PUREQCATiCfi MD mammmea OF PES'fiCifi Thesis for the Degree of Ph. D, E‘flCéétSAN STATE UMW‘ERSITY Pi 'G-CHUAN HU 1973 LIBRARY Michigan State University This is to certify that the thesis entitled PURIFICATION AND CHARACTERIZATION OF PESTICIN presented by Ping-chuan Hu has been accepted towards fulfillment of the requirements for Public Health MAMA/M Major professor / Date Aurfust 31; 1973 0.7639 ABSTRACT PURIFICATION AND CHARACTERIZATION OF PESTICIN By Ping-chuan Hu Pesticin was prepared from cell—free extracts of Yersinia pestis strain A1122 by a procedure involving fractionation with ammonium sul- fate and chromatography with Sephadex G-200, diethylaminoethyl cellulose, and calcium hydroxyapatite, respectively. A single antibacterial frac- tion (a-pesticin) was recovered from extracts of uninduced yersiniae whereas extracts of mitomycin C-induced cells yielded an additional active fraction (B'Pesticin) at the final step of purification. Anti- bacterial specificity and specific activity of the two fractions were identical, their absorption Spectra were typical of simple proteins and neither contained significant carbohydrate (measured as hexose or hexos- amine). Both pesticins were monomers with a §20,w of 4.4 and molecular weight of about 65,000 as estimated by velocity sedimentation, equi- librium sedimentation, and gel filtration in guanidine hydrochloride. In addition both preparations exhibited similar ratios of amino acids (both lacked detectable cysteine) and possessed common primary structure as judged by peptide analysis of tryptic digests. Isoelectric points of 5.49 and 5.87 were determined for n- and 3-pesticin, respectively. Although a- and B-pesticin were interconvertable in vitro, with Ping-chuan Hu equilibrium favoring of d-pesticin, both forms became rapidly altered to a third configuration, termed y-pesticin. The latter was eluted on calcium hydroxyapatite after B-pesticin and was without significant biological activity. The three forms of pesticin were antigenically identical as judged by immunodiffusion. These results indicate that pesticin, like certain cysteine-deficient colicins, can exist in con- former states. PURIFICATION AND CHARACTERIZATION OF PESTICIN By Ping-chuan Hu A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1973 ‘2 DEDICATION This thesis is dedicated to my wife, Shih-chin, who through her encouragement and support made my entire doctoral program possible. ii ACKNOWLEDGMENTS I would like to express my most sincere appreciation to Dr. Robert R. Brubaker, my major advisor, for his constant enthusiasm, encourage- ment and guidance throughout this investigation. Thanks also go to Drs. Harold L. Sadoff, Loren R. Snyder, John A. Boezi, and Robert L. Uffen for their interest in this project and helpful criticisms. The assistance of Prudence J. Hall in several experiments is also appreciated. A special acknowledgment is due to Drs. William C. Deal, Jr., Willis A. Wood, and Leland F. Velicer for their help in the completion of the ultracentrifuge studies, the amino acid analyses, and the gel filtration analyses, respectively. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . Vii INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . 4 I. Bacteriocins . . . . . . . . . . . . . . . . . . . . . . . 4 1. Classification of Bacteriocins . . . . . . . . . . . . 4 2. Inheritance of Bacteriocinogeny . . . . . . . . . . . . 6 3. Production of Bacteriocins . . . . . . . . . . . . . . 3 4. Mode of Action of Bacteriocins . . . . . . . . . . . . 9 Bacteriocins Affecting Energy Flux . . . . . . . . . . 10 Bacteriocins Affecting DNA Metabolism . . . . . . . . . 11 Bacteriocins Affecting Protein Synthesis . . . . . . . 11 Bacteriocins Causing Membrane Damage . . . . . . . . . 12 II. Pesticin . . . . . . . . . . . . . . . . . . . . . . . . . 13 REFERENCES Production and Host Range . . . . . . . . . . . . . . . 13 Factors Influencing Pesticin Activity and Pesticin Inhibitor . . . . . . . . . . . . . . . . . . 15 Pesticinogeny and Virulence . . . . . . . . . . . . . . 16 Genetic Determinant of Pesticin . . . . . . . . . . . . 17 Mode of Action . . . . . . . . . . . . . . . . . . . . 18 20 Page ARTICLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Characterization of Pesticin. P. C. Hu, and R. R. Brubaker. Manuscript to be submitted to Journal gf Bacteriology. APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Specificity, Induction, and Absorption of Pesticin. P. C. Hu, G. C. H. Yang, and R. R. Brubaker. Journal 9f Bacteriology, Vol. 112, Number 1, October, 1972. LIST OF TABLES TABLE Page 1. Molecular weight of pesticin . . . . . . . . . . . . . . . . 44 2. Amino acid compositions of pesticin . . . . . . . . . . . . 45 3. Interconversion of pesticin . . . . . . . . . . . . . . . . 46 vi FIGURE 10. 11. LIST OF FIGURES Page Elution profiles of induced and uninduced pesticin chromatographed on columns of calcium hydroxy- apatite . . . . . . . . . . . . . . . . . . . . . . . . . 48 Analysis of pesticin by polyacrylamide gel electrophoresis . 50 Velocity sedimentation of uninduced pesticin . . . . . . . . 52 Velocity sedimentation of induced a-pesticin and Brpesticin . 54 Determination of the molecular weight of pesticin by equilibrium sedimentation . . . . . . . . . . . . . . . . 56 Determination of the molecular weights of induced a-pesticin and B-pesticin by gel filtration in 6 M guanidine hydrochloride . . . . . . . . . . . . . . . . . 58 SDS-polyacrylamide gel electrOphoresis of uninduced pesticin . . . . . . . . . . . . . . . . . . . . . . . . 60 Two-dimensional peptide map of a tryptic digest of u-pesticin . . . . . . . . . . . . . . . . . . . . . . . 62 Two dimensional peptide map of a tryptic digest of B-pesticin . . . . . . . . . . . . . . . . . . . . . . . 64 Electrofocusing of a-pesticin and B-pesticin . . . . . . . . 66 Immunodiffusion analysis of pesticin . . . . . . . . . . . . 68 vii INTRODUCTION1 Pesticin is a bacteriocin produced by wild type cells of Yersinia pestis which is active against serogroup I strains of Yersinia pseudo- tuberculosis and certain isolates of Yersinia enterocolitica and Escher- ichia coli. Work by others indicated that this bacteriocin must be induced (by ultraviolet irradiation) to yield a significant titer of extracellular activity, but no pesticin could be found in sonic oscil- lated suSpensions of non-irradiated cells. Evidence was also presented which indicated that the production of pesticin following induction is a process of g§_ggzg synthesis. Discrepancies exist, however,which show that significant amounts of intracellular pesticin activity are detectable in cell extracts of non-irradiated cells. This "constitutive" activity may represent spontaneous synthesis by a small preportion of bacteria within the pesticinogenic population. Pesticin activity has been found to be inhibited by Fe+3 and hemin, but enhanced by Ca+2 and metal chelating agents. Another interesting finding was the discovery of a pesticin inhibitor which is produced by pesticinogenic cells as well as by pesticin-sensitive cells. The action of this inhibitor can be mimicked by bacterial lipopolysaccharides. The early failure to demonstrate antibacterial activity in cell extracts of non~irradiated organisms may have been due to the activity of 1A literature review which follows references the material present in this introduction. pesticin inhibitor or adverse effects of metal ions. It is not known if molecules of pesticin are released into the culture medium or re- tained inside the cell following induction. The mode of action of pesticin has been studied with crude or partially purified materials; and was claimed to be similar to that of colicin E2 (which degrades DNA, steps RNA synthesis, but causes little effect on protein synthesis or oxidative phosphorylation). However, the chemical methods of analysis employed by these workers did not per- mit the estimation of rates of synthesis or degradation of macromole- cules in pesticin-treated cells. Since crude or partially purified materials were used in this early study, the possibility that pesticin- ogenic cells may produce additional antibacterial molecules of distinct specificity can not be excluded. In addition, the inhibition of pesti- cin activity by Fe+3 and hemin has led to another possible mechanism of inhibition (that pesticin may interfere with iron-uptake by sensitive cells). If the latter is true, then the inhibitory action of pesticin on sensitive cells would more likely be bacteriostatic rather than bactericidal. To define the mode of action of pesticin with certainty, it would be necessary to use homogenous pesticin. Furthermore, little is known about the physical and chemical properties of pesticin, although some evidence indicates that it can be inactivated by proteolytic enzymes. Several related proteinaceous bactericins have been shown to be capable of reversible alteration in configuration leading to two physical "con- formers". It was of interest to see if pesticin molecules have a similar property. Thus, we decided, in this study to clarify the dis- crepancy of the distribution of pesticin activity following induction, to purify pesticin to homogeneity, and to characterize its chemical and physical properties. This thesis is organized into three sections. The first section is a literature review including a general discussion about related bacteriocins, and the production, host range, genetic regulation, and mode of action of pesticin. The second section is a manuscript pre- pared for publication concerning the purification and characterization of pesticin induced by mitomycin C. The third section is a published manuscript concerning the Specificity, induction, and absorption of pesticin; the latter is listed as an appendix. LITERATURE REVIEW I. Bacteriocins Bacteriocins are antibacterial substances long known to bacteri- ologists for their restricted host-range and wide distribution. In recent years bacteriocins have been characterized by molecular biolo- gists as proteins with peculiar modes of action. The first bacteriocin was described by Gratia in 1925, who showed that a strain of Escherichia 221i, strain V (for its virulence) produced a substance (named colicin V) which was bactericidal for g, £211, 1. Classification of Bacteriocins. Since the discovery of colicin V, many types of bacteriocins have been demonstrated and some have been purified from both gram-negative and gram-positive bacteria. Colicins, produced by enteric bacteria, have been extensively studied. The scheme of classification of colicins put forward by Fredericq (1948, 1963) has been used as a model in determining other bacteriocins, and has contribu- ted significantly to our understanding of these substances. Fredericq (1946) found that whenever sensitive cells are treated with a given colicin, resistant mutants appear, but these mutants remain sensitive to most but not all other colicins. By using such mutants, Fredericq grouped colicins into 17 types, each characterized by the fact that a mutant resistant to one colicin was resistant to all colicins of that type. Colicins, and other bacteriocins as well, are thought to kill upon initial adsorption onto a specific receptor, mutation to resistance evidently involves loss of this receptor (see below). Accordingly, the classification of colicins by Fredericq was based upon the presence of adsorption sites. However, not all bacteriocins, including certain colicins, can be classified by this criterion (Papavassiliou, 1961; Hamon and Peron, 1963). The catagories of various bacteriocins have been recently reviewed by Reeves (1972) and will not be mentioned. Bacteriocins which have been purified and well characterized such as colicins type E (Herschman and Helinski, 1967), type I (Konisky and Richards, 1970),and type D (Timmis, 1972) are known to be small soluble proteins; and colicin K (Goebel and Barry, 1958) and colicin V (Hutton and Goebel, 1961, 1962) have been shown to be associated with the 0 somatic antigens. However, electron microsc0pic examination of the supernatants of other bacteriocinogenic cultures following induction revealed that bacteriocins can also exist as large complex structures resembling phagelike particles or components (Bradley, 1967). These large bacteriocins differ from the small bacteriocins in that they are sedimentable, trypsin-resistant and thermolabile (Bradley, 1967). These two types of bacteriocins are not necessarily produced by different bac- terial Species. For example, type E colicins are all small molecules; on the other hand, colicin 15, has a molecular weight exceeding 200,000, and appears as a mixture of empty phage heads and intact tailed phage- like objects (Sandoval.e§.al, 1965). The discovery of this type of particulate bacteriocin, along with the fact that the production of bacteriocins and prOphages can both be l l l induced by ultraviolet irradiation and mitomycin C, and the demonstra- tion of plasmid DNA in certain bacteriocinogenic bacteria (DeWitt and Helinski, 1965; Hardy and Meynell, 1972) have led to the belief that some bacteriocins might be degenerated bacteriophages. Bacteriocins appearing as phage tails, such as pyocins (Ishii gt a1, 1965), have been explained as a temperate phage genome that has lost its ability to code for a head. This idea is supported by the observation that phage mutants can be isolated which produce phage particles with missing com- ponents or produce mixtures of a limited number of phage components (Epstein 2; 31, 1963). The evolutionary origin of small-molecule bac- teriocins is obviously more difficult to explain. Some of these have been shown to be associated with 0 somatic antigens (Goebel and Barry, 1958; Hutton and Goebel, 1961, 1962), thus their location is presumably within the cell wall. The plasmid giving rise to theSe substances nevertheless exists within the cytOplasm. Therefore, the implication is that the small bacteriocins are significantly different from large phagelike bacteriocins. 2. Inheritance of Bacteriocinogeny. In general, bacteriocins are thought to be products of nonchromosomal genes that exist on cytoplasmic replicating plasmids (Reeves, 1965), but the inheritance of bacteriocin- ogeny in most bacteriocin-producing organisms is not known. Our know- ledge of the inheritance of bacteriocinogeny is restricted therefore to colicinogeny, which is the system most familiar to geneticists. The first analysis of the genetic basis of colicinogeny was carried out by Fredericq and Betz-Barreau (1953 a, b, c). When these workers crossed colicinogenic strains with an auxotrophic F- strain of E, coli K12, only one of the crosses yielded recombinants (that using E. coli K30 as the colicinogenic strain). This cross gave only seven isolates which grew poorly on minimal medium and soon lost their ability to grow. An examination of their genetic characters showed them to resemble the K12 F- parent, but six of them produced colicin of type E. Thus, al- though the recombinants were unstable and reverted to the K12 parental type, the colicinogenic prOperty was transferred from K30 to the re- combinants and its inheritance was stable. This result indicated that colicinogency was inherited in a different way than that of the other genetic characteristics. Colicinogenic F+ recombinants were then ob- tained by crossing a K12 F+ strain with a K12 (El-K30) F' strain obtained from the first cross. When these colicinogenic F+ recombinants were crossed with an noncolicinogenic F' strain, 73% of the recombinants were colicinogenic. These results and subsequent findings (Fredericq, 1954, 1963; Kahn and Helinski, 1964) strongly suggested that an extra- chromosomal plasmid exists which confers the ability to express colicin and is able to be transferred from cell to cell, independently of the bacterial chromosome prOper. Silver and Ozeki (1962) showed that the transfer of colicinogenic factors I,Eflq and E2 from Salmonella typhimurium to E, 22;; is accompanied by the transfer of 14C-labeled DNA. Amati (1964) found that at the time of colicin induction, more DNA is synthesized per unit of cell mass in UV-induced E, 22;; colicinogenic for 1, El and E2 than in non-colicino- genic cells. DeWitt and Helinski (1965) identified satellite DNA in -Proteus mirabilis colicinogenic for El by equilibrium centrifugation in a CsCl density gradient. This satellite DNA has a heavier buoyant density than the chromosomal DNA, and was not observed in the parent non- colicinogenic strain of E. mirabilis. This DNA was assumed to be the colicinogenic factor El. DeWitt and Helinski also observed that when colicin E1 production was induced by mitomycin C, the buoyant density of the satellite DNA was not altered, but the increase in the amount of the col El DNA was approximately preportional to the magnitude of increase in colicin production. In contrast, Hardy and Meynell (1972) could not observe any increase in plasmid DNA accompanying the increased colicin E2 titer induced by mitomycin C in a E, 221; K12 strain carrying col E2-P9 factor. These workers concluded that induction of col E2-P9 does not resemble induction of prophage. The inheritance of the large bacteriocins is believed to be deter- mined by elements similar to defective prOphages (Bradley, 1967), but no such element has yet been identified. 3. Production of Bacteriocins. Irradiation with ultraviolet light or treatment with mitomycin C are the usual means for inducing synthesis of large quantity of bacteriocin; a normal culture of bacteriocinogenic bacteria always produces a small amount of bacteriocin. When colicino- genic bacteria are plated on a lawn of an indicator strain, small zones of inhibition or lacunae (Ozeki 25 El: 1959) are observed. The number of lacunae found corresponds to about 0.1% of the bacterial culture if E. gyphimurium LT2 (E2-P9) is used, but to over 50% after induction of the some strain by ultraviolet. Because these lacunae do not contain a central colony, it was concluded that normal colicin production re- SUlted from the occasional spontaneous synthesis of colicin by a small PrOFHDItion of the cells in the culture, and that these cells died in the process of synthesis. Bacteria producing other types of colicins were found also to form lacunae on lawns of sensitive cells (Ozeki _£ a_l_, 1959). 4. Mode of Action. Interest in bacteriocins is largely due to their remarkable mode of action. However, too few bacteriocins have been examined to enable us to draw any definite conclusions about their mode of action, but evidence such as the reversibility of the killing action of bacteriocins by trypsin (Nomura & Nakamura, 1962; Elgat and Ben-Gurion, 1969) and the restricted host range of each type (Reeves, 1972) of bacteriocin, certainly suggest that the active molecule is adsorbed to a specific receptor site on the bacterial surface. This event is presumably followed by the biochemical reactions which lead to eventual death of the cell. Different bacteriocins have different modes of action: some inhibit DNA synthesis or protein synthesis, some de- grade DNA or RNA, some inhibit a wide range of activities (such as macro- molecular synthesis together with permease function), and others are thought to have their primary effect on the deployment of energy of the bacterium (Reeves, 1972). The adsorption of molecules of bacteriocin by sensitive organisms can be quantitated by mixing bacterial cells and bacteriocin, removing the bacteria and bound bacteriocins by centrifugation, and assaying the residual antibacterial activity (Mayr-Harting, 1964). A more sensitive method is to mix purified iodinated molecules with the bacteria, incubate, centrifuge, and then assay the residual radioactivity in the supernatant (Konisky and Cowell, 1972). 10 Bacteriophage receptors have been reported to be cell wall structure or pili (Hayes, 1968), but bacteriocin receptors do not seem to have similar locations on the cell. Smarda (1965), and Nomura and Maeda (1965) showed that spherOplasts are sensitive to colicins and that certain colicins adsorb equally well to spherOplasts and to whole cells. They suggested that the receptor is on the membrane. Bacteriocins Affecting Energy Flux. Bacteriocins, including coli- cins type E1 (Jacob 95 El: 1952), type K (Nomura and Maeda, 1965), type A (Nagel de Zwaig, 1969), type Ia and lb (Levisohn ggngl, 1968), and pyocin (Jacob, 1954), have been shown to effect the energy flux of sensi- tive cells. Jacob _£__l (1952) showed that colicin E1 stopped nucleic acid synthesis and growth of sensitive cells, but had no effect on res- piration for at least 20 minutes. Luria (1964) showed that the same colicin inhibited the function of some permeases, and Bhattcharyya _£__l (1970) showed that colicin E1-K30 inhibited permease activity of isolated membrane vesicles. Colicin K.has been shown to inhibit nucleic acid and protein syn- thesis (Fields and Luria, 1969b) as well as potassium transport and permease function (Luria, 1964; Nomura and Media, 1965). The wide range of activities of these two colicins was thought to affect some aspect of energy metabolism (Luria, 1964). Fields and Luria (1969 a, b) showed that neither colicin El nor colicin K had significant effect on the accumulation of n-galactosides while both inhibited the accumulation of Ekgalactosides. This indicated that the effects are more specific than complete abolishment of energy metabolism. They attributed the different effects to a requirement for either ATP or PEP for the accumulation 11 phases of B-galactosides or a-galactosides permease respectively, and suggested the two colicins lead to a lowering of ATP levels. Colicin A has been shown to act similarly; it inhibits nucleic acid synthesis and accumulation of isoleucine and B-galactoside (Nagel de Zwaig, 1969). Both colicin la and lb were reported to inhibit macromolecular synthesis and ATP production (Levisohn g; 3;, 1968). A pyocin produced by Pseudomonas pyocyanea P10 has been shown to stop the growth of sensitive cells and to slow down the respiration rate. This pyocin also stOps multiplication of avirulent bacteriophage in sensitive bacteria and is thought to act on the utilization of energy in a fashion similar to that of colicin E1 (Jacob, 1954). Bacteriocins AffectinggDNA Metabolism. Colicin type E2 and megacin C were shown to affect the DNA metabolism of sensitive cells. Colicin E2 not only inhibited DNA synthesis but also degraded preexisting DNA (Nomura, 1963). Recent work of Obinata and Mizuno (1970), and Ringrose (1970) indicate that DNA degradation to acid soluble material is pre- ceeded by endonucleolytic cleavage of the DNA into large fragments. However, colicin E2 itself has no detectable lg_yl££g nuclease activity (Nomura, 1964; Ringrose, 1970), thus it is assumed that this substance activates a host nuclease. The nature of this nuclease has not been established. Megacin C stops DNA synthesis and greatly reduces RNA and protein synthesis, it also degrades DNA (Holland, 1963, 1965). Unlike colicin E2, megacin C does not completely inhibit RNA and protein synthesis. Bacteriocins Affect Protein Synthesis. Colicin E3 has been shown to affect protein synthesis (Nomura, 1963) but not DNA synthesis or 12 overall RNA synthesis. Recently, it has been shown that this colicin is active in cleaving l6S ribosomal RNA lg XIEEQ (Bowman 35 El 1971a; Boon, 1971). Another interesting finding is that the immunity of the colicinogenic strain has been shown to be due to an intracellular inhibitor (Bowman 2; El: 1971b). This inhibitor is present in crude preparations, which probably accounts for the failure of earlier at- tempts to demonstrate an lg 21239 effect of colicin E3 (Konisky and Nomura, 1967). Cloacin DE13 was found to have a mode of action similar to coli- cin E3, which inhibits protein synthesis «kaGraaf and Stourthamer, 1969). Pneumocins S6 and S8 were reported to inhibit protein synthesis also, but the action was slower (de Graaf and Stouthamer, 1971). Bacteriocins Causing Membrane Damage. There are two bacteriocins which cause membrane damage, megacin A-216 and enterococcin I-X14. When sensitive cells were treated with megacin A-216, the optical den- sity and respiration rate decreased (Ivanovics, gg El: 1959; Holland, 1962). It was found that the decrease in turbidity was due to leakage of cellular constituents, but the cell wall remained intact. This bacteriocin has now been shown to be a phospholipase A (Ozeki 25 El; 1966). Enterococcin I-Xl4 is produced by Streptococcus facecalis var. zymogens (Brock and Davie, 1963). Treatment of sensitive cells with this bacteriocin leads to rapid loss of viability and ability to take up 14C-labeled glycine (Davie and Brock, 1966). Similar effects were observed on spherOplasts, suggesting that the cell wall was not in- volved. Davie and Brock believed that this bacteriocin exhibits an l3 effect on cell membrane, but no direct evidence exists that the killing or loss of ability to take up glycine results from such membrane damage. II. Pesticin Pesticin, a bacteriocin produced by wild type Yersinia pestis, was first described by Ben-Gurion and Hertman (1958); it was found to in- hibit the growth of Yersinia pseudotuberculosis. This antibacterial substance was found to be inducible under conditions similar to those which induce the activity of other bacteriocins. Upon the further dis- covery of a second antibacterial activity in E. pestis and E. pseudo- tuberculosis, Brubaker and Surgalla (1962) named the original activity as pesticin I and the second pesticin II. However, the latter may not be a true bacteriocin (Brubaker 25 El: 1965), thus the original name pesticin is retained in this dissertation. 1. Production and Host Range. Pesticin activity can be reversed by trypsin; this finding indicates it is protein in nature. The anti- bacterial activity is stable between pH 6 and 8, but can be destroyed by extremes of pH and heat. Pesticin is much more active at 37 C than at 30 C while the Optimal temperature for its production is 26 C (Ben- Gurion and Hertman, 1958). Hertman and Ben-Gurion (1959) also presented convincing evidence that the expression of pesticin following irradiation with ultraviolet light represents Eg_ggyg synthesis of protein. This conclusion was based on the demonstration of requirements for certain amino acids during biosynthesis and by inhibition of synthesis by chloramphenicol. l4 Conflicting reports exist regarding the induction of pesticin. Ben-Gurion and Hertman (1958) obtained high titers of extracellular pesticin activity following irradiation of E. pestis with UV light, but were unable to detect activity in sonic oscillated cell suspensions. Therefore, these workers concluded that the induction of pesticin is similar to that of E type colicins. In contrast, Brubaker and Surgalla (1961, 1962) and Beesley g5 g; (1967) observed significant amounts of intracellular pesticin activity in cell extracts of non-irradiated or- ganisms. This constitutive activity may represent Spontaneous synthesis by a small proportion of bacteria within the pesticinogenic population. Similar phenomena have been observed in colicinogenic cells (DeWitt and Helinski, 1965). Most colicin molecules including those of the E type were reported to be released into the supernatant (Nomura, 1963; Nomura and Maeda, 1964) following induction; however the activities of colicins Ia and Ib were found associated with the cell mass following induction with mitomycin C (Konisky, 1971). It is not clear if pesticin, follow- ing induction, is released into the culture medium or retained inside the cells. In the study of the host range of pesticin, Burrows and Bacon (1960) showed that sensitivity to pesticin was confined to, and characteristic of, all serological group I strains of E. pseudotuberculosis. The latter are known to produce lipopolysaccharide containing paratose (Davis, 1961). Brubaker and Surgalla (1961) subsequently showed that certain strains of Escherichia coli, including the universal indicator strain 0, and some strains of non-pesticinogenic E. pestis are also sensitive to pesticin. 0f nine pesticin-resistant mutants of E. coli which were examined, 4 15 also lost their sensitivity to colicins B, D, I, and 81; the other five retained their sensitivity to the 16 colicins tested. These findings suggest that pesticin may be adsorbed onto receptors similar to those involved in colicin uptake. The host range of pesticin was subsequently extended to a few isolates of E. enterocolitica (R. R. Brubaker, Bacter- ial Proc. p. 18, 1966). 2. Factors InfluencingiPesticin Activity and Pesticin Inhibitor. Brubaker and Surgalla (1961, 1962) noted that the antibacterial activity +3, Mg+2, hemin and an acid soluble of pesticin can be suppressed by Fe metabolite (pesticin inhibitor, Egg Eglgw) produced by both pesticin producing cells and indicator cells. Whether these cations and hemin affect pesticin directly, or indirectly by activating the pesticin in- hibitor is still unclear. Suppression of the activity of pesticin by Fe+3, hemin, and pesticin inhibitor can be reversed by the addition of either Ca+2 or metal chelating agents (Brubaker and Surgalla, 1962). Many microbial exoenzymes are stabilized or activated by Ca+2 (Pollock, 1962), and a similar role for pesticin would be not surprising. The inhibitory effects of Fe+3 and hemin are more difficult to explain. By double-layer method, Brubaker and Surgalla (1961) and Brubaker 35 al (1965b) showed that when the concentrations of Fe+3 and Ca+2 in the overlayer agar were properly adjusted, indicator cells grew above and immediately adjacent to a streak of E. pestis, whereas growth of more distant indicator cells were suppressed. Indicator cells subcul- tured from colonies proximal to the producer colony remained sensitive to pesticin, and pesticin was not inhibited when indicator cells were overlayered in agar containing metal chelating agents. These findings 16 led Brubaker and his coworkers to predict that the lack of inhibition of growth in the immediate vicinity of the producer colony was due to the action of a slowly diffusing metabolite which they termed pesticin inhibitor. The chemical nature of pesticin inhibitor is unknown, but its effect can be mimicked by desferrioxamine, an iron carrier of Strepto- myces griseus (Brubaker 25 El, 1965a), and by lipOpolysaccharides (R, M, Soltysiak, personal communication). However, the role of pes- ticin inhibitor and the associated effect of iron remain obscure at the present time. 3. Pesticinogeny and Virulence. There are five determinants of virulence in E. pestis as summarized by Brubaker (1971); the abilities to: express pesticin activity (EEEI): produce V and W antigens (233+), elaborate envelope or fraction I antigen (EEQI), synthesize a surface structure which permits absorption of certain dyes and pigments (pgmf), and produce endogenous purines (pgg+). All of these factors can be estimated quantitatively Eg‘yiggg (Burrows, 1963; Surgalla 33 El: 1970) and undergo independent mutational loss resulting in the decrease of virulence in mice and guinea pigs (Burrows, 1963; Brubaker, 1971). Burrows (1965) found that among the strains of E. pestis examined in his laboratory, all fully virulent isolates were pesticinogenic and that all non-pesticinogenic strains were avirulent. The retention of pesticinogeny by virulent strains led Burrows to speculate that pesticin, or the ability to produce pesticin, is in some way associated with the pathogenic potential of E, pestis. Based on the observation that hemin and Fe+3 inhibit pesticin activity, Burrows suggested that hemin and l7 pesticin are able to complex in some way. The ability to accumulate hemin during growth on a defined medium differentiates Egmf from pgmf organisms. If pesticin was involved in this accumulation, then it would explain why Egg- strains were generally pgm-. However, some pggf strains are p§£+ thus the correlation between these genotypes is not absolute. Brubaker and his coworkers (1965, 1967) found that all pggf strains are capable of synthesizing coagulase and fibrinolysin, whereas Egg- strains are devoid of these two biological activities. Since coagulase and fibrinolysin are important invasive factors of pathogens, they assumed that the Egg? determinant is a Eggglglgg virulence determinant. This suggestion was confirmed in another report (Brubaker 23 El: 1965) upon isolation of an avirulent mutant (strain G 32) which was positive for all known virulent determinants except pggf and was negative for coagulase and fibrinolysin. Furthermore, simultaneous injection of mice with iron restored the full virulence of a pggf Egg- avirulent mutant in mice (Burrows, 1962), and suggested that pesticin is either not essential for virulence or that ferrous ion fulfilled a dual require- ment by replacing the products of both pggf and pgmf. However, the ultimate proof of pesticinogeny as a virulence determinant will depend upon the isolation of an avirulent mutant which is Egg- but positive for both coagulase and fibrinolysin. 4. Genetic Determinant of Pesticin. Compared to the intensive and fruitful studies on the inheritance of colicinogeny, there is little known about the inheritance of other bacteriocins. Although the mutation of pst+ to pst- occurs at low frequency (Brubaker, 1971), this gene, in 18 view of the nature of its product, is assumed to exist on an extra- chromosomal replicon (Brubaker and Surgalla, 1961). However, there is no evidence indicating that pesticinogeny is transferable. These studies have been hampered by lack of a suitable conjugation system in Yersinae; this type of system is necessary for the transfer of colicinogeny (Fredericq, 1954, 1963). The physical presence of plas- mids in Yersinae is also uncertain. Little and Brubaker (1972) were unable to detect plasmids in cells of E. pestis pst+, vwa+, fea+, pgmf by use of ethidium bromide-cesium chloride density gradient centri- fugation. These results, of course, do not eliminate the possibility that these genes are located on an integrated plasmid. 5. Mode of Action. Using crude or partially purified materials, Elgat and Ben-Gurion (1969), showed that viable cells of E, ggll strain 0 decreased exponentially in broth in proportion to the con- centration of added pesticin. Similar reduction of E. pseudotubercu- 1osis was observed only in the presence of 0.05 M CaCl These workers 2. claimed that the action of pesticin on macromolecular synthesis of sen- sitive cells of E, 22;; strain 0 was similar to that of colicin E2. After exposure to pesticin, DNA synthesis was arrested and RNA was de- graded, but little effect was observed on protein synthesis. Further- more, both pesticin and colicin E2 induced lysogenic E, 22;; strain (Pl), but pesticin, unlike colicin E2, was active in the presence of 2,4-dinitrophenol. The early findings of Nomura (1963) who employed radioactivity-tracing techniques, showed that colicin E2 degraded DNA and stopped RNA synthesis, thus the effect on RNA synthesis was probably a consequence of DNA degradation. Since the chemical methods of analysis 19 employed by Elgat and Ben-Gurion did not permit the estimation of rates of synthesis or degradation of macromolecules, further work will be necessary to define with certainty the mode of action of pesticin. REFERENCES Amati, P. (1964), J. Mol. Biol. 8, 239. Beesley, E. D., Brubaker, R. R., Janseen, W. A., and Surgalla, M. J. (1967), J. Bacteriol. 94, 19. Ben-Gurion, R., and Hertman, I. (1958), J. Gen. Microbiol. 19, 289. Bhattcharyya, P., wendt, L., Whitney, E., and Silver, 8. (1970), Science 168, 998. Boon, T. (1971), Proc. Natl. Acad. Sci. U.S.A. 68, 2421. Bowman, C. M., Dahlberg, J. E., Ikemura, T., Konisky, J., Nomura, M. (1971a), Proc. Natl. Acad, Sci. U.S.A. 68, 964. Bowman, C. M., Sidikaro, J., and NOmura, M. (1971b), Nature New Biology 234, 133. Bradley, D. E. (1967), Bacterial. Rev. 31, 230. Brock, T. D., and Davie, J. M. (1963), J. Bacteriol. 86, 708. Brubaker, R. R. (1969), J. Bacteriol. 98, 1404. Brubaker, R. R. (1971), Cur. Top. Microbiol. 57, 111. Brubaker, R. R., Beesley, E. D., and Surgalla, M. J. (1965a), Science 149, 422. Brubaker, R. R., and Surgalla, M. J. (1961), J. Bacteriol. 82, 940. Brubaker, R. R., and Surgalla, M. J. (1962), J. Bacteriol. 84, 539. Brubaker, R. R., Surgalla, M. J., and Beesley, E. D. (1965b), Zentr. Bakteriol. Parasitenk. Abt. I. Orig. 196, 302. Burrows, T. W. (1962), Brit. Med. Bull. 18, 69. 20 21 Burrows, T. W. (1963), Ergebn. Mikrobiol. 37, 59. Burrows, T. W. (1965), Zentra. Bakteriol. Parasitenk. Abt. I. Orig. 196, 315. Burrows, T. W., and Bacon, G. A. (1960), Brit. J. Exp. Path. 39, 278. Davie, J. M., and Brock, T. D. (1966), J. Bacteriol. 91, 595. Davis, D. A. L. (1961), Nature 191, 43. DeWitt, W., and Helinski, D. R. (1965), J. M01. Biol. 13, 692. Elgat, M., and Ben-Gurion, R. (1969), J. Bacteriol. 98, 359. Epstein, R. H., Bolle, A., Steinberg, C. M., Kellenberger, E., Boy De La Tour, E., Chevalley, R., Edgar, R. S., Susman, M,, Denhardt, G. H., and Lielausis, A. (1963), Cold Spring Harbor Symp. Quant. Biol. 28, 375. Fields, K. L., and Luria, S. E. (1969a), J. Bacteriol. 97, 57. Fields, K. L., and Luria, S. E. (1969b), J. Bacteriol. 97, 64. Fredericq, P. (1946), Compt. Rend. Soc. Biol. 140, 1189. Fredericq, P. (1948), Rev. Belge. Pathol. Med. Exp. 29 (Suppl. 4), 1. Fredericq, P. (1954), Compt. Rend. Soc. Biol. 148, 399. Fredericq, P. (1963a), J. Theoret. Biol. 4, 159. Fredericq, P. (1963b), Ergeb. Mikrobiol. Immunol. Exp. Therap. Berlin 37, 114. Fredericq, P. and Betz-Barreau, M. (1953a), Compt. Rend. Soc. Biol. 147, 1110. Fredericq, P. and Betz—Barreau, M. (1953b), Compt. Rend. Soc. Biol. 147, 1653. Fredericq, P., and Betz-Barreau, M. (1953c), Compt. Rend. Soc. Biol. 147, 2043. 22 Goebel, W. F., and Barry, C. T. (1958), J. Expl. Med. 107, 185. de Graff, F. K., and Stouthamer, A. H. (1969), J. Gen. Microbiol. 53, xiii. de Graff, F. K., and Stouthamer, A. H. (1971), J. Microbiol. Serol. 37, 1. Gratia, A. (1925), Compt. Rend. Soc. Biol. 93, 1040. Hamon, Y. and Peron, Y. (1963), Ann. Inst. Pasteur 104, 55. Hardy, K. G., and Meynell, G. G. (1972), J. Bacteriol. 112, 1007. Hayes, W. (1968), The Genetics of Bacteria and Their Viruses, John Wiley and Sons Inc. New York. Herschman, H. R., and Helinski, D. R. (1967), J. Biol. Chem. 242, 5360. Hertman, 1., and Ben-Gurion, R. (1959), J. Gen. Microbiol. 21, 135. Holland, I. B. (1962), J. Gen. Microbiol. 29, 603. Holland, 1. B. (1963), Biochem. BiOphys. Res. Communs. 13, 246. Hutton, J. J., and Goebel, W. F. (1961), Proc. Natl. Acad. Sci. 47, 1498. Hutton, J. J., and Goebel, W. F. (1962), J. Gen. Physiol. 45 (suppl.), 125. Ishii, S., Nishi, Y., and Egami, F. (1965), J, Mol. Biol. 13, 428. Ivanovics, G., Alfoldi, L., and Nagy, E. (1959), J. Gen. Microbiol. 21, 51. Jacob, F. (1954), Ann. Inst. Pasteur 86, 149. Jacob, F., Siminovitch, L., and Wollman, E. (1952), Ann. Inst. Pasteur 83, 295. Kahn, P. L., and Helinski, D. R. (1964), J. Bacteriol. 88, 1573. Konisky, J. (1972), J. Biol. Chem. 247, 3750. 23 Konisky, J., and Cowell, B. S. (1972), J. Biol. Chem. 247, 6524. Konisky, J., and Nomura, M. (1967), J. Mol. Biol. 26, 181. Konisky, J., and Richards, F. M. (1970), J. Biol. Chem. 245, 2972. Levisohn, R., Konisky, J., and Nomura, M. (1968), J. Bacteriol. 96, 811. Little, R. V., and Brubaker, R. R. (1972), Infect. Immun. 5, 630. Luria, S. E. (1964), Ann. Inst. Pasteur 107 (Suppl. to No. 5), 67. Mayr-Harting, A. (1964), J. Pathol. Bacteriol. 87, 255. Mindich, L. (1966), J. Bacteriol. 92, 1090. Nagel de Zwaig, R. (1969), J. Bacteriol. 99, 913. Nomura, M. (1963), Cold Spring Harbor Symp. Quant. Biol. 28, 315. Nomura, M. (1964), Proc. Natl. Acad. Sci. U.S.A. 52, 1514. Nomura, M., and Maeda, A. (1965), Zentr. Bakteriol. Parasitenk. Abt. I. Orig. 196, 216. Nomura, M., and Nakamura, M. (1962), Biochem. BiOphys. Res. Communs. 7, 306. Obinata, M., and Mizuno, D. (1970), Biochem. et BiOphys. Acta. 199, 330. Ozeki, M., Higashi, Y., Saito, H., An, T., and Amano, T. (1966), Bikens J. 9, 201. Ozeki, H., Stocker, B. A. D., and De Margerie, H. (1959), Nature 184, 337. Papavissiliou, J. (1961), Nature 190, 110. Pollock, M. R. (1962), 12_The Bacteria, Vol. 4, Gunsalus, I. C., and Stanier, R. Y., Eds., Academic Press, New York, p. 121. Reeves, P. (1965), Bacteriol. Rev. 29, 25. 24 Reeves, P. (1972), The Bacteriocins, Springer-Verlag, New York, Heidel- berg, Berlin. Ringrose, P. (1970), Biochem. et BiOphys. Acta 213, 320. Sandoval, H. K., Reilly, H. C., and Tandler, B. (1965), Nature 205, 522. Silver, 8., and Ozeki, H. (1962), Nature 195, 873. Smarda, J. (1965), Zentr. Bakteriol. Parasitenk. Abt. I. Orig. 196, 240. Surgalla, M. J., Beesley, E. D., and Albizo, J, M. (1970), Bull. Wld. Hlth. Org. 42, 993. Timmis, K. (1972), J. Bacteriol. 109, 12. ARTICLE CHARACTERIZATION OF PESTICIN P. C. Hun and R. R. Brubaker Manuscript to be submitted to Journal g_f_ Bacteriology FOOTNOTE * This work was submitted in partial fulfillment of the requirements for the degree of Doctor of PhilOSOphy. 26 ABSTRACT Pesticin was prepared from cell-free extracts of Yersinia pestis strain A1122 by a procedure involving fractionation with (NH4)2504 and chromatography with Sephadex G-200, diethylaminoethyl cellulose, and calcium hydroxyapatite, respectively. A single antibacterial fraction (a-pesticin) was recovered from extracts of uninduced yersiniae whereas extracts of mitomycin C-induced cells yielded an additional active fraction (g-pesticin) at the final step of purification. Anti- bacterial specificity and specific activity of the 2 fractions were identical; their absorption spectra were typical of simple proteins and neither contained significant carbohydrate (measured as hexose or hexosamine). Both pesticins were monomers with a E“ of 4.4 and 20,w molecular weight of about 65,000 as determined by sedimentation velocity centrifugation, equilibrium sedimentation, and gel filtration in guani- dine hydrochloride. In addition, both preparations exhibited similar ratios of amino acids (both lacked detectable cysteine) and possessed common primary structure as judged by peptide analysis of tryptic digests. Isoelectric points of 5.49 and 5.87 were determined for a- and B-pesticin, respectively. Although a- and§3épesticin were inter- convertable lg XIEEQJ with equilibrium favoring formation of a-pesticin, both forms became rapidly altered to a third configuration, termed y~pesticin. The latter was eluted on calcium hydroxyapatite afterta- pesticin and was without significant biological activity. The 3 forms of pesticin were antigenically identical as judged by immunodiffusion. These results indicate that pesticin, like certain cysteine-deficient colicins, can exist in conformer states. 27 INTRODUCTION Pesticin is a bacteriocin produced by wild type cells of Yersinia pestis which is active against serogroup I strains of Yersinia pseudo- tuberculosis and certain isolates of Yersinia enterocolitica and Escherichia coli (2,4). Although the concentration of pesticin in cell- free extracts of normal yersiniae is sufficient to permit their use as a source of material for purification (14), the specific activity can be increased by induction with mitomycin C (unpublished observations) or ultra-violet light (UV) (2,14). Upon attempting to purify large quantities of pesticin by use of extracts of mitomycin C-induced cells, we noted the presence of a new fraction of activity, termedtS-pesticin, at the final step of purifica- tion with calcium hydroxyapatite. This fraction was eluted after the activity common to uninduced and induced cells, termed G-pesticin, had been collected. Both fractions exhibited identical specific activities when tested against a variety of indicator bacteria (unpublished obser- vations) suggesting that the 2 activities were not products of distinct genes. The purpose of the study reported here was to prove this hypothesis and to determine the nature of the difference between a- and B-pesticin. Some possibilities considered were that the 2 molecules reflect either conformer states, distinct aggregates of subunits, or products of con- version by an epigenetic mechanism such as glucosylation or peptide Cleavage. The results indicate that G- and B-pesticin, and possibly a tilird biologically inactive form termed Yipesticin, are conformers. This cOnclusion is based upon comparison of physical, chemical, and inmuno- Egéirlic properties and the demonstration of interconversion lg XEEEQ, 28 MATERIALS AND METHODS Bacteria. Pesticin was prepared from cells of E, pestis strain A1122 and assayed with indicator cells of E, pseudotuberculosis strain PBl/+ by the procedure of Hu, Yang, and Brubaker (l4). Cultivation and induction. The medium and cultural conditions employed previously (14) were used to prepare uninduced bacteria al- though in some cases the cells were grown to stationary phase at 26 C in fermenter vessels (Model SF 305, New Brunswick Scientific Co., New Brunswick, N.J.). Induced cells were prepared by aeration at 26 C in 200 ml of medium per 2 liter flask on a shaker (Model R25, New Bruns- wick Scientific Co.). When the pOpulation reached a density of 5 x 108 cells per ml, the organisms were collected by centrifugation (27,000 x g) for 15 min and resuspended in fresh medium. After further aeration for 15 min, the cultures received mitomycin C (0.2 Ug/ml) and a mixture of amino acids equivalent to 50% of that present in the synthetic medium of Yang and Brubaker (26). This amino acid component was replaced by 200 uC of 14C-algal hydrolysate (Calatomic, Los Angeles, Calif.) per 200 ml of medium when radioactive-labeled pesticin was induced. Aeration was continued for an additional 8 h and the cells were then harvested by centrifugation, washed with 0.1 M tris (hydroxymethyl) aminomethane (Tris)-HCl buffer (pH 7.8), resuspended in the same buffer, and subjected to disruption as previously described. Purification 9E pesticin. The process of purification was slightly modified from that used previously (14) by including a step of fraction- ation on Sephadex G-200 (Pharmacia, Piscataway, N.J.). The procedure entailed dialyzing that fraction of crude extract which precipitated between 33 and 66% saturated (NH with 0.05 M Tris-HCl buffer, 4)2304 29 30 pH 7.8, containing 1.0 M NaCl. This preparation was then applied to a column (2.0 x 40 cm) of Sephadex G-200 and eluted with the same buffer. The fraction containing pesticin was concentrated by precipation with 70% saturated (NH4)28 and dialyzed against 0.05 M Tris-HCl buffer 04 (pH 7.8); the sample was then chromatogramed with DEAE cellulose and calcium hydroxyapatite as previously described. Ultracentrifugal analysis. Preparations were dialyzed for 24 h against 2 changes of 500 m1 potassium phosphate buffer, pH 7.0, con- taining 0.1 M KCl. The final dialysate was used for dilutions and in reference cells. Analytical sedimentations were performed in a Spinco Model E ultracentrifuge equipped with a RTIC temperature control unit and either Rayleigh interference optics with 12 mm double sectored cells or Schlieren Optics with 30 mm double sectored cells. Sedimenta- tion velocity analysis was conducted at 19.2 C for uninduced and at 6.5 C for induced preparations at a rotor speed of 56,100 rpm; the sedimentation coefficient was corrected to §20,w (4). High speed sedi- mentation equilibrium analysis was performed at 10.9 C at rotor speeds of 25,978 rpm and 25,943 rpm for uninduced and induced preparations, respectively. Calculations of molecular weight, according to the meniscus depletion technique (27), were based on the fringe displace- ment following achievement of equilibrium. Analytical gel filtration lg guanidine hydrochloride. Analytical gel filtration was performed in 1.2 x 100 cm chromaflex glass columns (Kontes Glass Co., Vineland, N.J.) containing 6% Agarose equilibrated with 6 M guanidine hydrochloride and 0.02 M sodium phosphate buffer, pH 6.5 (equilibration mixture) (8). The freshly prepared columns were examined for prOper packing by chromatography of 0.2% blue dextran 31 (Pharmacia) in equilibration mixture plus 10% sucrose. Upon demonstra- tion of suitable gel beds, the columns were treated with equilibration mixture containing 0.01 M dithiothreitol prior to application of samples. The marker proteins transferrin (MW 76,000), bovine serum albumin (MW 68,000), and ovalbumin (MW 45,000) were obtained from Sigma (St. Louis, Mo.) and iodinated (1311) by the procedure of Helmkamp et a1. (12). Prior to chromatography, these marker proteins and preparations of 14C-pesticin were solubilized in 0.5 m1 of a solution of 8 M guani- dine hydrochloride, 0.05 M dithiothreitol, 0.01 M ethylenediaminetetra- acetic acid, and 0.05 M Tris-HCl (pH 8.5) (dissociation mixture) for 45 min at 56 C. At this time, the dissosciation mixture containing proteins received 0.1 m1 of equilibration mixture containing 0.6% blue dextrose and 50% sucrose and was applied to the column. Polyacrylamide disc gel electrophoresis. This procedure was per- formed as previously described (14). Sodium dodecyl sulfate (SDS)1pglyacrylamide disc gel electr0phoresis. Samples were prepared at a concentration of 0.5 to 1 mg per ml and electrophoresed by the method of Shapiro, Vinuela, and Maizel (23) as modified by JOhnson, Debacker, and Bozei (15). Gels were stained for 12 h with 0.4% Coomassie blue in 10% trichloroacetic acid and 20% methanol and then destained by washing for 6 h in a solution of 10% trichloroacetic acid and 33% methanol followed by rinsing in 10% trichloroacetic acid. Standard proteins were RNA polymerase from Pseudomonas putida where the molecular weights for the a, 8, and 3‘ subunits are 44,000, 155,000, and 165,000, respectively (15). Also used were bovine serum albumin, 32 glutamate dehydrogenase, ovalbumin, and chymotrypsinogen A of respective molecular weights of 68,000, 53,000, 45,000, and 25,700. Isoelectric focusing. The small scale isoelectric focusing tech- nique of Massey et a1. (19) was employed. The focusing column con- sisted of a glass tube (14.5 x 0.8 cm) containing an enlargement 1 cm from the bottom to secure the 5% acrylamide plug. The plug, 2 cm in length, was poured, polymerized, and then soaked overnight in 3% H2804. A linerar gradient was then prepared from the heavy solution (0.15 ml carrier ampholine; pH 3 to 10, 1.0 gm sucrose; 1.95 ml water; and approx- imately 500 ug of pesticin) and light solution (0.05 ml carrier ampho- line, 0.3 gm sucrose, 2.67 ml water, and approximately 500 ug of pesticin); 3% ethylene diamine was layered onto the top of the gradient. The tube was then inserted into the gel electrOphoresis apparatus (Polyanalyst, Buchler Instrument, Fort Lee, N.J.) with 3% H2804 as the anode and 3% ethylene diamine as the cathode. Upon completion of the determination, 0.2 m1 fractions were collected and assayed for pH (pH.Meter 26, Radiometer, London Co., Westlake, Ohio) and biological activity. Amino acid analysis. Samples were dialyzed exhaustively against glass distilled water for 24 h and, following lyophilization, 0.5 mg of preparation were dissolved in 1 m1 of 6 N HCl in glass tubes. The tubes were sealed and placed in a drying oven at 110 C for 24 h. Following hydrolysis, portions of the sample equivalent to 25 ug of pro- tein were analyzed for amino acids by the procedure and apparatus described by Robertson, Hammerstedt, and Wood (20). Corrections for decomposition of serine and threonine were made as described by Moore and Stein (20). Trypt0phan was determined by the method of Beaven and Holiday (1). 33 Peptide analysis. Lyophilized samples of pesticin (8.0 mg) in tubes were dissolved in 2 m1 of deionized water. After denaturation at 90 C in a water bath for 4 min, the preparations received 2 m1 of 0.2 M NHAHCO3, pH 8.0. After mixing, 15 ul of a L-l-tosylamido-2-pheny1ethyl chloromethyl ketone-treated preparation of trypsin was added (5 mg per m1 of 0.001 N HCl) as described by Carpenter (5)rfliusa small drop of toluene (in order to maintain sterility). After incubation for 12 h at 37 C, 5 pl of trypsin solution was again added and incubation was continued for another 8 h. The preparations were then centrifuged at 27,000 x‘g for 15 min to remove insoluble cores and the supernatant fluids were transferred to 250 ml round bottom flasks. Each prepara- tion then received 2 ml of distilled water prior to shell freezing and 1y0philization (which was extended for 4 h after the samples had reached apparent dryness in order to assure removal of NH HCO3). The 4 samples were then dissolved in 1 m1 of deionized water, transferred to lyophilization tubes (1.3 x 10 cm) and again 1y0philized; this proce- dure was repeated twice. The residues were dissolved in deionized water to yield a concentration of approximately 25 mg per m1. A 40111 sample of each tryptic digest was carefully applied to thin-layer plates (Silica Gel 60 F-254, E. Merck, Darmstardt, Germany) and electrophoresis was performed with pyridine-acetic acid-water buffer, pH 6.5 (25:1:225), for 100 min at 900 V. Following electro- phoresis, the plates were air-dried and then chromatographed in the second dimension with n-butanol-acetic acid-pyridine-water (15:3:10:12) for 8.5 h. After air-drying followed by exposure to 80 C in an oven for 20 min, the cooled plates were sprayed with 0.4% ninhydrin in acetone and heated at 110 C for 15 min for the development of color. 34 Carbohydrate analysis. Hexose was determined with anthrone rea- gent (22) and amino sugars were assayed by the Elson-Morgan reaction (3). Interconversion. Equal concentrations of fresh a- and B-pesticin were mixed and then dialyzed overnight against 0.01 M sodium phosphate buffer, pH 7.0; samples of 3 ml (3.0 mg) were then added to 6 screwcap tubes (18 x 150 mm). A set of 3 tubes received 1.5 m1 of a freshly prepared and dialyzed sample of 14C-a-pesticin (1.5 mg; 15,000 cpm) and the remainder received a similar preparation of 14C-B-pesticin. The mixtures were saturated with CHCl3 vapour to maintain sterility and were rechromatogramed on calcium hydroxyapatite after storage at 26‘C for 1, 3 and 7 days. Fractions containing proteins, as detected with a UV monitor (Model UA-2, Instrumentation Specialties Co., Inc., Lincoln, Neb.), were collected, pooled, and precipitated with an equal volume of cold 10% trichloroacetic acid. The precipitates were col- lected by membrane filtration, (0.45 HA, Millipore Corp., Bedford, Mass.), rinsed with cold 5% trichloroacetic acid, dried, and immersed in 10 m1 of toluene base containing 0.4% 2,5-diphenyloxazole (PPO) and 0.005% 1,4-di-2-(5-phenyloxazolyl)-benzene(dimethyl POPOP). Radio- activity was determined in a Packard Tricarb scintillation counter. Immunodiffusion. Samples of pesticin were diffused in gel against a whole serum to E. pestis as previously described (14). RESULTS Purification. Pesticin from uninduced and induced cells was puri- fied as described in Materials and Methods. Preparations of uninduced cells yielded a single major fraction during the final step of chromato- graphy on calcium hydroxyapatite. However, a second fraction of activity appeared when preparations of induced cells were carried through the same process of purification (Fig. 1). The first fraction, termed u-pesticin, was evidently common to uninduced and induced organ- isms whereas only induced cells expressed significant concentrations of the second fraction, termed B-pesticin. The specific activities of both preparations against indicator cells of E. pseudotuberculosis were 18,000 units per mg of protein. Homogeneity. Samples of o- and B-pesticin yielded single bands which migrated together during electrophoresis in 8.0% acrylamide gels (Fig. 2). Similarly, single symmetrical Schlieren patterns were ob- tained during high speed velocity sedimentation (Fig. 3,4). Plots of fringe displacement versus the square of the radius of rotation during high speed equilibrium centrifugation were linear except for slight upward curvature witht3-pesticin due, presumably, to the presence of a minor contaminant of low molecular weight (Fig. 5). The preparations exhibited single bands of precipitate when diffused in gels against a complete antiserum (ggg_Eglgy). Stability. Preparations of this degree of purity were stable for no longer than a month in 0.15 M sodium phOSphate buffer, pH 7.0, at -20 C. Longer storage resulted in appearance of an additional band in polyacrylamide gels which was probably identical to the y-form (ggg below). 35 36 Spectra and composition. Absorption spectra of a- and B-pesticin were typical of simple proteins. At pH 7.0, maximum absorption for both preparations occurred at 278 nm and the absorption ratio (280 nm to 260 nm) was 1.92 and 1.94 for the a- andfg- forms, respectively. The preparations contained approximately 1% carbohydrate (as hexose) and lacked detectable hexosamine. No attempt was made to detect metals. Sedimentation coefficients. Using an averaged partial specific volume of 0.74 cm3/gm calculated from the amino acid composition (£23 E2193), an §20,w of 4.32, 4.36, and 4.37 was calculated for u-pesticin (uninduced), o-pesticin (induced), and typesticin (induced). Molecular weights. Molecular weights determined from sedimentation coefficients approximated 62,500. Fringe displacements obtained during sedimentation to equilibrium were plotted against the square of the distance of rotation (Fig. 5). Weight-average values calculated from these determinations were 66,000, 65,000, and 66,000 for a-pesticin (uninduced), o-pesticin (induced), and B-pesticin (induced), respect- ively. A molecular weight of 63,000 was determined for both o- and B-pesticin by gel filtration in guanidine hydrochloride (Fig. 6). Subunit composition. As shown in Fig. 7, a-pesticin migrated in SDS-polyacrylamide gel as a single band at a rate indistinguishable from that of the d-subunit of E. putida RNA polymerase (MW 44,000). In a separate experiment, both o- and B-pesticin migrated as a single band at rates identical to that of ovalbumin (MW 45,000). The para- meters described above are summarized in Table 1. Amino acid analysis. Results obtained from analyses of a- and B-pesticin are shown in Table 2. The preparations exhibited similar content of amino acids and lacked detectable cysteine. 37 Peptide analysis. No distinctions in peptide composition was detected between preparations of a- and B-pesticin (Fig. 8,9). At least 36 common peptides were present in tryptic digests. Isoelectric focusing. The isoelectric points of a- and B—pesticin were 5.49 and 5.87, respectively, as determined from the data shown in Fig. 10. Interconversion Eg'ylggg. A mixture of a- and B-pesticin was mixed, as carrier, with homogenous 14C-a-pesticin or 14C-B- pesticin. Upon rechromatography after incubation at 26 C for 1, 3, and 7 days, interconversion between a- and B-pesticin was observed with equilibrium favoring formation of the former (Table 3). However, the majority of radioactivity was converted to a new form, termed Y—pesticin, which appeared as a distinct fraction on calcium hydroxyapatite immediately after B-pesticin had been eluted. This third fraction arose with greater facility from B—pesticin than from o-pesticin and was devoid of antibacterial activity. Immunodiffusion. Homogenous preparations of a-, Br, and Y-pesticin exhibited a reaction of identity when diffused against antiserum in gel (Fig. 11). DISCUSSION We previously reported a simple procedure which permitted recovery of high yields of homogenous pesticin from extracts of normal yersiniae (14). Use of a minor modification of this method with extracts of mitomycin C—induced culls resulted in detection of a new fraction of activity (fipesticin) which remained associated with the apparent con- stitutive fraction (u-pesticin) until the final step of purification on calcium hydroxyapatite. These 2 fractions differed in isoelectric point but exhibited similar or identical antibacterial spectra, spe- cific activities, sedimentation coefficients, subunit structure, molecular weight, amino acid ratios, peptide composition, and antigenic properties. In addition, the 2 molecules were interconvertable lg XIEEQ upon prolonged incubation at 26 C although considerable activity was lost in the form of a third form which was biologically inactive (y-pesticin). These observations indicate that the difference between a- and f}pesticin does not involve differences in primary structure, subunit aggregation, or epigenetic modification. The 2 activities evidently reflect alternative configurations or "conformer" states (16) of the same protein. A similar phenomenon has been described for other proteins (6,16,17) including colicin E2 (13) and E3 (10). The latter, like pes- ticin, are deficient in cysteine and thus unable to form disulfide bonds which can stabilize tertiary structure. This amino acid is present in colicin D (25), but not colicins Ia or 1b (17), none of which have been reported to exist as conformers. The proportion of hydrophobic amino acids in pesticin was higher than that in the extracellular colicins E2 and E3 and more typical of that in the intracellular colicins Ia, 38 39 Ib, and D. Pesticin contained a high proportion of aspartic acid and low proportion of alanine relative to the colicins noted above. Although results obtained by guanidine hydrochloride-gel filtra- tion and SDS-polyacrylamide gel electrOphoresis clearly showed that pesticin consists of a single polypeptide chain, the molecular weight determined by the latter method was significantly lower than that ob- tained by other procedures. ElectrOphoresis in SDS-gels may yield erroneous results for proteins of low molecular weight (7). For ex- ample, the molecular weight of the subunit of staphylococcal lactose- specific phosphocarrier protein was 11,600 when determined by gel filtration in 6 M guanidine hydrochloride and 9,200 when assayed by gel electrophoresis in 1% SDS (11). The anomalous behavior of pesti- cin (MW 65,000) during electrophoresis in SDS is unusual and suggests that portions of the molecule remain stabilized in this detergent. The fact that significant concentrations of B-pesticin were only obtained following induction with mitomycin C is curious. It is con- ceivable that the physiological conditions of uninduced cells favor maintenance of the a-conformer whereas environmental changes within induced organisms permit accumulation of the B-conformer. Little is known about the effect of ionic strength and pH on pesticin although they can influence the equilibrium.lgfly$££g of conformers of other proteins (5,16). Alternatively, expression of the B-conformer may reflect a less probable variation of folding which does not become evident until large quantities of pesticin are synthesized following induction. Evidence favoring this hypothesis was the recovery of a minor fraction, now recognized as B-pesticin, from a crude extract of uninduced cells following direct chromatography on calcium hydroxyapatite 40 (14). This fraction is lost, possibly due to conversion to o- or Y-pesticin, when such extracts are subjected to the complete process of purification. A full assessment of the differences between the con- former states of pesticin may depend upon a better understanding of the mechanism of calcium hydroxyapatite chromatography, a process which may not depend exclusively upon electrostatic forces (9). LITERATURE CITED Beavan, G. H., and E. R. Holiday. 1952. Ultraviolet absorption spectra of proteins and amino acids. Advan. Protein Chem. ‘Z:319-386. Ben-Gurion, R., and Hertman, I. 1958. Bacteriocin-like material pro- duced by Pasteurella pestis. J. Gen. Microbiol. 125289-297. Boas, N. F. 1953. Method for the determination of hexosamine in tissues. J. Biol. Chem. 293:553-563. Brubaker, R. R. 1972. The genus Yersinia: biochemistry and genet- ics of virulence. Cur. TOp. Microbiol. Immun. _l;111-158. Carpenter, F. H. 1967. Treatment of trypsin with TPCK. p. 237. E3 Hirs, C. H. W. (ed.), Methods in Enzymology, Vol. XI. Academic Press Inc., New York. Citri, N., N. Garber, and M. Sela. 1960. The effect of urea and guanidine hydrochloride on activity and optical rotation of penicillinase. J. Biol. Chem. 222:3454-3459. Dunker, A. K. and R. R. Rueckert. 1969. Observations on molecular weight determinations on polyacrylamide gel. J. Biol. Chem. Eflfiz5074-5080. Fish, W. W., K. G. Mann, and C. Tanford. 1969. The estimation of polypeptide chain molecular weights by gel filtration in 6 M guanidine hydrochloride. J. Biol. Chem. Eggz4989-4994. Ginsberg, H. S. 1969. Adsorption chromatography for isolation of viral proteins: The use of substituted cellulose and calcium phosphate. p. 390-398. 13_Habel, K. and N. P. Salzman (ed.), Fundamental Techniques in Virology. Academic Press Inc., New York. 41 10. 11. 12. 13. 14. 15. 16. 17. 18. 42 Click, J. M., S. J. Kerr, A. M. Gold, and D. Shemin. 1972. Multiple forms of colicin E3 from Escherichia coli CA-38 (colE Biochemistry 11:1183-1188. 3; ColI). Hays, J. B., R. D. Simon, and S. Roseman. 1973. Sugar transport: a trimeric lactose-specific phosphocarrier protein of the Staphy- lococcus aureus phosphotransferase system. J. Biol. Chem. Efi§z94l-956. Helmkamp, R. W., R. L. Goodland, W. F. Bale, I. L. Spar, and L. E. Mutschler. 1960. High Specific activity iodination of l-globu- lin with iodine-13l-monochloride. Cancer Res. EQ;1495-1500. Herschman, H. R., and D. R. Helinski. 1967. Purification and characterization of colicin E2 and E3. J. Biol. Chem. EQE:5360- 5368. Hu, P. C., G. C. H. Yang, and R. R. Brubaker. 1972. Specificity, induction, and absorption of pesticin. J. Bacteriol. _lE:212-219. Johnson, J. C., M. DeBacker, and J. A. Boezi. 1970. Deoxyribo- nucleic acid-dependent ribonucleic acid polymerase of Pseudomonas putida. J. Biol. Chem. ggg;1222-1232. Jolly, R. L., and H. S. Mason. 1965. The multiple forms of mush- room tyrosinase. J. Biol. Chem. EEQ:PCl489-l49l. Kitto, G. B., P. M. Wassarman, and N. O. Kaplin. 1966. Enzymatically active conformers of mitochondrial malate dehydrogenase. Proc. Natl. Acad. Sci. Wash. 41:1498-1500. Konisky, J. 1972. Characterization of colicin Ia and colicin Ib: chemical studies of protein structure. J. Biol. Chem. 247:3750- 3755. 19. 20. 21. 22. 23. 24. 25. 26. 27. 43 Massey, T., J. Behnke, S. Dagger, and W. Deal. 1973. Extremely rapid isoelectric focusing in a Simple inexpensive apparatus: applications to eight proteins. Biochem. Biophys. Res. Communs. In press. Moore, 3., and W. H. Stein. 1963. Chromatographic determination of amino acids by the use of automatic recording equipment. p. 819- 831. 13 Colowick, S. P. and N. 0. Kaplan (ed.), Methods in Enzymology, Vol. VI. Academic Press Inc., New York. Robertson, D. C., R. H. Hammerstedt, and W. A. Wood. 1971. Structure of 2-keto-3-deoxy-6-phosphogluconate aldolase. J. Biol. Chem. 239:2075-2083. Roe, J. H. 1965. The determination of sugar in blood and Spinal fluid with anthrone reagent. J. Biol. Chem. EEE:335-343. Shapiro, A. L., E. Vinuela, and J. A. Maizel. 1967. Molecular weight estimation of polypeptide chains by electrophoresis in SDS-polyacrylamide gels. Biochem. BiOphyS. Res. Commun. E§,815- 820. Schachman, H. K. 1957. Ultracentrifugation, diffusion, and vis- cometry, p. 32-103. Eg_Colowick, S. P., and N. 0. Kaplan (ed.), Methods in Enzymology, Vol. IV. Academic Press Inc., New York. Timmis, K. 1972. Purification and characterization of colicin D. J. Bacteriol. 192:12-20. Yang, G. C. H., and R. R. Brubaker. 1971. Effect of Ca+2 on the synthesis of deoxyribonucleic acid in virulent and avirulent Yersinia. Inf. Immun. ‘E:59-65. thantis, D. 1964. Equilibrium ultracentrifugation of dilute solutions. Biochemistry E:297-317. TABLE 1. MOLECULAR WEIGHT OF PESTICIN Uninduced Induced Pesticin pesticin a-pesticin B-pesticin Native Estimated by equilibrium sedimentation 66,000 65,000 66,000 Estimated from E20,w 62,500 62,500 62,500 Dissociated Estimated by gel filtration a in 6 M guanidine hydrochloride --- 63,000 63,000 Estimated by SDS-acrylamide gel filtration 44,000 44,000 44,000 8Not determined 45 TABLE 2. AMINO ACID COMPOSITIONS OF PESTICINa Uninduced Induced by Mitomycin C Amino Acid pesticin a-pesticin B-pesticin Aspartic Acid 125.6 136.0 140.4 Threonine 47.6 43.2 41.2 Serine 69.2 69.2 68.4 Glutamic acid 62.4 80.8 72.4 Proline 33.9 21.0 30.8 Glycine 65.6 70.0 64.8 Alanine 36.2 40.8 36.3 Valine 51.2 53.6 49.6 Methionine 14.8 18.6 15.8 Isoleucine 50.0 55.2 50.0 Leucine 64.4 71.6 68.8 Tyrosine 29.6 33.1 30.16 Phenylalanine 42.0 43.6 40.8 Lysine 51.0 52.8 50.8 Histidine 13.2 13.8 14.5 Arginine 46.0 48.8 45.6 TryptOphanb 4.0 4.5 4.2 anMoleS per 100 Hg of protein. bDetermined by alkaline spectrophotometry. 46 .wcauouacoa uoHoa>muuas he manouoouoo uozw o.maus.~mno.sN onw.m ooa.~ oo~.~ ooo.mH --- same a m.~m N.mm o.“ oom.m ooe.a cos ooo.mH --- mama m e.eauo.owuo.m ooe.~ coo.HH omm coo.mH --- use H m.mmum.m "m.mm omm.e one ooo.~ --- ooo.nH same a a.m~”N.N "m.mk ooo.s omm ooe.au --- coo.mH mama m w.o u~.aa a--- ooH oom.~H --- coo.mH Ame H A>.m.av cwoaumoau» cauaumoaum uaoaumoouo afloaumoaum afloaumoauu . . . 0 em on hua>auomoavmu Asmov AEQUV coaquSQca mo oEHH moum>ooou mo caumm muw>auomofivmu woum>oomm Aua>auomoaemu Hmauaao ZHUHHmmm mo ZOHmmm>zoummHzH .m mdm.“qu ~53”.on .‘3xou on“. ca monfluommo mum mCOgufivcoo HouGoEHuoaxm .Gaofiumoaum was aaofiumoauo wo wofimooowouuooam Gama tmnEsz c269... 66 Om ON 0_ CM ON 0_ _ L _ _i4.lfll|l Null Ll senate t 5238.... _ _ _ e _ _ 9 co to ADV com .ofioaumoaum Amv .awofiummauo A