==== Front Sci Rep Sci Rep Scientific Reports 2045-2322 Nature Publishing Group UK London 78622 10.1038/s41598-020-78622-6 Article Phage lysin that specifically eliminates Clostridium botulinum Group I cells Zhang Zhen Lahti Meeri Douillard François P. Korkeala Hannu Lindström Miia miia.lindstrom@helsinki.fi grid.7737.40000 0004 0410 2071Department of Food Hygiene and Environmental Health, Faculty of Veterinary Medicine, University of Helsinki, P. O. Box 66, 00014 Helsinki, Finland 9 12 2020 9 12 2020 2020 10 215714 6 2020 23 11 2020 © The Author(s) 2020Open 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/.Clostridium botulinum poses a serious threat to food safety and public health by producing potent neurotoxin during its vegetative growth and causing life-threatening neuroparalysis, botulism. While high temperature can be utilized to eliminate C. botulinum spores and the neurotoxin, non-thermal elimination of newly germinated C. botulinum cells before onset of toxin production could provide an alternative or additional factor controlling the risk of botulism in some applications. Here we introduce a putative phage lysin that specifically lyses vegetative C. botulinum Group I cells. This lysin, called CBO1751, efficiently kills cells of C. botulinum Group I strains at the concentration of 5 µM, but shows little or no lytic activity against C. botulinum Group II or III or other Firmicutes strains. CBO1751 is active at pH from 6.5 to 10.5. The lytic activity of CBO1751 is tolerant to NaCl (200 mM), but highly susceptible to divalent cations Ca2+ and Mg2+ (50 mM). CBO1751 readily and effectively eliminates C. botulinum during spore germination, an early stage preceding vegetative growth and neurotoxin production. This is the first report of an antimicrobial lysin against C. botulinum, presenting high potential for developing a novel antibotulinal agent for non-thermal applications in food and agricultural industries. Subject terms Food microbiologyAntimicrobialsBacteriophageshttp://dx.doi.org/10.13039/501100002341Academy of Finland299700Walter Ehrström FoundationHelsinki Institute of Life Science (HiLIFE) Fellowshipissue-copyright-statement© The Author(s) 2020 ==== Body Introduction Clostridium botulinum is a Gram-positive, spore-forming anaerobic bacterium that produces botulinum neurotoxin (BoNT). BoNT causes botulism, a potentially fatal flaccid paralysis, in humans and animals1. The pathogenic process is initiated by spore germination and outgrowth of C. botulinum into vegetative, toxinogenic cultures in food or feed, or in the host gastrointestinal tract or in deep wounds2. Thus, botulism can manifest as an intoxication due to preformed BoNT or as a toxicoinfection from spores. C. botulinum is a heterogeneous taxon that comprises several genetically and physiologically distinct species with the common feature of botulinum neurotoxin production. Human botulism is predominantly associated with C. botulinum Group I and II. Sporadic cases of human botulism are also related to BoNT-producing Clostridium baratii (also called Group V) and BoNT-producing Clostridium butyricum (also called Group VI). Animal botulism is mainly associated with C. botulinum Group III. The link between botulism and Clostridium argentinense (also called Group IV) is unclear2. Botulism is a rare but extremely serious condition, with a single case activating a national alert and outbreak investigation, and causing enormous economic loss3. High-temperature processing in the food industry is the most effective method for inactivation of C. botulinum spores and the neurotoxin, but often compromises nutritional and sensory qualities of food products4. There is a growing interest in development of novel methods alternative to thermal process to control the botulism hazard5,6. Bacteriophage lysins are natural hydrolytic enzymes that lyse host bacterial cells by disrupting the peptidoglycan layer7. In Gram-positive bacteria, many phage lysins consist of an amino-terminal catalytic domain, which cleaves one of the major bonds in the peptidoglycan, and a carboxyl-terminal binding domain, which recognizes unique carbohydrate epitopes of cell wall polysaccharides. The synergistic action of both domains confers potent antimicrobial activity selectively against specific host bacteria, which is a distinctive advantage over classical antibiotics and chemical preservatives. Moreover, the biodegradable nature of proteinaceous phage lysins is a considerable benefit for potential applications in food and feed safety and in therapeutics8–10. A number of phage lysins from a variety of host bacteria have been characterized11,12. Commercial applications of phage lysins have also progressed rapidly in developing treatments for microbial infections and solutions to improve food safety13–15. Clostridium perfringens is one of the most extensively studied anaerobic Gram-positive bacteria with characterized phage lysins and identified application potential. To date, ten lysins have been characterized16–24. A series of further studies have significantly improved the application potential of these lysins in prevention of C. perfringens-associated food-borne illness and enteric diseases of domestic animals, or as a diagnostic tool in the detection of C. perfringens25,26. Two phage lysins from Clostridium difficile and one from each of Clostridium tyrobutyricum and Clostridium sporogenes have been identified, all showing specific antimicrobial effects against the host bacterial species27–30. There is no report on phage lysins against C. botulinum. None of the currently known lysins was reported to possess lytic activity against C. botulinum. A phage lysin acting specifically against C. botulinum would allow a novel non-thermal solution to control botulism hazards in foods or feeds. Here we report a putative phage lysin, CBO1751, identified in C. botulinum Group I strain ATCC3502. We demonstrate that recombinantly expressed CBO1751 has specific lytic activity against cells of C. botulinum Group I. We show that CBO1751 can effectively eliminate newly germinated cells, offering potential for early interruption of the pathogenic process of C. botulinum. Results Analysis of the putative phage lysin gene CBO1751 was identified by prophage region analysis of the genome of C. botulinum Group I strain ATCC3502 using the PHASTER web server31. cbo1751, encoding a putative phage lysin, was found in one of two intact prophage regions. The amino acid sequence of CBO1751 is predicted to encode an enzymatically active domain (EAD) of N-acetylmuramoyl-L-alanine amidase (cd02696) on the N-terminal side and a cell wall‐binding domain (CBD) of bacterial Src homology 3 (SH3b) on the C-terminal side (Fig. 1a), suggesting that CBO1751 is an amidase lysin. CBO1751 shared an overall identity in amino acid sequence of 39% with the CS74L lysin of C. sporogenes and < 30% identity in amino acid sequence with other known lysins in Clostridium spp. (Supplementary Fig. S1). The amino acid sequences of EADs on the N-terminal side between the CBO1751 and CS74L lysins showed 58% identity (Fig. 1b), whereas the CBDs on the C-terminal side presented only 11% identity, suggesting that only the EADs of CBO1751 and CS74L are phylogenetically related. We further compared the amidase EADs of CBO1751 and CS74L with PlyPSA, a Listeria endolysin with four characterized catalytic residues responsible for zinc-dependent amidase activity32. Although the EADs of CBO1751 and CS74L shared < 30% amino acid identity to the EAD of PlyPSA, the alignment showed that all the corresponding catalytic residues are conserved in CS74L and CBO1751 (His9, Glu24, His80 and Glu141; Fig. 1b), implying that CBO1751 might have amidase activity similar to CS74L and PlyPSA. The SH3b CBD of CBO1751 showed high sequence divergence to the SH3b of other known lysins, including the well-studied SH3b domain (< 10% pairwise sequence identity) of staphylolytic lysins33–35.Figure 1 Features of CBO1751. (a) Schematic representation of the two domains of CBO1751 and SDS-PAGE analysis of purified CBO1751. Left lane of gel image, Precision Plus Protein Dual Color Standards (Bio-Rad, Hercules, USA). Right lane, eluted his-tagged CBO1751. The original gel is shown in Supplementary Fig. S2. (b) Amino acid sequence alignment of amidase EAD of CBO1751 with Clostridium sporogenes CS74L and Listeria monocytogenes PlyPSA. The conserved catalytic residues are marked in yellow. We then investigated the prevalence of CBO1751 homologs in C. botulinum and in other bacteria. BLAST search found 45 CBO1751 homologs with > 80% amino acid identity in genomes of 152 C. botulinum Group I strains (Supplementary Table S1). BLAST search also identified 16 CBO1751 homologs with > 80% amino acid identity in genomes of 45 C. sporogenes strains. In addition, three CBO1751 homologs with > 80% amino acid identity were found in both C. botulinum Group I and C. sporogenes strains. BLAST analysis did not find any full-length CBO1751 homologs in C. botulinum Group II-IV or in other bacteria. The analysis indicates that CBO1751, a putative novel lysin, is highly conserved in a number of C. botulinum Group I and C. sporogenes strains. Characterization of lytic activity and host specificity of CBO1751 Through cloning of cbo1751 into pET21b vector and expression in Escherichia coli Rosetta 2(DE3) pLysS cells, we obtained the purified CBO1751 protein (yield approximately 15 mg/l of E. coli culture) with molecular weight of 29 kDa using affinity chromatography (Fig. 1a). To assess if CBO1751 exerts lytic activity on C. botulinum, turbidity reduction assay was performed using vegetative cell suspension of C. botulinum ATCC3502 with initial optical density at 600 nm (OD600) adjusted to approximately 1. In the cell suspension supplemented with protein-free dialysis buffer (DB), OD600 remained steady across all time points and was 0.94 for 60 min (Fig. 2a). However, a substantial reduction in OD600 was observed in cell suspensions supplemented with 0.2 to 5 µM of CBO1751, suggesting lytic activity of CBO1751 against C. botulinum ATCC3502 vegetative cells. The lytic activity of CBO1751 showed time and dose-dependency. The minimum OD600 (0.17) was reached at 60 min when treated with 5 µM of CBO1751 (Fig. 2a). To avoid variation in vegetative culture and cell suspension preparations between different batches, we prepared one stock of frozen cells of C. botulinum ATCC3502 and used this as substrate to determine the lytic activity of CBO1751. Unlike fresh vegetative cells, frozen cells with the control DB treatment showed marked drop of OD600, indicating autolysis during turbidity reduction assay (Fig. 2b). Nevertheless, CBO1751 still exhibited time and dose-dependent activity over a range of concentration from 0.02 to 5 µM. The OD600 values after subtraction of the control DB values were used to generate linear regression slopes (ΔOD600/min) of lysis curves. The enzymatic activity of CBO1751 was determined as 6,600 units/mg based on the linear equation of the slopes of the treatments of 0.02 to 0.3 µM CBO1751 (Fig. 2b).Figure 2 Lytic activity of CBO1751. (a) Lytic activity of CBO1751 as analyzed by measuring OD600 of Clostridium botulinum ATCC3502 vegetative cell suspension over 60 min after addition of DB or 0.2 to 5 µM of CBO1751. (b) Determination of lytic activity of CBO1751. Left: Lytic activity analysis by measuring OD600 of C. botulinum ATCC3502 frozen cell suspension over 30 min after addition of DB or 0.02 to 5 µM of CBO1751. Right: Linear regression plot of the slopes (ΔOD600/min) of lysis curves against the four CBO1751 concentrations (0.02, 0.05, 0.1 and 0.3 µM). All OD600 values of (a) and (b) were normalized to an initial OD600 of 1. (c) Lytic activity of CBO1751 against different Firmicutes strains as analyzed by comparing the ratio between OD600 values of 5 µM CBO1751-treated against DB-treated vegetative cell suspensions for 30 and 60 min. (d) Most probable number (MPN) enumeration of C. botulinum ATCC3502, NCTC2916 and 62A cells after treatment with 5 µM of CBO1751 or DB for 60 min. All results are presented as means of three replicates ± standard deviations. *, P < 0.05; **, P < 0.01; ***, P < 0.001. We then tested the specificity of the lytic activity of CBO1751 against vegetative cell suspensions of a range of C. botulinum strains and other Firmicutes strains by comparing the CBO1751/DB-treated OD600 ratio at 30 min and 60 min. Of the seven C. botulinum Group I strains tested, the CBO1751/DB-treated OD600 ratio decreased markedly at both timepoints and varied from 0.18 to 0.41 at 60 min, whereas the values for four C. botulinum Group II strains were above 0.88 at 30 min and decreased only slightly to 0.68 to 0.72 at 60 min (P < 0.001, Fig. 2c). No marked reduction in OD600 was observed in C. botulinum Group III strains. These results suggest that CBO1751 has significant lytic activity against C. botulinum Group I strains, mild lytic activity against C. botulinum Group II strains, and no lytic activity against C. botulinum Group III strains. CBO1751 also showed a mild lytic activity against C. sporogenes NINF45, C. baratii CCUG24033 and C. butyricum BL86/13 with the CBO1751/DB-treated OD600 ratios at 60 min of 0.60, 0.69 and 0.64, respectively. No reduction in OD600 was observed for C. perfringens ATCC13124, C. difficile CD-UN5/11–14, Bacillus cereus ATCC14579 or for Bacillus subtilis 1012M15, suggesting that CBO1751 has no lytic activity against these Firmicutes. We further performed viable cell enumeration of C. botulinum Group I strains ATCC3502, NCTC2916, and 62A cell suspensions treated with CBO1751 using the most-probable-number (MPN) method. Cell suspensions were equally distributed in two aliquots and treated with 5 µM of CBO1751 or the control DB for 60 min. To exclude the contribution of spontaneous cell death during the treatment, cell counts of the three strains after CBO1751 treatment were compared to those after DB treatment. While the cell counts of strains ATCC3502, NCTC2916, and 62A were estimated to be 5.5 × 107, 4.7 × 107 and 6.4 × 107 MPN/ml after the 60-min DB treatment, counts of 3.7 × 105, 1.6 × 104 and 5.5 × 103, respectively, were measured with the CBO1751 treatment (Fig. 2d), indicating about 2 to 4 log units difference in reduction of viable cell counts between CBO1751 treatment and the control DB treatment. The result provides further evidence for the lytic activity of CBO1751 against C. botulinum Group I strains. Effect of pH and salt on the lytic activity of CBO1751 We conducted turbidity reduction assay to evaluate the influence of pH and salt type and concentration on the lytic activity of CBO1751 using frozen cells of C. botulinum ATCC3502. To exclude the effect of autolysis during the assays, the lysis curves with 5 µM CBO1751 treatment were presented after subtraction of the corresponding OD600 values of DB treatment under the same reaction conditions. CBO1751 at 5 µM caused a marked reduction of OD600 at pH from 7.5 to 10.5. The slopes of the lysis curves were similar at pH from 8.5 to 10.5 and were slightly deeper than that at pH 7.5, suggesting that an alkaline pH is more favorable for the lytic activity of CBO1751 than a neutral pH. The CBO1751 treatment caused only moderate reduction in OD600 at pH 6.5 and no reduction at pH 4.5, 5.5, and 11.5 (Fig. 3a). The results suggest that CBO1751 is active over a pH range of 6.5 to 10.5, with optimal activity at alkaline pH. We also evaluated the effect of salt ions on the enzymatic activity of CBO1751. Due to the addition of 5 µl of CBO1751 (dialyzed against DB buffer comprising 500 mM NaCl) to the final reaction mixture (200 µl), the lowest NaCl concentration tested was 12.5 mM. The CBO1751 treatment only caused a minor OD600 reduction in the presence of 400 mM NaCl (Fig. 3b). By contrast, similar lysis curves were observed with 12.5 and 100 mM NaCl. Due to cell autolysis triggered by the high concentration of NaCl, lysis curve at 200 mM NaCl flattened after 5 µM CBO1751 treatment for 5 min, but kept a deep slope during the first 4 min of CBO1751 treatment, which is comparable to that observed with 12.5 or 100 mM NaCl. These results suggest that NaCl up to 200 mM (1.2%) does not strongly affect the lytic activity of CBO1751. The lytic activity of CBO1751 was sensitive to the presence of divalent salts CaCl2 and MgSO4. CBO1751 treatment showed only a minor reduction of OD600 in the presence of 50 mM CaCl2 and MgSO4, and even less reduction with 100 and 200 mM CaCl2 and MgSO4 (Fig. 3c,d). The results suggest that a small amount of divalent cations significantly attenuates the lytic activity of CBO1751.Figure 3 Effects of pH, NaCl, CaCl2 and MgSO4 on the lytic activity of CBO1751. Lytic activity analysis by measuring OD600 of Clostridium botulinum ATCC3502 frozen cells after addition of DB or 5 µM CBO1751 at pH from 4.5 to 11.5 (a) or in the presence of 12.5–400 mM NaCl (b), 0–200 mM CaCl2 (c) or 0–200 mM MgSO4 (d). All OD600 values were normalized to an initial OD600 of 1 and adjusted by subtraction of the corresponding DB control values. Results are indicated as means of three replicates ± standard deviations. Elimination of newly germinated cells by CBO1751 Since lysins kill bacterial cells by degrading the peptidoglycan cell wall, it is conceivable that CBO1751 may lyse newly germinated C. botulinum Group I cells once the peptidoglycan cell wall has been exposed. To confirm this, we prepared a germinating culture in a microfluidic system under anaerobic conditions and observed the effect of CBO1751 using time-lapse microscopy. After incubation of C. botulinum ATCC3502 spores in tryptone-peptone-glucose-yeast extract (TPGY) medium in a microfluidic chamber for 6 h, most of the phase-bright spores germinated and outgrew into thread-shaped vegetative cells that retained the empty shell of the spore coat at one end of the filament. Then, CBO1751 suspended in PBS (concentration optimized to 20 µM) was perfused into the chamber to replace TPGY. After CBO1751 perfusion for about 12 min, an intensive cell lysis was detected in newly germinated cells (Fig. 4 and Supplementary Video S1). Practically all cells were eliminated within 25 min. In contrast, germinated cells maintained their intact cell shape after perfusion of control DB suspended in PBS for 50 min, and no cell lysis was observed (Supplementary Fig. S3). These observations confirmed the lytic activity of CBO1751 against newly germinated C. botulinum Group I cells.Figure 4 CBO1751 eliminates newly germinated cells of Clostridium botulinum ATCC3502 as analyzed by microscopy. First image from left, C. botulinum ATCC3502 spores in microfluidic chamber. Other images, representative time-lapse (min:sec) shots of cell lysing with 20 µM of CBO1751. Scale bars, 2 μm. Square-shaped dots are intrinsic pillars of the microfluidic chamber. Discussion Bacteriophage lysins are the essential peptidoglycan-degrading proteins that are required by double-stranded DNA (dsDNA) phages for the lysis of host bacterial cell wall and release of their viral progenies to extracellular space. With dsDNA phages, lysins are exported by a holin-dependent or -independent mechanism through the bacterial cytoplasmic membrane into the periplasmic space and therefrom access the peptidoglycan substrate in the cell wall36. Therefore, phage lysins are primarily defined as endolysins. However, these enzymes can also degrade the peptidoglycan layer of Gram-positive bacterial cells exogenously and cause immediate cell lysis37. This feature underpins the broad application of phage lysins as natural antimicrobials with technical feasibility. In recent decades, the ever-increasing number of characterized phage lysins exhibits large antimicrobial potential against a list of pathogenic and food spoilage bacteria11,12. Novel approaches alternative to thermal treatment would be a welcome addition in the toolbox for controlling the serious food and feed safety risks caused by the neurotoxigenic C. botulinum. Moreover, the use of natural antimicrobials in food products would reduce the need for chemical preservatives without compromising food safety. Here we demonstrate for the first time an active phage lysin efficiently killing C. botulinum. The lytic activity of CBO1751, originally identified in C. botulinum Group I strain ATCC3502, was verified with 5 µM CBO1751 leading to more than 2 log units reduction in C. botulinum cells in an hour. We also demonstrated efficient elimination of C. botulinum cells newly germinated from spores by addition of CBO1751, suggesting a great potential for CBO1751 as an effective control strategy to interrupt the pathogenic process of C. botulinum at an early stage. pH is a major hurdle used in food preservation38. In general, pH < 4.6 is required to control C. botulinum spore germination and growth39. Food products and feeds with pH above 5 are favorable for C. botulinum growth40. Since some low-acid or non-acid foods can contain a relatively high prevalence of C. botulinum spores41–43, additional hurdles are needed to control the botulism hazard. Similar to phage lysins in other Clostridium spp.27–30, the lytic activity of CBO1751 exhibited a preference for neutral and alkaline pH, which allows CBO1751 to serve as an effective hurdle in these low-acid or non-acid foods to prevent production of neurotoxigenic cultures. Moreover, we observed robust cells lysis by CBO1751 in the presence of up to 200 mM (1.2%) NaCl, suggesting that the lytic activity of CBO1751 tolerates the NaCl levels added in many food products. While divalent metal cations were shown to be essential for the lytic activity of some zinc-dependent amidase lysins44, they exerted inhibitory effects on the lytic activity of some other amidase lysins24,28. Although CBO1751 has conserved zinc-binding active sites in the amidase EAD, we observed marked reduction of its lytic activity in the presence of 50 mM CaCl2 and MgSO4 (~ 0.6%). It is likely that an excess of divalent metal cations, once saturated the catalytic center of the amidase, affect the interaction between the lysin and the bacterial cell wall. The findings indicate that permitted levels of some divalent cation-based food additives might attenuate the effect of CBO1751 in elimination of C. botulinum. The proteinaceous nature of CBO1751 allows its safe application at high concentrations. Further studies are warranted to optimize the antibotulinal potency and synergistic effects of CBO1751 with other preservative hurdles in food products. The host spectrum of CBO1751 is restricted to C. botulinum Group I, and to a lesser extent to C. botulinum Group II, C. sporogenes, C. baratii and C. butyricum. A rather rigid host specificity of CBO1751 was evidenced by a relatively high lytic activity of CBO1751 against all seven tested C. botulinum Group I strains, and a substantially lower activity against C. botulinum Group II, C. sporogenes, C. baratii and C. butyricum strains, and no activity against other Firmicutes. This is consistent with the identification of highly conserved CBO1751 homologs (> 80% amino acid identity) in many C. botulinum Group I genomes but not in any strain of Group II, C. baratii and C. butyricum, and other Firmicutes. C. sporogenes is considered a nontoxigenic variant of C. botulinum Group I and forms a distantly related clade in the Group I phylogeny. Recent studies suggest that the C. sporogenes clade also contains some botulinum neurotoxigenic strains including CDC68016NT, AM370, AM1195, AM553, ATCC51387, Osaka05, Prevot594 and Prevot166245,46. CDC68016NT, AM370 and AM1195 carry putative phage lysins with > 80% amino acid identity to CBO1751 (Fig. 5 and Supplementary Table S1), implying that the three toxigenic C. sporogenes-like strains might be highly susceptible to CBO1751. However, strains AM553, ATCC51387, Osaka05, Prevot594 and Prevot1662 were found to possess putative lysins with only ~ 50% amino acid identity to CBO1751 (Fig. 5). Similarly, ATCC19397 and NCTC2916, two strains in the C. botulinum clade of Group I phylogeny, do not carry any highly conserved CBO1751 homolog, but possess CBO1751 homologs with 50% overall sequence identity and showed efficient lysis upon CBO1751 treatment. Intriguingly, the CBD of these CBO1751 homologs shared > 80% amino acid identity to the CBD of CBO1751 (Fig. 5). The CBD is generally responsible for host specificity of lysins by binding to unique carbohydrate epitopes of cell wall polysaccharides in select bacterial species. Therefore, CBO1751 seems to have a broad host spectrum encompassing both clades in the C. botulinum Group I phylogeny. Further tests with more Group I strains will offer a better understanding of the host specificity of CBO1751. Moreover, to combat the wide heterogeneity of all BoNT-producing species, it is important to characterize lysins that target specifically each phylogenetically distinct group. This will offer not only additional antibotulinal lysins, but also a set of active EADs and CBDs for further domain shuffling and engineering to enhance the lytic activity and broaden the host spectrum, ultimately providing an all-encompassing-tool to control the botulism hazard.Figure 5 Amino acid sequence alignment of CBO1751 and putative lysins of Clostridium botulinum CDC68016NT (LAGL01000000, contig00983, nt 2654–3412), AM553 (WP_061329857), ATCC51387 (LAGD01000000, contig00030, nt 5143–5913), Osaka05 (GAE03455), Prevot594 (AJD31758), Prevot1662 (LAGM01000000, contig00035, nt 5087–5857), ATCC19397 (ABS32796) and NCTC2916 (EDT81697). Sequences encoding putative CBDs are indicated in a green box. As an example of a lysin CBD, the SH3b domain is found in many bacteria and is proposed to recognize peptidoglycan stem peptides and cross-bridges of the bacterial cell wall33–35. In the case of the bacteriolysin lysostaphin, the SH3b domain specifically binds to pentaglycine cross-bridges that are a unique feature of staphylococci33. We found that the CBO1751 CBD was highly conserved (> 80% amino acid identity) in a large number of C. botulinum Group I strains, but not in C. botulinum Group II or III, or other BoNT-producing Clostridium species. Except for some putative amidases in Clostridiaceae, CBO1751 CBD shared < 50% amino acid identity to some hypothetical surface proteins or peptidases in the family Bacillaceae. We speculate that the CBD of CBO1751 might recognize the assumingly conserved peptidoglycan structure in C. botulinum Group I strains, thereby conferring its highest lytic activity against C. botulinum Group I. Further characterization of the CBO1751 CBD substrate will reveal the cell wall features of C. botulinum Group I. Due to specific binding, lysin CBDs show great potential for use as diagnostic tools in bacterial detection13. A variety of CBD-based methods were developed for rapid detection of L. monocytogenes47, Staphylococcus aureus48, B. cereus25,49, C. perfringens25, and C. tyrobutyricum50. The rigid host specificity of CBO1751 might favor the development of a CBO1751 CBD-based diagnostic tool for detection of C. botulinum Group I in food and clinical samples. In summary, our data show that the putative lysin CBO1751 efficiently lyses C. botulinum Group I cells and may provide a useful tool in non-thermal control of C. botulinum. Further studies are required to develop CBO1751-based antibotulinal applications in food products and rapid detection methods for C. botulinum, both approaches providing valuable novel contributions to controlling the botulism hazard. Materials and methods Computational analysis of phage lysin The genome of C. botulinum Group I strain ATCC3502 (Genbank accession number AM412317) was analyzed for prophage regions using the PHASTER web server. Sequence homology analysis was performed using BLASTp and NCBI Conserved Domain Search tool, and sequence alignment was done with ClustalW. Bacterial strains and culture conditions The bacterial strains used in this study are listed in Supplementary Table S2. All clostridia were cultured in anaerobic TPGY medium at 30 °C (C. botulinum Group II strains) or 37 °C in an anaerobic workstation with an atmosphere of 85% N2, 10% CO2, and 5% H2 (MK III; Don Whitley Scientific Ltd., West Yorkshire, UK). B. subtilis 1012M15, B. cereus ATCC14579 and E. coli Rosetta 2(DE3) pLysS cells (Merck Millipore, Darmstadt, Germany) were grown in Luria–Bertani (LB) medium at 37 °C. When appropriate, growth media were supplemented with 100 μg/ml ampicillin and 34 μg/ml chloramphenicol. Expression and purification of recombinant CBO1751 CBO1751 (GenBank accession number CAL83288) was commercially synthesized and cloned in the NheI/SalI restriction sites of pET21b to generate a C-terminal fusion construct with 6 × His tag (GenScript Biotech, Leiden, Netherlands). The vectors were transformed into E. coli Rosetta 2(DE3) pLysS cells (Merck Millipore). The expression of His-tagged CBO1751 was induced with 1 mM IPTG at 30 °C for 8 h. The expressed protein was purified using metal-chelate affinity chromatography with Ni–IDA resin (Merck Millipore) as previously described51. Eluted proteins were examined by SDS-PAGE prior to dialysis (MWCO 8 kDa, Spectrum Labs, New Brunswick, NJ, USA) against 500 ml of dialysis buffer (DB: 500 mM NaCl, 50% glycerol, 20 mM Tris–HCl, pH 7.9) overnight at 4 °C. Protein concentrations were determined using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific Oy, Vantaa, Finland) with bovine serum albumin (Merck Millipore) as a standard. Lytic activity analysis Clostridium strains were cultured to mid-exponential phase (OD600 0.6–0.7) and cells were collected by centrifugation at 5000 × g for 5 min. Pellets were washed gently, resuspended in PBS buffer and adjusted to approximately OD600 of 1. A stock of purified CBO1751 was concentrated by Amicon Ultra-15 centrifugal filter (Merck Millipore) and adjusted by DB to the concentration of 200 µM for lytic activity analysis. Turbidity reduction assay was made by measuring the OD600 of the cell suspensions immediately after adding CBO1751 in a 96-well plate in a volume of 200 µl/well using Multiskan™ Ascent microplate reader (Thermo Fisher Scientific). For each reaction, 195 µl of cell suspension was mixed with 5 µl of DB-diluted CBO1751 to reach final concentrations of 0.2 to 5 µM, or with 5 µl DB as control and incubated at room temperature. The OD600 of the cell suspensions was measured every 5 min over 60 min. Cell counts were determined after treatment for 60 min using the MPN method as described52. To determine the lytic activity, a frozen cell stock was prepared from C. botulinum ATCC3502 grown until OD600 0.6–0.7. Cells were washed gently, resuspended in 20% glycerol and flash frozen in liquid nitrogen. Turbidity reduction assay was conducted as above except that OD600 was measured every 30 s over 30 min. Data analysis was performed as previously described44,53. OD600 values were normalized to an initial OD600 of 1 and adjusted by subtracting the corresponding DB control values [adjusted value = normalized value + (1 − normalized DB control value)]. The deepest slopes (R2 > 0.995) of lysis curves were selected for linear regression analysis to calculate lytic activity. The activity unit was defined as the amount of enzyme resulting in a reduction of 0.001 OD unit/min in the OD600 of C. botulinum ATCC3502 frozen cells. To assess the effect of pH on the CBO1751 lysin activity, frozen C. botulinum ATCC3502 cells were equally suspended in universal buffers with pH ranging from 4.5 to 11.5 and adjusted to OD600 of approximately 1. The universal buffer was prepared by mixing 20 mM each of boric acid and phosphoric acid, followed by titration with sodium hydroxide54. For each reaction, 195 µl of cell suspension was mixed with 5 µl of DB-diluted CBO1751 (final concentration 5 µM) or with 5 µl DB as control. The final pH of each reaction was verified by pH meter. The turbidity reduction assay was carried out at room temperature with the measurement of OD600 every 1 min over 60 min. To evaluate the effect of salts on the CBO1751 activity, frozen C. botulinum ATCC3502 cells were equally distributed in 50 mM Tris–HCl buffer (pH 7.5) with NaCl (final concentration of 12.5, 100, 200 or 400 mM), CaCl2 (final concentration of 0, 50, 100 or 200 mM) or MgSO4 (final concentration of 0, 50, 100 or 200 mM) and adjusted to OD600 of approximately 1. Turbidity reduction assay was performed as above by adding 5 µl of DB-diluted CBO1751 (final concentration 5 µM) or 5 µl DB to cell suspensions. OD600 values were normalized and adjusted as described above. All experiments were conducted with three biological replicates. Student's t-test was used for statistical comparisons. Time-lapse imaging C. botulinum ATCC3502 spores were prepared as described55 and fixed into a microfluidic plate using CellASIC ONIX2 microfluidic system (Merck Millipore) according to the manufacturer's instructions in an anaerobic workstation. TPGY medium was perfused into the microfluidic chamber with the pressure of 13.8 kPa for 6 h to enable sufficient spore germination. After 20 µM of CBO1751 or control DB was perfused into the microfluidic plate to replace TPGY medium, phase-contrast images of newly germinated cells were taken every 20 s over 60 min using a Leica DMi8 inverted microscope with a 100-fold oil‐immersion lens (Leica Microsystems, Wetzlar, Germany). The images were processed using Metamorph (Universal Imaging, Bedford Hills, NY, USA). Supplementary information Supplementary Information 1. Supplementary Information 2. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary Information is available for this paper at 10.1038/s41598-020-78622-6. Acknowledgements We thank Thu Zar Myint and Hanna Korpunen for technical assistance. This work was supported by Helsinki Institute of Life Science (HiLIFE) Fellowship, the Academy of Finland (grant 299700), and the Walter Ehrström Foundation. Author contributions Z.Z. and M.L. designed the study. Z.Z., M.L. and F.P.D. performed experiments. Z.Z., M.L., F.P.D., H.K. and M.L. contributed to the interpretation of data. Z.Z. and M.L. wrote the manuscript. Z.Z., M.L., F.P.D., H.K. and M.L. reviewed the manuscript. M.L. and Z.Z. contributed to funding acquisition. Data availability All data are available from the corresponding authors upon reasonable request. Competing interests The authors declare no competing interests. ==== Refs References 1. Lindström M Korkeala H Laboratory diagnostics of botulism Clin. Microbiol. Rev. 2006 19 298 314 10.1128/CMR.19.2.298-314.2006 16614251 2. Peck MW Biology and genomic analysis of Clostridium botulinum Adv. Microb. Physiol. 2009 55 183 265 10.1016/S0065-2911(09)05503-9 19573697 3. Peck MW Stringer SC Carter AT Clostridium botulinum in the post-genomic era Food Microbiol. 2011 28 183 191 10.1016/j.fm.2010.03.005 21315972 4. Awuah GB Ramaswamy HS Economides A Thermal processing and quality: principles and overview Chem. Eng. Process. 2007 46 584 602 10.1016/j.cep.2006.08.004 5. Lopes RP Mota MJ Gomes AM Delgadillo I Saraiva JA Application of high pressure with homogenization, temperature, carbon dioxide, and cold plasma for the inactivation of Bacterial spores: a review Compr. Rev. Food Sci. Food Saf. 2018 17 532 555 10.1111/1541-4337.12311 6. Maier MB Schweiger T Lenz CA Vogel RF Inactivation of non-proteolytic Clostridium botulinum type E in low-acid foods and phosphate buffer by heat and pressure PLoS ONE 2018 13 e0200102 10.1371/journal.pone.0200102 29969482 7. Loessner MJ Bacteriophage endolysins—current state of research and applications Curr. Opin. Microbiol. 2005 8 480 487 10.1016/j.mib.2005.06.002 15979390 8. Fischetti VA Bacteriophage endolysins: a novel anti-infective to control Gram-positive pathogens Int. J. Med. Microbiol. 2010 300 357 362 10.1016/j.ijmm.2010.04.002 20452280 9. Schmelcher M Donovan DM Loessner MJ Bacteriophage endolysins as novel antimicrobials Future Microbiol. 2012 7 1147 1171 10.2217/fmb.12.97 23030422 10. Cattoir V Felden B Future antibacterial strategies: From basic concepts to clinical challenges J. Infect. Dis. 2019 220 350 360 10.1093/infdis/jiz134 30893436 11. Donovan DM Antimicrobial bacteriophage-derived proteins and therapeutic applications Bacteriophage 2015 5 e1062590 10.1080/21597081.2015.1062590 26442196 12. Gerstmans H Criel B Briers Y Synthetic biology of modular endolysins Biotechnol. Adv. 2018 36 624 640 10.1016/j.biotechadv.2017.12.009 29248682 13. Schmelcher M Loessner MJ Bacteriophage endolysins: applications for food safety Curr. Opin. Biotechnol. 2016 37 76 87 10.1016/j.copbio.2015.10.005 26707470 14. Love MJ Bhandari D Dobson RCJ Billington C Potential for bacteriophage endolysins to supplement or replace antibiotics in food production and clinical care Antibiotics 2018 7 17 10.3390/antibiotics7010017 15. Abdelkader K Gerstmans H Saafan A Dishisha T Briers Y The preclinical and clinical progress of bacteriophages and their lytic enzymes: the parts are easier than the whole Viruses 2019 11 96 10.3390/v11020096 16. Zimmer M Vukov N Scherer S Loessner MJ The murein hydrolase of the bacteriophage φ3626 dual lysis system is active against all tested Clostridium perfringens strains Appl. Environ. Microbiol. 2002 68 5311 5317 10.1128/AEM.68.11.5311-5317.2002 12406719 17. Simmons M Donovan DM Siragusa GR Seal BS Recombinant expression of two bacteriophage proteins that lyse Clostridium perfringens and share identical sequences in the C-terminal cell wall binding domain of the molecules but are dissimilar in their N-terminal active domains J. Agric. Food Chem. 2010 58 10330 10337 10.1021/jf101387v 20825156 18. Schmitz JE Ossiprandi MC Rumah KR Fischetti VA Lytic enzyme discovery through multigenomic sequence analysis in Clostridium perfringens Appl. Microbiol. Biotechnol. 2011 89 1783 1795 10.1007/s00253-010-2982-8 21085950 19. Nariya H Identification and characterization of a putative endolysin encoded by episomal phage phiSM101 of Clostridium perfringens Appl. Microbiol. Biotechnol. 2011 90 1973 1979 10.1007/s00253-011-3253-z 21484204 20. Tillman GE Simmons M Garrish JK Seal BS Expression of a Clostridium perfringens genome-encoded putative N-acetylmuramoyl-l-alanine amidase as a potential antimicrobial to control the bacterium Arch. Microbiol. 2013 195 675 681 10.1007/s00203-013-0916-4 23934074 21. Seal BS Characterization of bacteriophages virulent for Clostridium perfringens and identification of phage lytic enzymes as alternatives to antibiotics for potential control of the bacterium Poult. Sci. 2013 92 526 533 10.3382/ps.2012-02708 23300321 22. Gervasi T Expression and delivery of an endolysin to combat Clostridium perfringens Appl. Microbiol. Biotechnol. 2014 98 2495 2505 10.1007/s00253-013-5128-y 23942878 23. Tamai E X-ray structure of a novel endolysin encoded by episomal phage phiSM 101 of Clostridium perfringens Mol. Microbiol. 2014 92 326 337 10.1111/mmi.12559 24674022 24. Ha E Son B Ryu S Clostridium perfringens virulent bacteriophage CPS2 and its thermostable endolysin lysCPS2 Viruses 2018 10 251 10.3390/v10050251 25. Kretzer JW Use of high-affinity cell wall-binding domains of bacteriophage endolysins for immobilization and separation of bacterial cells Appl. Environ. Microbiol. 2007 73 1992 2000 10.1128/AEM.02402-06 17277212 26. Kazanavičiūtė V Misiūnas A Gleba Y Giritch A Ražanskienė A Plant-expressed bacteriophage lysins control pathogenic strains of Clostridium perfringens Sci. Rep. 2018 8 10589 10.1038/s41598-018-28838-4 30002425 27. Mayer MJ Narbad A Gasson MJ Molecular characterization of a Clostridium difficile bacteriophage and its cloned biologically active endolysin J. Bacteriol. 2008 190 6734 6740 10.1128/JB.00686-08 18708505 28. Wang Q Euler CW Delaune A Fischetti VA Using a novel lysin to help control Clostridium difficile infections Antimicrob. Agents Chemother. 2015 59 7447 7457 10.1128/AAC.01357-15 26392484 29. Mayer MJ Payne J Gasson MJ Narbad A Genomic sequence and characterization of the virulent bacteriophage ΦCTP1 from Clostridium tyrobutyricum and heterologous expression of its endolysin Appl. Environ. Microbiol. 2010 76 5415 5422 10.1128/AEM.00989-10 20581196 30. Mayer MJ Gasson MJ Narbad A Genomic sequence of bacteriophage ATCC 8074–B1 and activity of its endolysin and engineered variants against Clostridium sporogenes Appl. Environ. Microbiol. 2012 78 3685 3692 10.1128/AEM.07884-11 22427494 31. Arndt D PHASTER: a better, faster version of the PHAST phage search tool Nucleic. Acids Res. 2016 44 W16 21 10.1093/nar/gkw387 27141966 32. Korndörfer IP The crystal structure of the bacteriophage PSA endolysin reveals a unique fold responsible for specific recognition of Listeria cell walls J. Mol. Biol. 2006 364 678 689 10.1016/j.jmb.2006.08.069 17010991 33. Mitkowski P Structural bases of peptidoglycan recognition by lysostaphin SH3b domain Sci. Rep. 2019 9 5965 10.1038/s41598-019-42435-z 30979923 34. Gonzalez-Delgado LS Two-site recognition of Staphylococcus aureus peptidoglycan by lysostaphin SH3b Nat. Chem. Biol. 2020 16 24 30 10.1038/s41589-019-0393-4 31686030 35. Lu JZ Fujiwara T Komatsuzawa H Sugai M Sakon J Cell wall-targeting domain of glycylglycine endopeptidase distinguishes among peptidoglycan cross-bridges J. Biol. Chem. 2006 281 549 558 10.1074/jbc.M509691200 16257954 36. Sturino JM Klaenhammer TR Engineered bacteriophage-defence systems in bioprocessing Nat. Rev. Microbiol. 2006 4 395 404 10.1038/nrmicro1393 16715051 37. Fischetti VA Bacteriophage lysins as effective antibacterials Curr. Opin. Microbiol. 2008 11 393 400 10.1016/j.mib.2008.09.012 18824123 38. Leistner L Basic aspects of food preservation by hurdle technology Int. J. Food Microbiol. 2000 55 181 186 10.1016/S0168-1605(00)00161-6 10791741 39. Advisory Committee on the Microbiological Safety of Food. Report on vacuum packing and associated process. https://acmsf.food.gov.uk/sites/default/files/mnt/drupal_data/sources/files/multimedia/pdfs/acmsfvacpackreport.pdf (1992). 40. Austin JW Dodds KL Blanchfield B Farber JM Growth and toxin production by Clostridium botulinum on inoculated fresh-cut packaged vegetables J. Food Prot. 1998 61 324 328 10.4315/0362-028X-61.3.324 9708304 41. Hyytiä E Hielm S Korkeala H Prevalence of Clostridium botulinum type E in Finnish fish and fishery products Epidemiol. Infect. 1998 120 245 250 10.1017/S0950268898008693 9692602 42. Lindström M Myllykoski J Sivelä S Korkeala H Clostridium botulinum in cattle and dairy products Crit. Rev. Food Sci. Nutr. 2010 50 281 304 10.1080/10408390802544405 20301016 43. Pernu N Keto-Timonen R Lindström M Korkeala H High prevalence of Clostridium botulinum in vegetarian sausages Food Microbiol. 2020 91 103512 10.1016/j.fm.2020.103512 32539985 44. Schmelcher M Waldherr F Loessner MJ Listeria bacteriophage peptidoglycan hydrolases feature high thermoresistance and reveal increased activity after divalent metal cation substitution Appl. Microbiol. Biotechnol. 2012 93 633 643 10.1007/s00253-011-3372-6 21720825 45. Weigand MR Implications of genome-based discrimination between Clostridium botulinum group I and Clostridium sporogenes strains for bacterial taxonomy Appl. Environ. Microbiol. 2015 81 5420 5429 10.1128/AEM.01159-15 26048939 46. Williamson CH Comparative genomic analyses reveal broad diversity in botulinum-toxin-producing Clostridia BMC Genom. 2016 17 180 10.1186/s12864-016-2502-z 47. Schmelcher M Rapid multiplex detection and differentiation of Listeria cells by use of fluorescent phage endolysin cell wall binding domains Appl. Environ. Microbiol. 2010 76 5745 5756 10.1128/AEM.00801-10 20622130 48. Yu J Zhang Y Li H Yang H Wei H Sensitive and rapid detection of Staphylococcus aureus in milk via cell binding domain of lysin Biosens. Bioelectron. 2016 77 366 371 10.1016/j.bios.2015.09.058 26433070 49. Kong M Na H Ha NC Ryu S LysPBC2, a novel endolysin harboring a Bacillus cereus spore binding domain Appl. Environ. Microbiol. 2019 85 e02462-18 10.1128/AEM.02462-18 30552194 50. Gómez-Torres N Development of a specific fluorescent phage endolysin for in situ detection of Clostridium species associated with cheese spoilage Microb. Biotechnol. 2018 11 332 345 10.1111/1751-7915.12883 29160025 51. Zhang Z Two-component signal transduction system CBO0787/CBO0786 represses transcription from botulinum neurotoxin promoters in Clostridium botulinum ATCC 3502 PLoS Pathog. 2013 9 e1003252 10.1371/journal.ppat.1003252 23555260 52. Mascher G Mertaoja A Korkeala H Lindström M Neurotoxin synthesis is positively regulated by the sporulation transcription factor Spo0A in Clostridium botulinum type E Environ. Microbiol. 2017 19 4287 4300 10.1111/1462-2920.13892 28809452 53. Briers Y Lavigne R Volckaert G Hertveldt K A standardized approach for accurate quantification of murein hydrolase activity in high-throughput assays J. Biochem. Biophys. Methods 2007 70 531 533 10.1016/j.jbbm.2006.10.009 17169435 54. Yang H Zhang H Wang J Yu J Wei H A novel chimeric lysin with robust antibacterial activity against planktonic and biofilm methicillin-resistant Staphylococcus aureus Sci. Rep. 2017 7 40182 10.1038/srep40182 28067286 55. Brunt J Carter AT Pye HV Peck MW The orphan germinant receptor protein GerXAO (but not GerX3b) is essential for L-alanine induced germination in Clostridium botulinum Group II Sci. Rep. 2018 8 7060 10.1038/s41598-018-25411-x 29728678