LIBRARY Michigan State University w- AIIIIIHIIHIIIIIIIIIIIIIIIIIINWIM L 3 1293 01093 9795 This is to certify that the thesis entitled STUDIES ON A BACTERIOCIN-LIKE ACTIVITY PRODUCED BY PEDIOCOCCUS ACIDILACTICI EFFECTIVE AGAINST GRAN-POSITIVE ORGANISMS presented by PAUL WALTER RUECKERT has been accepted towards fulfillment of the requirements for M.S. 433mm MICROBIOLOGY 5 PUBLIC HEALTH l l ,. ’ fill/"v Major professor Date ST’Q T767 0-7639 OVERDUE FINES ARE 25¢ PER DAY _ PER ITEM Return to book drop to remove this checkout from your record. STUDIES ON A BACTERIOCIN-LIKE ACTIVITY PRODUCED BY PEDIOCOCCUS ACIDILACTICI EFFECTIVE AGAINST GRAM-POSITIVE ORGANISMS By Paul walter Rueckert A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1979 ABSTRACT STUDIES ON A BACTERIOCIN-LIKE ACTIVITY PRODUCED BY PEDIOCOCCUS ACIDILACTICI EFFECTIVE AGAINST GRAMPPOSITIVE ORGANISMS By Paul walter Rueckert Inhibitory activity by cultures of 3, acidilactici, FBB-61, was studied in solid and liquid media using Lactobacillus plantarum, FBB- 12, as the indicator organism. Conditions favoring optimal yield on solid media were determined, and production in liquid media was studied following the development of a standardized bioassay. The activity was found to be bactericidal, non-dialyzable across a semipermeable membrane, and stable to heat (100°C for 60 minutes) and freezing. The inhibitory agent is pronase-sensitive, and is destroyed by treatment with chloroform and ether. Certain of these data justify classifica- tion of the agent(s) responsible for the Pediococcus inhibitory activity as a bacteriocin. A 2X concentration of activity was achieved by the use of dialysis cultures, but all attempts to concentrate and/or purify the bacteriocin by techniques commonly used in the purification of pro- teins were unsuccessful. Several explanations for the apparent failure to achieve significant concentration in broth are discussed. ACKNOWLEDGMENTS I wish to express my sincere gratitude to my major professor, Dr. R. N. Costilow, for his guidance and encouragement throughout the course of this investigation and the preparation of this thesis. I would also like to thank Dr. H. L. Sadoff and Dr. R. R. Brubaker for the use of their laboratory facilities. ii TABLE OF CONTENTS ACKNOWLEDGMENTS . . . . . . . . . . . . . LIST OF TABLES O O O O O O O O O O O 0 LIST OF FIGURES . . . . . . . . . . . . . LIST OF PLATES . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . Cultures and growth conditions . . . . Agar overlay technique . . . . . . Standard assay for bacteriocin activity RESULTS 0 O O O O O O O O O O O O O I O O Demonstration of Activity . . . . . . Solid media . . . . . . . . . . . . Liquid media . . . . . . . . . . . Production . . . . . . . . . . . . . . Solid media . . . . . . . . . . . . Liqu1d media 0 O O O I O O O O O O Cell-Free Preparations . . . . . . . . Physical and Chemical Characteristics . Stability in broth as a function of DialySiS O O O I O O O O O O O O 0 Chemical characteristics Concentration . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . iii Page vii 19 19 19 20 21 21 21 22 25 25 25 32 33 33 36 36 39 41 (TABLE OF CONTENTS) Page LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . . . . 44 iv LIST OF TABLES Table 1. Inhibitory activity of P. acidilactici on TSA as a function of inoculum size. Table 2. Inhibitory activity of P, acidilactici on TSA as a function of incubation time and temperature. Table 3. Effect of dialysis on inhibitory activity in broth. Figure Figure Figure Figure Figure LIST OF FIGURES Mixed culture growth curves of g. acidilactici and L. plantarum in TSB. Inhibitory activity of P, acidilactici cultures as a function of producer cell density in TSB containing glucose at con- centration of 0.25, 0.5, 1.0, and 2.0%. Inhibitory activity of P, acidilactici cultures in TSB con- taining glucose at 0.25, 0.5, 1.0, and 2.02 as a function of incubation time. Loss of inhibitory activity of 2, acidilactici culture supernatant with time at incubation temperatures at 25°C, 30°C, and 37°C. Stability of P, acidilactici inhibitory activity in TSB to heat treatment at 100°C. vi LIST OF PLATES Plate 1. Deferred antagonism of L. plantarum by previous growth of P. acidilactici on TSA. vii INTRODUCTION Lactic acid fermentations of brined cucumbers were studied exten- sively for many years in search of a means by which the frequent parti- cipation of undesirable organisms might be eliminated. Successful fer- mentations may involve up to five species of lactic acid bacteria. These are Lactobacillus plantarum, L. brevis, Pediococcus acidilactici (formerly P, cervisiae (2, 30)), Leuconostoc mesenteroides and Strep- tococcus faecalis (67). The latter two species are usually active only when the growth of P, acidilactici is delayed or absent (66, 67). A study of pure culture fermentations of brined cucumbers (24) dem- onstrated that P. acidilactici inhibited the growth of L. plantarum. Simultaneous inoculation of the two organisms in cucumber brine result- ed in a two day delay in the initiation of growth by the Lactobacillus. Both organisms initiated rapid growth in cucumber brines when inoculat- ed individually. Further investigation revealed the lag before the growth of L, plantarum in mixed cultures may last a week or more, as indicated by acid production (23). Certain strains of P, acidilactici are also inhibitory toward the growth of Staphylococcus aureus in broth cultures (34). However, the antagonism is relieved by the addition of yeast extract, tryptone, bio- tin, niacin or catalase to the medium. These investigators concluded that competition for vital nutrients, and the production of hydrogen peroxide contribute to the phenomenon. Isolates of the Pediococcus genus from various sources were also tested for inhibitory activity against L. plantarum and other microor- ganisms by a seeded agar overlay technique (26). Two strains were found which consistently inhibited L, plantarum and other Pediococcus isolates, but not each other. Many gram-positive organisms were sensi- tive to the activities of the inhibitory strains, including L, 22222? teroides 42, S, faecalis, Micrococcus luteus, S. aureus ATCC 10537 and Bacillus cereus T. None of the gramrnegative bacteria or yeasts tested were suppressed. Addition of catalase to the growth medium had no ef- fect on the inhibitory activity, eliminating hydrogen peroxide as the causative agent. A great number of microbial toxins have been described which are responsible for the suppression of one population by another (1). In addition to hydrogen peroxide, other simple metabOlic products includ- ing hydrogen sulfide, ammonia, nitrite, carbon dioxide, lactate, vola- tile fatty acids and ethanol are inhibitory in nature if they are al- lowed to accumulate to high levels in the surrounding medium (1). These differ from the "classical" antibiotics in that the latter are antagonistic to microbial growth when present at very low concentra- tions (76). Most antibiotics are organic compounds of low molecular weight and are typically the products of minor and secondary metabolic pathways (71). Bacteriolytic or mycolytic enzymes are produced by some micro- organisms, including certain species of Bacillus, Staphylococcus, Flavobacterium, Pseudomonas,‘Myxococcus, Streptomyces and Chalaropsis (1). Bacterial cultures are also lysed by bacteriophage infection, and some strains are lysed by defective bacteriophages; i.e., by phages which are unable to complete a full growth cycle (1). Many investiga- tors prefer to classify the latter entities with a group of substances loosely referred to as the bacteriocins (69). Bacteriocins are highly potent antimicrobial agents which usually have an essential, biologi- cally active protein moiety and a bactericidal mode of action (78). However, exceptions to classical criteria based on the properties of colicins (bacteriocins produced by Escherichia coli) are encountered frequently (see Literature Review), and a precise definition of the term is not available. Previous studies on the antimicrobial activities of P, acidilactici strains FBB-6l and L-7230 have been limited to those by Fleming gt a; (26) and Haines and Harmon (34), already described above. As this an- tagonism.may be at least partially responsible for the rapid develop- ment of P, acidilactici prior to that of L, plantarum in natural fer- mentations of brined cucumbers and Spanish-type green olives (26), it » is of interest to describe the phenomenon more fully. The present in- vestigation was designed to determine the nature of the agent(s) re- sponsible for the inhibitory activity; to develop a reproducible assay for quantitative analysis of cell-free supernatants; to describe the growth conditions which result in optimum yield; and to concentrate the activity for the future development of a purification scheme. As mentioned above, the antimicrobial activity of P, acidilactici has been shown to be caused by something other than hydrogen peroxide (26), a common metabolic product of lactic acid bacteria. Results of the present study indicate that the activity is non-dialyzable across a semipermeable membrane and is bactericidal in nature. Agar cut from inhibitory zones, suspended in broth and spotted on indicator lawns produced no plaques, which demonstrated the absence of self-replicating, infective entities (85). The agent is at least partially degraded by pronase, indicating the presence of a proteinaceous moiety on the mole- cule. These results led this investigator to believe the inhibitory activity of P, acidilactici is due to the release of a bacteriocin(s) into the growth medium. Thus, the Literature Review is confined to studies of bacteriocins, with emphasis on those bacteriocins produced by gram-positive organisms. LITERATURE REVIEW Definition The study of bacteriocins began in 1925 when Gratia published his observations on a highly specific antibiotic produced by Escherichia 321; V and effective against E, ggli_¢ (69). Gratia characterized the "antibiotic" as to its dialysibility, its stability to heat and chloro- form, and found it was precipitated by acetone. With the discovery that this was but one of a group of similar antibiotics produced by the Enterobacteriaceae, Gratia and Fredericq suggested the agents be called "colicines" (69). Other organisms were also found to produce "colicine- 1ike" antibiotics, and Jacob gtual (62) proposed the more general term "bacteriocine" to include all highly specific antibacterial proteins which are effective mainly against strains of the producer species. Today, both terms are spelled without the final "e". Mayr-Hartingjgt a; (62) distinguish three overlapping periods in the history of bacteriocin studies. The first was largely descriptive, beginning with the Gratia publication and continuing until about 1950. Much of the technology for detection and assay were developed during this period. The second period began in the early 1950's and consisted of numerous studies on the genetics of bacteriocinogeny. These studies have shown most bacteriocinogenic factors are plasmid-borne (78), and many of these plasmids also determine the regulation of bacteriocin 5 synthesis, their release from the cells, and the host cell immunity. Certain bacteriocinogenic factors promote their own transfer on conju- gation of the producer organisms with compatible recipient strains (46, 69, 81). A third field of investigation includes studies of the chemi- cal nature of bacteriocins, their biosyntheses, and their modes of ac- tions. Principle lesions usually occur in energy production, macro- molecular synthesis, or membrane transport and permeability (78). How- ever, bacteriocins of gram-positive bacteria may often promote bacteri- olysis (3, ll, 39), sporostasis (54), and spheroplast formation (59). A great deal of knowledge concerning the nature of bacteriocins has been obtained through these studies, and the field has been review- ed often (28, 37, 44, 62, 65, 68, 69). Until recently, the majority of the investigators have emphasized the bacteriocins of gram-negative bacteria, and especially the colicins. Thus, the classical definitions of bacteriocins were originally based on the characteristics of colicins, and generally included up to six criteria (78): a) a narrow inhibitory spectrum of activity centered about the homologous species; b) the pres- ence of an essential, biologically active protein moiety; c) a bacteri- cidal mode of action; d) attachment to specific cell receptors; e) plas- mid-borne genetic determinants of bacteriocin production and of host cell bacteriocin immunity; f) production by lethal biosynthesis (i.e., commitment of the bacterium to produce a bacteriocin will ultimately lead to cell death). However, the continued use of the term bacteriocin as defined above has been seriously questioned in a recent review on the bacterio- cins of gram-positive bacteria (78). Relatively few bacteriocins, es- pecially those produced by gram-positive organisms, are found to meet the classical criteria which define the colicins. Thus, gram-positive bacteriocins often have activity spectra which extend to a number of organisms of different species and genera, the majority of these being gramepositive. A less solid host cell immunity is often associated with gram-positive bacteriocin production. As a result of such dis- crepancies, there is at present no universally accepted definition of the term bacteriocin. Personal preference often determines if a given antibacterial agent is labelled as a bacteriocin, a lytic enzyme, or a defective bacteriophage. A number of substances which have been classified as bacteriocins have not been adequately characterized so as to justifiably place them under any classification at this time. Tagg gt gl_(78) have urged re- straint on the use of the term bacteriocin on incompletely defined an- tagonistic substances until they have at least been shown to have an essential protein moiety and bactericidal activity. Many investigators have already termed inhibitory agents under study as "bacteriocin-like" (7, 36, 44, 51) prior to further characterizations of the agents. Demonstration of Bacteriocin Activity The methods most commonly used to demonstrate bacteriocinogeny are performed on solid media and are generally referred to as the simulta- neous and the deferred antagonism procedures (78). Both tests rely on the inhibition of an indicator organism caused by the bacteriocin re- ‘leased during the growth of a producer strain. Simultaneous antagonism; This method involves the simultaneous growth of a producer strain and an indicator strain on or in a solid medium. The technique most often used is to spot a small volume of producer culture onto the surface of a plate that has been freshly seeded with the indicator organism so as to yield confluent growth upon incubation (80). A zone of inhibition appears around the growth of the bacteriocinogenic culture. The procedure has been modified for faculta- tive anaerobes by inoculating the plate as a stab into the agar rather than as a surface spot (5, 18). Another variation employes the use of wells in freshly seeded agar which are filled with agar containing the producer strain (72). Some investigators have sprayed the indicator culture onto the plate after the producer strain had been inoculated (43, 49). Each of these methods relies upon the early production of the bacteriocin and upon its diffusion through solid media. Deferred antgggnism: This procedure allows for the accumulation of the inhibitor prior to testing. The producer organism is grown on agar medium for a designated period of time and then killed by exposure to chloroform or heat (15). An agar overlay freshly seeded with the indicator strain is poured onto the surface of the plate, and the plate is incubated for a suitable time. Alternatively, the indicator strain may be streaked across the killed cells of the producer organism (62). Deferred antagonism techniques have often been proven to be more sen- sitive than simultaneous antagonism procedures, and allow independent variation of time and conditions of incubation of the producer and in- dicator cultures (78). It is imperative that the investigator deter- mines the stability of the activity against the method of sterilization used. Bacteriocins vary widely in their heat stability (69), and at least one has been found to be chloroform-labile (9). Either method of detection is dependent upon both the production of the bacteriocin and its diffusion through the medium employed (62). The composition of the medium has been shown to directly or indirectly affect the diffusion of colicin (63) and the sensitivity of an indicator strain to a bacteriocin of Streptococcus mutans (70). It has been sug- gested (62) that a wide variety of conditions be applied in a search for bacteriocinogenic properties of microorganisms. Assays for Quantitation of Bacteriocin Activities There are no specific chemical or physical reactions for the assays of bacteriocin activity, which has led to the common use of microbiolog— ical assays for the quantitation of these compounds (62). Jacob 35 al_ (62) suggest that while such assays often have the disadvantages of low precision, lengthy incubation times, and difficulties in comparing yields of different types of bacteriocin, the assays do have a high specificity and a high sensitivity. The titers of bacteriocin assays are usually defined as the reciprocal of the highest dilution to cause inhibition of an indicator organism under standardized conditions, and is expressed as Arbitrary Units (A;U.) per ml of sample. It is neces- sary to consider if non-specific inhibitory activity is contributing to the titer and to eliminate viable producer cells from the sample before performing the assay (62). The latter is commonly accomplished by heat (13, 16, 79, 80); filtration (6, 11); or chloroform (20, 28, 32). The investigator must make certain that the method of sterilization does not reduce the titer either by inactivation of the bacteriocin or by interference of the assay. 10 Critical dilution.method: This simple method is the most commonly used for both qualitative detection and for quantitative assay of a number of bacteriocins (62). Serial dilutions (usually two-fold) are made of the sample and uniform drops of each dilution are transferred to the surface of plates seeded with an appropriate indicator organism. The plates are incubated for a designated time interval and the degree of inhibition due to each drop is examined. The end-point is usually considered to be the highest dilution which resulted in complete inhi- bition, i.e., no visible growth, of the indicator organism. The titer (A.U./ml) is the reciprocal of this dilution. All conditions of the assay must be carefully standardized, including the size of the drop, the indicator inoculum size and the incubation time. The obvious limi- tation of this assay is the subjectivity in determining the end-point, but the simplicity of the procedure usually outweighs the inherent lack of precision (62). The method is highly reproducible if care is taken to maintain the appropriate set of defined conditions (32). Survivor count method: This method assumes that under certain idealized conditions, the reaction of bacteriocin with a large excess of indicator cells will result in a distribution of bacteriocin parti— cles among the cells according to the Poisson Distribution (62). The bacteriocin concentration is expressed in Lethal Units (L.U.), i.e., the minimum amount of bacteriocin which will kill a single sensitive indicator cell, and is derived from the equation: L0 = -N1 x In (NZ/N1) where L0 = L.U. per ml of sample; N1 - viable cells per ml at time of mixing; N2 8 viable cells at end of experiment. An indicator suspension is equilibrated at the temperature of reaction, after which a sample is 11 withdrawn for viable counting; this gives an estimate for N1. A sample containing bacteriocin is added in known proportion to the cell suspen— sion and mixed. After an adsorption period, a sample is again withdrawn to determine viable counts (N2). The time allowed for adsorption and the temperature at which reaction mixtures are incubated have varied with the organism and bacteriocin under study. Reaction conditions have included 10 minutes at 36°C (62); 20 minutes at 37°C (40); 45 min- utes at 36°C, followed by l - 2 hours in the refrigerator (61); 45 min- utes at 37°C (56); and 2 hours at 0°C (56). The assay is reasonably precise and highly sensitive if the counts are made with a sufficient degree of replication (62). Modified survivor count method: This assay substitutes the meas- urement of light-absorbing properties of the indicator suspension for the determination of viable counts in the above bioassay, thus increas— ing its rapidity. The procedure described by Jetten gt §1_(49) is per- formed by mixing serial dilutions of bacteriocin in sterile tubes with tryptcase soy broth and 107 bacteria from an exponential phase culture of the indicator strain. The tubes are incubated at 37°C for 4 hours and the optical density is measured at 600 nm. A control tube contain- ing no bacteriocin is also inoculated, from which the optical density corresponding to 1002 survival is determined. An S-shaped curve results when absorbsnce is plotted arithmetically against the dilution of bac- teriocin. The titer (A.U./ml) is determined from the graph as the reci- procal of the dilution yielding 50% increase in absorbance. A drawback of this assay might be the contribution to the light- absorbancy of the suspension made by the cells killed by the bacteriocin (62). Several variations to the assay have been described which are 12 attempts to eliminate this problem. One involves the use of the redox indicator 2,3,5-triphenyl tetrazolium chloride which is reduced to its red-colored formazan intracellularly by metabolizing cells (73). Anoth- er depends upon the measurement of ultraviolet-absorbing material re- leased from killed cells (52). Production of Bacteriocin Conditions of culture: The ability to produce a bacteriocin is usually genetically stable in bacteriocinogenic strains, but synthesis does not occur all the time or under all conditions (68). It is im- portant to identify the optimum medium and conditions of incubation prior to the development of a purification scheme. Some bacteriocins of gram-negative bacteria (55, 68) and many bac- teriocins of gram-positive organisms (8, 29, 36, 41, 53, 74, 77, 82, 84) are produced at significant levels only on solid media. Increasing the viscosity of liquid media by the addition of agar, dextran, glycerol or starch has been shown in one study to increase the yield of bacteriocin (78). The yield of staphylococcin 1580 may be up to twenty times great— er in a semisolid medium than in the corresponding liquid medium (49). In general, complex media appear to support higher levels of bac- teriocin than do simple media (62). However, large fluctuations in yield have been noted with different batches of these types of media (25, 79). Tagg gt a; (79) recovered streptococcin A from cultures in tryptic soy broth and Difco Todd-Hewitt broth, but not from Oxoid Todd- Hewitt broth, trypticase soy broth, tryptose phosphate broth or brain- heart infusion. Similarly, Rodgers (70) found that some strains of 13 S, mutans produced bacteriocin in Oxoid brain-heart infusion agar but not in BBL brain-heart infusion agar, while the reverse was true for other strains. The primary carbon source and various supplements to the growth media have also been shown to affect bacteriocin production. Glucose inhibits the synthesis of colicin K (60), streptococcin B-74628 (77) and staphylococcin 462 (35), but stimulates the production of strepto- coccin A-FF22 (80). The addition of mannitol enhances the yield of staphylococcin 462 but decreases that of staphylococcin 414 (35). Yeast extract enhances bacteriocin production of certain strains of S, mutans (70) and group B streptococci (77), but represses the synthesis of bacteriocin from B, stearothermgphilus strain NU-lO (85). Production of bacteriocin by Clostridium butyricum NCIB 7423 in a semi-defined medium is directly proportional to the amount of casein hydrolysate added (11). The pH of the growth medium may be critical for optimal production of many bacteriocins. Goebel gt §1_(3l) found it essential for E, 321$ K 235 L O cultures to be maintained at pH 7.0 for maximum recovery of colicin K. Production of streptococcin APFFZZ on Todd-Hewitt agar was found to be best at pH 6.0 to 6.5, and no bacteriocin was detected at pH 7.5 or above (80). Production of staphylococcin 1580 in trypticase soy broth is optimum between pH 6.5 and 8.0 (49). Temperature of incubation is also important for optimum yield. Generally, the temperature optimal for growth of the organism will re- sult in the maximum production of bacteriocin (78). Growth at elevated temperatures often inhibits bacteriocin synthesis completely (14, 77), and may lead to an irreversible loss of activity (14, 78). The latter l4 phenomenon is usally associated with the loss of the bacteriocinogenic factor (78). Aeration of cultures greatly increases the yield of certain staph- ylococcal bacteriocins (13, 49), and no activity was found in anaero- bically grown g. epidermidis cultures (49). Conversely, production of bacteriocin by group A streptococcus strain FF22 was completely suppres- sed if the cultures were shaken (79). Inhibitors and inactivators: Many bacteriocinogenic cultures ex- crete substances which are antagonistic to the bacteriocins they pro- duce. Some mutants of E, 221$ release a specific inhibitor of colicin B into the growth media (33). Release of a specific teichoic acid in the latter stages of the growth cycle of S. faecalis subsp. zympgenes strain X14 results in the diSappearance of bacteriocin activity in the cultures (17). Dialysis of culture supernatants containing staphylo- coccin 1580 greatly increases bacteriocin activity, which has led in- vestigators to believe that a low molecular weight inhibitor may be present in the preparation (49). Protease produced by the bacteriocino- genic strain may inactivate bacteriocins of Serratia marcescens (27), Clostridium botulinum (22) and group A streptococci (79). Some of these proteases are denatured by boiling, thus protecting those bacteriocins which are heat stable (22, 79). Bacteriocin inactivators may be respon- sible for low activities under conditions which are conducive to their production and release into the growth media. Production as a function of the stage of growth: Maximum bacteri- ocin yields may occur at different phases of the growth cycle, and the timing of harvest must then be determined empirically for each organism and set of conditions. Streptocin STH1 production is optimal during 15 exponential phase, but levels drop sharply before the culture enters stationary phase (75). Staphylococcin C55 levels increase throughout the exponential phase, reaching a maximum between 24 and 48 hours of growth, and then gradually decline (l6). Butyricin 7423 is released during late exponential phase, while perfringocin 11105 only appears at the onset of stationary phase (11). Some strains of Cornebacterium diphtheriae release bacteriocin continuously whereas others appear to produce it in bursts (78). Several studies have reported substantial losses of bacteriocin levels on prolonged incubation of cultures, which may be due to enzymatic degradation, appearance of specific inactiva- tors, or readsorption to the producer cells (78). Induction: Bacteriocin yields may often be increased many-fold by induction in a manner analogous to prophage induction. The most commonly employed methods of induction are treatments with ultraviolet irradiation or with mitomycin C (78). Other procedures that have been used include treatment with nitrogendmustard, hydrogen peroxide, N- methyl-N'-nitro-N—nitrosoguanidine and cold shock treatment (62). How- ever, cases are known where attempts to induce bacteriocin production have reduced the yield (13, 58). ‘While a majority of gramrnegative bac- teriocins are reported to be inducible, a significantly smaller propor- tion of gram-positive bacteriocins have been successfully induced by the procedures mentioned above (78). Properties of Bacteriocins Chemical composition: Although bacteriocins are a chemically di- verse group of substances, the presence of an essential protein component 16 provides a unifying property. Some bacteriocins of gram-positive bac- teria appear to be simple proteins (38, 39), while others, including certain staphylococcal (29, 35), clostridial (78), and lactobacillal (19, 83) bacteriocins seem to be quite complex with lipid and carbo- hydrate components in addition to protein. The composition of staphy- lococcin 414 has been likened to that of the staphylococcal cell mem- brane (29). Physical properties: The bacteriocins range in size from molecular weights of 8,000 to complex defective phage particles with molecular weights in excess of 106. Certain preparations of staphylococcin (29) and megacin Cx-337 (21) were found to contain ring-like structures of diameters 1.0 to 6.4 pm similar to membrane vessicles. Many bacterio- cins produced by gram-positiVe organisms are believed to exist in two or more distinct physical forms (11, 22, 29, 35, 49, 75, 77, 83). The different molecular components of some of these bacteriocins appear to exist in equilibrium, with the degree of association and dissociation being influenced by pH and the ionic strength of the preparation (78). Stability: The stability of a bacteriocin preparation is an imr portant factor to consider during the development of a purification scheme. Increased purification often results in decreased stability (22, 57, 77, 79). Bovine serum albumen has been used successfully by some investigators to prevent excessive inactivation of the bacteriocin (29, 64). The stability of bacteriocins to pH extremes is extremely varied, but most of these compounds seem considerably more tolerant of acid than alkaline pH treatments (78). Thus, megacin A-216 and streptococcin ArFF22 are both stable at pH 2 - 7 (45, 79), and staphylococcin 1580 is 17 stable at pH 3.5 - 8.5 (47). Both butyricin 7423 and perfringocin 11105 are stable from pH 2 - 12 (11), and boticin E-SS is stable from pH 1.1 - 9.5 (22). Other bacteriocins have a much narrower range of pH values over which they are stable. Thus, boticin P is stable only at pH 6.5 - 7.5 (54). Heat stability of bacteriocins is more difficult to define, as this property is dependent upon the state of purification, pH ionic strength and the presence or absence of protective molecules (78). The range of thermostability among the bacteriocins is wide. Boticin P is sensitive to a treatment of 60°C for 30 minutes (54), and streptocin STH is inactivated in 10 minutes at this temperature (75). However, 1 closticin A is resistant to a 30 minute treatment at 100°C (42), and lactocin LP27 can withstand 60 minutes at this temperature (83). Staph- ylococcin 1580 is still bioactive after a 15 minute exposure to 120°C (49). Isolation and Purification Isolation: Cells and cell debris of the producer organism are often removed by centrifugation or membrane filtration, though loss of activity sometimes occurs with the latter method (62). However, many bacteriocins are predominantly cell-bound, in which case separation from the cell mass may be made by a variety of physical, chemical, or enzymatic means (78). These include homogenization of the cells, me- chanical disruption, elution with urea and sodium chloride, acid ex- traction, heat treatment and trypsin-lysozyme treatment. Alternatively, the activity may be diffused into a solid medium, which is usually 18 extracted with a suitable eluant or via means of a "freeze-thaw" tech- nique (55). Purification: Where the initial source of bacteriocin is insuf- ficiently concentrated, large volumes of material are required and an initial concentration step is necessary (62). Among the successful methods of concentration are evaporation in a rotary evaporator at 40°C, lyophilization, and ultrafiltration (62). The purification process for non-dialyzable bacteriocins may be aided by the use of diffusates of the nutrient medium (l6, 19, 49). The starting material is then free of all high molecular weight, non-bacterial substances, which simplifies subsequent operations. Following concentration, the impure material is subjected to a suitable combination of purification methods, including fractional pre- cipitation, fractional absorption, column chromatography (ion exchange and/or molecular exclusion), centrifugation, electrophoresis, and iso- electric focusing (62). The choice as to the combination of these methods and the order in which they are employed is largely empircal. A common problem in bacteriocin purification is the loss of activity as purification progresses (78). Thus, the specific activity (units of bacteriocin per milligram protein) is usually monitored at each step, and the procedure is modified where possible where there is excess loss of activity. The purified bacteriocin preparation is tested for homo- geneity using more than one criterion. Common techniques used include gel electrophoresis, ultracentrifugal analysis, and/or immunological methods (62). MATERIALS AND METHODS Cultures and growth conditions: The producer organism used through- out this study was P. acidilactici EBB-61 (formerly P. cerevisiae FBB- 61 (2, 30)). The indicator organism was L, plantarum FEB-12. Both organisms are gram-positive, homofermentative lactic acid bacteria and were isolated from commercial cucumber brines. 'Stock cultures were maintained in stabs in LBS agar (BBL), and were transferred regularly at 3~month intervals. The basal growth medium employed was trypticase soy broth (BBL) buffered with potassium phosphate (pH 7.2) to 0.1 M (TSB). The phos- phate was added aseptically to the medium after autoclaving to prevent the formation of precipitate. Additional glucose was added aseptically after autoclaving to a concentration of 12 unless otherwise indicated. All cultures were incubated at 30°C without shaking. Agar overlay technique: This procedure was used most often to study inhibitory activity on solid media. The effector organism was grown on the surface of trypticase soy agar (TSA) as colonies or in streaks for an appropriate time interval. The plate surfaces were sterilized by a 3~minute exposure to vapors from 0.3 ml of chlorofOrm on filter paper strips (2.0 cm square). The petri plates were then ventilated briefly and overlayed with 4 ml of soft agar medium freshly seeded with L, plantarum (106 cells/ml). After 18 - 24 h of incubation at 30°C, the plates were examined for zones of inhibition in the agar l9 20 overlays. The width of each zone was measured from the edge of the colony to the outermost point of inhibiton (clearing). Standard assay for bacteriocin activity: Bacteriocin activity in broth media was quantitated by use of an assay adapted from that of Reeves (68). Cells were removed from broth cultures of the producer organism by centrifugation (103 x g, 4°C). Two-fold serial dilutions were made of the culture supernatant in fresh TSB. Each tube ( 5 ml volume) was then placed in a boiling water bath for 6 min to destroy any remaining producer cells. After cooling to room temperature, each dilution was inoculated (1%) with late log-phase L, plantarum cells diluted to an O.D. of 0.1. All tubes were incubated at 30°C for 600 16 h, at which time the optical densities were read at 600 nm. The reciprocal of the dilution yielding a 502 increase in optical density over that of the undiluted sample was taken to be the titer expressed as Arbitrary Units (A.U.) per ml. RESULTS Demonstration of Activity Solid media: Growth of L, plantarum was consistently inhibited on TSA either by the simultaneous growth of P, acidilactici or by a product(s) of the latter organism remaining in agar after killing or removing the viable cells. Simultaneous antagonism.was demonstrated by inoculating the surface of agar media with log—phase cultures of the two organisms in perpendicular streaks and allowing the plates to incu- bate at 30°C for 4 to 5 days. The development of L, plantarum was or- dinarily inhibited within zones of l - 2 mm from the edge of the pedio- coccal growth. Deferred antagonism was demonstrated in two ways. First, the agar overlay technique (Materials and Methods) consistently resulted in in- hibitory zones in the soft agar of 4 - 6 nm from the producer culture. However, these zones were sometimes hazy, and their clarity was inverse- ly related to both the amount of chloroform used in sterilization and to the duration of exposure to the vapors. Plates exposed to vapors from 0.3 ml chloroform for 5 min developed zones of inhibition as usual, but plates prepared in an identical manner but exposed to the vapors for 10 min or more were entirely free from inhibitory zones. These data indicate the bacteriocin activity is chloroform labile. Much wider zones were obtained by complete removal of producer cells. 21 22 Colonies on the agar surface were cut out using a cork bore of appro- priate size, and the resulting plug of agar was removed with the tip of a spatula. A 10-fold dilution of exponential-phase indicator cells was streaked up to the edge of the hole in the agar, and the plate was in- cubated overnight. A clearly defined zone of inhibition developed which extended up to 20 mm from the edge of the hole (Plate 1). Liquid media: Inhibitory activity was demonstrated in broth cul- tures by following viable counts in a mixed species inoculation of P, acidilactici and L, plantarum. Each organism was inoculated at approxi- mately 105 cells per ml in TSB. Total viable counts were determined by plating samples on TSA. Viable counts 0f.L: plantarum were determined on TSA containing 0.22 sucrose in place of glucose. The PediOcoccus culture does not utilize sucrose and grows much slower than the £32327 bacillus on this medium. Viable counts of P, acidilactici were obtained by taking the difference between the two counts. The results are pre- sented in Fig. 1. Early development of the L, plantarum culture was followed by a rapid decline of viable counts at approximately 18 h, while P, acidilactici appeared to develop normally. The L, plantarum 2 cells/ml for several days. This organism population remained below 10 made a much more rapid comeback against the pediococci in an earlier study in which the investigators used cucumber juice broth for the growth medium (24). 23 Plate 1. Deferred antagonism of L. plantarum by previous growth of P. acidilactici on TSA. LOG VIABLE CELLS/MI 24 l Figure l. i; 40 60 80 TIME. HOURS Mixed culture growth curves of P, (o) and L, plantarum (0) in TSB. 100 1‘20 acidilactici 25 Production Solid media: Bacteriocin production on solid media was studied as a function of carbohydrate addition, pH, inoculum size and incuba— tion time and temperature. Activity was determined using the agar overlay technique and measuring the widths of the resulting zones of inhibition. Substitutions of maltose, mannose, fructose and yeast ex- tract for glucose in TSA had little or no effect on the size of the in- hibitory zones. Similarly, supplementing a 1% glucose medium with car- bohydrates and yeast extract at concentrations of 1% and 2% had no ef- fect on the expression of activity. The inhibition of indicator growth was most pronounced at neutral pH values, and the activity decreased as the inoculum size of the effector strain was decreased (Table 1). Growth of the effector organism at 25°C resulted in smaller zones of inhibition than those produced at 30°C or 37°C. The zones reached their maximum diameters after 5 days of incubation (Table 2). Liquid media: Prior to studying the production of the bacteriocin in broth media, it was necessary to develop a suitable assay and an appropriate method for sterilization of the spent broth. The critical dilution assay (see Literature Review) was unsatisfactory for this study, as no activity was observed when culture supernatant was spotted on either a lawn of indicator cells or on overlays seeded with the in- dicator strain. Thus, an assay similar to the Modified Survival Count Assay described by Jetten gt EL (49) was developed and used throughout this study. This assay is described in detail in the Materials and Methods section. Plastic caps were used on test tubes after it was ob- served that a substance toxic to L, plantarum was extracted from 26 Table l. Inhibitory activity of P, acidilactici on TSA as a function of inoculum.size Cells Imla inhibi‘gigyhzgfxe, mm 108 4.0 107 3.0 10° 3.0 105 3 . O 10" 2 . O 103 1.0 a Log-phase g, acidilactici cells in TSB were harvested, washed once in 0.1 M potassium phosphate (pH 7.0) and diluted to approximate cell concentrations listed above. Aliquots of 0.05 ml of each dilution were spotted on TSA and were incubated 4 days at 30°C. b Inhibitory activity of the effector culture was determined by the agar overlay technique as described in Materials and Mbthods. Results are the averages of two experiments. 27 Table 2. Inhibitory activity of g, acidilactici on TSA as a function of incubation time and temperature Width of inhibitory zones, mmb Incubation time, daysa 25°C 30°C 37°C 1 1 1 1 2 12 12 9 3 10 9 15 4 1O 14 17 5 7 17 20 6 1 17 19 a Log-phase P, acidilactici cells in TSB were harvested, washed once in 0.1 M potassium phosphate (pH 7.0) and resuspended in one volume of fresh broth. A series of TSA plates was spot inoculated with 0.05 ml cell suspension and incubated at the designated temperatures. b Inhibitory activities of the effector cultures on days 1 through 6 were determined by the agar overlay technique as described in Materials and Methods. Results are the averages of two experiments. 28 polyurethane stoppers during autoclaving. This phenomenon has been reported in an earlier study (4) and the toxic materials were isolated and characterized as volatile fatty amines. Sterilization methods tested included saturation of the active broth with chloroform and subsequent aeration for solvent removal, fil- tration through a membrane filter (Nalgene Filter Apparatus, pore size 0.22 um), and heat treatment. Activity was absent in those samples treated with chloroform and in those which were filter sterilized, but no loss of activity occurred when the test tubes with broth were allow- ed to stand in a boiling water bath for 5 - 6 min (assayed by standard procedure). This treatment effected the killing of all producer cells, and was subsequently used as the method of sterilization prior to each bioassay. Bacteriocin production was studied as a function of producer cul— ture optical density at 600 nm (O.D. ) in TSB supplemented with glu- 600 case to concentrations of 0.252, 0.502, 1.02 and 2.02. Log-phase cells grown in TSB were harvested by centrifugation, washed once in saline (0°C) and suspended in TSB (glucose omitted) to a final O.D. of 0.1. 600 This suspension was used as the inoculum (102) for each of the four media described above, and the cultures were incubated at 30°C. Same ples were withdrawn at time intervals ranging between 6 and 12 h through 48 h and assayed according to standard procedure (Materials and Methods). The results are presented in Fig. 2. Bacteriocin production appears to commence either immediately or shortly after inoculation of the pedio- cocci into fresh media, and is directly related to the cell density up to an approximate O.D.600 of 2.0. Further increases in optical density resulted in no or slight increases in bacteriocin titers. The maximum ACTIVITY, A.U./ml 29 I l I 12- . 8- - 11» - l I 00 I 2 3 4 O.D. 600 OF PRODUCER STRAIN Figure 2. Inhibitory activity of P, acidilactici cultures as a function of producer cell density in TSB containing glucose at con— centrations of 0.25 (A), 0.5 (0), 1.0 (D), and 2.0% (o). 30 cell density attained was dependent upon glucose levels in the media up to 12, but where glucose was not limiting, bacteriocin titer was inde- pendent of glucose concentration. ‘ Fig. 2 also shows rapid decreases in bacteriocin titers which occur as the cultures enter the stationary phases of their growth cycles. The change in bacteriocin titers with incubation time observed in the same experiment are depicted in Fig. 3. The initial loss of activity in the cultures, which may be as high as 50% within a 6 h interval in some in- stances, is followed by a leveling off period where the titers remain relatively constant throughout the remainder of the experiment. Such a dramatic decrease in bacteriocin levels might be due to the release of proteases into the medium, but efforts to demonstrate such activity in the culture supernatant by a method described by Costilow and Coulter (12) were unsuccessful. The fact that the bacteriocin titer levels off before maximum cell density is attained suggests the possibility of a nutrient other than glucose being a limiting factor in bacteriocin production. However, supplementation of the medium with yeast extract at concentrations of 0.1%, 0.252, or 1.0% resulted in no significant increase in titer. No inhibitory activity was produced when the Pediococcus culture was grown in LBS broth (BBL), a complex medium used for isolation and cultivation of lactobacilli (71), even though cell numbers were frequently greater than those reached in TSB. Growth of the producer culture under a vari- ety of conditions including incubation at 37°C and 26°C, incubation at 30°C on a rotary shaker, and incubation at 37°C in an anaerobic chamber had little or no effect on the production of bacteriocin in broth media. ACTIVITY, A.U./ml 31 T I I I I I I o 12 ~ 0 l - '4’ . (3- ' I ‘J A I V II— - A A o 1 1 L L 1 1 l 0 12 24 36 48 TIME. HOURS Figure 3. Inhibitory activity of P. acidilactici cultures in TSB containing glucoge at 0.25 (A), 0.5 (0), 1.0 (U) and 2.0% (o) as a function of incubation time. 32 Cell-Free Preparations Cell-free preparations of bacteriocins are necessary for meaning— ful studies of their production, their physical and chemical character- istics, and for operations leading to their concentration and purifica- tion. The procedure used most often during this study to obtain the activity in a cell-free state was a simple heat treatment of broth cul- ture supernatant. Cultures of L, acidilactici were grown in TSB to late exponential-phase in flasks incubated statically at 30°C, and were centrifuged at 103 x g for 10 min to remove most of the cells. The re- sulting supernatant was sterilized in 5 - ml aliquots (18 x 125 mm test tubes) by submerging the tubes in a boiling water bath for 5 - 6 min. The preparations made by this procedure were then assayed according to standard procedure, and commonly had titers of 6 to 8 Arbitrary Units per m1 (A.U.lml). A second method by which cell-free preparations of bacteriocin was obtained was a freeze-thaw extraction of solid media (80). TSA plates were spread with 0.1 ml samples of a lOO-fold dilution of log-phase pro- ducer culture and incubated 5 days at 30°C. The agar medium was col- lected and transferred to small Erlenmeyer flasks, which were then near- ly submerged in dry ice-acetone baths for 10 min, or until the media had frozen. The flasks were removed from the baths and inverted at room temperature. The extracts produced as the media thawed were collected in test tubes and assayed according to standard procedure (Materials and Methods). The titers varied between 6 and 8 A.U./m1. In contrast, all attempts to demonstrate inhibitory activity in extracts of solid media which had supported growth of P, acidilactici without first 33 freezing the agar met with failure. A number of methods were attempted to obtain bacteriocin activity either from producer cells or cell extracts. Log-phase cells grown in TSB were harvested, washed once in saline and eluted with 6M urea, 2M NaCl, or with solutions of triton X-100 (0.1%, 0.52 and 1.02) in 2M NaCl containing 1 mM ethylene diamine-tetra-acetate. No inhibitory activity against L, plantarum.was observed in cell extracts following disruption by sonication (MSE Sonicator), by a French pressure cell, or by a tissue homogenizer using the method of Dajani gt aL_(15) for the isolation of viridins from alpha-hemolytic streptococci. Physical and Chemical Characteristics Stability in broth as a function of temperature: Loss of bacteri- ocin activity in TSB was a temperature-dependent phenomenon (Fig. 4). Thus, little decrease in titer occurred when active broth (sterilized by heat) was allowed to stand over a period Of 10 days at 4°C, but a significant loss of activity occurred over the same period of time at 30°C. Inactivation of the bacteriocin at 37°C was only slightly more rapid than at 30°C. In view of these data, it is interesting that the bacteriocin is resistant to inactivation by heat treatment at 100°C (Fig. 5). In this experiment, 10 mls of supernatant broth was transferred to each of 6 test tubes and placed in a boiling water bath for varying time inter— vals. Each tube was then allowed to cool to room temperature, and the broths were assayed according to standard procedure. There was only a slight decrease in titer in the broth that had received a 30 minute heat ACTIVITY. A. U./ml 34 I l I l 0 2 4 6 8 IO TIME. DAYS Figure 4. Loss of inhibitory activity of P. acidilactici culture supernatant with time A? incubation temperatures of 25°C (0), 30°C (0), and 37°C (0). 4: ACTIVITY. A.U/ml ha 35 << I I I < < I ' fl . 1 . .