key: cord-0299176-cdomrsxc
authors: Hilpert, Kai; Gani, Jurnorain; Rumancev, Christoph; Simpson, Nathan; Lopez-Perez, Paula Matilde; Garamus, Vasil M.; von Gundlach, Andreas Robert; Markov, Petar; Scocchi, Marco; Mikut, Ralf; Rosenhahn, Axel
title: Rational Designed Hybrid Peptides Show up to a 6-fold Increase in Antimicrobial Activity and Demonstrate Different Ultrastructural Changes as the Parental Peptides measured by BioSAXS
date: 2021-11-05
journal: bioRxiv
DOI: 10.1101/2021.11.05.467440
sha: 20f59a4d5ce2740669a237cf25045a3fea90efe6
doc_id: 299176
cord_uid: cdomrsxc

Antimicrobial peptides (AMPs) are a promising class of compounds being developed against multi-drug resistant bacteria. Hybridization has been reported to increase antimicrobial activity. Here, two proline-rich peptides (consP1: VRKPPYLPRPRPRPL-CONH2 and Bac5-v291: RWRRPIRRRPIRPPFWR-CONH2) were combined with two arginine-isoleucine-rich peptides (optP1: KIILRIRWR-CONH2 and optP7: KRRVRWIIW-CONH2). Proline-rich antimicrobial peptides (PrAMPs) are known to inhibit the bacterial ribosome, shown also for Bac5-v291, whereas it is hypothesized a “dirty drug” model for the arginine-isoleucine-rich peptides. That hypothesis was underpinned by transmission electron microscopy and biological small-angle X-ray scattering (BioSAXS). The hybrid peptides showed a stronger antimicrobial activity compared to the proline-rich peptides, except when compared to Bac5-v291 against E. coli. The increase in activity compared to the arginine-isoleucine-rich peptides was up to 6-fold, however, it was not a general increase but was dependent on the combination of peptides and bacteria. BioSAXS experiments revealed that proline-rich peptides and arginine-isoleucine-rich peptides induce very different ultrastructural changes in E. coli, whereas a hybrid peptide (hyP7B5GK) shows changes, different to both parental peptides and the untreated control. These different ultrastructural changes indicated that the mode of action of the parental peptides is different from each other as well as from the hybrid peptide hyP7B5GK. All peptides showed very low haemolytic activity, some of them showed a 100-fold or larger therapeutic window, demonstrating the potential for further drug development.

In the light of the current headlines of the COVID-19 pandemic, the problem of resistance against antibiotics is still progressing and is frequently called a silent pandemic. Especially the beginning of the COVID-19 pandemic demonstrated severely the multitude of effects of an infectious disease where no effective treatment is available. The development of microbial resistance against natural occurring or human-made antibiotics is a natural selection process and was soon discovered after the use of the first antibiotic. With the emergence of multi-drug resistant strains and a dwindling development of antibacterial drugs with new modes of action, the health care systems worldwide are under threat to face a more severe pandemic if no novel antibiotics are developed. Many operations, transplantations, and immunosuppressant therapies may not be performable any longer, leaving modern medicinal care devastated. Hence, it is important and urgent to expand the discovery and translation of new alternatives to treat bacterial infections, especially with new modes of action.

According to the world health organization (WHO, 2019), more than 700,000 people die worldwide each year because of antibiotic-resistant infections. If nothing is done about this antimicrobial-resistant crisis, it was estimated in the O'Neil Report, that the numbers will increase to 10,000,000 by 2050, (https://amr-review.org/Publications.html). Already we are facing extreme multi-drug resistant bacteria like Mycobacterium tuberculosis, Pseudomonas aeruginosa, Neisseria gonorrhoeae, and Staphylococcus aureus.

Antimicrobial peptides (AMPs) are an extremely diverse group of compounds, found in all kingdoms of life that show various biological functionalities, for example, antibacterial, antifungal, antiviral, antiparasitic, anticancer, and immunomodulatory (Mahlapuu et al., 2016) . Several databases are dedicated to AMPs (Kang et . Furthermore, a study investigated the effects of AMP's on blood components, since most peptides have low bioavailability and therefore the most probable manner of administration is intravenously (Yu et al., 2015) . Consequently, the safety and efficacy of some AMPs have been investigated in clinical trials (Czaplewski et al., 2016; Greber and Dawgul, 2017) .

Bac5 (43mer) is a proline-rich AMP (PrAMP) from the cathelicidin class, isolated from bovine neutrophils more than 30 years ago (Gennaro et al., 1989) . Recently, fragments of Bac5 were investigated and the mode of action was described (Mardirossian et al., 2018a) . Bac5 fragments inhibit bacterial protein synthesis as other PrAMPs and are most active against Gram-negative pathogens (Mardirossian et al., 2019b; Tokunaga et al., 2001) . The N-terminal 1-17 amino acid fragment of Bac5, called Bac5(1-17), retained antimicrobial activity and the same overall mechanism of action (Mardirossian et al., 2019b) . Bac5(1-17) was optimized using spot synthesis and variant 291 (RWRRPIRRRPIRPPFWR-CONH2) showed improved antimicrobial activity while displaying low toxicity and the same mode of action as the parental peptide (Mardirossian et al., 2019a) .

We have developed a novel prediction method (unpublished results) for AMPs with a low haemolytic activity that was based on our peptide library screen of 3,000 members (Mikut et al., 2016) . Two arginine-isoleucine-rich peptides optP1 (KIILRIRWR) and optP7 (KRRVWIIW), were identified from this new prediction strategy, showing broad-spectrum activity. The peptide optP7 was used as a lead and selected for studying lipidation, glycosylation, cyclisation, and grafting into a cyclotide (Grimsey et al., 2020; Koehbach et al., 2021) . Hybridization of different AMPs has been investigated to further improve their performance (Shang et al., 2020; Wade et al., 2018 Wade et al., , 2019 . Here we report the creation of different hybrid molecules using peptides optP1 (KIILRIRWR) and optP7 (KRRVWIIW) in combination with the Bac5(1-17) variant 291 (RWRRPIRRRPIRPPFWR-CONH2) and a consensus sequence from proline-rich antimicrobial peptides, called consP1 (VRKPPYLPRPRPRPL-CONH2). The antimicrobial activity and haemolytic activity of the parental and hybrid peptides were determined. Small-angle X-ray scattering (SAXS) measurements, performed at the BioSAXS beamline, showed differences in the ultrastructural changes of E. coli for two selected hybrid peptides compared to the parental peptides.

Automated solid-phase peptide synthesis (SPPS) on a MultiPep RSI peptide synthesizer (Intavis, Tuebingen; Germany) was used to produce the AMPs. We applied 9-fluorenyl-methoxycarbonyl-tertbutyl (Fmoc/tBu) strategy and used N,N,N′,N′-Tetramethyl-O-(1H-benzotriazole-1-yl)uronium hexafluorophosphate (HBTU) activation. Crude peptides were analysed and purified to the homogeneity of >90% by liquid chromatography-electrospray ionisation mass spectrometry (LC-ESI-MS). For a detailed description please see (Grimsey et al., 2020) . The peptide hyP7B5Cys and consP1 were purchased at Synpeptide Co., Ltd (Shanghai, China).

In this project the following bacterial strains were used: a) methicillin- 

To grow the bacterial strains Mueller Hinton broth (MHb) (Merck, Life Science UK Limited, Dorset, UK)) was used, which was prepared concurring to the manufacturer's directions. Liquid bacterial cultures were incubated at 37 o C for 18-20 h on a shaker, bacterial culture on solid media was incubated for 18-24 h at 37 o C.

A broth microdilution assay was applied to determine the minimum inhibitory concentrations (MIC) in MHb for both, using 5 x 10 5 CFU/ml (MICs in Table 1 ) and 10 8 CFU/ml (BioSAXS and electron microscopy, values in bracket in Table 1 

The protocol to determine the haemolytic activity of the peptides and HC50 values were described previously (Grimsey et al., 2020) . Briefly, human blood was washed and diluted with PBS and incubated with a pre-made dilution series of the peptides and incubated for 1 h at 37 °C. The 100% haemolysis value. Triton X (1% final concentration) was used for Haemoglobin was measured at 414 and 546 nm using an ELISA plate reader. HC50 values were determined using the Prism software (GraphPad). 

Based on the success of the hybridization strategy reported by others, for example (Fox et Firstly, we investigated E. coli cells treated with optP7 by transmission electron microscopy (TEM) to get a better understanding of the mode of action/ultrastructural changes of the peptides on E. coli, see Figure 1 . In order to achieve a sufficient number of bacteria to form a pellet, a bacterial concentration of 1x10 8 CFU/mL was used. The MIC of optP7 against E. coli at 1x10 8 CFU/mL was determined at 14 µM, see Table 1 . The peptide was incubated with the bacteria for 60 minutes at twice the MIC (28µM). The vast majority of the cells showed no sign of lysis, see Figure 1C . The outer membrane of E. coli was damaged and blebbing was visible, where lipopolysaccharides (LPS) form small vesicles and are shed from the cell, see Figure 1 . The inner membrane seemed to remain intact. The DNA in untreated bacterial cells comprised dispersed fibrils in the cytoplasm, see Figure 1A . The bacterial DNA is concentrated in a 'DNA-plasm' called the nucleoid (Ohniwa et al., 2007) . The shape of the nucleoid depends on the ratio of opposing forces (compacting and expanding). From the alignment, a consensus sequence was derived (consP1: VRKPPYLPRPRPRPL-CONH2), see Figure 2 . Various linkers were tested to combine these peptides, see Table 1 . E)

All peptides were tested against the gram-negative bacteria Pseudomonas aeruginosa PA01 and Escherichia coli UB1005, and the gram-positive methicillin-resistant bacteria Staphylococcus aureus HO 5096 0412 (MRSA). The peptides optP1 and optP7 showed strong activity against all tested bacteria, see Table 1 . The peptide consP1 showed only activity against E. coli and Bac5-v291 showed a medium activity against MRSA and P. aeruginosa, and strong activity against E. coli. The first four hybrid peptides (hyP1CoG1/G2 and hyP7CoG1/G2) are combinations of optP1 and optP7 with consP1. A flexible glycine linker was used and the influence of the order of peptides within the hybrid molecule was investigated, see Table 1 . The combination of consP1 and optP1 resulted in hyP1CoG1 and hyP1CoG2. Both hybrid peptides showed a strongly increased activity against all bacteria when compared to consP1. When compared to optP1 they showed a decrease in activity against MRSA. The antimicrobial activity against E. coli remains unchanged in comparison to optP1. In contrast, hyP1CoG1 shows a 6-fold improvement in antimicrobial activity against P. aeruginosa and hyP1CoG2 shows a 2.4-fold improvement compared to optP1. The combination consP1 and optP7 resulted in the hybrid peptides of hyP7CoG1 and hyP7CoG2, see Table 1 . Both peptides demonstrate a strong improvement in activity when compared to consP1. When compared to optP7, nearby the same antimicrobial activity against MRSA was determined. The peptide hyP7CoG1shows a similar MIC against E. coli, hyP7CoG2 shows a 1.7-fold decrease compared to optP7. Activity against P. aeruginosa dropped 1.7-fold for hyP7CoG1and was very similar for hyP7CoG2 when compared to optP7.

The next set of peptides was created by combining Bac5-v291 with optP7 using four different linkers.

The first was the GGG-linker for flexibility (hyP7B5G), the second a KKK-linker to improve charge (hyP7B5K), the third a GKGKG-linker to combine flexibility and charge (hyP7B5GK). The fourth linker was similar to the third, however, the hybridisation was achieved by a cysteine bridge (hyP7B5Cys). The cysteine bridge is expected to be cleaved in the reducing environment of the bacterial cytosol and both peptides could then act independently once inside the bacterial cytosol.

All the hybrid peptides in this set were 13.5-fold more active against P. aeruginosa when compared to Bac5-v291 and showed very similar MIC values when compared to optP7. Peptides hyP7B5G, hyP7B5K, and hyP7B5GK show also very similar MIC values when compared to Bac5-v291 and optP7. The exception is peptide hyP7B5Cys, which showed a 2.5-fold increase in activity compared to optP7 and a 2.8-fold increase compared to Bac5-v291. All the peptides in this set are more active against MRSA when compared to both, Bac5-v291 and optP7. Peptides hyP7B5G, hyP7B5K showed a 13.5-fold improvement, peptides hyP7B5GK and hyP7B5Cys a 27-fold improvement compared to Bac5-v291. Similarly, peptides hyP7B5G, hyP7B5K showed a 3-fold improvement, peptides hyP7B5GK and hyP7B5Cys a 6-fold improvement compared to Bac5-v291. All peptides showed very low haemolytic activity (Table 1 ). The peptides with the highest haemolytic activity still showed at least a 10-fold therapeutic window, some of them 100-fold or larger. Small-angle X-ray scattering for biological samples (BioSAXS) on the P12 beamline can be used to 8 measure ultrastructural changes within the bacteria in the range between 2 and 157 nm, from which we 9 used the range from 20 to 120 nm for analysis of the bacteria after treatment with the antimicrobial 10 peptides. In recent years we used this method to classify the mode of action of various antibiotics and 11

antimicrobial peptides against E. coli and MRSA (Von Gundlach et al., 2016a, 2016b, 2019). Here we 12

used BioSAXS to investigate the ultrastructural changes of E. coli due to the exposition of the selected 13 parental peptides optP7, Bac5-v291, and the hybrid peptides hyP7B5GK and hyP7B5Cys. BioSAXS 14 measurements were performed using bacteria at 1x10 8 CFU/mL to have high enough bacteria densities 15

for the scattering experiments. The MICs for the selected peptides against E. coli at 1x10 8 CFU/mL 16 were determined (Table 1 , values in brackets). BioSAXS measurements were performed at 2x MIC. 17

The corresponding scattering curves of the peptide treated bacteria are presented in Figure 3 . To 18 quantify the subtle changes between the different treatments we used a PCA analysis using SciXMiner 19 and the "Peptide Extension" tool (Mikut, well in the PCA from the short peptide optP7. This shows that the ultrastructural changes induced by 27 these two peptides are very different. This is in agreement with our hypothesis that these two peptides 28

have a different mode of action. The hybrid peptide hyP7B5GK is in between the parent peptides on 29 PC1 and also shows an increase in PC2. The other hybrid peptide hyP7B5Cys is closer to the parent 30

optP7. This hybrid peptide is designed to fall apart in the cytosol due to the reducing environment and 31

it seems that optP7 in this case, drives the ultrastructural changes. 32 33 34 Antimicrobial resistance is a silent pandemic that needs urgent solutions in order to be able to provide 49 appropriate health care. Antimicrobial peptides are a very broad and diverse class of compounds that 50

can kill multi-drug resistant bacteria and therefore have the potential to become the next generation of 51 antibiotics. Some hybrid AMPs have been shown to possess improved activity (Shang et al., 2020; 52 Wade et al., 2018, 2019) and here we explored the combination of two different classes of peptides, 53

proline-rich peptides, and arginine-isoleucine-rich peptides. While the mode of action of the proline-54 rich peptides is known, the mode of action of the arginine-isoleucine-rich peptides has yet to be 55 discovered. We hypothesized a "dirty drug" model and performed TEM studies on one of the arginine-56

isoleucine-rich peptides (optP7). The peptides induced outer membrane stress and created an enlarged 57 and spherical nucleoid with some small denser foci, typically seen in DNA-dependent RNA 58 polymerase inhibitors and protein synthesis inhibitors. In addition, a non-lytic mode of action was 59 observed. 60

To discuss the obtained results in the context of conventional antibiotics, compared the peptides to 61 three antibiotics, tigecycline (bacteriostatic activity by binding to the 30S ribosomal subunit of the 62 bacterial ribosome), chloramphenicol (bacteriostatic activity by binding to the 30S ribosomal subunit 63 of the bacterial ribosome), and rifampicin (bactericidal activity by inhibiting the DNA-dependent RNA 64 polymerase activity), see Figure 5 . This data was measured at the same beamtime session as the peptide 65 set. As the antibiotics have slower kill kinetics than the AMPs, 240min incubation time was used. The 66

concentration of the standard antibiotics (tigecycline, MIC 4µg/ml, chloramphenicol MIC 8µg/ml, 67 rifampicin MIC 15µg/ml) was 3x their MIC against E. coli. The PCA showed that all peptides induced 68 ultrastructural changes were distinctively different from conventional antibiotics. The PCA values of 69

Bac5-v291 and hyP7B5GK deviate strongly from the untreated controls and the standard antibiotics. 70

The PCA values of peptides hyP7B5Cys and optP7 are nearest to chloramphenicol. with optP7 for 60 min, at 32 µg/ml, C) and D) treated with chloramphenicol for 240 Acknowledgements 157

We acknowledge the use of the Image Resource Facility, St George's University. We thank Thomas 158

Bruckdorfer from Iris Biotechnology for the discussion and recommendations of creating a peptide 159 that will be cleaved in the cytosol. We acknowledge the team of P12 BioSAXS beamline at EMBL, 160

Hamburg for excellent support. KH thanks the universe/life for the opportunity to continue on and 161 despite the odds to be able to keep researching. 

Attacin, an antibacterial protein 171 from Hyalophora cecropia, inhibits synthesis of outer membrane proteins in Escherichia coli by 172 interfering with omp gene transcription

Buforins: Histone H2A-derived antimicrobial 175 peptides from toad stomach

Alternatives to antibiotics-a pipeline portfolio review

Functional 181 interaction of human neutrophil peptide-1 with the cell wall precursor lipid II

Design and 184 characterization of novel hybrid antimicrobial peptides based on cecropin A, LL-37 and 185 magainin II

Automated acquisition and analysis of small 187 angle X-ray scattering data

Purification, composition, and activity of two 190 bactenecins, antibacterial peptides of bovine neutrophils

Antimicrobial Peptides Under Clinical Trials

The effect of lipidation and 196 glycosylation on short cationic antimicrobial peptide

Alternative mechanisms of action of cationic antimicrobial 199 peptides on bacteria

Sequence Requirements and an Optimization Strategy for Short Antimicrobial Peptides

Short 204 cationic antimicrobial peptides interact with ATP

Role of RNA polymerase and transcription in the 207 organization of the bacterial nucleoid

DRAMP 2.0, an updated data 209 repository of antimicrobial peptides

Antimicrobial Peptides Optimized for Binding to Escherichia coli Chaperone 212

Interactions of the 215 novel anfimicrobial peptide buforin 2 with lipid bilayers: Proline as a translocation promoting 216 factor

Comparison of a short linear 218 antimicrobial peptide with its di-sulfide-cyclized and cyclotide-grafted variants against clinically 219 relevant pathogens

Influence of the yjiL-mdtM Gene Cluster on the 221

Antibacterial Activity of Proline-Rich Antimicrobial Peptides Overcoming Escherichia coli 222 Resistance Induced by the Missing SbmA Transporter System

Antimicrobial Peptides Inhibit Either the Bacterial 70S Ribosome or the Assembly of its Large 226 50S Subunit

Insect-228 Derived Proline-Rich Antimicrobial Peptides Kill Bacteria by Inhibiting Bacterial Protein 229 Translation at the 70 S Ribosome

Antimicrobial Peptides: An 232 Emerging Category of Therapeutic Agents

Covalent binding of the natural antimicrobial peptide indolicidin to DNA abasic sites. Nucleic 236 Acids Res

Fragments of the nonlytic proline-rich antimicrobial peptide Bac5 kill Escherichia coli cells by 239 inhibiting protein synthesis

The 241 Host Antimicrobial Peptide Bac71-35 Binds to Bacterial Ribosomal Proteins and Inhibits 242 Protein Synthesis

The Dolphin Proline-Rich Antimicrobial Peptide Tur1A Inhibits Protein Synthesis by 245 Targeting the Bacterial Ribosome

Proline-rich peptides with improved antimicrobial activity against E. coli, K. pneumoniae and 248 A. baumannii

Search for Shorter Portions of 254 the Proline-Rich Antimicrobial Peptide Fragment Bac5(1-25) That Retain Antimicrobial 255 Activity by Blocking Protein Synthesis

Role 258 of the Escherichia coli SbmA in the antimicrobial activity of proline-rich peptides

Computer-Based Analysis, Visualization, and Interpretation of Antimicrobial 261 Peptide Activities

The MATLAB Toolbox SciXMiner: User's Manual and Programmer's Guide

Improving short 267 antimicrobial peptides despite elusive rules for activity

Transcription-coupled nucleoid architecture in bacteria

2: An enhanced database of structure and antimicrobial/cytotoxic activity of natural 274 and synthetic peptides

Insight 276 into Invertebrate Defensin Mechanism of Action

Hybrid Antimicrobial Peptide 279 Targeting Staphylococcus aureus and Displaying Anti-infective Activity in a Murine Model

Antibacterial activity of 282 bactenecin 5 fragments and their interaction with phospholipid membranes

BioSAXS Measurements Reveal That Two Antimicrobial Peptides Induce Similar 286 Molecular Changes in Gram-Negative and Gram-Positive Bacteria

BioSAxS measurements reveal that two antimicrobial peptides induce similar molecular 290 changes in Gram-negative and Gram-positive bacteria

Small angle X-ray scattering as a high-throughput method to classify antimicrobial 294 modes of action

Use of small-angle X-ray scattering to resolve intracellular structure 298 changes of Escherichia coli cells induced by antibiotic treatment

Systematic Analysis of Hybrid

Hybrids made from antimicrobial peptides 303 with different mechanisms of action show enhanced membrane permeabilization

APD3: the antimicrobial peptide database as a tool for 306 research and education

Effect of proline position on the 308 antimicrobial mechanism of buforin II

Design of embedded-hybrid antimicrobial 311 peptides with enhanced cell selectivity and anti-biofilm activity

Interaction of 314 blood components with cathelicidins and their modified versions