Stable isotope analysis confirms substantial changes in the fatty acid composition of bacteria treated with antimicrobial random peptide mixtures (RPMs) | Scientific Reports

2022-10-25 02:40:21 By : Ms. Alice Li

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Scientific Reports volume  12, Article number: 11230 (2022 ) Cite this article Methyl J Acid

Stable isotope analysis confirms substantial changes in the fatty acid composition of bacteria treated with antimicrobial random peptide mixtures (RPMs) | Scientific Reports

Resistance of plant-pathogenic bacteria to classic antibiotics has prompted the search for suitable alternative antimicrobial substances. One promising strategy could be the use of purposely synthesized random peptide mixtures (RPMs). Six plant-pathogenic bacteria were cultivated and treated with two RPMs previously found to show antimicrobial activity mainly by bacterial membrane disruption. Here, we show that bacteria treated with RPMs showed partly remarkable changes in the fatty acid pattern while those unaffected did not. Quantitative changes could be verified by compound specific isotope analysis of δ13C values (‰). This technique was employed due to the characteristic feature of stronger bonds between heavier isotopes in (bio)chemical reactions. As a proof of concept, the increase in abundance of a fatty acid group after RPM treatment was accompanied with a decrease in the 13C content and vice versa. We propose that our findings will help designing and synthesizing more selective antimicrobial peptides.

Since time immemorial, pathogenic bacteria have been a worldwide problem in the food sector1,2. Namely, these harmful bacteria can compromise agricultural yields2 and also cause food spoilage followed by health risks to consumers3. Hence, combating their occurrence by antibiotics is an important task. However, such efforts are getting increasingly catchier due to the resistance of pathogenic bacteria against classic antibiotics1. Therefore, one of the major challenges of our time is to track down alternative ways in combatting bacteria by active substances4.

One new and promising approach to substitute classic antibiotics could be the random antimicrobial peptide mixtures (RPMs)5,6,7. Synthesized dedicatedly in the laboratory, antimicrobial RPMs are similar in structure and they resemble the properties of naturally occurring antimicrobial peptides5,6, which are part of the immune response of animals and plants7,8. Conceptually, RPMs are produced from two purposefully selected amino acids. Namely, one amino acid has to be positively charged while the second amino acid should carry a hydrophobic residue5. Both features enable RPMs to attack the lipid bilayer of bacteria5,6. This antimicrobial peptide mixture has controlled chain length (20-mer) but the sequences are totally random5 (Fig. S1).

Despite notable antibacterial activity of RPMs with both different amino acids (including enantiomers), some of the results are difficult to predict and interpret5,6,7. For instance, some pathogenic bacteria were strongly affected while others showed no response to the RPMs and the mode of action remained unclear9. Arguably, RPMs may provoke stress on (the growth of) bacteria similarly to changes in temperature, or more basically, threats on lipid fluidity10,11,12. Under such pressure, bacteria but also higher organisms were shown to react by changes in the fatty acid (FA) composition11,13. In agreement with that, treatment of Legionella pneumophila with the antimicrobial peptide warnericin RK was followed by an increase in resistance which correlated with increasing shares of branched chain fatty acids (BCFAs) and a depletion of the average fatty acid chain length in the bacterial membranes14. Likewise, palmitic acid was more abundant in antibiotic-resistant variants of Pseudomonas fluorescens than in the ancestor strain15. Indications for changes in the fatty acids of bacteria in culture were also observed in a preliminary experiment in our laboratories with two pathogenic bacteria after treatment with RPMs.

The present study aimed to verify these indications by means of a cultivation experiment with six plant-pathogenic bacteria that are known to cause severe damage in agriculture (Table 1). The RPMs used in this study consisted of 20-mer peptide composed of either L-phenylalanine and L-lysine (FK20) or phenylalanine with D-lysine (FdK20) in 1:1 molar ratio and in random order (Fig. S1). Previous studies indicated an inactivation effect of FK20 on four of the bacterial samples (Table 1), while treatment with FdK20 had an inactivation effect on only three pathogenic bacteria (Table 1)9.

The six bacteria were cultivated to higher optical densities and aliquots were individually treated with the two RPMs. Effects of RPMs on bacterial growth was measured by means of the optical density at 600 nm (OD600)9. Then, lipids were extracted from freeze-dried samples, transesterified and the resulting fatty acid methyl esters (FAMEs) were analysed by gas chromatography coupled with mass spectrometry (GC/MS)21,22. Fatty acid patterns were related to those of the (untreated) control samples23. Evaluation was mainly based on four groups of FA, i.e. saturated (SFAs) and monoenoic ones (MUFAs), as well as n-2-methyl-branched or iso- (iFAs) and n-3-methyl-branched or anteiso-fatty acids (aFAs). Changes in the FA pattern were aimed to be verified by determination of stable carbon isotope ratios (δ13C values [‰]) of fatty acids as methyl esters by means of gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS) according to Krauß and Vetter24. The rationale for involvement of GC-IRMS is that chemical bonds are stronger between heavy isotopes24,25,26. Accordingly, the remaining share of fatty acids that decrease in abundance must get enriched in 13C while fatty acids that increase in abundance will get lighter in carbon24,25,26. Hence, measuring such changes in the δ13C values (‰) of fatty acids are direct and independent proof of alterations in the fatty acid profile in response to RPMs.

Nutrient broth (NB) was bought from Difco (Detroit, MI, USA). Methanol (99.85%), cyclohexane (99.5%) n-hexane (> 95%), and iso-propanol (99.8%) were ordered from Th. Geyer (Renningen, Germany). Sulfuric acid (96%) was bought from Carl Roth (Karlsruhe, Germany). Ethyl acetate (distilled, 99.5%), 3-hydroxyhexadecanoic acid (3-OH-16:0 (98%)), hydrocinnamic acid (3-phenylpropanoic acid, 3-Ph-3:0 (99%)), N,N-diisopropylcarbodiimide (DIC) and UltraPure water were purchased from Sigma Aldrich (Steinheim, Germany). N,N-Dimethylformamide (DMF) and piperidine were bought from Bio-Lab (Jerusalem, Israel). Undecanoic acid (> 97%) was ordered from Fluka (Steinheim, Germany). 8-Phenyloctanoic acid (8-Ph-8:0 (97%)) was bought from Alfa Aesar (Haverhill, MA, USA). 3-(Methylthio)propionic acid (3-MeS-3:0) was ordered from Santa Cruz Biotechnology (Dallas, TX, USA). Standards of iFAs and aFAs for identification were purchased from Larodan (Malmö, Sweden)27 while all other fatty acids were determined using the Supelco 37 component FAME mix (Sigma Aldrich, Steinheim, Germany) as reference standard27. The dicarboxylic FAME mix (di-7:0diME–di-12:0diME) was a standard mix isolated with countercurrent chromatography (CCC)28. The working gas CO2 (quality 4.5; δ13C value − 30.5‰) used for stable isotope analysis (IRMS), which was calibrated with the secondary reference material USGS40, was ordered from Westfalen company (Münster, Germany). The secondary reference material USGS40 (δ13C value (‰), − 26.39 ± 0.04‰ relative to Vienna Pee Dee Belemnite (VPDB)) was from the Reston Stable Isotope Laboratory (Reston, VA, USA)24. Tin capsules for liquids (3.5 × 5.5 mm, 0.04 mL volume, Sn purity 99.9%) were ordered from IVA Analysentechnik (Meerbusch, Germany).

RPMs were synthesized using the method of Hayouka et al.5 In brief, RPMs were synthesized, with a Liberty Blue™ automated microwave peptide synthesizer (CEM Corp., Matthews, NC, USA), by standard Fmoc-based solid-phase peptide synthesis (SPPS) on Rink amide resin (0.6 mmol/g substitution, 0.1 mmol scale). N,N-Diisopropylcarbodiimide and OxymaPure (Chem-Impex INTL) dissolved in N,N-dimethylformamide (DMF) were used for activation steps, and 20% (v/v) piperidine in DMF was used for Fmoc-deprotection steps. Each coupling step was conducted with binary combinations of protected amino acids, with a stock solution containing a 1:1 ratio between L-phenylalanine (F) and L/D-lysine (K/dK) (2 equivalents from each amino acid, 0.2 mmol). After cleavage from the resin, peptides were lyophilized and kept at − 20 °C. Evaluation of molecular weight was done using MALDI-TOF (Fig. S2).

Compounds were solubilized in 100% UltraPure water in a concentration of 100 mg/mL and stored at − 20 °C. These active substances consisted of RPMs of 20 components FK and FdK.

Fatty acid patterns of untreated strains of six bacteria (controls, Table 1) were recently studied by Wiedmaier-Czerny et al.23. In brief, cultivation was performed using existing standard protocols at the Institute of Biochemistry, Food Science and Nutrition (Hebrew University of Jerusalem, Israel)9. In brief, bacteria were grown in four 50 mL batches using nutrient broth medium (30 °C, with 180 rpm shaking) over night. Then, batches of the same bacteria were united, diluted to an OD600 of 0.1 in fresh nutrient broth medium (~ 108 CFU/mL) and divided into different aliquots of 0.5 L each. To each bacterial suspension RPMs were added in approximately 0.5 × [Minimal inhibitory concentration] (MIC) except bacteria with MIC value of 200 µg/mL. For these bacteria the concentration was as maximum we have used in the assay (Table 2). In these conditions we were able to observe the damaged and the undamaged bacterial cells.

MIC values determined by using sterile 96-well plates (Corning 3650) by a broth microdilution method in accordance to Hayouka et al.5 were adopted from Topman et al.9 were used to select appropriate RPMs concentrations for each bacterium. MICs (Table 2) were used in order to determine the RPMs concentrations for each bacterium in preparing samples for the fatty acid analysis.

Due to the larger amount of bacteria material required for further analysis, the amounts of RPMs were reduced in order to prevent complete standstill (Table 2). Evaluation was based by normalization of RPM-treated bacteria to % of controls (100%) as measured by OD6009. Namely, FK20 induced a reduction of the MIC of four bacteria (i.e. X. campestris, X. perforans, C. michiganensis and P. syringae) to between 4 and 21% of the corresponding untreated control samples (> 200 µg/mL) (Table 2). Two pathogenic bacteria, S. scabies and A. citrulli, were not affected by the treatment with FK20 (Table 2). Effects of FdK20 on the bacteria were very similar with the exception of P. syringae which was not affected by FdK20 (Table 2).

MICs (Table 2) were used in order to determine the RPM concentrations for each bacterium in preparing samples for the fatty acid analysis. Conditions of the previous MIC assay and the present treatments were different but main factors (i.e. RPM concentration, starting OD600 value, medium, incubation temperature) were the same. Hence, it can be assumed that the inhibition activity remained the same. After 24 h in a shaker (28 °C, 180 rpm), bacteria were centrifuged (15 min, 8000 rpm), washed 3 times with phosphate buffered saline (PBS) and followed by two additional washing steps with sterile double-distilled water. Then, bacteria were re-suspended in double-distilled water and freeze-dried before proceeding with lipid analysis.

Fatty acids present in about 10 mg of freeze-dried bacteria (pellet) or about 100 mg of freeze-dried supernatant were converted into the corresponding methyl esters (FAMEs) according to Wendlinger et al.22 with slight modifications23. In short, 2 mL of 1% sulfuric acid in methanol was added and the samples were heated first for 90 min to 80 °C, then sonicated for 10 min and finally heated again for 30 min to 80 °C. Demineralized water (1 mL), aqueous saturated NaCl solution (1 mL) and n-hexane (2 mL) were added after cooling on ice. The hexane phase was transferred into a 1.5-mL amber glass vial after shaking and phase separation. The final concentration for measuring was ~ 70 µg FAMEs per mL n-hexane. Internal standard (ISTD) solution (10 µL of 100 µg/mL 11:0 ethyl ester (11:0-EE) for 1 mL, prepared according to Wiedmaier-Czerny et al.23) was added to each sample vial. Two initially measured samples (“first treatment”, samples labelled with “p” in subscript, namely X. campestrisp, A. citrullip) were extracted by accelerated solvent extraction (ASE) by three different solvents23. These solvent systems in detail were 1.) 40 mL n-hexane/iso-propanol (3:2, v/v), 2.) 40 mL of the azeotropic mixture of cyclohexane/ethyl acetate (46:54, w/w) and 3.) 40 mL of the azeotropic mixture of methanol/ethyl acetate (44:56, w/w)23. The resulting extracts were transesterified and converted into sample solutions as shown above.

FAME solutions were measured with a 5890 Series II Plus/5972 GC/MS system (system I) using the parameters of Wiedmaier-Czerny et al.23. Standard or sample solutions (1 µL) were injected in splitless mode onto a 60 m × 0.25 mm i.d. capillary column coated with Rtx-2330 (Restek, Bad Homburg, Germany). All samples were measured twice both in full scan mode (m/z 50–550) and in selected ion monitoring (SIM) mode (m/z 74, 79, 81, 87, 88 and 101) according to Thurnhofer et al.21. Individual FAMEs, the 37 K component FAME mix and an iso-/anteiso-FAME mix were used as reference standards27,29. In GC/MS-SIM mode, saturated FAMEs were determined with m/z 87 and monounsaturated FAMEs with m/z 74 using the ratio of both ions for verification30. Also hydroxy-FAMEs (OH-FAMEs) were detected in some of the samples23. However, since concentrations of OH-FAMEs were very low, they were not evaluated in this study.

Fatty acids are presented as contribution to the sum of FAMEs (= 100%) and were divided into four classes (iFAMEs, aFAMEs, saturated FAMEs and monoenoic FAMEs). In the following, control samples were indicated with “C”, samples treated with FK20 with “FK” and samples treated with FdK20 with “FdK” (all in superscript and placed in front of the corresponding value). In the following, fatty acids will be mostly denoted “FAs” although FAMEs were measured.

Uncommon FAs in the supernatant were quantified with reference standards ordered upon tentative identification by GC/MS. Dicarboxylic fatty acids were determined using the response factor of the shortest chain length present in the standard (Di7:0-diME) and quantified by m/z 7428. S-containing fatty acids were analysed relative to the standard 3-MeS-3:0. In GC/MS-SIM mode, 3-MeS-3:0-ME was determined with m/z 61. 3-Ph-FAMEs were determined with m/z 104 and 2-Ph-FAMEs with m/z 91. These ions (m/z 61, 91 and 104) were determined in an additional GC/MS-SIM run.

Measurements were performed with a 6890 Series II GC (Agilent, Waldbronn, Germany) coupled via a ConfloIV interface to a Delta plus XP IRMS (Thermo Finnigan MAT, Bremen, Germany) system31 using the parameters of Krauß and Vetter24. A 1.5 m × 0.53 mm i.d. Rxi guard column (Restek, Bad Homburg, Germany) was linked by a SGE (Trajan) SilTite μ-union (0.32 − 0.53 mm; BGB Analytik, Rheinfelden, Germany) with a 30 m × 0.32 mm i.d. Optima 5MS column (Macherey–Nagel, Düren, Germany). The GC oven temperature was initially set for 1 min to 60 °C and then raised at 10 °C/min to 120 °C followed by 2 °C/min to 190 °C. After 1 min at this level, the temperature was raised with 20 °C/min to 250 °C (hold time 5 min). One microliter aliquots of each sample solution were injected.

GC-C-IRMS analyses were only performed with samples of the second treatment (X. campestris, A. citrulli, P. syringae, S. scabies, X. perforans, C. michiganensis) because available amounts of the first treatment (X. campestrisp, A. citrullip) were too low. For exact GC-C-IRMS measurements, FAs were anticipated to have concentrations of 20–40 ng/µL24. Due to the wide concentration ranges of individual FAs in the samples, bacterial FA solutions were stepwise diluted (Table S1) and 11:0-EE was added at 20 ng/µL as ISTD. The resulting dilutions were subsequently measured (1 µL injected, respectively) and δ13C values (‰) were determined for peaks which were in the anticipated concentration range, respectively. Each sample solution was analysed in triplicate.

Percent contributions of individual fatty acids to total fatty acids (%) were derived from conventional GC/MS (“Gas chromatography with mass spectrometry (GC/MS) analysis” section). Typically, all individual FAs contributing > 2% to the total fatty acids could be determined by GC-C-IRMS (labelled “Bi” in Tables S2–S4). These represented 75–94% of the total fatty acids in the samples (ΣBi (%), Tables S2–S4) (i.e. control as well as FK20- and FdK20-treated samples). The Σδ13C value (‰) of the FAs in a sample was calculated by summing up the measured individual δ13C values (‰) (δ13Cind) multiplied with the corresponding share (%) of the considered fatty acids followed extrapolation to 100% according to Eq. (1):

with Ai being the measured δ13C value (‰) of an individual fatty acid (δ13Cind) and Bi being its contribution to the total fatty acids (%/100%) that could be measured followed by extrapolation to 100%.

Then, the δ13Cind values (‰) of × FAs belonging to the same class of fatty acids (i.e. iFAs, aFAs, saturated (SFA) and monoenoic FAs (MUFA)) were summed up as exemplarily shown for iFAs in Eq. (2):

Visualisation of relative changes between RPM-treated samples compared to the (untreated) control samples required a further normalization step. In a first step, the factor f between Σδ13C value (‰) of control (Σδ13Ccont) of the sum of all fatty acid groups and Σδ13C value (‰) of treated samples (Σδ13Ctreat—either FK20 or FdK20) of the sum of all fatty acid groups (Tables S3, S4) was determined via Eq. (3):

Finally, Σδ13Cind values (‰) of the four fatty acid groups (e.g. Σδ13Cind (iFAs)) were normalized by multiplication with factor f (Eq. 4) (Tables S3, S4).

In the following the Σδ13Cind, norm values (‰) of FK20- or FdK20-treated samples were compared to the Σδ13Cind values (‰) of the control sample. For the sake of simplicity, the δ13C values (‰) were only described as δ13Cind, norm values (‰).

Effects of the methylation of fatty acids on δ13C values (‰) were very small. Although MeOH was more depleted in 13C than the fatty acids (δ13C: − 43.2‰), this effect was diluted by the many unaffected carbons in the acyl chain which additionally were of very similar chain length. Initial tests with MUFAs in the control sample of X. campestris and in the FK20 treatment sample differed only by 0.1‰ (i.e. − 1.7‰ with MeOH correction and − 1.6‰ without MeOH correction). As a consequence, δ13C values (‰) of fatty acids were not corrected due to the methylation.

Bulk δ13C values (‰) of (transesterified) FAs were determined with controls and FK20- and FdK20-treated samples of X. campestris, X. campestrisp and X. perforans. Aliquots of sample solutions corresponding with between 2 and 80 µg FAMEs were placed in tin capsules for liquids in triplicate (except X. campestrisp treated with FK20 due to insufficient sample material). The solvent was evaporated over a few minutes and capsules were sealed, weighed and introduced into the EA-IRMS system consisting of a Euro EA 3000 elemental analyser (Hekatech, Wegberg, Germany) and the IRMS system mentioned above using the parameters of Eibler et al.31. USGS40 (Reston Stable Isotope Laboratory, Reston, VA, USA) was used as secondary reference material. Ion currents m/z 44, m/z 45 and m/z 46 were determined relative to the working gas CO2 (δ13C value − 30.5‰) which was measured thrice per run for standardization24. Next to bulk δ13C values (‰) of FAs, some lyophilisates of control and FdK20-treated samples of X. campestris and X. campestrisp were also measured in triplicate by EA-IRMS. For this purpose, between 0.2 and 2.5 mg of lyophilisate were weighed into tin capsules for solids.

The resulting Σδ13Cind, sum value (‰) of the sum of all FA groups was also compared with the bulk EA-IRMS values of the corresponding FA fraction. For example, the Σδ13Cind, sum value (‰) in the X. perforans control sample was − 25.1‰ (Table 3) compared to − 24.3‰ obtained after EA-IRMS measurement (Table 3). The intrinsic differences between EA- and GC-C-IRMS data are in agreement with literature reports24,25,31. However, it was noticeable that the differences between EA-IRMS and GC-C-IRMS were larger when the percentage of FAs of the determined Σδ13Cind values (‰) with GC-C-IRMS was lower (Table 3). Exemplarily, the EA bulk Σδ13CEA values (‰) of X. campestris (control, FK20-- and FdK20-treated) and X. perforans (control, FK20 and FdK20 treated) were determined. Since these were almost identical to the bulk Σδ13Cind values (‰) determined via GC-C-IRMS, it was assumed that the other bacteria would behave similarly.

Sample preparations and measurements of aliquots of freeze-dried materials of all main samples (X. campestris, A. citrulli, P. syringae, S. scabies, X. perforans, C. michiganensis) were carried out in duplicate (n = 2). The relative standard deviation between the duplicates was < 1.3%, with one exception (2.2% for aFAs of X. campestris in FdK20 sample 1). In the following, resulting mean values of % contribution of individual FAs and FA groups will be reported. Samples of the preliminary treatments (X. campestrisp, A. citrullip), prepared half a year earlier, were only analysed once. Previous analysis indicated that first and second cultivations of X. campestris and A. citrulli resulted in slightly different FA patterns23. Hence, these samples had to be treated independently.

Relative standard deviations of δ13C values (‰) determined by GC-C-IRMS in triplicate were generally < 3.8% and mostly < 1%. All EA-IRMS measurements in triplicate showed relative standard deviations of < 0.35%.

A thorough statistical analysis could not be carried out due to the mostly independent samples of this study. However, considering our hypothesis that changes of > 3% between RPM-treated and untreated control samples were significant was verified by means of a t-test (p < 0.05). Exceptions were iFAs and aFAs in FdK20 treated S. scabies (Δ ~ 2%, p = 0.04) and aFAs in FdK20 treated C. michiganensis (Δ = 0.1%, p = 0.01). As will be shown below, the strongest effect was observed for samples containing iFAs. Namely, when present iFAs were affected in 75% of the bacterial samples by more than 3% after treatment with FK20 (Fig. S3). By contrast, saturated FAs and aFAs were only affected by > 3% in 25–38% of bacterial samples, respectively.

In order to exclude that high shares of bacterial lipids were released from the bacterial biomass into the cultivated medium, the entire supernatant of the centrifuged samples was separated and analysed on FAMEs. In total, between ~ 70 and 2000 µg FAMEs were detected but only 6–265 µg FAMEs originated from the four FA groups detected in the untreated control samples (i.e. iFAs, aFAs, saturated and monoenoic FAs). This amount which represented < 1% compared to 20–76 mg FAMEs in the whole freeze-dried pellet was considered negligible. Vice versa, the supernatant featured several uncommon FAs which were not detected in the pellets. GC/MS analysis of methylated supernatant enabled to identify sulphur-containing FAs (dominated by 3-MeS-3:0, Fig. S4a), dicarboxylic FAs (succinic acid (Di4:0) and glutaric acid (Di5:0), Fig. S4b), and n-aromatic FAs (2-Ph-2:0, and 3-Ph-3:0, Fig. S4c). Since the growth medium featured amino acids, they were likely generated by the bacteria from methionine, aspartic acid and glutamic acid, and phenylalanine, respectively. However, the uncommon FAs occurred arbitrarily, and effects due to treatment with RPM could not be observed. Most importantly, the bacterial pellets contained virtually all of the four FA groups which will be studied.

Possible effects of FK20 were evaluated by comparing the abundance of the four FA groups (iFAs, aFAs, saturated and monoenoic FAs) and partly individual FAs in pellets of centrifuged (untreated) controls and affected bacteria. With the exception of some very low abundant FAs, the same variety of FAs was detected in untreated controls and FK20-treated samples.

S. scabies and A. citrulli were not affected in their growth by the treatment with FK20. In agreement with that the FA patterns of treated samples and the untreated control were virtually the same. In S. scabies, iFAs (C39.1% vs. FK38.7%) and saturated FAs (C9.8% vs. FK9.7%) were virtually the same in untreated control and treated sample (Fig. S5). Only the share of monoenoic FAs was slightly lower (C9.6% and FK7.3%) in favour of aFAs (C41.5% and FK44.5%). Likewise, the most prominent FAs (a15:0 (C26.1% vs. FK27.5%), i16:0 (C22.2% vs. FK20.6%) and a17:0 (C16.0% vs. FK17.6%), Table S5) were barely affected by the treatment with FK20.

Also in A. citrulli, saturated FAs (C50.5% and FK49.5%) and monoenoic FAs (C48.5% vs. FK49.6%) were predominant and the slight differences were deemed irrelevant and considered as natural variations (Fig. S5, Table S6). In accordance, p values (> 0.05) did not indicate significant changes. As mentioned before, the control sample of the preliminary treatment (A. citrullip) showed a slightly different fatty acid pattern (slightly richer in monoenoic FAs)23. However, the FA composition of A. citrullip was also almost the same before and after the treatment (C47.4% vs. FK48.3% saturated FAs and C52.5% vs. FK51.6% monoenoic FAs) (Fig. S5).

Based on these samples, changes in the share of the four FA groups by less than 3% were considered to be subject to natural variations, while changes by > 3% in one FA group in RPM-treated bacteria will be treated as an effect caused by the treatment (“Quality control” section). This was taken into account when bacterial samples which were affected by FK20 were studied in the following.

Remarkably enough, all four bacteria samples (X. campestris, X. perforans, C. michiganensis and P. syringae) whose growth was impaired by FK20 showed higher changes of > 3% in the abundance of at least one of the four FA groups (Fig. 1, Table 4).

Percentage distribution of fatty acid groups (Σiso-FAs, Σanteiso-FAs, Σsaturated FAs and Σmonoenoic FAs) of (a) Xanthomonas campestris pathovar (pv) campestris (X. campestris), (b) Xanthomonas perforans (X. perforans), (c) Clavibacter michiganensis (C. michiganensis), and (d) Pseudomonas syringae pv. tomato (P. syringae) samples without treatment (control) and with FK20 treatment.

Treatment of X. campestris with FK20 reduced the share of iFAs by ~ 20% compared to the control (C38.4% → FK31.6%, Fig. 1a). Also, the share of saturated FAs was slightly decreased from C19.7% to FK17.8% (Fig. 1a). This decline was compensated for by a strong increase of monoenoic FAs by 8.2% from C23.2% to FK31.4% (Fig. 1a). Only the share of aFAs remained nearly the same (Fig. 1a). The observed changes were almost entirely due to a decrease in i15:0 by 5% (C23.7% → FK18.8%) in favour of 16:1n-7 which almost doubled its share (C11.1% → FK20.7%) (Table S7). Still, the same variety of FAs was detected before and after FK20 treatment (Table S7). Although the FA pattern of the control sample of X. campestris and the preliminary treatment X. campestrisp differed slightly, relative changes in X. campestrisp induced by the treatment with FK20 were on a similar level as in X. campestris (Fig. S6). Hence, these measurements produced strong evidence that the observed changes were meaningful.

In accordance with that, FK20 treatment of X. perforans also caused a remarkable shift in the FA pattern compared to the untreated control sample. Specifically, the enormous increase of iFAs by more than 20% (C41.3% → FK62.4%, Fig. 1b) was compensated for by notable decreases of the abundance in monoenoic FAs (C20.1% → FK10.6%) > saturated FAs (C23.3% → FK15.8%) > aFAs (C14.5% → FK11.0%) (Fig. 1b). Both i15:0 (C19.1% → FK27.2%) and i17:0 (C10.3% → FK16.6%) similarly increased in relative abundance, while great reduction was observed for 16:0 (C14.8% → FK9.7%), 16:1n-7 (C13.8% → FK6.3%) and a15:0 (C11.2% → FK7.6%) (Table S8).

Treatment of C. michiganensis with FK20 also effected the relative abundance of iFAs but in the opposite way as found above for X. perforans. Namely, FK20 treatment of C. michiganensis caused a decrease in iFAs (C50.6% → FK45.8%) (Fig. 1c) in favour of slight increases of saturated (C23.0% vs. FK24.5%) and monoenoic FAs (C14.0% vs. FK17.0%, Fig. 1c). Only the share of aFAs remained the same (C11.8% → FK11.7%, Fig. 1c). Changes in iFAs were mainly due a decrease in i15:0 (C37.8% → FK33.1%) (Table S9).

P. syringae only featured saturated FAs (which decreased from C41.0% to FK38.4%) and monoenoic FAs (which increased from C57.7% to FK60.7%) (Fig. 1d). Assumedly, the lower magnitude of changes in the FA pattern compared to X. campestris, X. perforans, and C. michiganensis was mainly due to the absence of iFAs in P. syringae (Table S10). Second, decreasing MIC values in the order P. syringae > X. perforans > X. campestris > C. michiganensis also indicated the least vulnerability of this bacterium (Table 2). Although changes in FAs of P. syringae caused by the treatment of FK20 were comparably small, they still exceeded those observed with A. citrulli whose growth was not affected by FK20.

Despite deviations in amount and response to MIC values, all examples indicated that changes in their FA pattern could serve as indicators for bacterial response on RPMs. For further support of this hypothesis, the experiments were repeated with another RPM, FdK20 with the only difference between them beeing the stereocenter of the cationic amino acid (lysine).

FdK20 treatment had no effect on the bacterial growth of three bacteria (S. scabies, both A. citrulli and A. citrullip, as well as P. syringae) also the FA patterns were not changed compared to the corresponding controls (Fig. S5). Differently to FK20 (“Effect of treatments with the RPM FK20 on the fatty acid pattern of the six bacteria samples” section), FdK20 did not impair the growth of P. syringae (Table 2). However, the absence of response to P. syringae was in agreement with no changes in FA composition.

Growth of the remaining three bacteria was similarly affected by FdK20 as by FK20. MIC values were slightly different but still in the order X. perforans > X. campestris > C. michiganensis (Table 2). Also, FdK20 induced similar changes in the FA patterns of X. perforans as FK20 (Table 5) but effects of FdK20 on X. campestris and C. michiganensis were different to FK20.

Treatment with FdK20 also strongly changed the FA pattern of X. campestris but in opposite direction as FK20. Namely, iFAs (C38.4% → FdK47.8%) and monoenoic FAs were increasing while saturated FAs dropped by ~ 15% (Fig. 2). Still, the direction of changes in the share of iFAs was also found to be non-uniform in the case of FK20 (“Effect of treatments with the RPM FK20 on the fatty acid pattern of the six bacteria samples” section). However, a noticeable exception was observed in the case of C. michiganensis. Although FdK20 was found to be active (see low MIC value in Table 2) and rich in iFAs, no changes were observed in the FA composition. These findings prompted us to determine δ13C values (‰) of individual FAs (as methyl esters) with GC-C-IRMS in order to verify potential relationships between growth effects and FA patterns.

Differences of percentage distribution and δ13C values (‰) of the fatty acid groups of the control and the FK20- or FdK20-treated samples of Xanthomonas campestris pathovar (pv) campestris (X. campestris). The left axis (orange) shows the difference of the percentage distribution of the fatty acid groups between FK20 or FdK20 – control and the right axis (blue) shows the difference of Σδ13Cind, norm values (‰) of the fatty acid groups of FK20 or FdK20—control.

Stronger bonds between heavier isotopes24,25,26 indicated that a decrease in the share of a FA was linked with an enrichment in 13C (less negative δ13C value (‰)). Vice versa, an increase in FA abundance should lead to a depletion in 13C (more negative δ13C value (‰)). For GC-C-IRMS measurements, FA to be studied must not co-elute with others and the peak area must be within a greater range. Both prerequisites were taken into account by modifying the GC oven program and measuring samples in different dilutions. However, only FAs contributing > 2% to the total fatty acids could be analysed by GC-C-IRMS (“Stable carbon isotope analysis (δ13C values (‰)) of FAMEs via gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS)” section). In dependence of the sample, 75–94% of the bacterial FAs could be studied by GC-C-IRMS (“Stable carbon isotope analysis (δ13C values (‰)) of FAMEs via gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS)” section). Since, a few major fatty acids showed most changes (“Effect of treatments with the RPM FK20 on the fatty acid pattern of the six bacteria samples”), the determinations were considered representative. However, in order to make measurements comparable with each other, shares measured by GC-C-IRMS were normalized to 100% in a first step (Eq. 1, “Calculation mode for δ13C values (‰) of FAs and groups of FAs via GC-C-IRMS” section). Then, shares of individual FAs (δ13Cind values (‰)) were summed up to give the Σδ13Cind value (‰) of the sample (Eq. 2). The correctness of both steps was tested with sample X. perforans after FdK20 treatment in which 94% could be measured by GC-C-IRMS (“94% sample”). Exclusion of several FAs allowed to reduce the considered FA pool to 75% (“75% sample”; only major fatty acids were considered). After normalization to 100% and addition of individual δ13Cind values (‰), the “94% sample” showed a Σδ13Cind value (‰) of − 26.4‰ compared to − 26.2‰ in the “75% sample”. This small difference verified that the normalization procedure was adequate. Also low standard deviations of < 1%, that means + /- 0.1‰ on the δ scale, showed that the differences were negligible.

GC-C-IRMS evaluation was also based on the four groups of fatty acids (i.e. iFAs, aFAs, saturated and mono-enoic FAs). In either case, GC-C-IRMS measurements of Σδ13Cind, norm values (‰) verified the changes in the FA pattern caused by FK20 or FdK20 treatments of the bacteria (relative to the control samples) (Table 4, Fig. 2). As predicted, a decrease in the share of a FA group was accompanied with an increase in the 13C content which is represented by more positive Σδ13Cind, norm values (‰). Last but not least, FA groups that did not change in abundance by the RPM treatment showed no changes in Σδ13Cind, norm values (‰).

For example, in X. perforans the share of iFAs increased from C45% to FK66% and FdK65%, respectively (Table 5). Accordingly, the Σδ13Cind, norm value (‰) of iFAs (C − 12.0‰) was getting more negative after the treatment (FK− 16.6‰ and FdK− 16.5‰) which corresponds with a depletion in 13C (Table 5). Vice versa, saturated FAs decreased in relative abundance from C21% to FK18% and FdK6%, in X. campestris while the corresponding Σδ13Cind, norm values (‰) verified enrichment in 13C (C − 4.1‰ to FK − 3.5‰ or FdK − 1.2‰) (Table 5). Finally, A. citrulli showed negligible changes in the only two FA groups (saturated and monoenoic FAs) and also the Σδ13Cind, norm values (‰) of saturated (C − 10.8‰, FK − 10.6‰ and FdK − 10.5‰) and monoenoic FAs (C − 11.6‰, FK − 11.8‰ and FdK − 11.8‰) remained unchanged (Table 5).

Most remarkably, the reversed change in the iFA share of X. campestris after treatment with FK20 and FdK20, respectively, could be verified by GC-C-IRMS (Table 5, Fig. 2). Namely, the increase of iFAs in the FdK20-treated sample was paired with a depletion in 13C (more negative Σδ13Cind, norm values (‰)) while FK20 caused a decrease of iFAs and an increase in the Σδ13Cind, norm value (‰) (more positive Σδ13Cind, norm values (‰)) (Fig. 2). Although it could not be clarified in this study, the strong changes in the fatty acid composition indicated that the enzyme functionality of fatty acid synthesis was possibly changed by the organisms.

In all experiments, the sum δ13C values (‰) of the fatty acids before and after the treatment agreed very well. Within the four groups of fatty acids, changes were most striking in the case of iFAs and MUFAs. Smaller changes in the δ13C values (‰) were observed for SFA while aFAs were barely affected by the treatments. Typically, relevance and δ13C values (‰) of iFAs and MUFAs changed in opposite direction. Namely, if iFAs increased in abundance, the MUFAs content was decreased. This may indicate that the mode of fatty acid synthesis of the organisms was changed due to the impact of RPMs.

Stable carbon isotope analysis invariably verified the changes in the FA groups of bacteria caused by treatment with FK20 and FdK20. Also, growth inhibition of bacteria by FK20 and FdK20 was generally associated with changes in the FA groups except for FdK20 in the case of C. michiganensis (inhibiting effect but no changes in the FAs). Nevertheless, also this observation could be fully verified by GC-C-IRMS analysis.

For all bacteria that showed an effect, differences were particularly noticeable when iFAs were present which is a rather common feature of bacteria. Although each bacterium responded differently and the effects could not be explained mechanistically this time, congruency of changes in FA groups and δ13C values (‰) observed in this study may contribute fundamentally to the understanding of the effects of RPMs on bacterial lipid membranes. Similarly, it would be interesting to test if corresponding changes in the fatty acid pattern and stable isotope ratio will also take place in treatments with classic antibiotics such as polymyxin B. In either case the approach used in this study could be helpful in the search not only of novel antimicrobial RPMs but also other antibiotics.

The datasets used and/or analysed during the current study available from authors upon request.

Fuchs, G. et al. Allgemeine Mikrobiologie, 9., vollst. überarb. und erw. Auflage (Thieme, Stuttgart, 2014).

Mahlein, A. K., Oerke, E. C., Steiner, U. & Dehne, H. W. Recent advances in sensing plant diseases for precision crop protection. Eur. J. Plant Pathol. 133, 197–209. https://doi.org/10.1007/s10658-011-9878-z (2012).

Holban, A. M. & Grumezescu, A. M. (eds) Microbial Contamination and Food Degradation. Handbook of Food Bioengineering (Academic Press, Elsevier, 2018).

Martens, E. & Demain, A. L. The antibiotic resistance crisis, with a focus on the United States. J. Antibiot. 70, 520–526. https://doi.org/10.1038/ja.2017.30 (2017).

Hayouka, Z. et al. Interplay among subunit identity, subunit proportion, chain length, and stereochemistry in the activity profile of sequence-random peptide mixtures. J. Am. Chem. Soc. 135(32), 11748–11751. https://doi.org/10.1021/ja406231b (2013).

Article  CAS  PubMed  Google Scholar 

Hayouka, Z. et al. Binary encoding of random peptide sequences for selective and differential antimicrobial mechanisms. Angewandte Chemie (Int. Ed.) 56(28), 8099–8103. https://doi.org/10.1002/anie.201702313 (2017).

Bauer, T. S., Menagen, B., Avnir, D. & Hayouka, Z. Random peptide mixtures entrapped within a copper-cuprite matrix: New antimicrobial agent against methicillin-resistant Staphylococcus aureus. Sci. Rep. 9(1), 1–8. https://doi.org/10.1038/s41598-019-47315-0 (2019).

Brogden, K. A. Antimicrobial peptides: Pore formers or metabolic inhibitors in bacteria. Nat. Rev. Microbiol. 3(3), 238–250. https://doi.org/10.1038/nrmicro1098 (2005).

Article  CAS  PubMed  Google Scholar 

Topman, S. et al. Random peptide mixtures as new crop protection agents. Microb. Biotechnol. 11(6), 1027–1036. https://doi.org/10.1111/1751-7915.13258 (2018).

Article  CAS  PubMed  PubMed Central  Google Scholar 

Nielsen, L. E., Kadavy, D. R., Rajagopal, S., Drijber, R. & Nickerson, K. W. Survey of extreme solvent tolerance in gram-positive cocci: Membrane fatty acid changes in Staphylococcus haemolyticus grown in toluene. Appl. Environ. Microbiol. 71(9), 5171–5176. https://doi.org/10.1128/AEM.71.9.5171-5176.2005 (2005).

Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

Li, J., Chikindas, M. L., Ludescher, R. D. & Montville, T. J. Temperature-and surfactant-induced membrane modifications that alter Listeria monocytogenes nisin sensitivity by different mechanisms. Appl. Environ. Microbiol. 68(12), 5904–5910. https://doi.org/10.1128/AEM.68.12.5904-5910.2002 (2002).

Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

Margesin, R. Cold-adapted Organisms: Ecology, Physiology, Enzymology and Molecular Biology (Springer, 1999).

de Mendoza, D. & Cronan, J. E. Thermal regulation of membrane lipid fluidity in bacteria. Trends Biochem. Sci. 8(2), 49–52. https://doi.org/10.1016/0968-0004(83)90388-2 (1983).

Verdon, J. et al. Fatty acid composition modulates sensitivity of Legionella pneumophila to warnericin RK, an antimicrobial peptide. Biochimica et Biophysica Acta (BBA)-Biomembranes 1808(4), 1146–1153. https://doi.org/10.1016/j.bbamem.2010.12.011 (2011).

Naghmouchi, K., Le Lay, C., Baah, J. & Drider, D. Antibiotic and antimicrobial peptide combinations: Synergistic inhibition of Pseudomonas fluorescens and antibiotic-resistant variants. Res. Microbiol. 163(2), 101–108. https://doi.org/10.1016/j.resmic.2011.11.002 (2012).

Article  CAS  PubMed  Google Scholar 

da Silva, A. C. R. et al. Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417(6887), 459–463. https://doi.org/10.1038/417459a (2002).

Article  ADS  PubMed  Google Scholar 

Jones, J. B., Lacy, G. H., Bouzar, H., Stall, R. E. & Schaad, N. W. Erratum to “Reclassification of the Xanthomonads associated with bacterial spot disease of tomato and pepper”. Syst. Appl. Microbiol. 27, 755–762. https://doi.org/10.1078/0723202042369884 (2004).

Article  CAS  PubMed  Google Scholar 

Burdman, S., Kots, N., Kritzman, G. & Kopelowitz, J. Molecular, physiological, and host-range characterization of Acidovorax avenae subsp. citrulli isolates from watermelon and melon in Israel. Plant Dis. 89(12), 1339–1347. https://doi.org/10.1094/PD-89-1339 (2005).

Article  CAS  PubMed  Google Scholar 

Cuppels, D. A. Generation and characterization of Tn5 insertion mutations in Pseudomonas syringae pv. tomato. Appl. Environ. Microbiol. 51(2), 323–327. https://doi.org/10.1128/AEM.51.2.323-327.1986 (1986).

Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

Meletzus, D. & Eichenlaub, R. Transformation of the phytopathogenic bacterium Clavibacter michiganense subsp. michiganense by electroporation and development of a cloning vector. J. Bacteriol. 173(1), 184–190. https://doi.org/10.1128/jb.173.1.184-190.1991 (1991).

Article  CAS  PubMed  PubMed Central  Google Scholar 

Thurnhofer, S., Lehnert, K. & Vetter, W. Exclusive quantification of methyl-branched fatty acids and minor 18: 1-Isomers in foodstuff by GC/MS in the SIM mode using 10, 11-dichloroundecanoic acid and fatty acid ethyl esters as internal standards. Eur. Food Res. Technol. 226(5), 975–983. https://doi.org/10.1007/s00217-007-0619-7 (2008).

Wendlinger, C., Hammann, S. & Vetter, W. Various concentrations of erucic acid in mustard oil and mustard. Food Chem. 153, 393–397. https://doi.org/10.1016/j.foodchem.2013.12.073 (2014).

Article  CAS  PubMed  Google Scholar 

Wiedmaier-Czerny, N. et al. Detailed analysis of the fatty acid composition of six plant-pathogenic bacteria. J. Chromatogr. B 1162, 122454. https://doi.org/10.1016/j.jchromb.2020.122454 (2021).

Krauß, S. & Vetter, W. Stable carbon and nitrogen isotope ratios of red bell pepper samples from Germany, The Netherlands, and Spain. J. Agric. Food Chem. 67(14), 4054–4063. https://doi.org/10.1021/acs.jafc.9b01631 (2019).

Article  CAS  PubMed  Google Scholar 

Rieley, G. Derivatization of organic compounds prior to gas chromatographic–combustion–isotope ratio mass spectrometric analysis: Identification of isotope fractionation processes. Analyst 119(5), 915–919. https://doi.org/10.1039/AN9941900915 (1994).

Article  ADS  CAS  Google Scholar 

Brand, W. A. & Coplen, T. B. Stable isotope deltas: Tiny, yet robust signatures in nature. Isot. Environ. Health Stud. 48(3), 393–409. https://doi.org/10.1080/10256016.2012.666977 (2012).

Eibler, D., Seyfried, C. & Vetter, W. Enantioselectivity of anteiso-fatty acids in hitherto uninspected sample matrices. J. Chromatogr. B 1061, 233–239. https://doi.org/10.1016/j.jchromb.2017.07.029 (2017).

Lehnert, K., Rashid, M. M., Barman, B. K. & Vetter, W. Identification and quantification of dicarboxylic fatty acids in head tissue of farmed Nile tilapia (Oreochromis niloticus). Eur. Food Res. Technol. 247(7), 1771–1780. https://doi.org/10.1007/s00217-021-03747-5 (2021).

Eibler, D., Krüger, S., Skirnisson, K. & Vetter, W. Combined thin layer chromatography and gas chromatography with mass spectrometric analysis of lipid classes and fatty acids in malnourished polar bears (Ursus maritimus) which swam to Iceland. J. Chromatogr. B 1046, 138–146. https://doi.org/10.1016/j.jchromb.2017.01.043 (2017).

Thurnhofer, S. & Vetter, W. A gas chromatography/electron ionization− mass spectrometry− selected ion monitoring method for determining the fatty acid pattern in food after formation of fatty acid methyl esters. J. Agric. Food Chem. 53(23), 8896–8903. https://doi.org/10.1021/jf051468u (2005).

Article  CAS  PubMed  Google Scholar 

Eibler, D., Seyfried, C., Kaffarnik, S. & Vetter, W. Anteiso-fatty acids in Brussels Sprouts (Brassica oleracea var. gemmifera L.): Quantities, enantioselectivities, and stable carbon isotope ratios. J. Agric. Food Chem. 63(40), 8921–8929. https://doi.org/10.1021/acs.jafc.5b03877 (2015).

Article  CAS  PubMed  Google Scholar 

We are grateful for financial support by the Ministry of Science, Research and Arts, Baden-Württemberg (Germany) in the framework of the partnership program between the Robert H. Smith Faculty of Agriculture, Food and Environment of the Hebrew University of Jerusalem and the Faculties of Agricultural and Natural Sciences of the University of Hohenheim.

Open Access funding enabled and organized by Projekt DEAL.

Department of Food Chemistry (170B), Institute of Food Chemistry, University of Hohenheim, 70593, Stuttgart, Germany

Nina Wiedmaier-Czerny, Dorothee Schroth, Stephanie Krauß & Walter Vetter

Institute of Biochemistry, Food Science and Nutrition, The Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Rehovot, 7610001, Israel

Shiri Topman-Rakover, Aya Brill & Zvi Hayouka

Department of Plant Pathology and Microbiology, Institute of Environmental Sciences, The Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Rehovot, 7610001, Israel

Shiri Topman-Rakover, Aya Brill & Saul Burdman

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

N.W.-C. performed the sample cleanup, evaluated the results, generated Figures and Tables, and wrote the first draft of the manuscript. W.V. designed the study, supervised the fatty acid and IRMS analysis, and was involved in the writing of the first draft of the manuscript. D.S. was involved in the fatty acid analysis. S.K. performed the IRMS measurements. S.T.-R. and A.B. cultivated and treated the bacteria. S.B. selected and provided the bacterial strains. Z.H. designed the study, supervised the bacterial part and was involved in the writing. All authors read the final draft of the manuscript.

Correspondence to Zvi Hayouka or Walter Vetter.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Wiedmaier-Czerny, N., Schroth, D., Krauß, S. et al. Stable isotope analysis confirms substantial changes in the fatty acid composition of bacteria treated with antimicrobial random peptide mixtures (RPMs). Sci Rep 12, 11230 (2022). https://doi.org/10.1038/s41598-022-13134-z

DOI: https://doi.org/10.1038/s41598-022-13134-z

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Scientific Reports (Sci Rep) ISSN 2045-2322 (online)

Stable isotope analysis confirms substantial changes in the fatty acid composition of bacteria treated with antimicrobial random peptide mixtures (RPMs) | Scientific Reports

2-Amino-4-Nitrophenol Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.