Susceptibility in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 59, 7320–7326 (2015). Table 1 Table 1 is available in the Supplementary Files section. Additional Declarations There is NO Competing Interest. Supplementary Files MAD1RISupplementaryInformationv.3.docx Supplementary Information Table1.docx Cite Share Download PDF Status: Published Journal Publication published 07 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6497899","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":448059914,"identity":"eab0265a-5d0f-4bd6-ac11-5157a339a83b","order_by":0,"name":"Scott Medina","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuUlEQVRIiWNgGAWjYDACCQaGAx8qbCAcHiK1MB6ccSaNNC3Mh3lbDpOgRX52d8IB3obzefIzEhgfvG0jQovBnbMbDkjuuF1scCOB2XAuUVokcjccMDxzO3GDRAKbNC8xWuRnALUktp1LnD8jgf03UVoYbgC1HGw7kNhwI4GNmSgtBkAtBxvOJBcbnHnYLDnnHHEO2/z5T4Vdnnx78sEPb8qIcRgUJDAwMDaQoB6iZRSMglEwCkYBDgAAlnI/68sxQi4AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-5441-2164","institution":"Pennsylvania State University","correspondingAuthor":true,"prefix":"","firstName":"Scott","middleName":"","lastName":"Medina","suffix":""},{"id":448059915,"identity":"a2817f31-e0cc-4425-9cbc-c3eb53bf69bf","order_by":1,"name":"Hugh Glossop","email":"","orcid":"","institution":"Penn State University","correspondingAuthor":false,"prefix":"","firstName":"Hugh","middleName":"","lastName":"Glossop","suffix":""},{"id":448059916,"identity":"98cf8c4f-841e-4094-88a1-f1add169f328","order_by":2,"name":"Gebremichal Gebretsadik","email":"","orcid":"","institution":"University of Minnesota","correspondingAuthor":false,"prefix":"","firstName":"Gebremichal","middleName":"","lastName":"Gebretsadik","suffix":""},{"id":448059917,"identity":"2e564220-9153-47ef-ab7d-325c12b3659d","order_by":3,"name":"Sabiha Sultana","email":"","orcid":"","institution":"Penn State University","correspondingAuthor":false,"prefix":"","firstName":"Sabiha","middleName":"","lastName":"Sultana","suffix":""},{"id":448059918,"identity":"76aa1c95-48ba-4791-a56e-0fd9b8590144","order_by":4,"name":"Nathan Schacht","email":"","orcid":"","institution":"University of Minnesota","correspondingAuthor":false,"prefix":"","firstName":"Nathan","middleName":"","lastName":"Schacht","suffix":""},{"id":448059919,"identity":"c6af2b86-4807-4d25-a697-fef0009b5cc8","order_by":5,"name":"Neela Yennawar","email":"","orcid":"","institution":"Penn State University","correspondingAuthor":false,"prefix":"","firstName":"Neela","middleName":"","lastName":"Yennawar","suffix":""},{"id":448059920,"identity":"968d735e-6dd6-48c9-8fb6-b016054ef629","order_by":6,"name":"Diptomit Biswas","email":"","orcid":"","institution":"The Pennsylvania State University","correspondingAuthor":false,"prefix":"","firstName":"Diptomit","middleName":"","lastName":"Biswas","suffix":""},{"id":448059921,"identity":"5df78f87-d0b8-41ab-ba00-80f88c540734","order_by":7,"name":"Anthony Baughn","email":"","orcid":"https://orcid.org/0000-0003-1188-4238","institution":"University of Minnesota","correspondingAuthor":false,"prefix":"","firstName":"Anthony","middleName":"","lastName":"Baughn","suffix":""}],"badges":[],"createdAt":"2025-04-21 17:15:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6497899/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6497899/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-67162-0","type":"published","date":"2025-12-07T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81964440,"identity":"9d85cf4a-e1e0-4c8d-85bf-3b1dc3cdb105","added_by":"auto","created_at":"2025-05-05 11:24:00","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":89487,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eProteolytic stability of MAD1 analogues.\u003c/strong\u003e (\u003cstrong\u003ea\u003c/strong\u003e) High-pressure liquid chromatograms of the indicated peptide (100 μM) following a 1 hour incubation in the absence (black) and presence (red) of proteinase K. (\u003cstrong\u003eb\u003c/strong\u003e,\u003cstrong\u003ec\u003c/strong\u003e) Stability of MAD1 (b) and MAD1-RI over two weeks of incubation in the proteolytic Mtb (mc\u003csup\u003e2\u003c/sup\u003e 6230) supernatant.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/a107950be285b7c4c910c562.png"},{"id":81964444,"identity":"af398192-4465-4ed8-8e92-cc1aa266f8f0","added_by":"auto","created_at":"2025-05-05 11:24:00","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":152638,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConformational dynamics of MAD1 analogues.\u003c/strong\u003e (\u003cstrong\u003ea\u003c/strong\u003e,\u003cstrong\u003eb\u003c/strong\u003e) CD spectra of MAD1 and MAD1-I (a), or MAD1-R and MAD1-RI (b), in phosphate buffer. (\u003cstrong\u003ec\u003c/strong\u003e) Secondary structure assignment for the MAD1 (top) and MAD1-RI (bottom) peptides. Structural classification was performed using the Beta Structure Selection (BeStSeL) algorithm. (\u003cstrong\u003ed\u003c/strong\u003e-\u003cstrong\u003ei\u003c/strong\u003e) \u003cem\u003eLeft\u003c/em\u003e: Full CD spectra and \u003cem\u003eRight\u003c/em\u003e: Ellipticity changes for the β-sheet (θ\u003csub\u003e212nm\u003c/sub\u003e) and tryptophan-zipper (θ\u003csub\u003e228nm\u003c/sub\u003e) structural features, for MAD1 (d,f,h) and MAD1-RI (e,g,i) as a function of SDS concentration (d,e), NaCl concentration (f,g) and temperature (h,i).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/cc07282d5f8c3ce555da2a1b.png"},{"id":81964457,"identity":"08864a4c-207d-4313-98e9-baeca78e7875","added_by":"auto","created_at":"2025-05-05 11:24:02","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":329784,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMycomembrane dynamics of MAD1-RI. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Fluorescent micrographs of \u003cem\u003eMtb\u003c/em\u003e (mc\u003csup\u003e2\u003c/sup\u003e 6230) co-stained with auramine-O (green) and Cy5-labeled MAD1-RI (red). Individual channel and merged image shown. Scale bar = 25 µM. (\u003cstrong\u003eb\u003c/strong\u003e) Fluorescent intensity line plot of the mycolic acid and MAD1-RI signals from the region delineated by the dashed yellow line in panel a, merged. (\u003cstrong\u003ec\u003c/strong\u003e) Scanning electron micrographs of \u003cem\u003eMtb\u003c/em\u003e (mc\u003csup\u003e2\u003c/sup\u003e 6230) treated with varying concentrations of MAD1-RI (as indicated at the top left corner). Scale bar = 1µM. (\u003cstrong\u003ed\u003c/strong\u003e) False colored scanning electron micrograph highlighting MAD1-RI assembled structures (tan) at the \u003cem\u003eMtb\u003c/em\u003e cell surface. (\u003cstrong\u003ee\u003c/strong\u003e,\u003cstrong\u003e f\u003c/strong\u003e) SAXS data with model fit (e) and P(r) (f) plots for MAD1 and MAD1-RI. (\u003cstrong\u003eg\u003c/strong\u003e) Simulated molecular weights of MAD1 and MAD1-RI species obtained by AUC. (\u003cstrong\u003eh\u003c/strong\u003e) Putative arrangement, based on SAXS and AUC simulations, of MAD1 and MAD1-RI within a scaled schematic of the mycobacterial outer membrane.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/e2a2a5c24985886d41649f83.png"},{"id":81964451,"identity":"83d0fc0a-6b16-41f4-91ec-cb118113c087","added_by":"auto","created_at":"2025-05-05 11:24:01","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":149703,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMAD1-RI antimycobacterial activity against Mtb. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Time-dependent kill kinetics of MAD1 and MAD1-RI in proliferating Mtb (H37Rv) cultures. (\u003cstrong\u003eb\u003c/strong\u003e) Concentration-dependent antimycobacterial efficacy of MAD1-RI (MBC = 80 µM). (\u003cstrong\u003ec\u003c/strong\u003e) Fractional inhibitory concentration plots of MAD1-RI and the indicated antibiotic. Independent, Additive and Synergistic fractional inhibitory concentration (FIC) scoring regions are highlighted. (\u003cstrong\u003ed \u003c/strong\u003e- \u003cstrong\u003ef\u003c/strong\u003e) Time-dependent kill kinetics of MAD1-RI, MOX or combination in (d) proliferating, (e) nutrient starved and (f) hypoxic \u003cem\u003eMtb\u003c/em\u003e cultures (H37Rv for panels d and e; H37Ra for panel f).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/52064bcb0c3b93f55be4e801.png"},{"id":81965755,"identity":"fdaaef13-13c8-4ffa-a012-1f8f50c51ea8","added_by":"auto","created_at":"2025-05-05 11:32:02","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":74483,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMAD1-RI mechanism of action.\u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eTranscriptional responses of \u003cem\u003eMtb \u003c/em\u003eto MAD1-RI. RNA-seq analysis was performed on \u003cem\u003eMtb \u003c/em\u003eH37Rv treated with MAD1-RI. (b) Increased sensitivity of MAD1-RI (160 µM) and Lysocin E (6 µg/mL) treated \u003cem\u003eMtb \u003c/em\u003eto SDS. (\u003cstrong\u003ec\u003c/strong\u003e) Membrane potential changes of DiOC2(3)-labeled \u003cem\u003eMtb \u003c/em\u003etreated in the absence (black) or presence of the indicated MAD1-RI (blue) or MAD1 (green) concentration. CCCP (100 μM) was used as the positive control. Statistical significance determined relative to untreated controls (blank) using Student’s t-test, with p values shown.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/b7e74b2c16de31d174a785aa.png"},{"id":100213950,"identity":"77ee39e2-9b06-4919-a123-933c8c6356ab","added_by":"auto","created_at":"2026-01-14 08:10:20","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1775721,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/34d7bb62-bedc-4e67-80d3-be92e546bcf3.pdf"},{"id":81964462,"identity":"e941d204-8941-4f19-ab52-85c6019ba2c2","added_by":"auto","created_at":"2025-05-05 11:24:02","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":489407,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"MAD1RISupplementaryInformationv.3.docx","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/11b62dec3de0996175ead79f.docx"},{"id":81964454,"identity":"310de78f-035b-43ce-9483-f9ade2bad9f7","added_by":"auto","created_at":"2025-05-05 11:24:01","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":20467,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6497899/v1/3c576b6b7b659665da077728.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Retro-Inversion Imparts Antimycobacterial Specificity to Host Defense Peptides","fulltext":[{"header":"Main","content":"\u003cp\u003eMycobacterial lung and disseminated infections arise from several species of environmental and pathogenic mycobacteria. While \u003cem\u003eM. tuberculosis\u0026nbsp;\u003c/em\u003e(\u003cem\u003eMtb\u003c/em\u003e), the causative agent of tuberculosis, has been studied for centuries, non-tuberculous mycobacteria are emerging pathogens that also cause pulmonary and soft tissue infections, particularly in cystic fibrosis and immunocompromised patients. Increasing global incidence of both tuberculous and non-tuberculous mycobacterial infections, in addition to their acquisition of pan-drug resistance,\u003csup\u003e1\u003c/sup\u003e has urged the development of novel therapeutic strategies.\u003c/p\u003e\n\u003cp\u003ePeptide antibiotics, particularly host-defense peptides (HDPs), are promising alternatives to traditional small molecule antibiotics in treating drug-resistant infections.\u003csup\u003e2\u003c/sup\u003e For example, colistin (polymyxin E) is now a drug of last resort against pan-resistant, gram-negative bacteria in clinical settings.\u003csup\u003e3\u003c/sup\u003e Due to the diversity of HDP modes of action, which are mechanistically distinct from classical small molecule antibiotics, phenotypic resistance towards HDPs is expected to be rare. In fact, despite millennia of exposure to HDPs as part of innate immune responses, widespread acquisition of resistance has not yet been documented. Indeed, antimicrobial peptides have co-evolved with bacterial pathogens, leading to a diverse array of membrane-lytic HDPs produced across species,\u003csup\u003e4,5\u003c/sup\u003e habitats,\u003csup\u003e6\u003c/sup\u003e and time.\u003csup\u003e7,8\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYet, natural HDPs are limited to the twenty canonical,\u0026nbsp;\u003cem\u003eL\u003c/em\u003e-chiral, amino acid building blocks, thereby making them proteolytically unstable. Sequence retro-inversion, where both backbone direction (retro) and amino acid chirality (inversion) are reversed, represents an attractive strategy to impart innate protease resistance to HDPs, while maintaining side chain topology.\u003csup\u003e9,10\u003c/sup\u003e However, nearly four decades of research have shown that gains in peptide stability from retro-inversion often come with a penalty of decreased biologic activity due to altered secondary structure and self-assembly propensity.\u003csup\u003e11-13\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eHere, we show that, contrary to the prevailing paradigm observed in non-mycobacterial pathogens, retro-inversion of HDPs improves their potency by nearly an order of magnitude towards mycobacteria. Bacteriologic assays suggest this enhanced potency is only realized for mycobacterial targets, and that retro-inversion does not appreciably change HDP toxicity towards non-mycobacterial species. Consequently, the lead HDP, identified from a library of screened retro-inverted sequences, demonstrates a 4-fold increase in selectivity towards \u003cem\u003eMtb\u003c/em\u003e over non-mycobacteria, and a greater than 30-fold reduction in host toxicity, compared to the native peptide. Additional bacteriologic testing of this lead candidate, referred to as MAD1-RI, shows the sequence is capable of sterilizing replicating and dormant \u003cem\u003eMtb\u003c/em\u003e cultures, synergistically enhances the activity of clinical TB antibiotics, and is effective towards drug-resistant clinical \u003cem\u003eMtb\u003c/em\u003e isolates. Finally, using a battery of biochemical, biophysical and transcriptomic analyses we elucidate the mechanistic basis of this enhanced activity, and demonstrate that gains in potency operate through altered mycomembrane thermodynamics of retro-inverted sequences. Thus, we identify retro-inversion as a potentially unique driver of anti-mycobacterial specificity in the design of HDPs that utilize mechanisms distinct from the traditional dogma of increased proteolytic stability.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003eBacteriologic screening of retro-inverted sequences\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo explore the bacteriologic effects of retro-inversion we began by chemically modifying the model antimycobaterial peptide, MAD1 (\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003em\u003c/span\u003eycomembrane-\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003ea\u003c/span\u003essociated \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003ed\u003c/span\u003eisruption sequence 1).\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e MAD1 exerts antitubercular activity through disruption of the mycolic acid rich outer mycomembrane, putatively due to its structural mimicry of the mycobacteria specific porin MspA. Inhibition assays utilizing a series of \u003cem\u003eMtb\u003c/em\u003e complex strains (\u003cem\u003eM.\u003c/em\u003e bovis, mc\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e 6230, mc\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e 7901, H37Ra, and H37Rv strains) and non-tuberculous mycobacteria (\u003cem\u003eM. smegmatis\u003c/em\u003e and \u003cem\u003eM. abscessus\u003c/em\u003e) cultures showed that retrosynthesis (backbone reversal, MAD1-R) or chiral inversion (enantiomerization, MAD1-I) of MAD1 did not, individually, alter minimum inhibitory concentration (MIC) of the analogues\u003c/p\u003e\u003cp\u003erelative to the parent sequence (\u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e; see Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e for peptide mass and purity). Conversely, the combination of these features, producing MAD1-RI, resulted in an approximate order of magnitude improvement in MIC against several of the tested mycobacteria relative to native MAD1. For example, single digit micromolar MICs were recorded for MAD1-RI against \u003cem\u003eM. smegmatis\u003c/em\u003e, \u003cem\u003eM. bovis\u003c/em\u003e and \u003cem\u003eMtb\u003c/em\u003e (H37Ra). Against the virulent \u003cem\u003eMtb\u003c/em\u003e strain H37Rv, MAD1-RI demonstrated an MIC of 40\u0026micro;M, while the native MAD1 was weakly active (MIC\u0026thinsp;\u0026gt;\u0026thinsp;40 \u0026micro;M). MAD1-RI was similarly effective against a panel of drug-resistant clinical \u003cem\u003eMtb\u003c/em\u003e isolates (PMID: 30908506), with MICs\u0026thinsp;\u0026le;\u0026thinsp;40\u0026micro;M (Supplementary Table S2).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAll four MAD1 sequences were subsequently counter screened against the non-mycobacterial pathogens, methicillin-resistant \u003cem\u003eS. aureus\u003c/em\u003e (MRSA), \u003cem\u003eP. aeruginosa\u003c/em\u003e, and \u003cem\u003eA. baumanii\u003c/em\u003e to assess mycobacterial selectivity. A selectivity index (SI, see Table\u0026nbsp;1) was then calculated to quantitively compare microbial specificity. Remarkably, retro-inversion yielded a\u0026thinsp;\u0026gt;\u0026thinsp;16-fold increase in mycobacterial selectivity (MAD1-RI SI\u0026thinsp;=\u0026thinsp;32) relative to the MAD1, MAD1-R, and MAD1-I peptides (SI\u0026thinsp;=\u0026thinsp;1\u0026ndash;2). Finally, safety of the MAD1 and MAD1-RI sequences were compared by measuring their half-maximal inhibitory concentration (IC\u003csub\u003e50\u003c/sub\u003e) against A549 human lung epithelial cells, and subsequently calculating a toxicity index (TI, see Table\u0026nbsp;1). Gratifyingly, retro-inversion decreased the peptide\u0026rsquo;s toxicity, increasing the TI from 5.3 for the native sequence to 368.4 for MAD1-RI; nearly two orders of magnitude improvement.\u003c/p\u003e\u003cp\u003eTo evaluate the impact of proteolytic stability on MAD1-RI activity we performed degradation assays in the presence of proteinase K, a particularly promiscuous and aggressive protease (\u003cb\u003eFig.\u0026nbsp;1a\u003c/b\u003e). Chromatographic results show that, as expected, backbone reversal alone (retrosynthesis, MAD1-R) did not enhance the proteolytic stability of the parent sequence, with both showing near complete digestion after 1 hour. Conversely, amino acid enantiomerization (inversion, MAD1-I) yielded a sequence that was highly resistant to proteolysis under the tested conditions, which was conserved when combined with retrosynthesis to produce MAD1-RI. Additional time-dependent stability studies in the presence of \u003cem\u003eMtb\u003c/em\u003e-secreted proteases indicate that native MAD1 was completely proteolyzed after 14 days (\u003cb\u003eFig.\u0026nbsp;1b\u003c/b\u003e), while MAD1-RI generally remained stable up to two weeks in culture (\u003cb\u003eFig.\u0026nbsp;1c\u003c/b\u003e). Taken in context with our MIC results, these findings indicate that enhanced proteolytic stability, alone, does not explain the significantly greater antimycobacterial potency of the retro-inversed MAD1-RI peptide, as typical dogma suggests. Instead, later studies indicate altered thermodynamic stability of retro-inverted sequences in the unique mycolate membrane of \u003cem\u003eMtb\u003c/em\u003e may play a causative role.\u003c/p\u003e\u003cp\u003eWe next set out to assess whether these mycobacterial-specific effects represent a conserved phenomenon of retro-inversion across membrane-active HDPs. To test this, we prepared five pairs of native peptides and their retro-inverted counterparts (see Table\u0026nbsp;1), which included two analogues of MAD1 identified through artificial intelligence guided screening (MAD2.2, MAD5.1) and three antimicrobial sequences isolated from insect venoms (Latarcin 3a,\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e Euminitin F,\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Pinipesin\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e). Three of the retro-inverted peptides, MAD2.2-RI, Latarcin 3a-RI and Euminitin F-RI, showed a significant enhancement in activity relative to the native sequence across the majority of mycobacterial cultures studied; in some cases improving activity by 16 times. MAD5.1, on the other hand, showed no improvement following sequence retro-inversion. This may be a consequence of its high formal charge (+\u0026thinsp;8), where cationic membrane depolarization, and subsequent detergent-like destabilization, by the peptide may be insensitive to retro-inversion mediated conformational changes. Additionally, we observed no significant antibacterial activity of the centipede derived Pinipesin peptide across all tested species. Performing additional experiments under nutrient poor, and reduced temperature conditions (30\u0026deg;C), as were employed in the original publication,\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e did not yield improved potency (data not shown). Nevertheless, collectively our studies suggest that retro-inversion may represent a broadly applicable design strategy to enhance the antimycobacterial activity of HDPs, which operates through mechanism(s) independent of enhanced proteolytic stability. We further explore this assertion through a series of conformational assays performed in both model systems and mycobacterial membranes.\u003c/p\u003e\u003cp\u003e\u003cb\u003eConformational consequences of peptide retro-inversion\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eCircular dichroism was employed to establish the secondary structure of the four conformational MAD1 variants in pH 7.4 phosphate buffer (\u003cb\u003eFig.\u0026nbsp;2a,b\u003c/b\u003e). As previously reported, MAD1 displays mixed α-helical (minima at 206nm and 219nm) and β-sheet (minima at 212nm) conformations under physiologic conditions,\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e with an unusual exciton band at 228 nm indicative of a tryptophan zipper.\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e As expected, MAD1 and MAD1-I (Fig.\u0026nbsp;2a), as well as MAD1-R and MAD1-RI (Fig.\u0026nbsp;2b), congeners display mirrored CD spectra as a result of residue enantiomerization. Unexpectedly, although the spectra of MAD1 and MAD1-I show the putative tryptophan zipper signal at 228nm, the retro-synthetic MAD1-R and MAD1-RI analogues have lost this feature. This suggests backbone reversal may alter residue side chain orientation and disrupt tryptophan-mediated peptide assembly. Additional \u003cem\u003ein silico\u003c/em\u003e conformational analyses utilizing the BeStSel algorithm\u003csup\u003e\u003cspan additionalcitationids=\"CR21\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e show that MAD1 and MAD1-RI possess similar anti-parallel β-sheet content, with accompanying turn-like arrangements, in low ionic strength conditions (\u003cb\u003eFig.\u0026nbsp;2c\u003c/b\u003e). However, nearly 50% of the spectra for both sequences are ascribed to disordered conformations (other), suggesting significant conformational flexibility that may potentiate varied and environmentally sensitive self-assembly phenotypes.\u003c/p\u003e\u003cp\u003eBased on this observation, we next compared the structural dynamics of MAD1 and MAD1-RI in the presence of increasing concentrations of sodium dodecyl sulfate (SDS) micelles, a common model for anionic lipid membranes (\u003cb\u003eFig.\u0026nbsp;2d,e\u003c/b\u003e). In the absence of SDS, MAD1 displays intense exciton-coupled bands at 212 and 228 nm, indicative of long-range assembly. This is significantly disrupted in the presence of SDS, with plots of the concentration-dependent change in ellipticity showing that MAD1\u0026rsquo;s β-sheet (θ\u003csub\u003e212nm\u003c/sub\u003e) and tryptophan-zipper (θ\u003csub\u003e228nm\u003c/sub\u003e) structures are nearly completely abolished in the presence of \u0026ge;\u0026thinsp;1mM SDS (Fig.\u0026nbsp;2d, right). MAD1-RI, on the other hand, was significantly more conformationally rigid, showing only minor changes in structural states up to 90mM SDS. Parallel studies performed in the presence of the \u003cem\u003eMtb\u003c/em\u003e cell envelope glycolipid, trehalose dimycolate, showed a comparable trend, with MAD1 becoming more disordered and MAD1-RI conversely adopting greater structural order (Supplementary Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Finally, CD experiments conducted under varying ionic strength (\u003cb\u003eFig.\u0026nbsp;2f,g\u003c/b\u003e) and temperature (\u003cb\u003eFig.\u0026nbsp;2h,i\u003c/b\u003e) conditions displayed a similarly contrasting conformational behavior. Remarkably, MAD1-RI maintains its secondary structure even at temperatures as high as 95\u0026deg;C (Supplementary Fig. S2), suggesting significant secondary structural stability. Taken together, our \u003cem\u003eex cellulo\u003c/em\u003e results suggest that retro-inversion leads to greater conformational rigidity of the MAD1 peptide in physiologic solutions. This, in turn, may lower the entropic barrier to interpolation and assembly of the sequence within mycomembranes, thereby enhancing its antimycobacterial potency. Using a series of cell based and biophysical assays, we next test this assertion in the context of living mycobacterial cells.\u003c/p\u003e\u003cp\u003eFluorescent confocal and electron microscopy were initially utilized to probe MAD1-RI localization to the \u003cem\u003eMtb\u003c/em\u003e cell surface, as well as investigate peptide-mediated morphologic changes in myco-envelope architecture. Fluorescence micrographs shown in \u003cb\u003eFig.\u0026nbsp;3a\u003c/b\u003e demonstrate that the peptide rapidly engages the \u003cem\u003eMtb\u003c/em\u003e cell wall and co-localizes with outer membrane mycolic acids (\u003cb\u003eFig.\u0026nbsp;3b\u003c/b\u003e). When employed at sub-lethal concentrations (0.5x MIC), these outer membrane interactions cause the typically smooth \u003cem\u003eMtb\u003c/em\u003e surface to adopt a ruffled morphology (\u003cb\u003eFig.\u0026nbsp;3c\u003c/b\u003e). Increasing concentration to 5x MIC led to a ruffling of the pathogen outer membrane, which was complemented by the presence of extracellular aggregates; presumably released intracellular contents. Close inspection of the \u003cem\u003eMtb\u003c/em\u003e cell surface revealed the formation of MAD1-RI nano-assemblies (\u003cb\u003eFig.\u0026nbsp;3d\u003c/b\u003e), indicating the pore-formation capabilities of MAD1 have been maintained after retro-inversion. To further interpret the nature of these assemblies, and better understand how conformational plasticity of the peptides may contribute to their membrane-dependent assembly phenomena, small angle x-ray scattering (SAXS) and analytical ultracentrifugation studies were designed. SAXS curves shown in \u003cb\u003eFig.\u0026nbsp;3e\u003c/b\u003e demonstrate that MAD1-RI adopts a more extended and enlarged conformation relative to MAD1, with a radius of gyration (R\u003csub\u003eg\u003c/sub\u003e) and maximum dimension of 8.2 and 22.2 \u0026Aring; for MAD1 and 10.2 and 30.6 \u0026Aring; for MAD1-RI, respectively. This further supports our assertion that MAD1-RI is more conformationally rigid than the parent sequence. A double maximum observed in the P(r) SAXS curves (\u003cb\u003eFig.\u0026nbsp;3f\u003c/b\u003e) suggests that both peptides adopt an ensemble of homodimeric conformers. Area under the curve (AUC) analyses confirmed this observation and demonstrated a dimeric species for both peptides in physiologic solution (\u003cb\u003eFig.\u0026nbsp;3g\u003c/b\u003e). Curiously, the predicted molecular weight of the MAD1-RI dimer was ~\u0026thinsp;300 Da larger than the MAD1 dimer, despite their identical true molecular weight. It is possible this difference is due to changes in the hydration shell and bound ions for the MAD1-RI dimer assembly. Further SAXS experiments performed in the presence of SDS yielded a model of peptide assembly that suggests MAD1 adopts a head-to-head dimer, while MAD1-RI associates head-to-tail (\u003cb\u003eFig.\u0026nbsp;3h\u003c/b\u003e). This MAD1-RI dimer conformation would span approximately 3 nm in its longest dimension, 1 nm larger than the prediction for MAD1 dimers. Importantly, the head-to-tail alignment of MAD1-RI dimers may allow propagation of these conformers to completely span the ~\u0026thinsp;8 nm thickness of the outer \u003cem\u003eMtb\u003c/em\u003e mycomembrane. Thus, compared to MAD1, MAD1-RI may more readily associate into a pore that spans the \u003cem\u003eMtb\u003c/em\u003e outer membrane and potentiate the leakage of intracellular contents; further corroborating the intracellular aggregates observed during SEM microscopy (Fig.\u0026nbsp;3d).\u003c/p\u003e\u003cp\u003e\u003cb\u003eAntitubercular activity of MAD1-RI\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe next evaluated time dependent killing of \u003cem\u003eMtb\u003c/em\u003e by MAD1-RI, in both proliferating and dormant TB cultures. Initial studies confirmed that, at equivalent concentrations, MAD1-RI showed a\u0026thinsp;~\u0026thinsp;4-log greater anti-TB efficacy relative to MAD1 (\u003cb\u003eFig.\u0026nbsp;4a\u003c/b\u003e). Additional kinetic growth assays demonstrated that the Minimum Bactericidal Concentration (MBC, \u0026gt;\u0026thinsp;99.9% reduction in colony forming units) of MAD1-RI towards \u003cem\u003eMtb\u003c/em\u003e was twice its MIC at ~\u0026thinsp;80 \u0026micro;M (\u003cb\u003eFig.\u0026nbsp;4b\u003c/b\u003e) and complete culture sterilization occurred at 4x MIC (160 \u0026micro;M) between 7 and 14 days of exposure (Fig.\u0026nbsp;4a,b).\u003c/p\u003e\u003cp\u003eNext, combinatorial studies were performed to identify potential synergistic pairings of MAD1-RI and clinical TB antibiotics (\u003cb\u003eFig.\u0026nbsp;4c\u003c/b\u003e, Supplementary Fig. S3). Fractional inhibitory concentration (FIC) scores show that MAD1-RI did not antagonize the activity of any of the tested antibiotics, and for the majority of candidates acted independently or in an additive fashion. Two antibiotics, Moxifloxacin (MOX) and Meropenem, displayed synergistic activity with MAD1-RI (FIC score\u0026thinsp;\u0026le;\u0026thinsp;0.5). MOX is a second-line fluoroquinolone commonly used to treat drug-resistant TB, and therefore this drug was prioritized for follow up combinatorial kill kinetic studies. In proliferating \u003cem\u003eMtb\u003c/em\u003e (\u003cb\u003eFig.\u0026nbsp;4d\u003c/b\u003e), the co-incubation of MAD1-RI (40 \u0026micro;M) and MOX (0.25 \u0026micro;g/mL) lead to complete sterilization of the culture after two weeks of incubation, which represents a\u0026thinsp;~\u0026thinsp;6-log enhancement in activity relative to the monotherapy controls. This activity is also superior to the front-line drugs Isoniazid and Rifampicin, which, despite a\u0026thinsp;~\u0026thinsp;2-log reduction in the first week of exposure, permitted regrowth of treated \u003cem\u003eMtb\u003c/em\u003e cultures after 3 and 10 days, respectively (Supplementary Fig. S4).\u003c/p\u003e\u003cp\u003eFinally, we performed similar \u003cem\u003eMtb\u003c/em\u003e kill kinetics studies under nutrient starvation (\u003cb\u003eFig.\u0026nbsp;4e\u003c/b\u003e) and hypoxic (\u003cb\u003eFig.\u0026nbsp;4f\u003c/b\u003e) conditions. This was done as \u003cem\u003eMtb\u003c/em\u003e adopts a quiescent state in the phagosomes of infected macrophages within granulomatous lesions. This metabolically inactive state promotes persistence of the pathogen by rendering it phenotypically resistant to most antibiotics. This was confirmed in our experimental models, where under nutrient starvation and hypoxic conditions Isoniazid was inactive (Supplementary Fig. S5), and the efficacy of MOX significantly blunted (Fig.\u0026nbsp;4e,f). Conversely, we hypothesized that MAD1 would remain active under these conditions due to the insensitivity of its mycomembrane destabilizing mechanism of action to the pathogen\u0026rsquo;s metabolic state. Results shown in Supplementary Fig. S6 confirm this assertion, and again show enhanced activity of MAD1-RI relative to MAD1 toward persister \u003cem\u003eMtb\u003c/em\u003e. As expected, the combination of MAD1-RI and MOX showed more effective killing than the individual therapies in nutrient starved \u003cem\u003eMtb\u003c/em\u003e cultures (Fig.\u0026nbsp;4e). However, under hypoxic conditions, the combinatorial formulation was equal in its efficacy to MOX alone, and was significantly less active than the anaerobic positive control drug Metronidazole (Supplementary Fig. S5b). Notably, although effective \u003cem\u003ein vitro\u003c/em\u003e, metronidazole lacks efficacy in \u003cem\u003ein vivo\u003c/em\u003e TB models due, at least in part, to insufficient anaerobic conditions to allow reductive activation of the drug.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e In sum, our results show that MAD1-RI is effective towards dormant \u003cem\u003eMtb\u003c/em\u003e persisters, and suggest that modulation of particular metabolic pathways during dormancy may blunt the antimycobacterial activity of the peptide; an assertion further explored through genomic studies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eMAD1-RI antimycobacterial mechanisms\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo characterize the transcriptional response to MAD1-RI exposure, H37Rv \u003cem\u003eMtb\u003c/em\u003e cultures were treated for 24 hours with 25 \u0026micro;M of the peptide before performing RNA sequencing. Results in \u003cb\u003eFig.\u0026nbsp;5a\u003c/b\u003e show that 16 genes were induced and 7 genes repressed (\u0026gt;\u0026thinsp;1-fold, q\u0026thinsp;\u0026lt;\u0026thinsp;0.05, principal components analysis shown in Supplementary Figure S7). Within the upregulated data set, the five-gene operon \u003cem\u003eespACD-Rv3613c-Rv3612c\u003c/em\u003e is particularly notable given its required role in ESX-1 secretion, which is essential for \u003cem\u003eMtb\u003c/em\u003e virulence and host cell survival.\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e Additionally, induction of the \u003cem\u003ePE20\u003c/em\u003e gene encodes for the PPE20 protein, which forms a complex with PE15 to facilitate molecular transport across the \u003cem\u003eMtb\u003c/em\u003e cell envelope.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Finally, MAD1-RI also induced \u003cem\u003eRv1057\u003c/em\u003e, which is proposed to be upregulated in response to envelope stress induced by sodium dodecyl sulfate (SDS) exposure and stabilizes MmpL3 complexes that shuttle lipid components to the cell wall.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e Thus, it was hypothesized that MAD1-RI-treated \u003cem\u003eMtb\u003c/em\u003e would exhibit increased sensitivity to membrane-targeting agents, such as detergents like SDS. Indeed, after 24 hours of drug treatment, survival differences were observed between treated and untreated cultures upon SDS exposure. Here, the viability of cultures pre-treated with MAD1-RI and lysocin E, a peptide that causes membrane disruption,\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e declined more rapidly and to a greater extent than untreated cultures, indicating heightened sensitivity to SDS (\u003cb\u003eFig.\u0026nbsp;5b\u003c/b\u003e). Together, these results suggest that MAD1-RI compromises the integrity of the \u003cem\u003eMtb\u003c/em\u003e cell envelope by disrupting cell wall secretion, transport and biosynthetic processes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo empirically test this assertion, we utilized the lipophilic dye DiOC2(3) to monitor MAD1-RI mediated changes in \u003cem\u003eMtb\u003c/em\u003e membrane potential. This fluorophore intercalates into the mycobacterial outer membrane and displays decreased red fluorescence as membrane potential is reduced. As expected, increasing MAD1-RI concentration from 1.25 to 37.5 \u0026micro;M led to a systematic reduction in DiOC2(3) fluorescence, producing similar membrane depolarization at the highest tested peptide concentration to the carbonyl cyanide 3-chlorophenylhydrazone (CCCP) protonophore positive control (\u003cb\u003eFig.\u0026nbsp;5c\u003c/b\u003e).\u003c/p\u003e\u003cp\u003eFinally, it is interesting to note that MAD1-RI exposure also repressed \u003cem\u003eprpD\u003c/em\u003e and \u003cem\u003eRv1129c\u003c/em\u003e (\u003cem\u003eprpR\u003c/em\u003e) transcription in \u003cem\u003eMtb\u003c/em\u003e. Both genes encode for key enzymes in the methylcitrate cycle utilized by \u003cem\u003eMtb\u003c/em\u003e to process fatty acids as a carbon source during survival within macrophages.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Additionally, \u003cem\u003ebfrB\u003c/em\u003e encodes the iron storage protein ferritin that regulates iron homeostasis in \u003cem\u003eMtb\u003c/em\u003e during survival and proliferation within infected host cells.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e Thus, repression of these genes suggests that, in addition to its membrane destabilizing effects, MAD1-RI can disrupt fatty acid metabolism and lipid biosynthesis to compromise \u003cem\u003eMtb\u003c/em\u003e fitness \u003cem\u003ein vivo\u003c/em\u003e. This may explain the reduction in MAD1-RI activity observed in our \u003cem\u003ein vitro\u003c/em\u003e dormant models relative to the proliferating cultures (Fig.\u0026nbsp;4). Here, although lipid metabolism may be downregulated by the peptide, the pathogen is rescued by other carbon sources available in the rich media. However, evidence from animal studies\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e indicating that \u003cem\u003eMtb\u003c/em\u003e virulence \u003cem\u003ein vivo\u003c/em\u003e is dependent on a shift to lipid catabolism suggests that, under nutrient depravation conditions within an infected host, MAD1-RI\u0026rsquo;s anti-TB activity may be significantly enhanced.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eWhile peptide retro-inversion is an established approach to improve the proteolytic stability of therapeutic peptides, we show here that it also enhances the antimycobacterial potency and specificity of HDPs through mechanisms distinct from enzymatic digestion. MAD1-RI exhibits a particularly strong effect compared to other tested sequences, likely due to a gramicidin-like ion-leaking antimycobacterial mechanism. Biophysical studies suggest that this mode of action differs from general membrane destabilization, potentially involving a head-to-tail arrangement forming intercalating pores or a β-helical pore with two copies of the retro-inversed MAD1-RI sequence in the mycomembrane. We show this activity not only leads to rapid killing of replicating \u003cem\u003eMtb\u003c/em\u003e, but potently synergizes with clinically relevant TB antibiotics. These findings, in combination with its regulation of genes important to \u003cem\u003eMtb\u003c/em\u003e virulence and cell wall biosynthesis, suggest that MAD1-RI may be a potent addition to the anti-TB therapeutic arsenal. In addition to its clinical potential, the rapid binding of MAD1-RI to mycobacterial cells may also enable its use as a diagnostic probe to monitor pathogens in wastewater and agricultural settings. More broadly, our findings suggest that retro-inversion may represent a unique chemical approach to design narrow-spectrum bactericidal peptides with anti-mycobacterial specificity.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003ePeptide synthesis, purification and characterization\u003c/h2\u003e \u003cp\u003eAll peptides were synthesized following conventional Fmoc/tBu solid phase peptide synthesis (SPPS) using an automated Liberty Blue 2.0 microwave synthesizer (CEM Corp., North Carolina) following a high-efficiency SPPS protocol.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e Crude peptides were purified by HPLC and purity and mass spectra were confirmed by LCMS on a LCMS-2020 instrument (Shimadzu Corp.). Fluorescence images were collected with a Biotek Cytation 3 plate reader and Zeiss LSM 880 confocal laser scanning microscope and scanning electron micrographs were obtained with a Zeiss Sigma scanning electron microscope.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eCircular Dichroism (CD) spectroscopy\u003c/h2\u003e \u003cp\u003eCD measurements were conducted using a JASCO J-1500 spectrometer equipped with a Peltier-controlled PTC-517 thermostat cell holder. Spectra were recorded from 260 nm to 185 nm at a scan speed of 50 nm/min, with a bandwidth of 1 nm at 25\u0026deg;C. A 1 mm pathlength quartz cuvette was used, and the peptide concentration was 100 \u0026micro;M. A buffer blank was measured before the sample for baseline subtraction, and the data were converted to molar ellipticity. Analysis was performed using the Jasco Multivariate SSE program along with BeStSel algorithm\u003csup\u003e\u003cspan additionalcitationids=\"CR21\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e for single-spectrum analysis and secondary structure prediction. Temperature-dependent CD experiments were carried out from 20\u0026deg;C to 95\u0026deg;C in 5\u0026deg;C increments, with a heating rate of 1\u0026deg;C/min. Spectra were recorded at 50 nm/min from 260 nm to 185 nm, with a data pitch of 1 nm, a DIT of 4 seconds, and a bandwidth of 1 nm. Wavelength-dependent changes at 208 nm and 220 nm were plotted against temperature. All CD experiments utilized phosphate buffer except for data reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef and Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg in which SDS in pure water was used to determine the effect of SDS concentration on secondary structure.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAUC Methods\u003c/h3\u003e\n\u003cp\u003ePeptides, dissolved in phosphate buffer (pH 7.4) at a concentration of 33 \u0026micro;M, were loaded into 12 mm Epon-charcoal centerpieces within AUC cells featuring sapphire windows. These cells were placed into an An50 titanium rotor, which had been pre-equilibrated to 37\u0026deg;C, the experimental temperature. The rotor was then inserted into the chamber of a Beckman-Coulter Optima multiwavelength AUC equipped with absorbance optics. A full vacuum was applied in the chamber, and the rotor was allowed to re-equilibrate for 2 hours. A method scan was created using the UltraScan III software and transferred to the instrument, where the experiment began after temperature equilibration. The rotor was spun at 40,000 RPM for 11 hours, capturing radial scans every three minutes at 280 nm. Once the AUC experiment was complete, the data were imported into UltraScan III. Reference scans were automatically chosen to convert the raw radial intensity data into pseudo-absorbance. The air-liquid meniscus was manually selected for each sector. The dataset was also manually cropped, typically between 6.1 cm and 7.1 cm, and the first 5\u0026ndash;10 scans, as well as those following complete sedimentation, were excluded from analysis. The edited data were then processed using the LIMS supercomputer at Penn State, with an S-value range set from 1 to 10, a resolution of 100, and a frictional ratio range from 1 to 4, with a resolution of 64. Time-invariant noise was also accounted for during the initial analysis. When the residuals fell below 0.003, the data were refitted, this time incorporating both time and radially invariant noise, along with 11 meniscus fits to ensure precise determination of the meniscus. Once the correct meniscus was identified, a final time and radial invariant noise fit was performed using an iterative method. Lastly, the data were analyzed using a genetic algorithm with Monte Carlo simulations (selecting 1\u0026ndash;2 species per sample). 32 Monte Carlo simulations were run using 16 processors, and the resulting pseudo-3D plots were analyzed for final calculations of the s-value, frictional ratio, and molecular weight.\u003c/p\u003e\n\u003ch3\u003eBioSAXS\u003c/h3\u003e\n\u003cp\u003eSmall angle X-ray scattering (BioSAXS) data were collected on peptides at a concentration of 4.5 mg/ml using X-rays with a wavelength of 1.54 \u0026Aring; from an in-house Rigaku MM007 rotating anode X-ray source. This system was coupled with the BioSAXS2000nano Kratky camera, which features OptiSAXS confocal max-flux optics specifically designed for SAXS experiments and a HyPix-3000 Hybrid Photon Counting detector for high sensitivity. The sample was positioned in a capillary with a detector-to-sample distance of 495.5 mm, calibrated using silver behenate powder from The Gem Dugout (State College, PA). The scattering vector q-space (q\u0026thinsp;=\u0026thinsp;4πsin(θ)/λ, where 2θ is the scattering angle) ranged from q_min\u0026thinsp;=\u0026thinsp;0.008 \u0026Aring;⁻\u0026sup1; to q_max\u0026thinsp;=\u0026thinsp;0.6 \u0026Aring;⁻\u0026sup1;. The X-ray beam energy was 1.2 keV, with a Kratky block attenuation of 22% and a beam diameter of approximately 100 \u0026micro;m. Peptide samples were automatically loaded onto a quartz capillary flow cell via a Rigaku autosampler, which was cooled to 4\u0026deg;C and aligned with the X-ray beam. The sample cell and entire X-ray flight path, including the beam stop, were kept in a vacuum (below 1 \u0026times; 10⁻\u0026sup3; torr) to eliminate air scattering. The Rigaku SAXSLAB software controlled automated data collection for each peptide, incorporating thorough cleaning cycles between samples. Data reduction, including image integration, normalization, and background subtraction, was also handled by SAXSLAB software. Six ten-minute images, along with three replicates of both protein and buffer samples, were collected, averaged, and inspected to confirm that no radiation damage occurred. Overlays of the SAXS data confirmed no radiation decay or sample loss during the 60-minute collection period. Following this, buffer subtraction was performed to isolate the raw SAXS scattering curve of the peptide. The forward scattering intensity (I(0)) and the radius of gyration (Rg) were calculated using the Guinier approximation, which assumes that at very small angles (q\u0026thinsp;\u0026lt;\u0026thinsp;1.3/Rg), the intensity follows I(q)\u0026thinsp;=\u0026thinsp;I(0)exp[\u0026minus;\u0026thinsp;1/3(qRg)\u0026sup2;]. The results were consistent with the expected size of peptide dimers. Further analysis of the data, including radius of gyration, Dmax, Guinier fits, Kratky plots, and pair distance distribution functions, was carried out using the ATSAS software. Solvent envelopes were calculated with DAMMIF, an algorithm for deriving ab initio bead models directly from solution scattering data.\u003c/p\u003e\n\u003ch3\u003eFluorescence and electron microscopy\u003c/h3\u003e\n\u003cp\u003eFor fluorescence microscopy, \u003cem\u003eMtb\u003c/em\u003e mc\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e6230 was grown to an OD of 0.02 in supplemented Middlebrook 7H9 broth. The bacteria was isolated by centrifugation and re-suspended in 7H9 broth containing cyanine-5 tagged MAD1-RI (10 mM) for 30 min prior to drying and heat fixation on a glass microscopy slide. If necessary, Hoechst 33343 stain (1 mg/mL was also added to the bacteria prior to fixation). A glass cover slip was then adhered to the top of the sample with a dab of ProLong\u0026trade; Diamond Antifade Mountant (Invitrogen) overnight in the dark.\u003c/p\u003e \u003cp\u003eFor SEM samples, bacteria were grown identically to the fluorescence microscopy protocol and incubated in the presence of 0.5 x MIC, 1 x MIC, and 5 x MIC of MAD1-RI for 30 min before being passed through a 0.2 micron pore filter disc. The bacteria trapped on the filter disc were then fixed with glutaraldehyde (2.5% v/v) in phosphate buffer for 30 min followed by step-wise dehydration with 10, 25, 50, 75, 85, 95, and 2 x 100% solutions of ethanol. The discs were then dehydrated with a Leica EM CPD300 Critical Point Dryer (Leica) prior to being mounted on titanium stubs and sputter coated with a 4.5 nm layer of iridium prior to viewing.\u003c/p\u003e\n\u003ch3\u003eBacteriologic testing\u003c/h3\u003e\n\u003cp\u003eAntimicrobial activity was determined by standard broth microdilution minimal inhibitory assays wherein peptide was serially diluted across a 96-welled plate and bacteria adjusted to an OD\u003csub\u003e600\u003c/sub\u003e of 0.002 was added to each concentration of peptide.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e All bacteria were grown and assayed in specialized broth at 37\u0026deg;C. \u003cem\u003eM. smegmatis\u003c/em\u003e (mc\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e155) and \u003cem\u003eMtb\u003c/em\u003e complex lab strains (\u003cem\u003eM. bovis\u003c/em\u003e, mc\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e6230, H37Rv and H37Ra) were grown in Middlebrook 7H9 and supplemented 7H9 broth, respectively. \u003cem\u003eMtb\u003c/em\u003e clinical isolates were grown in Middlebrook supplemented 7H9 containing 40 mM pyruvate. All virulent strains of \u003cem\u003eMtb\u003c/em\u003e were handled under BSL-3 conditions following institutionally approved protocols. \u003cem\u003eS. aureus\u003c/em\u003e (USA300), \u003cem\u003eP. aeruginosa\u003c/em\u003e (PAO1), and \u003cem\u003eA. baumanii\u003c/em\u003e (ATCC19606) were grown in Mueller-Hinton broth (MHB). 7H9 broth for the auxotrophic mc\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e6230 strain was prepared with glycerol (0.2% v/v), oleic acid-albumin-dextrose catalase (OADC; 10% v/v), Tween-80 (0.02% v/v), and pantothenic acid (50 mg/L). For colony growth, the mycobacteria species were grown on Middlebrook 7H10 agar with the same supplementation except for Tween-80. The other bacteria were all grown on MHB agar.\u003c/p\u003e \u003cp\u003eMinimum inhibitory concentration (MIC) assays were performed by dissolving peptides to 320 \u0026micro;M in sterile nuclease-free water and 2-fold serially diluting the peptide stock in 7H9 broth (50 \u0026micro;L volume after dilution) in a sterile 96-well round bottom plate in triplicate. Bacterial growth was measured via OD\u003csub\u003e600\u003c/sub\u003e, and the culture was diluted to an OD\u003csub\u003e600\u003c/sub\u003e value of 0.002 for the mCherry expressing \u003cem\u003eMtb\u003c/em\u003e H37Rv and other \u003cem\u003eMtb\u003c/em\u003e clinical isolates. 50 \u0026micro;l of bacterial suspension was added to the treated wells in a 1:1 v/v ratio (100 \u0026micro;L final volume).\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e H37Rv and H37Ra growth were assessed by measuring fluorescence at 570/610 nm using a plate reader after incubating at 37\u0026deg;C for 96 hours. MTT (3- (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays were employed to evaluate the MIC against \u003cem\u003eMtb\u003c/em\u003e clinical isolates. The growth of clinical isolates was assessed by measuring fluorescence at 570 nm using a plate reader after 7 days of incubation. The MIC for each drug was defined as the lowest concentration that reduced fluorescence by 90%. Assays were repeated with at least 2 independent replicates for each strain (n\u0026thinsp;\u0026ge;\u0026thinsp;6).\u003csup\u003e33\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eMinimum bactericidal concentration (MBC) for replicating cultures was determined by adjusting exponentially growing \u003cem\u003eMtb\u003c/em\u003e to OD\u003csub\u003e600\u003c/sub\u003e of 0.1 and treating bacilli with 40, 80, 160 \u0026micro;M of MAD1-RI. Culture plating was done on 7H10 agar at day 0, 1, 3 and 7. The MBC was defined as the lowest concentration which reduced CFU count by 99% relative to the time-zero inoculum on day 7. Twenty-one-day kill kinetics were assessed to evaluate the bactericidal activity of MAD1-RI in combination with other drugs by CFU count. For this purpose, a mid-log-phase \u003cem\u003eMtb\u003c/em\u003e culture was diluted in fresh medium (OD\u003csub\u003e600\u003c/sub\u003e of 0.1). After aliquoting the culture to 30 ml square bottles, the indicated concentration of anti-TB drugs was added to each sample, and cultures were incubated for 21 days. Each culture dilution was plated at selected time intervals on 7H10 agar plates, and the \u003cem\u003eMtb\u003c/em\u003e CFU on each plate were enumerated after incubating at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e enriched atmosphere for 3\u0026ndash;4 weeks.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eDrug synergy\u003c/h2\u003e \u003cp\u003eCheckerboard assays were used to test for interactions between MAD1-RI and drugs with known anti-TB activities by the broth microdilution method in 96-well microtiter plates. After preparing serially diluted drugs on two different plates, 25 \u0026micro;l from each well of the MAD1-RI plate (plate 1) and another drug plate (plate 2) were transferred to the corresponding wells of a new plate (plate 3) and mixed (the final drug volume of each well was 50 \u0026micro;l). Then, like the MIC assay, 50 \u0026micro;l of mCherry expressing \u003cem\u003eMtb\u003c/em\u003e H37Rv suspension (OD\u003csub\u003e600\u003c/sub\u003e of 0.002) was added to the wells of each plate considered for testing.\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e The MIC for each drug or combination was defined as the lowest concentration that reduced fluorescence by 90% after incubating the plates at 37\u0026deg;C for 4 days. To evaluate the effect of each combination, the obtained MIC values were used to calculate the fractional inhibitory coefficient index (FICI) as follows: FICI = (MIC of drug A in combination/MIC of drug A alone) + (MIC of drug B in combination/MIC of drug B alone). The FIC index (FICI) calculated for each drug combination was categorized based on FICI: \u0026lt;0.5 as synergistic, \u0026gt;\u0026thinsp;0.5 to \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e\u0026le;\u003c/span\u003e\u0026thinsp;1 as additive; and \u0026gt;\u0026thinsp;2 as antagonistic.\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003e \u003cb\u003eKill curves with non-replicating\u003c/b\u003e \u003cb\u003eMtb\u003c/b\u003e\u003c/p\u003e \u003cp\u003eTo study the bactericidal activity of MAD1-RI against non-replicating mycobacteria (NRP), nutrient-starved and hypoxia-induced nonreplicating \u003cem\u003eMtb\u003c/em\u003e were used. To obtain the hypoxia-induced, NRP \u003cem\u003eMtb\u003c/em\u003e phenotype, H37Ra cultures in the early log phase were transferred to an anaerobic chamber (atmosphere of 5% CO\u003csub\u003e2\u003c/sub\u003e, 10% H\u003csub\u003e2\u003c/sub\u003e, 85% N\u003csub\u003e2\u003c/sub\u003e) and grown for 3 weeks at 100 rpm, 37\u0026deg;C. Then, to evaluate the bactericidal activities of MAD1-RI against NRP \u003cem\u003eMtb\u003c/em\u003e, H37Ra cultures were transferred to 15 ml test tubes and treated with different concentrations of the indicated drugs. The cultures were then incubated for two weeks and plated on 7H10 agar at selected time intervals.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e To obtain nutrient-starved NRP \u003cem\u003eMtb\u003c/em\u003e, H37Rv bacilli were grown to the exponential phase. Bacteria were then harvested by centrifugation (3,200 rpm; 4\u0026deg;C; 5 min) and washed twice with PBS containing 0.025% Tween 80. The bacteria were diluted to a final OD\u003csub\u003e600\u003c/sub\u003e of 0.2 in PBS and incubated for 3 weeks at 37\u0026deg;C. Clumps were removed by low-speed centrifugation (2000 x \u003cem\u003eg\u003c/em\u003e; 3 min) just before treatment. To evaluate the bactericidal activities of the peptides against nutrient-starved NRP, \u003cem\u003eMtb\u003c/em\u003e H37Rv was transferred to 30 ml square bottles, and treated with MAD1-RI, isoniazid, or moxifloxacin at the indicated concentration. Each culture dilution was plated at selected time intervals on 7H10 agar plates, and the \u003cem\u003eMtb\u003c/em\u003e CFU on each plate was enumerated after incubating at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e enriched atmosphere.\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eTranscriptional profiling and analysis\u003c/h3\u003e\n\u003cp\u003eFor RNA isolation, \u003cem\u003eMtb\u003c/em\u003e H37Rv was cultured in 7H9 media to an OD\u003csub\u003e600\u003c/sub\u003e of 0.1 before being treated with 25 \u0026micro;M MAD1-RI for 24 h.\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e After treatment, mycobacterial RNA was extracted using a Zymogen RNA Miniprep Kit (Sigma-Aldrich, USA) with minor modifications to the manufacturers protocol. Briefly, cells were resuspended in trizol reagent and 1% Polyacyl carrier and disrupted using Zirconia beads (Biospec, Bartlesville, OK, USA) by beating two times for 1 minute and cooling on ice for two minutes in between. The samples were spun down at 14,000 RPM for 10 min at 4\u0026deg;C and the supernatant was transferred to clean tubes. Each sample was vortexed after 50 \u0026micro;l of BCP (bromo 3-chloro propane) was added. The samples were then harvested into a fresh RNase-free E-tube and one volume of ethanol (95\u0026ndash;100%) was added directly to one volume sample (1:1). Samples were mixed well by vortexing. The sample mixtures were then loaded into a Zymo-Spin Column in a collection tube, and centrifuged at 14,000rpm for 1 minute.\u003c/p\u003e \u003cp\u003eTo clean up the RNA, the columns were washed with 400 \u0026micro;l RNA wash buffer by centrifuging for 30 seconds. To decontaminate DNA, DNase I Reaction Mix was added directly to the column matrix which was then incubated at RT (20\u0026ndash;30℃) for 15 minutes, followed by centrifugation for 30 seconds. The samples were then washed with 400 \u0026micro;l Direct-zol RNA PreWash twice by centrifuging for 1 minute. Samples were washed once again by adding 700 \u0026micro;l RNA Wash Buffer to the column and centrifuging for 1 minute. Finally, 50 \u0026micro;l of Dnase/Rnase-Free Water was directly added to the column matrix and centrifuged for 1 minute. The eluted RNA was stored at -80℃ and sent for processing and sequencing. cDNA libraries were prepared by SeqCenter (Pittsburg, PA) with a Stranded Total RNA Prep using the Ribo-Zero Plus 563 Microbiome kit (Illumina Inc). cDNA libraries were sequenced using the Illumina Novaseq platform optimized for 150 bp paired-end reads and producing approximately 12\u0026nbsp;million reads.\u003c/p\u003e \u003cp\u003eRNA sequencing data was analyzed by preprocessing the raw .fastq output files using the pipeline available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/MDHowe4/RNAseq-Pipeline\u003c/span\u003e\u003cspan address=\"https://github.com/MDHowe4/RNAseq-Pipeline\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. FastQC was utilized to measure the quality control of the reads. Read lengths and t-overhangs were trimmed via Cutadapt, the minimum read length for which was set to 30 bp. Alignment to the \u003cem\u003eMtb\u003c/em\u003e H37Rv genome (NC_000962.3) was performed by the STAR aligner with spliced alignment detection disabled (--alignIntronMax 1). The read counts for given genes were procured via featureCounts, with exclusion criteria defined as genes with fewer than 10 reads across all samples. A negative binomial generalized linear model was created with DESeq2 to quantify the differential expression of genes. Significance was reserved for genes which met both a log2-fold change\u0026thinsp;\u0026ge;\u0026thinsp;1 or \u0026le; -1 and an adjusted p-value of \u0026lt;\u0026thinsp;0.05 criteria. The ggplot2 package was used to generate volcano plots for these data.\u003c/p\u003e\n\u003ch3\u003eSDS sensitivity assay and assessment of membrane potential\u003c/h3\u003e\n\u003cp\u003e \u003cem\u003eMtb\u003c/em\u003e H37Rv was grown in 7H9 medium supplemented with OADC and 0.05% Tween 80 to an OD600 of 0.5. Cultures were then treated with 160 \u0026micro;M MAD1-RI and 6 \u0026micro;g/ml Lysocin E for 24 hours. Following treatment, bacterial pellets were washed twice with sterile PBS and resuspended to an OD600 of 0.1 in PBS. SDS was added to a final concentration of 0.005%, and cultures were plated on 7H10 agar at 0 and 3 hours post-SDS exposure. After incubation at 37\u0026deg;C for three to four weeks, CFUs were enumerated, and percent survival at 3 hours was calculated relative to the starting CFUs at 0 hours.\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eMembrane potential was assessed as previously described,\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e with modifications. Briefly, mid-exponential- phase \u003cem\u003eMtb\u003c/em\u003e H37Ra cells grown in supplemented 7H9 medium were centrifuged (3,000 rpm for 10 min) and resuspended to a final OD\u003csub\u003e600\u003c/sub\u003e of 0.1 in 3 ml supplemented 7H9 medium at pH 7.0 in 30-ml square Nalgene bottles. Peptides and ionophores were added to the final concentrations indicated in the graph and incubated at 37\u0026deg;C. At 30 min, 180 \u0026micro;l was removed, 20 \u0026micro;l of 150 \u0026micro;M 3,3-diethyloxicarbocianide chloride [DiOC2(3)] was added, and the mixture was incubated at room temperature for 30 min. Cells were then washed, resuspended in supplemented 7H9 medium, transferred to a black-walled 96-well plate, and analyzed in a Molecular Devices SpectraMax M2 plate reader, and fluorescence was assessed by excitation of samples at 488 nm and recording the emissions at 530 nm and 610 nm. Ratios of the emission at 610 nm/emission at 530 nm were calculated.\u003c/p\u003e"},{"header":"Declarations","content":" \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData Availability\u003c/h2\u003e \u003cp\u003eThe data that support the findings of this study are available from the corresponding author upon reasonable request. Primary RNA-seq data are publicly available through the National Center for Biotechnology Information via SRA link \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/sra/PRJNA1157802\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/sra/PRJNA1157802\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eResearch reported here was supported by SIG S10 of the National Institutes of Health under award number # S10-OD028589 for the small angle X-ray scattering, NIH grant S10 OD032215-01 for the Optima AUC and S10 OD030490 for the Wyatt SEC-MALS-DLS system to Dr. Neela Yennawar. We thank William R. Jacobs, Jr and Michael Berney of Albert Einstein College of Medicine for gifting the various mycobacterial strains used in this study. We also wish to thank the assistance of Ms. Julia Fecko at the X-ray Crystallography core at the Penn State Huck Institutes of the Life Sciences and the UMN BioSafety Level 3 Program for facility management. Funding for this work was provided by NIH R01-AI165996 to S.H.M.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eZhang, Y. \u0026amp; Yew, W. Mechanisms of drug resistance in Mycobacterium tuberculosis: update 2015. Int. J. Tuberc. Lung Dis. 19, 1276\u0026ndash;1289 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRima, M., Rima, M., Fajloun, Z., Sabatier, J. M., Bechinger, B. \u0026amp; Naas, T. Antimicrobial Peptides: A Potent Alternative to Antibiotics. Antibiotics 10 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLandman, D., Georgescu, C., Martin, D. A. \u0026amp; Quale, J. Polymyxins revisited. Clin. Microbiol. Rev. 21, 449\u0026ndash;465 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRezaei Javan, R., Van Tonder, A. J., King, J. P., Harrold, C. L. \u0026amp; Brueggemann, A. B. Genome sequencing reveals a large and diverse repertoire of antimicrobial peptides. \u003cem\u003eFront. Microbiol.\u003c/em\u003e 9, 2012 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLazzaro, B. P., Zasloff, M. \u0026amp; Rolff, J. Antimicrobial peptides: Application informed by evolution. Science 368, eaau5480 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantos-J\u0026uacute;nior, C. D. \u003cem\u003eet al.\u003c/em\u003e Discovery of antimicrobial peptides in the global microbiome with machine learning. Cell (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaasch, J. R. M. A., Torres, M. D. T., Melo, M. C. R. \u0026amp; de la Fuente-Nunez, C. Molecular de-extinction of ancient antimicrobial peptides enabled by machine learning. Cell Host Microbe 31, 1260\u0026ndash;1274.e1266 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWan, F., Torres, M. D. T., Peng, J. \u0026amp; de la Fuente-Nunez, C. Deep-learning-enabled antibiotic discovery through molecular de-extinction. Nat. Biomed. Eng. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChorev, M. \u0026amp; Goodman, M. A dozen years of retro-inverso peptidomimetics. Acc. Chem. Res. 26, 266\u0026ndash;273 (1993).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoodman, M. \u0026amp; Chorev, M. On the concept of linear modified retro-peptide structures. Acc. Chem. Res. 12, 1\u0026ndash;7 (1979).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDoti, N., Mardirossian, M., Sandomenico, A., Ruvo, M. \u0026amp; Caporale, A. Recent Applications of Retro-Inverso Peptides. Int. J. Mol. Sci. 22 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, C. \u003cem\u003eet al.\u003c/em\u003e Limitations of peptide retro-inverso isomerization in molecular mimicry. J. Biol. Chem. 285, 19572\u0026ndash;19581 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePreston, G. W. Different directions for retro-inverso peptides. J. Pept. Sci. 28, e3384 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSimonson, A. W. \u003cem\u003eet al.\u003c/em\u003e Pathogen-specific antimicrobials engineered de novo through membrane-protein biomimicry. Nat. Biomed. Eng. 5, 467\u0026ndash;480 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Moraes, L. \u003cem\u003eet al.\u003c/em\u003e First generation of multifunctional peptides derived from latarcin-3a from Lachesana tarabaevi spider toxin. Front. Microbiol. 13, 965621 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRangel, M. \u003cem\u003eet al.\u003c/em\u003e Chemical and biological characterization of four new linear cationic α-helical peptides from the venoms of two solitary eumenine wasps. Toxicon 57, 1081\u0026ndash;1092 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChaparro-Aguirre, E., Segura-Ram\u0026iacute;rez, P. J., Alves, F. L., Riske, K. A., Miranda, A. \u0026amp; Silva J\u0026uacute;nior, P. I. Antimicrobial activity and mechanism of action of a novel peptide present in the ecdysis process of centipede Scolopendra subspinipes subspinipes. Sci. Rep. 9, 13631 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCochran, A. G., Skelton, N. J. \u0026amp; Starovasnik, M. A. Tryptophan zippers: Stable, monomeric β-hairpins. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 98, 5578\u0026ndash;5583 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, J., Yong, W., Deng, Y., Kallenbach, N. R. \u0026amp; Lu, M. Atomic structure of a tryptophan-zipper pentamer. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 101, 16156\u0026ndash;16161 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMicsonai, A. \u003cem\u003eet al.\u003c/em\u003e Accurate secondary structure prediction and fold recognition for circular dichroism spectroscopy. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 112, E3095-E3103 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMicsonai, A. \u003cem\u003eet al.\u003c/em\u003e BeStSel: webserver for secondary structure and fold prediction for protein CD spectroscopy. Nucleic Acids Res. 50, W90-W98 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMicsonai, A. \u003cem\u003eet al.\u003c/em\u003e Disordered\u0026ndash;Ordered Protein Binary Classification by Circular Dichroism Spectroscopy. Front. Mol. Biosci. 9 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlinkenberg, L. G., Sutherland, L. A., Bishai, W. R. \u0026amp; Karakousis, P. C. Metronidazole Lacks Activity against Mycobacterium tuberculosis in an In Vivo Hypoxic Granuloma Model of Latency. J. Infect. Dis. 198, 275\u0026ndash;283 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHunt, D. M. \u003cem\u003eet al.\u003c/em\u003e Long-Range Transcriptional Control of an Operon Necessary for Virulence-Critical ESX-1 Secretion in Mycobacterium tuberculosis. J. Bacteriol. 194, 2307\u0026ndash;2320 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoradia, V., Frando, A. \u0026amp; Grundner, C. The Mycobacterium tuberculosis PE15/PPE20 complex transports calcium across the outer membrane. PLoS Biol. 20, e3001906 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePang, X., Cao, G., Neuenschwander, P. F., Haydel, S. E., Hou, G. \u0026amp; Howard, S. T. The β-propeller gene Rv1057 of Mycobacterium tuberculosis has a complex promoter directly regulated by both the MprAB and TrcRS two-component systems. Tuberculosis 91 Suppl 1, S142\u0026ndash;149 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWiegand, I., Hilpert, K. \u0026amp; Hancock, R. E. Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat. Protoc. 3, 163\u0026ndash;175 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang, S., Hicks, N. D., Cheng, Y. S., Silva, A., Fortune, S. M. \u0026amp; Sacchettini, J. C. Structural and functional insight into the Mycobacterium tuberculosis protein PrpR reveals a novel type of transcription factor. Nucleic Acids Res. 47, 9934\u0026ndash;9949 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKurthkoti, K. \u003cem\u003eet al.\u003c/em\u003e The mycobacterial iron-dependent regulator IdeR induces ferritin (bfrB) by alleviating Lsr2 repression. Mol. Microbiol. 98, 864\u0026ndash;877 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSavvi, S., Warner, D. F., Kana, B. D., McKinney, J. D., Mizrahi, V. \u0026amp; Dawes, S. S. Functional characterization of a vitamin B12-dependent methylmalonyl pathway in Mycobacterium tuberculosis: implications for propionate metabolism during growth on fatty acids. J. Bacteriol. 190, 3886\u0026ndash;3895 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins, J. M., Porter, K. A., Singh, S. K. \u0026amp; Vanier, G. S. High-Efficiency Solid Phase Peptide Synthesis (HE-SPPS). Org. Lett. 16, 940\u0026ndash;943 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeberetsadik, G. \u003cem\u003eet al.\u003c/em\u003e Lysocin E Targeting Menaquinone in the Membrane of Mycobacterium tuberculosis Is a Promising Lead Compound for Antituberculosis Drugs. Antimicrob. Agents Chemother. 66, e0017122 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin, A. \u003cem\u003eet al.\u003c/em\u003e Multicenter study of MTT and resazurin assays for testing susceptibility to first-line anti-tuberculosis drugs. Int. J. Tuberc. Lung Dis. 9, 901\u0026ndash;906 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOdds, F. C. Synergy, antagonism, and what the chequerboard puts between them. J. Antimicrob. Chemother. 52, 1 (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGengenbacher, M., Rao, S. P. S., Pethe, K. \u0026amp; Dick, T. Nutrient-starved, non-replicating Mycobacterium tuberculosis requires respiration, ATP synthase and isocitrate lyase for maintenance of ATP homeostasis and viability. Microbiology 156, 81\u0026ndash;87 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJeon, A. B. \u003cem\u003eet al.\u003c/em\u003e 2-aminoimidazoles potentiate \u0026szlig;-lactam antimicrobial activity against Mycobacterium tuberculosis by reducing \u0026szlig;-lactamase secretion and increasing cell envelope permeability. PLoS One 12, e0180925 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeterson, N. D., Rosen, B. C., Dillon, N. A. \u0026amp; Baughn, A. D. Uncoupling Environmental pH and Intrabacterial Acidification from Pyrazinamide Susceptibility in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 59, 7320\u0026ndash;7326 (2015).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Table 1","content":"\u003cp\u003eTable 1 is available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6497899/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6497899/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAntimicrobial host defense peptides are promising alternatives to resistance prone small molecule antibiotics. To overcome the poor physiologic stability of these therapeutic candidates it is common to prepare proteolytically resistant retro-inverso analogues, where sequence backbone direction and amino acid chirality are reversed. However, in many cases, gains in stability are offset by altered assembly propensities and reduced biologic potency. Here, we show that, contrary to the dogma for non-mycobacterial pathogens, retro-inversion of antimycobacterial host defense peptides improves their potency, specificity and host safety by an order of magnitude. Biophysical assays suggest that altered mycomembrane thermodynamics, instead of improved proteolytic stability, plays a causative role in retro-inverso mediated potency gains. Additional bacteriologic assays using a lead retro-inversed candidate, MAD1-RI, demonstrates this analogue rapidly sterilizes both replicating and dormant cultures of \u003cem\u003eMycobacterium tuberculosis\u003c/em\u003e, is effective towards drug-resistant clinical isolates of the pathogen, and synergistically enhances the activity of co-incubated antibiotics. Transcriptomic studies uncover complementary membrane destabilizing and metabolic mechanisms of antitubercular action for MAD1-RI, and in doing so identify sequence retro-inversion as a simple, but powerful, modality in the \u003cem\u003ede novo\u003c/em\u003e design of non-natural antimycobacterial peptides.\u003c/p\u003e","manuscriptTitle":"Retro-Inversion Imparts Antimycobacterial Specificity to Host Defense Peptides","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-05 11:23:47","doi":"10.21203/rs.3.rs-6497899/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"f5cba137-6e15-48ee-84cf-fffe5cbcb4ec","owner":[],"postedDate":"May 5th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47691878,"name":"Biological sciences/Biochemistry/Peptides"},{"id":47691879,"name":"Biological sciences/Biotechnology/Molecular engineering/Antimicrobials/Antibiotics"}],"tags":[],"updatedAt":"2026-01-14T08:10:11+00:00","versionOfRecord":{"articleIdentity":"rs-6497899","link":"https://doi.org/10.1038/s41467-025-67162-0","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-12-07 05:00:00","publishedOnDateReadable":"December 7th, 2025"},"versionCreatedAt":"2025-05-05 11:23:47","video":"","vorDoi":"10.1038/s41467-025-67162-0","vorDoiUrl":"https://doi.org/10.1038/s41467-025-67162-0","workflowStages":[]},"version":"v1","identity":"rs-6497899","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6497899","identity":"rs-6497899","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}