ABSTRACT _Salmonella enterica_ causes an estimated 1 million illnesses in the United States each year, resulting in 19,000 hospitalizations and 380 deaths, and is one of the four major

global causes of diarrhoeal diseases. No effective treatments are available to the food industry. Much attention has been given to colicins, natural non-antibiotic proteins of the

bacteriocin class, to control the related pathogen _Escherichia coli_. We searched _Salmonella_ genomic databases for colicin analogues and cloned and expressed in plants five such proteins,

which we call salmocins. Among those, SalE1a and SalE1b were found to possess broad antimicrobial activity against all 99 major _Salmonella_ pathovars. Each of the two salmocins also showed

FOR THE CONTROL OF FOODBORNE _SALMONELLA ENTERICA_ SEROVAR TYPHIMURIUM Article Open access 02 March 2023 PHENOTYPIC CHARACTERIZATION AND GENOMIC ANALYSIS OF A SALMONELLA PHAGE L223 FOR

BIOCONTROL OF _SALMONELLA_ SPP. IN POULTRY Article Open access 03 July 2024 CHARACTERIZATION OF TWO NOVEL _SALMONELLA_ PHAGES HAVING BIOCONTROL POTENTIAL AGAINST _SALMONELLA_ SPP. IN

GASTROINTESTINAL CONDITIONS Article Open access 29 May 2024 INTRODUCTION _Salmonella_ is a rod-shaped Gram-negative bacterium of the _Enterobacteriaceae_ family. _Salmonella enterica_ is the

type subspecies and is further divided into six subspecies that include over 2500 serovars. _S. enterica_ infections are common and are the leading cause of gastroenteritis worldwide1.

_Salmonella_ causes an estimated 1 million illnesses in the United States each year, resulting in an estimated 19,000 hospitalizations and 380 deaths2

(https://www.cdc.gov/salmonella/index.html). Over the last five years (2012–2017), 51 _Salmonella_ outbreaks have been recorded in USA. Most of the food poisonings were due to contaminated

poultry or vegetables and fruits, but also red meats and fish (https://www.cdc.gov/salmonella/outbreaks.html). Worldwide, _Salmonella_ contamination in the food chain is one of four key

global causes of diarrhoeal diseases, with 550 million annual illnesses (https://www.who.int/mediacentre/factsheets/fs139/en/), and with exposure routes for non-typhoidal _Salmonella_

detected in all food categories3. The prevention and treatment of _Salmonella_ infections, and the reduction of contamination of food and feed, have traditionally relied on generic methods

such as continuous cold chains, thorough cooking, pasteurization and proper animal farm and food processing plant hygiene. The advantage of this approach is that the interventions do not

require much prior knowledge about the specific _Salmonella_ strains needing control. However, physicochemical interventions such as heating or treating food with oxidizing agents, acids or

salts involve occupational risks and could change the properties or flavor of the treated food in undesirable ways4. Clinically, options for treating non-typhoidal _Salmonella_ infections

with existing chemistries are limited due to the potential for increased levels of excretion and the development of antibiotic resistance by _Salmonella_5,6. Therefore, there exists an

urgent need for effective yet safe approaches for preventing _Salmonella_ infections, and for compatible methods for preventing or minimizing contamination of food and feed with

_Salmonella_. In search for new alternatives to antibiotics to control pathogenic strains of the related Gram-negative bacterium _Escherichia coli_, much attention has recently been given to

colicins. These are natural non-antibiotic antibacterial proteins (bacteriocins), produced by certain strains of _E. coli_ (hence the name colicins) that kill or inhibit the growth of other

strains of _E. coli_ to establish ecological dominance7,8,9. Colicin-like bacteriocins are structurally organized in three domains: an N-terminal translocation domain responsible for

transfer across the cell envelope mediated by bacterial translocator proteins, a central receptor-binding domain responsible for interaction with bacterial outer membrane receptor proteins

and a C-terminal cytotoxic domain responsible for antibacterial cytotoxic activity. Depending on the translocation system employed, colicins are divided into two groups: group A (Tol system,

e.g. TolABQR proteins involved in colicin import) and group B (Ton system, e.g. TonB, ExbBD proteins involved in colicin import). Based on the bactericidal activity provided by the

cytotoxic domain, colicins are classified as pore-forming, nucleases, or inhibitors of peptidoglycan synthesis. To protect bacterial cells producing colicins from their cytotoxic activity,

specific inhibitors called immunity proteins are simultaneously produced7,8. Recently, we have expressed in plants and characterized 12 colicins and found that they are excellent candidates

as food additives for controlling _E. coli_ in food10. The U.S. Food & Drug Administration (FDA) twice granted our plant-produced colicins GRAS (Generally Recognized As Safe) status as

antimicrobials for application to fruits and vegetables (GRN 593, https://www.accessdata.fda.gov/scripts/fdcc/index.cfm?set=GRASNotices&id=593) and meat products (GRN 676,

https://www.accessdata.fda.gov/scripts/fdcc/?set=GRASNotices&id=676), thus paving the way to commercialization of colicins as food additives or food processing aides for control of

foodborne _E. coli_ infections. The two genera, _Escherichia_ and _Salmonella_, are closely related, and _Salmonella_ strains have been known to sometimes harbor colicin genes11,12,13,14;

therefore, as a part of this study, we evaluated almost all (21) known colicins against major pathogenic strains of _Salmonella_, and found that Group B colicins provided moderate control,

which, however, was insufficient for practical applications. Since _Salmonella_-specific colicin analogues have never been looked into, we then searched _Salmonella_ genomic databases for

colicin analogues and cloned and expressed five such proteins, which we called salmocins (for _Salmonella_ bacteriocins). All plant-expressed salmocins showed high antibacterial activity,

and included different modes of action. Two porin-type bacteriocins in particular, SalE1a and SalE1b, were each found to possess broad antimicrobial activity against all 99 major pathogenic

_Salmonella_ strains tested with remarkably high potency (average >106 arbitrary activity units (AU)/µg protein, up to 3 orders of magnitude higher than colicins). We also evaluated

salmocins as antibacterial agents on meat matrices, and compared the intra- and inter-specific activities of salmocins and colicins. Our results show for the first time the practical

application of salmocins as food safety interventions as well as their potential as novel non-antibiotic bactericides. RESULTS GREEN PLANTS ARE EFFICIENT EXPRESSION HOSTS FOR PRODUCTION OF

FUNCTIONAL COLICINS AND SALMOCINS Screening of 23 bacteriocins (colE2, colE3, colE5, colE6, colE7, colE8, colE9, colD, colIa, colIb, colN, colK, colB, colA, colR, colY, colM, col5, col10,

colS4, cloacin DF13, colU, col28b), most of them _E. coli_ colicins, for antimicrobial activity against _Salmonella enterica_ ssp. _enterica_ serovars revealed that Group A bacteriocins

tested (colicins E2, E3, E5, E6, E7, E8, E9, A, N, K, R, Y, U, 28b and cloacin DF13), which utilize the Tol translocation machinery for import to susceptible bacteria, were non-effective

(Supplementary Fig. 1). Conversely, activity against _Salmonella_ was confirmed for colicins M, Ia, Ib, 5, 10 and S4, in agreement with literature reports. _Salmonella_ serovar Typhimurium

was reported to be insensitive to Group A colicins E1, E2 and E315,16 and it was demonstrated that the TolQRA region of _Salmonella_ serovar Typhimurium has differential levels of

expression, means of regulation, and, likely, functions from its corresponding region in _E. coli_17, indicating that although the BtuB receptor utilized by these colicins is functional, but

translocation is impaired. Sensitivity of _Salmonella_ to colM was described decades ago16 and NCBI database searches showed that analogues of _E. coli_ colicins Ia, Ib, M, and B seem to be

widely distributed in _Salmonella_ with 99–100% identity in amino acid sequence. Due to relatively low overall antimicrobial activity, we concluded that the efficient control of all

selected disease-relevant _S. enterica_ serovars by colicins M, Ia, Ib, 5, 10 and S4 was not feasible. Therefore, putative bacteriocin genes from _Salmonella_ were selected from the NCBI

database on the basis of homology to the activity domains of _E. coli_ colicins but with differences in amino acid composition in translocation and receptor protein domains of Group A

colicins (Supplementary Fig. 2). The cut-off was set to not less than 70% identity on amino acid level for cytotoxicity domain to ensure sufficient similarity. For the translocation domain,

the cut-off was set to not more than 80% identity to ensure sufficient difference. The latter rule had one exception: the translocation domain of SalE3 shared 95% identity to that of colE3

which could explain the lowest degree of anti-_Salmonella_ activity of SalE3 among salmocins. Five _Salmonella_ bacteriocins, which we call salmocins, representing 3 activity groups (DNase,

RNase, pore-forming) of analogous proteins, were selected and designated Sal, followed by additional letters designating the highest similarity to its corresponding _E. coli_ colicin.

Salmocin nucleotide sequences optimized for _N. benthamiana_ codon-usage were cloned into a tobacco mosaic virus (TMV)-based assembled expression vector (Table 1, Supplementary Fig. 3). To

prevent potential toxicity of constructs in _E. coli_ for salmocins with potential nuclease activity, an intron (from _Ricinus communis cat 1_ gene) was introduced. The coding sequences of

immunity proteins corresponding to nuclease-active salmocins were cloned into potato virus X (PVX)-based assembled viral vectors for co-expression (Supplementary Table 1, Supplementary Fig. 

3), as described10. Salmocins can be expressed at very high levels in plants such as _Nicotiana benthamiana_ (Fig. 1a,b) as well as in edible plant hosts such as spinach (Fig. 1d). Acidic

extraction resulted in efficient recovery of soluble salmocins from plant material with the concomitant elimination of native plant proteins (Supplementary Fig. 4). The recombinant protein

yields in _N. benthamiana_ upon soluble extraction with neutral buffer ranged from 18–37% of total soluble protein (TSP) or 1.2–1.7 g salmocin/kg fresh weight of leaf biomass, depending on

the particular protein (Table 2, Fig. 1b). PRODUCTION OF SALMOCINS IN STABLE TRANSGENIC HOSTS Stable transgenic _Nicotiana benthamiana_ plants containing the genomic insertion of TMV-based

viral vector double-inducible with ethanol for SalE1b expression (Supplementary Fig. 5) exhibited normal growth and development, and selected transgenic lines accumulated salmocins upon

induction with ethanol to the expected levels (Fig. 1c). SALMOCINS SALE1A AND SALE1B ARE THE MOST BROADLY AND HIGHLY ACTIVE SALMOCINS FOR CONTROL OF _S. ENTERICA_ To determine the salmocin

antimicrobial activity spectrum, 109 strains representing 105 _S. enterica_ ssp. _enterica_ serotypes were selected and screened (Supplementary Table 2). The screen included one strain each

of all serotypes (except serotypes Typhi and I4,5:12:r:-) that are documented at the U.S. Centers for Disease Control and Prevention (CDC)

(https://www.cdc.gov/nationalsurveillance/pdfs/salmonella-annual-report-2013-508c.pdf) as having caused at least 100 incidences of human _Salmonella_ infection from 2003–2012, two strains of

serotypes Typhimurium, Enteritidis and Javiana and 6 serotypes causing less than 100 incidences or not reported to CDC. In order to estimate the breadth of the activity spectrum, all

strains were tested at least once and 36 or 35 strains were subsequently re-screened in triplicate experiments with salmocins and colicins, respectively (Fig. 2). The broadest antimicrobial

activity spectrum was identified for salmocins SalE1a and SalE1b, which showed positive antibacterial activity against 100% and 99% of all strains evaluated, respectively. Significant

breadth of activity was also observed for salmocins SalE2 (94%), SalE3 (70%) and SalE7 (95%) as reflected by their activity on the subset of 36 strains represented in Fig. 2e. The five

salmocins analysed were divided into four groups based on their ability to control major pathogenic _Salmonella_ strains. Salmocins SalE1a and SalE1b were universally active, each being able

to kill all tested pathovars and showing the highest average activity of higher than 105 AU/µg recombinant protein on all tested strains (Fig. 2a) and in most cases higher than 103 AU/µg

protein against individual strains (Fig. 3, Supplementary Fig. 9). The remaining salmocins fell into two groups, with salmocins SalE2 and SalE7 in one group having a 100-fold lower average

activity (<105 AU/µg protein, Fig. 2a, Supplementary Figs 6, 8), and SalE3 in another group showing substantially lower average activity (102 AU/µg, Fig. 2a, Supplementary Fig. 7). In

contrast to the high potencies of salmocins in inhibiting enteropathogenic _S. enterica_ strains, the specific activities of colicins Ia, Ib, M, 5, 10 and S4 (Supplementary Table 3) were 2–4

orders of magnitude lower (2–3 logs AU/µg, Fig. 2b), although most of the 109 strains were inhibited by colicins Ia (92%) and Ib (90%) and about one third of strains by colicins S4 (45%), 5

(25%), 10 (29%) and M (34%), as also reflected in the susceptibility pattern of the subset of 35 strains (Fig. 2f). In general, salmocins demonstrated higher and broader activity against

_Salmonella_ than _E. coli_ colicins. Conversely, salmocins showed low (below 102 AU/µg) inter-specific and narrrow activity against _E. coli_ STEC (Supplementary Table 4) strains (Fig. 

2c,g, Supplementary Fig. 10). SALMOCIN SALE1A CONTROLS _SALMONELLA_ ON CONTAMINATED CHICKEN MEAT MATRICES The bactericidal efficacy of plant-produced individual salmocin SalE1a as well as

salmocin blends for control of _Salmonella_-contaminated meat surfaces was modeled in a simulation study. Efficacy of salmocin treatment was assessed for the extent of reduction in the

pathogenic bacterial population level on salmocin-treated (individual SalE1a at an application rate of 3 mg/kg meat and salmocin blend consisting of SalE1a + SalE1b + SalE2 + SalE7 applied

at 3 + 1 + 1 + 1 mg/kg meat, respectively), both in relation to plant extract control-treated, meat samples and statistically significant net reductions in viable counts of 2–3 logs CFU/g

meat at all timepoints analysed were found (Fig. 4). The highest level of reduction of bacterial populations was observed for the 4-salmocin blend (concentration of 3 + 1 + 1 + 1 mg/kg meat)

with up to 3.39 mean log reduction vs. carrier treatment upon 48 h of storage, which corresponds to a 99.6 mean percent reduction of bacteria. A single salmocin, SalE1a (applied at 3 mg/kg

meat), was able to control _Salmonella_ contamination on meat with similar efficacy to the blend of four salmocins applied at double the concentration (6 mg/kg meat total salmocin). Even a

treatment with salmocins at very low dose (total salmocin 0.6 mg/kg meat; 0.3 + 0.1 + 0.1 + 0.1 mg/kg meat for a blend of SalE1a + SalE1b + SalE2 + SalE7) produced statistically significant

reductions of bacterial populations of about 1 log CFU for up to 48 h of storage. Upon initial reduction of bacterial contamination, re-growth of viable bacteria was observed after 72 h,

indicating that salmocins act quickly but have no prolonged technical effect on food. PLANT-PRODUCED SALMOCINS ARE DIGESTED BY GASTRIC AND INTESTINAL PROTEASES Although salmocins applied to

raw food will be degraded during cooking, it is important to show that ingested salmocins not degraded by food processing will be degraded in the human gastrointestinal tract to prevent

influencing the indigenous GI tract microflora, as well as to avoid the potential development of allergic reactions. A database search of all salmocin amino acid sequences

(http://www.Allergenonline.org) revealed distal relationships but no exact matches to known allergens. To verify the low risk of allergenicity and residual activity upon ingestion, the

digestibility of salmocin proteins was analysed. A complete loss or negligible residual antimicrobial activity was observed for SalE1b and SalE7 or SalE1a, respectively (Supplementary Fig. 

11a). For SalE1b and SalE7, the full-length proteins disappeared upon 1 h incubation at gastric conditions. Full-length SalE1a protein was found in negligible amount after only 20 min of

gastric digestion and was completely degraded upon additional incubation at duodenal conditions for 5 min. (Supplementary Fig. 11b). RECOMBINANT SALMOCINS ARE CORRECTLY EXPRESSED BY PLANTS

The primary structure including post-translational modifications of the plant-expressed recombinant salmocins contained in plant TSP extracts was analysed by mass spectrometry-based

sequencing. Search results of each MS/MS dataset from proteolytic peptides of salmocins against the UniProt/SwissProt database confirmed the identity of each of the analysed salmocins

(Supplementary Table 5). The integrity of purified salmocins SalE1a, SalE1b and SalE7 was further analysed by MS-based sequencing of protein termini using ISD and molecular mass

determination methods, which confirmed that all salmocin proteins were intact upon plant expression. Post-translational modifications observed were restricted to cleavage of N-terminal

methionine in case of SalE2, SalE7, SalE1a and SalE1b and N-terminal acetylation for SalE7 and SalE1a (Supplementary Table 5). DISCUSSION Perhaps the most significant outcome of our study is

proof that _Salmonella_ harbors genes for its own active species-specific colicin-like bacteriocins, much like other Gram-negative bacteria including _Escherichia coli_10 and _Pseudomonas

aeruginosa_18,19,20, in addition to exploiting _E. coli_ colicins. The two most interesting salmocins SalE1a and SalE1b are 69% identical at the amino acid level with most differences found

in the protein region responsible for receptor binding. It is expected that both proteins use the same translocation machinery but different receptors for entry into susceptible bacterial

cells. This is supported by the finding that one of the _Salmonella_ strains (serotype Kintambo) was killed by SalE1a but not by SalE1b, and that for SalE1a no activity was detected against

any of the _E. coli_ strains tested whereas SalE1b showed the highest antibacterial effect of all the salmocins tested against _E. coli_. By salmocin SalE3, only about 60% of tested strains

were inhibited at low average activity which could be explained by SalE3 having the highest identity to Group A colicins of the protein region responsible for translocation. The other

significant conclusion of this study is that plant-expressed salmocins are excellent candidate biocontrol agents for pathogenic _Salmonella_. To our knowledge, this work is the first study

devoted to the novel recombinant bacteriocins, salmocins, and has demonstrated that salmocins, applied singly or as mixtures, efficiently control all major pathogenic serotypes of

_Salmonella enterica_ ssp. _enterica_ not only _in vitro_ but also in simulation studies involving contaminated meat matrices. This work clearly demonstrates the specificity, potency, and

efficacy of salmocins against major enteropathogenic serotypes of _Salmonella enterica_ ssp. _enterica_. In contrast to the seven major _E. coli_ pathotypes known as the “Big 7” classified

by the Food Safety and Inspection Service (FSIS) branch of the U.S. Department of Agriculture (USDA) as adulterants on meat based on historical analysis of _E. coli_ food poisonings

(https://www.gpo.gov/fdsys/pkg/FR-2011-09-20/pdf/2011-24043.pdf), a list of major foodborne _Salmonella_ strains requiring control has not been fully defined by regulatory agencies,

primarily due to higher diversity of the pathovars responsible for the outbreaks. Lacking further guidance, we included in our tests 99 of 101 _S. enterica_ ssp. _enterica_ serotypes known

to have caused at least 100 incidences of human _Salmonella_ infections reported to CDC during 2003–2012

(https://www.cdc.gov/nationalsurveillance/pdfs/salmonella-annual-report-2013-508c.pdf). This number is 15 times higher than the number of _E. coli_ pathovars defined by FDA (seven). Our

finding that two salmocins, SalE1a and SalE1b, each possessed broad antimicrobial activity against all 99 major pathogenic _Salmonella_ strains tested as well as remarkably high activity

(average > 106 AU/µg), is unexpected. For comparison, colicins, which are salmocin analogues produced by _E. coli_ cells, exhibit much narrower selectivity against seven _E. coli_

pathovars, and mixtures of three to five colicins had to be used to efficiently inhibit strains of all “Big 7” _E. coli_ STEC serotypes. Colicins also demonstrated much lower average

activity against _E. coli_ “Big 7” STEC strains (average < 103 AU/µg), although higher activity has been observed on a strain of serotype O104:H4 (>105 AU/µg)10 that caused major

outbreaks in 2011 in Europe, and a common laboratory strain _E. coli_ DH10B (>105 AU/µg). Similar limited intra-specific activity and efficacy have been observed also with _Pseudomonas_

pyocins17. Our analysis of cross-specific activity of salmocins and colicins on _E. coli_ and _Salmonella_, respectively, confirmed the expected low activity against bacteria of different

genera. In particular, the activity of salmocins against “Big 7” STEC strains was negligible (less than 102 AU/µg) although some salmocins (such as SalE2, SalE7 and SalE1b, but not SalE1a)

were found active against _E. coli_ O104:H4 (103AU/µg) and laboratory strain DH10B (105AU/µg) (data not shown). Similarly, the activity of colicins on _Salmonella_ pathovars was found to be

low, with colicins Ia and Ib being active on over 80% of strains, but with average activity of only colIa being higher than 3 × 103 AU/µg (or three to four orders of magnitude less than

salmocins SalE1a and SalE1b). The practical conclusion of this study is that, to combat pathotypes of both genera simultaneously, or with a single intervention, mixtures of colicins and

salmocins will be required. The salmocins SalE1a and SalE1b have been tested at different concentrations and used to simulate the effect of salmocin treatment on poultry meat matrices spiked

with _Salmonella_ and stored at 10 °C, providing additional evidence for the practical importance of these two salmocins as intervention agents for food safety. The issue of performance at

low temperature is important because much of the food is processed and distributed using cold chains, and any new candidate antimicrobial needs to be active in applications that are

compatible with the existing industrial infrastructure and process parameters. Our data demonstrate that salmocins can be expressed at very high levels in the plant _Nicotiana benthamiana_,

a standard manufacturing host for multiple biopharmaceuticals currently undergoing clinical trials21,22,23,24,25, as well as in edible plant hosts such as spinach, and that the antibacterial

proteins expressed in either host are fully active. The expression levels in most cases reached 20–37% of total soluble protein or 1.2–1.7 g product/kg of fresh leaf biomass without process

optimization. These findings show that salmocins are not toxic to plants, and suggest that future optimization of industrial procedures for transfection or induction in transgenic hosts and

downstream recovery could be developed that are both economical and scalable to meet market demand. In sharp contrast, attempts to express bacteriocin proteins in bacterial production

systems usually met with generalized host toxicity even in species other than the bacteriocins’ homologous bacterial hosts (e.g.26,27). Thus, plants are excellent hosts for manufacturing not

only phage endolysins28,29, but also multiple bacteriocins including colicins10, pyocins20, and now also salmocins. Salmocins are natural non-antibiotic antibacterial proteins that offer

multiple potential applications including treatment of food to eliminate pathogenic _Salmonella_, as well as use as human and animal therapeutic alternatives to antibiotics. The use of

salmocins as food additives or food processing aids is especially attractive because of the magnitude of current food safety issues worldwide, and because these product candidates can be

approved relatively quickly using the GRAS regulatory pathway in the USA. METHODS BACTERIAL STRAINS AND GROWTH CONDITIONS _Escherichia coli_ DH10B, _Salmonella enterica_ ssp. _enterica_

(Supplementary Table 2) and STEC (Supplementary Table 4) cells were cultivated at 37 °C in LB medium (lysogeny broth30). _Agrobacterium tumefaciens_ ICF32031 cells were cultivated at 28 °C

in LBS medium (modified LB medium containing 1% soya peptone (Duchefa)). PLASMID CONSTRUCTS The methods used to construct and apply TMV- and PVX- based magnICON® vectors and TMV-based

vectors for EtOH-inducible expression (Supplementary Figs 3,5) were as described previously10. Coding sequences of genes of interest (salmocins (Table 1) and salmocin immunity proteins

(Supplementary Table 1)) were codon-optimized for _Nicotiana benthamiana_, synthesized by Thermo Fisher Scientific and were cloned into the BsaI sites of the respective destination vectors.

To avoid toxicity to bacteria, for some salmocin sequences an intron of _Ricinus communis_ _cat1_ gene was introduced for gene synthesis (Supplementary Fig. 3), correct intron splicing was

verified as described10. PLANT MATERIAL AND INOCULATIONS _Nicotiana benthamiana_ and _Spinacia oleracea_ cv. Frühes Riesenblatt plants were grown in the greenhouse (day and night

temperatures of 19–23 °C and 17–20 °C, respectively, with 12 h light and 35–70% humidity). Six-week-old plants were used for inoculations. Plant transfection was done as described10. STABLE

PLANT TRANSFORMATION AND REGENERATION AND ETHANOL-INDUCTION OF TRANSGENE EXPRESSION _N. benthamiana_ was transformed by _Agrobacterium_-mediated leaf disk transformation using vectors for

EtOH-inducible transgene expression and induction of detached leaves of T0 generation transgenic plants for salmocin expression was done as described10. PROTEIN ANALYSIS Plant leaf material

was ground in liquid nitrogen and protein extracts were either prepared with 5 vol. 2× Laemmli buffer (crude extracts) or different buffers, e.g. 50 mM HEPES pH 7.0, 10 mM K acetate, 5 mM Mg

acetate, 10% (v/v) glycerol, 0.05% (v/v) Tween-20, 300 mM NaCl (total soluble protein (TSP) extracts). The protein concentration of TSP extracts was determined by Bradford or BCA assay

using Bio-Rad Protein Assay (Bio-Rad Laboratories) or Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific) and BSA (Sigma-Aldrich) as a standard. For analysis by SDS-PAGE and

Coomassie-staining using PageBlue™ Protein Staining Solution (Thermo Fisher Scientific), protein extracts were denatured at 95 °C for 5 min before loading. The estimation of the percentage

of recombinant colicins of TSP was done by comparison of TSP extracts to known amounts of BSA standard (Sigma-Aldrich) on Coomassie-stained SDS-PAA gels. SALMOCIN PURIFICATION Plant TSP

extracts were prepared by supplementation of leaf material ground in liquid nitrogen with 5 vol. pre-chilled extraction buffers as 20 mM citric acid pH 4, 20 mM NaH2PO4, 30 mM NaCl, 0.05%

Tween-80 for SalE1a and SalE7, or 20 mM citric acid pH 5.5, 20 mM NaH2PO4, 30 mM NaCl, 0.05% Tween-80 for SalE1b. Homogenates were incubated for 15 min on ice. Extracts were clarified by

centrifugation for 15 min at 3515 × g and filtration using Miracloth. For SalE7 and SalE1a, clarified extracts were supplemented with 10 mg/ml diatomaceous earth. All extracts were incubated

for 30 min at room temperature with constant agitation and were clarified again by centrifugation for 15 min at 3515 × g and filtered through filter discs of 8–12 µm pore size before

loading for column-purification by cation exchange chromatography (CIEX) using SP-SepharoseFF for SalE1a and SalE1b and by CIEX using CM-Sepharose for SalE7. SalE1a was step-eluted with 20 

mM citric acid pH 4, 20 mM Na2HPO4, 1 M NaCl, 0.05% Tween-80 upon column wash with extraction and 45% elution buffer. Buffer exchange of SalE1a eluate was performed using U-tube

concentrators (Sartorius) at molecular cut off of 10 kDa with 20 mM Na2HPO4, 10 mM citric acid, 50 mM NaCl, pH 6. SalE1b was step-eluted with 10 mM citric acid pH 7.5, 20 mM Na2HPO4, 100 mM

NaCl upon column-wash with extraction buffer and 8% elution buffer. SalE7 was step-eluted using 50 mM Na2HPO4, 10 mM citric acid, 50 mM NaCl upon column-wash with extraction buffer and 50%

elution buffer. Eluted fractions of SalE1b and SalE7 or of SalE1a upon buffer exchange were snap-frozen in liquid nitrogen and freeze-dried. SALMOCIN ANTIMICROBIAL ACTIVITY DETERMINATIONS

Semi-quantitative determination of salmocin antimicrobial activity was done by spot-on-lawn/radial diffusion assay on serial dilutions of plant TSP extracts containing salmocins as employed

in Schulz _et al_.10. Salmocin specific antimicrobial activity was calculated in arbitrary activity units (AU) per µg recombinant protein using the reciprocal of the highest dilution with

visible growth reduction effect on bacterial cells and the recombinant protein content of solutions analysed. REDUCTION OF BACTERIAL POPULATIONS ON FOOD Chicken breast fillet was purchased

from a local supermarket. Nalidixic acid resistant mutants of strains of _S. enterica_ ssp. _enterica_ serovars Enteritidis (strain ATCC®13076™*), Typhimurium (strain ATCC®14028™*), Newport

(strain ATCC®6962™*), Javiana (strain ATCC®10721™*), Heidelberg (strain ATCC®8326™*), Infantis (strain ATCC®BAA-1675™*) and Muenchen (strain ATCC®8388™*) were individually grown in LB medium

supplemented with 25 µg/ml nalidixic acid to stationary phase, diluted with fresh LB and grown to exponential phase. For contamination of poultry, bacterial cultures were diluted with LB

medium to OD600 = 0.001 (~2 × 105 cfu/ml) and mixed 1:1:1:1:1:1:1. A pool of chicken breast fillets cut into pieces of about 20 g weight was inoculated with 1 ml of a mixture of 7 _S.

enterica_ strains at ~2 × 105 CFU/ml density per 100 g of meat at room temperature resulting in an initial contamination level of meat matrices of about 3 log CFU/g of a 7-serotype mixture

of pathogenic _S. enterica_; attachment of bacteria to meat surfaces was allowed for 30 min at room temperature. Subsequently, chicken breast trims were treated by spraying (10 ml/kg) with

either plant extract control (TSP extract of WT _N. benthamiana_ plant material with no salmocins, prepared with 50 mM HEPES pH 7.0, 10 mM K acetate, 5 mM Mg acetate, 10% (v/v) glycerol,

0.05% (v/v) Tween-20, 300 mM NaCl), or salmocin solutions (either individual or mixtures of TSP extracts of _N. benthamiana_ plant material expressing salmocins SalE1a, SalE1b, SalE2 and

SalE7 prepared with the same buffer as the plant extract control) at concentrations of 3 mg/kg SalE1a, or 3 mg/kg SalE1a, 1 mg/kg SalE1b, 1 mg/kg SalE2, 1 mg/kg SalE7 or 0.3 mg/kg SalE1a,

0.1 mg/kg SalE1b, 0.1 mg/kg SalE2, and 0.1 mg/kg SalE7. Treated meat trims were further incubated at room temperature for 30 min. Aliquots of meat trims corresponding to ~40 g were packed

into BagFilter®400 P sterile bags (Interscience) and stored for 1 h, 1 d and 3 d at 10 °C, which represents realistic industrial meat processing conditions that are permissive but suboptimal

for bacterial growth. In total, meat samples were incubated at room temperature for 1.5 h during salmocin treatment before they were sealed and stored at 10 °C. For analysis of bacterial

populations, poultry aliquots were homogenized with 4 vol. peptone water using Bag Mixer®400CC® homogenizer (settings: gap 0, time 30 s, speed 4; Interscience) and colony forming units (CFU)

of _S. enterica_ were enumerated on XLD medium (Sifin Diagnostics) supplemented with 25 µg/ml nalidixic acid upon plating of serial dilutions of microbial suspensions. Samples were analysed

in quadruplicate. STATISTICAL ANALYSIS OF DATA The efficacy of the salmocin treatment in reducing the number of viable pathogenic _Salmonella_ in the experimentally contaminated meat

samples was evaluated by comparing the data obtained with the carrier-treated control samples and salmocin-treated samples by two-tailed unpaired parametric t-test with 6 degrees of freedom

using GraphPad Prism v. 6.01. SIMULATED GASTRO-DUODENAL DIGESTION _IN VITRO_ Gastric (phase I) and duodenal (phase II) digestion _in vitro_ was performed with plant-produced lyophilized

purified SalE7, SalE1a or SalE1b dissolved in Millipore water by the method described previously10. Briefly, individual salmocins were incubated in simulated gastric fluid (SGF) and

simulated intestinal fluid (SIF) containing pepsin or trypsin and chymotrypsin in physiological concentrations, respectively. Incubation with pepsin in SGF for up to 60 min was followed by

incubation with trypsin and chymotrypsin in SIF for up to 3 h. Aliquots of the reactions were evaluated for antimicrobial activity and protein degradation pattern by SDS-PAGE and

Coomassie-staining upon different intervals of incubation. For SDS-PAGE analysis, pre-cast 4-20% Mini-PROTEAN® TGX™ gels (Bio-Rad Laboratories) with loading corresponding to 1.5 µg SalE1a

and SalE7 or 1 µg SalE1b proteins per lane were used. MATRIX-ASSISTED LASER DESORPTION/IONIZATION (MALDI) TIME-OF-FLIGHT (TOF) MASS SPECTROMETRY (MS) For proteolytic digestion, TSP extracts

prepared from plant material expressing salmocins with 5 vol. 20 mM Na citrate, 20 mM NaH2PO4, 30 mM NaCl, pH 5.5 were subjected to SDS-PAGE and Coomassie-stained SDS gel bands containing 5 

µg of protein were excised and destained by consecutive washing with 100 mM NH4HCO3 and 100 mM NH4HCO3 in acetonitrile (ACN)/H2O (50; 50, v/v). Disulfide bonds were reduced with 10 mM DTT

for 45 min at 50 °C followed by alkylation with 10 mg/ml of iodoacetamide for 60 min. Destained and alkylated gel bands were then subjected to proteolytic digestion with different sequencing

grade endoproteinases (Promega, Madison, USA). Protease:protein ratio in the digestion solutions was adjusted to 1:20 (w/w) and digestions were carried out for 12 h at 25 °C (chymotrypsin)

or 37 °C (Asp-N, Glu-C, Lys-C, trypsin). Proteolytic peptides were extracted by consecutive washing with H2O, ACN/H2O/trifluoroacetic acid (50; 45; 5, v/v/v) and ACN, respectively.

Extraction solutions were combined, concentrated in a vacuum centrifuge and resolubilized in H2O/acetic acid (90; 10, v/v). Proteolytic salmocin peptides obtained as described above or

purified intact plant-produced salmocin SalE1a, SalE1b and SalE7 proteins were purified for mass spectrometry by solid-phase extraction using C4 or C18 bonded silica material (ZipTip®,

Millipore, Darmstadt, Germany) and elution solutions were co-crystallized on a MALDI ground steel target with 2,5-dihydroxyacetophenone as well as 2,5-dihydroxybenzoic acid matrix (Bruker

Daltonics, Bremen, Germany). Mass spectra were acquired on a MALDI-TOF/TOF mass spectrometer (Autoflex SpeedTM, Bruker Daltonics, Bremen, Germany) with positive polarity in linear mode for

molecular mass determination and in reflector mode for protein sequencing by In-source decay (ISD) analysis. The matrix crystals were irradiated with a Nd:YAG laser (Smart beam-IITM, Bruker

Daltonics, Bremen, Germany) at an emission wavelength of 355 nm and set to a pulse rate of 1 kHz. MS and MS/MS spectra were recorded with flexControl (version 3.4, Bruker Daltonics, Bremen,

Germany) by accumulation of at least 5000 or 10000 laser shots (per sample spot), respectively. Laser energy was set slightly above the threshold for MS experiments and set to maximum for

MS/MS analyses. Spectra processing was carried out with flexAnalysis (version 3.4, Bruker Daltonics, Bremen, Germany) by applying baseline subtraction with TopHat algorithm, smoothing with

Savitzky-Golay algorithm and peak detection with SNAP algorithm. The mass spectrometer was calibrated using a set of standard peptides and proteins with known masses (Peptide Calibration

Standard II, Protein Calibration Standard I and II, Bruker Daltonics, Bremen, Germany). Determination of the intact molecular mass was based on the mass-to-charge-ratios (_m/z_) of single

and multiple charged molecular ions. Sequencing of protein termini was carried out by ISD analysis. The annotation of ISD fragment spectra was carried using BioTools (version 3.2, Bruker

Daltonics, Bremen, Germany) by in silico generation of _m/z_ values for fragment ions and their comparison with the _m/z_ values of the fragment signals observed within the acquired ISD

spectra. This approach enabled the identification of the terminal amino acid sequences as well as of present modifications. For protein sequencing analysis, only fragment (MS/MS) spectra

were used for the identification of proteolytic peptides and the annotation was carried out with PEAKS Studio (version 7.5, Bioinformatics Solutions Inc., Waterloo, Canada). Identification

of proteins and verification of their amino acid sequences was performed by searching the MS/MS data against the NCBI nr database and the UniProt/SwissProt database to which the sequences of

the salmocins were appended, respectively. Database search was performed with a parent mass error tolerance of 50 ppm and a fragment mass error tolerance of 0.5 Da. The maximum number for

both missed cleavages as well as post-translational modifications for one proteolytic fragment was set to 3. Non-specific cleavage was allowed for both protein termini. DATA AVAILABILITY The

authors declare that all data supporting the findings of this study are available within the article and supplementary information. REFERENCES * Majowicz, S. E. _et al_. The Global Burden

of Nontyphoidal Salmonella Gastroenteritis. _Clin. Infect. Dis._ 50, 882–889 (2010). Article  PubMed  Google Scholar  * Scallan, E. _et al_. Foodborne Illness Acquired in the United

States—Major Pathogens. _Emerg. Infect. Dis._ 17, 7–15 (2011). Article  PubMed  PubMed Central  Google Scholar  * Hoffmann, S. _et al_. Attribution of global foodborne disease to specific

foods: Findings from a World Health Organization structured expert elicitation. _PLoS ONE_ 12, e0183641 (2017). Article  PubMed  PubMed Central  Google Scholar  * Sohaib, M., Anjum, F. M.,

Arshad, M. S. & Rahman, U. U. Postharvest intervention technologies for safety enhancement of meat and meat based products; a critical review. _J. Food Sci. Technol._ 53, 19–30 (2016).

Article  CAS  PubMed  Google Scholar  * Diard, M. _et al_. Antibiotic Treatment Selects for Cooperative Virulence of _Salmonella_ Typhimurium. _Curr. Biol._ 24, 2000–2005 (2014). Article 

CAS  PubMed  Google Scholar  * Wen, S. C., Best, E. & Nourse, C. Non-typhoidal _Salmonella_ infections in children: Review of literature and recommendations for management. _J. Paediatr.

Child Health_ 53, 936–941 (2017). Article  PubMed  Google Scholar  * Cascales, E. _et al_. Colicin Biology. _Microbiol. Mol. Biol. Rev._ 71, 158–229 (2007). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Yang, S.-C., Lin, C.-H., Sung, C. T. & Fang, J.-Y. Antibacterial activities of bacteriocins: application in foods and pharmaceuticals. _Front. Microbiol._ 5,

241 (2014). PubMed  PubMed Central  Google Scholar  * Callaway, T. R. _et al_. Colicin Concentrations Inhibit Growth of _Escherichia coli_ O157:H7 _in Vitro_. _J. Food. Prot._ 67, 2603–2607

(2004). Article  CAS  PubMed  Google Scholar  * Schulz, S. _et al_. Broad and efficient control of major foodborne pathogenic strains of _Escherichia coli_ by mixtures of plant-produced

colicins. _Proc. Natl. Acad. Sci. USA_ 112, E5454–E5460 (2015). Article  CAS  PubMed  PubMed Central  Google Scholar  * Barker, R. M. Colicinogeny in Salmonella typhimurium. _J. Gen.

Microbiol._ 120, 21–26 (1980). CAS  PubMed  Google Scholar  * Vicente, A. C. & de Almeida, D. F. Identification of multiple-resistance (R) and colicinogeny (Col) plasmids in an epidemic

Salmonella agona serotype in Rio de Janeiro. _J. Hyg._ 93, 79–84 (1984). Article  CAS  PubMed  PubMed Central  Google Scholar  * Campos, L. C. & Hofer, E. Colicinogeny in Salmonella

serovars isolated in Brazil. _Mem. Inst. Oswaldo Cruz._ 83, 189–192 (1988). Article  CAS  PubMed  Google Scholar  * Nedialkova, L. P. _et al_. Inflammation Fuels Colicin Ib-Dependent

Competition of _Salmonella_ Serovar Typhimurium and _E. coli_ In _Enterobacterial_ Blooms. Galán J. E., ed. _PLoS Pathog._ 10, e1003844 (2014). Article  PubMed  PubMed Central  Google

Scholar  * Guterman, S. K., Wright, A. & Boyd, D. H. Genes affecting coliphage BF23 and E colicin sensitivity in Salmonella typhimurium. _J. Bacteriol._ 124, 1351–1358 (1975). CAS 

PubMed  PubMed Central  Google Scholar  * Graham, A. C. & Stocker, B. A. D. Genetics of Sensitivity of _Salmonella_ Species to Colicin M and Bacteriophages T5, T1, and ES18. _J.

Bacteriol._ 130, 1214–1223 (1977). CAS  PubMed  PubMed Central  Google Scholar  * Prouty, A. M., Van Velkinburgh, J. C. & Gunn, J. S. _Salmonella enterica_ Serovar Typhimurium Resistance

to Bile: Identification and Characterization of the _tolQRA_ Cluster. _J. Bacteriol._ 184, 1270–1276 (2002). Article  CAS  PubMed  PubMed Central  Google Scholar  * Smith, K. _et al_.

Activity of Pyocin S2 against _Pseudomonas aeruginosa_ Biofilms. _Antimicrob. Agents Chemother._ 56, 1599–1601 (2012). Article  CAS  PubMed  PubMed Central  Google Scholar  * McCaughey, L.

C. _et al_. Lectin-Like Bacteriocins from _Pseudomonas_ spp. Utilise D-Rhamnose Containing Lipopolysaccharide as a Cellular Receptor. _PLoS Pathog._ 10, e1003898 (2014). Article  PubMed 

PubMed Central  Google Scholar  * Paškevičius, Š. _et al_. Plant-expressed pyocins for control of _Pseudomonas aeruginosa_. _PLoS ONE._ 12, e0185782 (2017). Article  PubMed  PubMed Central 

Google Scholar  * Qiu, X. _et al_. Reversion of advanced Ebola virus disease in nonhuman primates with ZMappTM. _Nature._ 514, 47–53 (2014). Article  ADS  CAS  PubMed  PubMed Central  Google

Scholar  * The PREVAIL II Writing Group. A Randomized, Controlled Trial of ZMapp for Ebola Virus Infection. _N. Engl. J. Med._ 375, 1448–1456 (2016). Article  Google Scholar  * Yao, J.,

Weng, Y., Dickey, A. & Wang, K. Y. Plants as Factories for Human Pharmaceuticals: Applications and Challenges. _Int. J. Mol. Sci._ 16, 28549–28565 (2015). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Lomonossoff, G. P. & D’Aoust, M.-A. Plant-produced biopharmaceuticals: A case of technical developments driving clinical deployment. _Science._ 353, 1237–1240

(2016). Article  ADS  CAS  PubMed  Google Scholar  * Tusé, D. _et al_. Clinical Safety and Immunogenicity of Tumor-Targeted, Plant-Made Id-KLH Conjugate Vaccines for Follicular Lymphoma.

_BioMed. Res. Int._ 2015, 648143 (2015). Article  PubMed  PubMed Central  Google Scholar  * Medina, C., Camacho, E. M., Flores, A., Mesa-Pereira, B. & Santero, E. Improved Expression

Systems for Regulated Expression in _Salmonella_ Infecting Eukaryotic Cells. _PLoS ONE._ 6, e23055 (2011). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Diaz, E., Munthali,

M., de Lorenzo, V. & Timmis, K. N. Universal barrier to lateral spread of specific genes among microorganisms. _Mol. Microbiol._ 13, 855–861 (1994). Article  CAS  PubMed  Google Scholar

  * Oey, M., Lohse, M., Scharff, L. B., Kreikemeyer, B. & Bock, R. Plastid production of protein antibiotics against pneumonia via a new strategy for high-level expression of

antimicrobial proteins. _Proc. Natl. Acad. Sci. USA_ 106, 6579–6584 (2009). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Starkevič, U. _et al_. High-yield production of a

functional bacteriophage lysin with antipneumococcal activity using a plant virus-based expression system. _J. Biotechnol._ 200, 10–16 (2015). Article  PubMed  Google Scholar  * Bertani, G.

Studies on lysogenesis. I. The mode of phage liberation by lysogenic Escherichia coli. _J. Bacteriol._ 62, 293–300 (1951). CAS  PubMed  PubMed Central  Google Scholar  * Bendandi, M. _et

al_. Rapid, high-yield production in plants of individualized idiotype vaccines for non-Hodgkin’s lymphoma. _Ann. Oncol._ 21, 2420–2427 (2010). Article  CAS  PubMed  Google Scholar  Download

references ACKNOWLEDGEMENTS The project was partially financed by Investitionsbank Sachsen-Anhalt, Magdeburg, Germany (Grant #1704/00088). We thank Dr. Mirko Buchholz (Fraunhofer Institute

for Cell Therapy and Immunology, Department of Drug Design and Target Validation, Halle, Germany) for MALDI-TOF/TOF-based analytical services, Dr. Antje Breitenstein, Lia Bluhm and Anja

Banke (BioSolutions Halle GmbH, Halle, Germany) for their support in evaluating salmocins’ antimicrobial activity. We also thank Dr. Kristi Smedley (Center for Regulatory Services,

Woodbridge, VA, USA) and Prof. Chad Stahl (University of Maryland, College Park, MD, USA) for valuable advice and guidance on regulatory strategy. AUTHOR INFORMATION Author notes * Tobias

Schneider and Simone Hahn-Löbmann contributed equally to this work. AUTHORS AND AFFILIATIONS * Nomad Bioscience GmbH, Biozentrum Halle, Weinbergweg 22, D-06120, Halle (Saale), Germany Tobias

Schneider, Simone Hahn-Löbmann, Anett Stephan, Steve Schulz, Anatoli Giritch & Yuri Gleba * Fraunhofer Institute for Cell Therapy and Immunology, Department of Drug Design and Target

Validation, Biozentrum Halle, Weinbergweg 22, D-06120, Halle (Saale), Germany Marcel Naumann & Martin Kleinschmidt * DT/Consulting Group, 2695 13th Street, Sacramento, CA, 95818, USA

Daniel Tusé Authors * Tobias Schneider View author publications You can also search for this author inPubMed Google Scholar * Simone Hahn-Löbmann View author publications You can also search

for this author inPubMed Google Scholar * Anett Stephan View author publications You can also search for this author inPubMed Google Scholar * Steve Schulz View author publications You can

also search for this author inPubMed Google Scholar * Anatoli Giritch View author publications You can also search for this author inPubMed Google Scholar * Marcel Naumann View author

publications You can also search for this author inPubMed Google Scholar * Martin Kleinschmidt View author publications You can also search for this author inPubMed Google Scholar * Daniel

Tusé View author publications You can also search for this author inPubMed Google Scholar * Yuri Gleba View author publications You can also search for this author inPubMed Google Scholar

CONTRIBUTIONS Author contributions: S.H.-L., T.S., A.S., A.G., M.N., M.K. and Y.G. designed research; T.S., S.H.-L., A.S., M.N. and St.S. performed research; T.S., S.H-L., A.S., A.G., M.N.,

M. K., D.T. and Y.G. analysed data; and S.H.-L., A.G., D.T. and Y.G. wrote the paper. CORRESPONDING AUTHOR Correspondence to Anatoli Giritch. ETHICS DECLARATIONS COMPETING INTERESTS The

authors declare no competing interests. ADDITIONAL INFORMATION PUBLISHER'S NOTE: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional

affiliations. ELECTRONIC SUPPLEMENTARY MATERIAL SUPPLEMENTARY INFORMATION RIGHTS AND PERMISSIONS 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 license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons

license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license 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 license, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Schneider, T., Hahn-Löbmann, S., Stephan, A. _et al._ Plant-made _Salmonella_

bacteriocins salmocins for control of _Salmonella_ pathovars. _Sci Rep_ 8, 4078 (2018). https://doi.org/10.1038/s41598-018-22465-9 Download citation * Received: 11 January 2018 * Accepted:

22 February 2018 * Published: 06 March 2018 * DOI: https://doi.org/10.1038/s41598-018-22465-9 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content:

Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative