/ . Biochem. 83, 727-736 (1978)
Bacteriolytic Enzyme Induced from Pyocinogenic Pseudomonas aeruginosa Purification and Characterization of PRl-Lysozyme 1 Norimichi OCHI, Masao AZEGAMI, and Shin-ichi ISHH Department of Biochemistry, Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-ku, Sapporo, Hokkaido 060 Received for publication, August 31, 1977
A bacteriolytic enzyme, PRl-lysozyme, has been purified from the lysate of mitomycin Cinduced pyocinogenic Pseudomonas aeruginosa, by acrinol treatment, Amberlite CG-50 chromatography, ammonium sulfate fractionation, Sephadex G-100 gel filtration and two cycles of SP-Sephadex C-50 chromatography. Homogeneity of the preparation was demonstrated by three electrophoretic techniques. PRl-lysozyme is a basic protein (pi, 9.4) and consists of a single polypeptide chain having a molecular weight of 24,000. The amino acid composition of the protein was analyzed, and no cysteine residue was found among more than 210 amino acid residues. The optimum pH for enzymatic activity was 6.4 and the enzyme exhibited about 50 to 70 times greater specific activity than hen egg-white lysozyme when assayed with chloroform-killed P. aeruginosa as a substrate. By analyzing the products of enzymatic action on purified peptidoglycan of P. aeruginosa, the enzyme was identified as an ,/V-acetylmuramidase, i.e., the same classification as hen egg-white lysozyme. PRl-lysozyme did not show any activity towards intact cells of gram-positive and gram-negative bacteria tested. However, the enzyme was able to lyse chloroform-killed gram-negative and grampositive bacteria.
Although many bacteriophage lytic enzyme systems have been investigated (1-10), few studies have been reported on the lytic enzymes produced by bacteriocinogenic bacteria which produce phagetail-like bacteriocin (IT). The fact that pyocin1
This work was supported in part by a grant for
ogenic Pseudomonas aeruginosa .PI 5 can produce a bacteriolytic enzyme together with pyocin Rl is interesting in view of the evolutionary relationship with the bacteriophage lytic enzyme systems. It seemed valuable to clarify the physico-chemical and biological characteristics of PRl-lysozyme by using a purified enzyme preparation for comparison with
^UTuHuVonapt Abbreviations: GIcNAc, W-acetylglucosamine; MurNAc, W-acetylmuramic acid; A,pm, /n«0-diaminopimelic acid; BSA, bovine serum albumin; MES 2(A'-morpholino)ethanesulfonic acid; BES, N,N-b\s(2hydroxyethyl>2-aminoethanesulfonic acid; SDS, sodium dodecyl sulfate; BPB, bromphenol blue. Vol. 83, No. 3, 1978
727
organisms. The purposes of the work described in this article were first, to establish a method of purification, and second, to characterize the enzyme , . . . „ . . -. • 8 at 0.5 mA/gel (0.5x8 cm) until pyronin G marker dye li i entered the resolving gel, and at 1.5 mA/gel until the < ~ U marker reached 1 cm from the bottom of the gel. After * Hojf -50 / - 50 % electrophoresis, gels were fixed in 10% trichloroacetic acid, stained with Coomassie brilliant blue R-250 and destained in a mixture of methanol, acetic acid, and 1 water ( 1 : 1 : 8 , v/v/v). SDS-gel electrophoresis was ^t—•—• -0 0 m performed at 3 mA/gel (0.5 x 7 cm) until bromphenol 10 20 30 40 blue (BPB) entered the gel, and then at 6 mA/gel for Fraction Number Fig. 1. Second SP-Sephadex C-50 chromatography of 3.5 h. Before electrophoresis, a lyophilized sample was PRl-lysozyme. Fraction VI was adsorbed on an SP- denatured in 0.01 M sodium phosphate buffer, pH7.1, Sephadex C-50 column (1.3x7 cm) equilibrated with containing 1% SDS and 5% 2-mercaptoethanol in a 0.05 M potassium phosphate buffer, pH 6.0. Elution boiling-water bath for 5 min. Staining and destaining was carried out at 4°C with a linear gradient of pH were carried out as described above. Isoelectric focus(6.0 to 7.2) and concentration of potassium phosphate ing was carried out at 100 volts for 1 h and then at 150 (0.05 to 0.175 M). Theflowrate was 20 ml/h, and 3.0 ml volts for 1.5 h. After electrophoresis, the gels (0.5x8 fractions were collected. , Absorbance at 280 ran; cm) were fixed in 3.5% perchloric acid for 30 min at • , lytic units per ml; , phosphate concentration; 45°C, stained with Coomassie brilliant blue G-250 in the same solvent. x x x , specific activity (lytic u
Ml
-
f\ ur 1 t
E
ty(unitsi
Is
TABLE I. Summary of the purification of PRl-lysozyme. Fraction I II m IV V VI Vn
Lysate Acrinol supernatant Amberlite CG-50 ( N H J ^ 45-75% Sephadex G-100 SP-Sephadex C-50 SP-Sephadex C-50
Volume (liters)
Total
100 135 6.6 0.15 0.57 0.115 0.033
— — 2,185 383 128 34.5 28.4
Total units 12,090 12,950 10,670 6,600 4,990 2,730 2,150
Yield (%) — — 4.89 17.3 39.0 78.3 75.7
100
107 88.3 54.6 41.3 22.6 17.8
J. Biochem.
PURIFICATION AND CHARACTERIZATION OF PR1-LYSOZYME electrophoresis at p H 4 . 3 , electrophoresis in 0 . 1 % SDS and analytical gel isoelectric focusing in a p H gradient from 3 to 10. The results indicate that the preparation was electrophoretically homogeneous. //. Characterization of PRl-Lysozyme—Molecular weight and isoelectric point: By slab gel electrophoresis in 0.1 % SDS, the molecular weight of the enzyme was estimated to be 24,000, based on a calibration curve obtained by using five proteins of known molecular weight as standards. The purified enzyme preparation also exhibited homogeneity on gel filtration through Sephadex G-100 (2.2 x 90 cm column; 0.05 M potassium phosphate buffer, p H 6.0, containing 0.15 M NaCl). A slightly higher
731
molecular weight value of 26,000 to 27,000 was obtained by the method of Andrew (27). The isoelectric point of the enzyme was estimated to be 9.4 by analytical gel isoelectric focusing with 2 % Ampholine, p H range 3 to 11. The results indicate that PRl-lysozyme consists of a single polypeptide chain and is a basic protein, as is hen egg-white lysozyme or T4-lysozyme, even though the molecular weight is higher than those of the latter lysozymes. Amino acid composition of PRl-lysozyme: Table H shows the amino acid composition of PRl-lysozyme. The values listed are presented in terms of molar ratios on the basis of 34 mol of alanine residues per mol of the enzyme. From
TABLE II. Amino acid composition of PRl-lysozyme. Samples were hydrolyzed with 6 N H O in evacuated, sealed tubes at 110°C for 24, 48, and 86 h, and with 4 N methanesulfonic acid at 115°C for 24 h. Values listed are expressed in terms of molar ratios, on the basis of 34 alanine residues per molecule of the enzyme. Values for serine and threonine were obtained by extrapolation to zero time of hydrolysis, and the maximum values were taken for valtne and isoleucine. Residues/Molecule Amino acid
Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Half-cystine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Tryptophan
24 h
Hydrolysis for 48 h 86 h
18.4 7.24 6.64 29.7 9.96 19.0 34.0 7.98 1.24 0 5.90 24.5 3.82 4.96 1.83 2.13 21.8 —
19.3 7.88 6.48 29.3 10.7 19.9 34.0 8.27 1.15 0 6.09 25.4 3.99 5.03 1.99 2.34 22.6 —
18.6 7.18 5.80 29.0 10.1 19.4 34.0 8.73 1.14 0 6.28 24.1 4.42 4.95 1.83 2.33 21.4 —
Total * Values obtained after hydrolysis with 4 N methanesulfonic acid. Vol. 83, No. 3, 1978
24 h» 19.2 7.83 6.68 30.9 11.0 20.3 34.0 8.32 1.71 0 6.39 25.5 4.41 5.04 2.06 2.01 23.0 4.86
Nearest integer
T4-Lysozyme (24)
19 8 7 30 10-11 20 34 9 2 0 7 25-26 5 5 2 2
22 11 6 13 3 11 15 9 5 2 10 16 6 5 1 13 13 3
23 5 213-215
164
732
the sum of the integer values, the total number of amino acid residues per molecule of the enzyme is estimated to be 213 to 215. From these values, the molecular weight of the enzyme was calculated to be 23,400 to 23,600, which is compatible with the value 24,000 determined by gel electrophoresis rather than the value obtained by gel filtration. Table II also indicates that there was no cysteine residue among more than 210 amino acid residues, and the number of basic amino acid residues was 27, accounting for more than 12% of the total number of residues. Glucosamine, galactosamine, and other amino sugars were not detected. Effect of pH on enzyme activity: The activity of PRl-lysozyme at various pH values was measured in potassium phosphate buffer (pH 5.0 to 8.0), Tris-HCl buffer (pH 7.5-9.0), 2-4) glycosidic linkages between TV-acetylmuramic acid and iV-acetylglucosamine ia
100
SO
v ^
\
1 10 : *
5
:
\
\
A
I
30 40 50 60 Temperature (°C)
00
Fig. 5. Stability of the enzyme to heat. The enzyme (0.32 units/ml) was incubated in 0.05 M potassium phosphate buffer, pH 6.5 containing 0.01% BSA for: • , 5min; O, lOmin; A, 20min; A, 60min at the indicated temperatures. After incubation, samples were chilled to 0°C and the residual activity was measured. Each activity is expressed as a percentage of the activity of a control sample which had been incubated at 0°C for the same time.
20
40 60 80 100 120 Incubation Time(min)
Fig. 6. Mode of action of PRl-lysozyme. Peptidoglycan (1.5 mg) was hydrolyzed with PRl-lysozyme (0.14 fig) in 1.7 ml of 0.02 M sodium acetate buffer, pH 6.0, at 37°C. At fixed intervals, 50 fi\ samples were withdrawn, and assayed for the indicated characteristics. Reducing groups and free amino groups were determined by the method of Park and Johnson (18) and by ninhydrin colorimetry (19), respectively. • , Reducing groups; O, free amino groups.
TABLE 111. Specificity of PRl-lysozyme. Peptidoglycan (1.5 mg) was incubated with or without 0.2 fig of PRllysozyme for 3 h at 37°C in 1.7 ml of 0.02 M sodium acetate buffer, pH 6.0. Digestion with 10 fig of hen egg-white lysozyme was also carried out in the same manner. The reaction mixtures were incubated with 2 mg of sodium borohydride at about pH 10 for another 48 h at 37°C. After reduction, the samples were dried and hydrolyzed with 6 N HC1. The hydrolysates were analyzed with an automatic amino acid analyzer. Other experimental details are given in the text. /imol/mg of peptidoglycan (molar ratio1)
Compound
Without enzyme Muramic acid Muramicitol Glucosamine Glucosaminitol Glutamic acid Alanine Diaminopimelic acid » Glutamic acid is taken as 100. Vol. 83, No. 3, 1978
0.71 0.05 0.81 0.05 0.95 1.53 0.86
(75) (5.3) (85) (5.3) (100) (161) (91)
Hen egg-white lysozyme 0.11 0.69 0.75 0.06 0.89 1.45
(12) (76) (84) (6.7) (100) (163)
0.86
(97)
PRl-lysozyme 0.03 0.82
(3.4) (94)
0.78 0.05 0.87 1.49 0.89
(90) (5.7) (100) (170) (102)
734
N. OCHT, M. AZEGAMI, and S. ISHII strates (Table IV). Values listed are expressed in terms of specific activity (units/^jjo), and values obtained with hen egg-white lysozyme are also listed for comparison. Neither PRl-lysozyme nor hen egg-white lysozyme showed any activity toward intact cells of gram-negative bacteria, while both
the glycan chain of peptidoglycan, as illustrated in Fig. 7. Bacteriolytic spectrum of PRl-lysozyme: The bacteriolytic spectrum of PRl-lysozyme was investigated by using intact, EDTA-treated, and chloroform-killed cells of various bacteria as sub-
PRl-lysozyme, T2, T4 6
1
)
' "^"
G1 cNAc
Mu r NAc
T3, T5, T7
G1 cNAc —
Vi-III. HM7
D-Glu
I I
L-Ala MurNAc PRl-lysozyme, T 2 , T4-
Fig. 7. Sites of peptidoglycan attacked by PRl-lysozyme and by other phage lytic enzymes. A part of the structure of peptidoglycan of gram-negative bacteria is illustrated. TABLE IV. Bacteriolytic spectrum of PRl-lysozyme. Bacteria listed in the table were grown in Nutrient BrothNaCl medium (Nutrient Broth 15 g, NaCl 5 g/liter) and collected at the late log phase (about 2x10* cells/ml) by centrifugation. Intact, EDTA-treated, and chloroform-killed cells of each bacterium were prepared as substrates. Intact and EDTA-treated cells were prepared by two cycles of suspension and centrifugation in 0.05 M potassium phosphate buffer, pH 6.4, and 0.05 M potassium phosphate buffer, pH 6.4, containing 5 mM EDTA, respectively. Chloroform-killed cells were obtained as described by Black and Hogness (2). Activities of enzymes were 0.31 units/ml and 0.36 units/ml for PRl-lysozyme and hen egg-white lysozyme, respectively. Specific activity (units//* Hen egg-white lysozyme EDTAtreated cells
Chloroformkilled cells
0 0 0 0
0.32 0.35 0.23 0.21
1.66 1.82 1.75 1.65
0
0.35
1.47
2.59 2.38
2.49 2.26
2.64 2.35
Intact cells Gram-negative bacteria Escherichia coll C Escherichia coli K12W6 Pscudomonas aeruginosa PI 5 Pseudomonas aeruginosa PI 1 Klebsiella pneumoniae Salmonella typhimurium Aerobacter aerogerus Serratia marcescens Proteus vulgaris Gram-positive bacteria Micrococcus lysodeikticus Staphyhcoccus aureus Bacillus subtilis
PRl-lysozyme Intact cells
EDTAtreated cells
Chloroformkilled cells
0 0 0 0 0 0 0 0 0
5.36 6.90 6.82 5.36
68.2 79.3 69.6 64.0 58.5 71.0 61.2 50.1 54.3
0 0
0 0
5.64 4.04 3.83 4.38
32.0 20.2 13.9
/ . Biochem.
PURIFICATION AND CHARACTERIZATION OF PR1-LYSOZYME
735
the enzymes were able to lyse chloroform-killed the hypothesis that these enzymes recognize not cells of these bacteria. However, hen egg-white only the sites which they cleave, but also structures lysozyme lysed intact cells of gram-positive bac- near them. teria such as M. lysodeikticus and S. aureus. These Composition analyses of PRl-lysozyme findings indicate that there is a distinct difference showed the absence of cysteine and cystine residues. in affinity for bacterial cells between hen egg-white T2-(23), T4-lysozyme (24), lambda endolysin (25), lysozyme and PRl-lysozyme, though the two en- and N20F' lytic enzyme (4) also contain no disulfide zymes have the same specificity against the pep- bond in their molecules. The lack of disulfide tidoglycan of P. aeruginosa. bonds may be a common feature of lytic enzymes related to bacteriophages and bacteriocins. This would distinguish them from bacterial or fungal DISCUSSION lytic enzymes, which are known to have one disulPRl-lysozyme was purified to homogeneity from fide bond in the molecule (26, 27), and from the a mitomycin C-induced lysate of pyocinogenic enzymes of higher organisms, such as hen eggP. aeruginosa by a reasonably convenient proce- white lysozyme, which has four disulfide bonds dure. PRl-lysozyme was shown to be an N- (28). The lack of disulfide bonds may partly acetylmuramidase, as is hen egg-white lysozyme, account for the heat lability of PRl-lysozyme. since both enzymes hydrolyzed ./V-acetylmuramide Our preliminary experiments suggest that the NH,linkages only, leaving N-acetylglucosaminide link- terminus and COOH-terminus of the enzyme ages intact, with peptidoglycan of P. aeruginosa molecule were methionine and leucine, as deteras a substrate. However, PRl-lysozyme hydro- mined by the dansylation (29) and carboxypeptidase lyzed Af-acetylmuramide linkages almost com- Y methods (30), respectively. If this is the case, pletely in the peptidoglycan, while the extent of PRl-lysozyme has the same terminal amino acid hydrolysis by hen egg-white lysozyme only reached residues as T4-lysozyme (24). almost 90% even though the amount was increased It is of interest to compare the lysozymes of up to 50 times that of the former enzyme (Table higher organisms and lytic enzymes related to ni). This superiority of PRl-lysozyme over hen phages and phage-tail-like bacteriocins as regards egg-white lysozyme was also observed when the structural differences and biological significance. activities of the two enzymes were compared with Such differences may be a result of evolutionary chloroform-killed gram-negative bacteria as sub- selective pressures related to enzyme stability, or strates by turbidimetric measurement (Table IV). the rate of enzyme synthesis, or both. Rapid Presumably, the chloroform treatment removes the synthesis of lytic enzyme could be advantageous lipid components, at least in part, from the bac- to phage or bacteriocin reproduction, because lysis terial cell surface and makes the cell wall susceptible of the host cells must occur within a short time, to attack by lytic enzymes. However, it is note- whereas a high degree of stability might not be worthy that hen egg-white lysozyme is able to lyse significant. It is known that T2- and T4-lysozymes gram-positive bacteria such as M. lysodeikticus and act upon the host cell from inside the cell (31) and S. aureus. Thesefindingsmay reflect differences of this may also be the case with PRl-lysozyme, peptidoglycan structure, especially that of the since the only biological function for these enzymes peptide portion, or differences of components on may be to lyse host cells from inside to release the the cell surface, between gram-positive and gram- newly formed phages or bacteriocins. In the case negative bacteria. In addition, Arima et al. (22) of higher organisms, the opposite situation may have reported that P. aeruginosa phage 95-induced exist, i.e., high stability is a significant factor and lytic enzyme, LE95, is an L-alanyl-D-glutamic acid rapid formation is not particularly relevant. Lysoendopeptidase and does not hydrolyze various zymes of higher organism may have to act against small fragments of peptidoglycan even though they bacteria which have invaded the organism, from contain the peptide moiety attacked by LE95, outside the bacterial cell. suggesting that the presence of the glycan chain In the present experiments, the number of of the peptidoglycan is essential for enzymatic PRl-lysozyme molecules produced from a single activity. These observations may be explained by cell appeared to be about 4x10*. This number Vol. 83, No. 3, 1978
736
N. OCHI, M. AZEGAMI, and S. ISHH
9. Jastrzemski, K.B. (1975) Ada Biochim. Pol. 22, 297-304 10. Taylor, A. (1971) Nature New Biol. 234, 144-145 11. Kato, F., Ogata, S., & Hongo, M. (1976) Agric. Biol. Chem. 40, 1101-1105 12. Yanai, A., Kato, K., Beppu, T., & Arima, K. (1976) Agric. Biol. Chem. 40, 1505-1508 13. Ornstein, L. (1964) Ann. N.Y. Acad. Sci. 121, 321349 14. Davis, B.J. (1964) Ann. N.Y. Acad. Sci. 121, 404-^27 15. Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, 4406-4412 16. Wrigley, C.W. (1971) in Methods in Enzymology (Jakoby, W.B., ed.) Vol. 22, pp. 559-564, Academic Press, New York 17. Simpson, R.J., Neuberger, M.R., & Liu, T.Y. (1976) J. Biol. Chem. 251, 1936-1940 18. Park, J.T. & Johnson, M.J. (1949) / . Biol. Chem. 181, 149-151 19. Yemm, E.W. & Cocking, E.C. (1955) Analyst 80, 209-213 20. Ghuysen, J.M., Tipper, D.J., & Strominger, J.L. (1966) in Methods in Enzymology (Neufeld, E.F. & Ginsburg, V., eds.) Vol. 8, pp. 685-711, Academic Press, New York 21. Andrew, P. (1964) Biochem. J. 91, 222-233 22. Yanai, A., Kato, K., Beppu, T., & Arima, K. (1976) Biochem. Biophys. Res. Commun. 68, 1146-1152 23. Inouye, M. & Tsugita, A. (1968) / . Mol. Biol. 37, 213-223 24. Tsugita, A. & Inouye, M. (1968) J. Mol. Biol. 37, 201-212 25. Black, L.W. & Hogness, D.S. (1969) J. Biol. Chem. 244, 1976-1981 26. Shih, J.W. & Hash, J.H. (1971) J. Biol. Chem. 246, 994-1006 27. Jackson, R.L. & Wolfe, R.S. (1968) J. Biol. Chem. 243, 879-888 We wish to thank Miss M. Fujimoto for her cooperation 28. Imoto, T.. Johnson, L.N., North, A.C.T., Phillips, D.C., & Rupley, J.A. (1972) in The Enzymes (Boyer; in amino acid analyses, and Dr. R. Tirrell for his advice P.D., ed.) Vol. 7, pp. 665-868, Academic Press, during the preparation of this paper. New York 29. Hartley, B.S. (1970) Biochem. J. 119, 805-822 30. Hayashi, R., Moore, S., & Stein, W.H. (1973) REFERENCES /. Biol. Chem. 248, 2296-2302 1. Tsugita, A., Inouye, M., Terzaghi, E., & Streisinger, 31. Emrich, J. & Streisinger, G. (1968) Virology 36, G. (1968) J. Biol. Chem. 243, 391-397 387-391 2. Black, L.W. & Hogness, D.S. (1969) / . Biol. Chem. 32. Kageyama, M. (1964) / . Biochem. 55, 49-53 244,1968-1975 33. Jshii, S., Nishi, Y., & Egami, F. (1965) / . Mol. 3. Taylor, A. (1970) Biochem. Biophys. Res. Commim. Biol. 13,428-431 41, 16-24 34. Ito, S. & Kageyama, M. (1970) / . Gen. Appl. 4. Moo-Penn, W., Mowry, C , & Wiesmeyer, H. (1969) Microbiol. 16, 231-240 Biochim. Biophys. Ada 178, 330-346 35. Kageyama, M. (1974) / . Gen. Appl. Microbiol. 20, 5. Arima, K., Yanai, A., & Bcppu, T. (1976) Agric. 269-275 Biol. Chem. 40, 1301-1306 . 36. Kageyama, M. (1970) / . Gen. Appl. Microbiol. 16, •6. Inouye, M., Arnheim, N., & Sternglanz, R. (1973) 523-530 J. Biol. Chem. 248, 7247-7252 37. Kageyama, M. & Inagaki, A. (1974) / . Gen. Appl. Microbiol. 20, 257-267 "7. Hongo, M.,Tahara, Y., & Ogata, S. (1974) Agric. Biol. Chem. 38, 755-761 8. DeMartini, M., Halegoua, S., & Inouye, M. (1975) / . Virol. 16, 459-461
may be compared with about 100 to 200 pyocin Rl particles produced per induced cell under the same conditions. These numbers are compatible with those of a temperate phage, lambda, in which the number of lambda endolysin molecules per cell is about 6x10*, and that of matured phage particles is about 100 (2). The optimum pH for the activity of PR1lysozyme was found to be near 6.4 with chloroformkilled P. aeruginosa cells as a substrate. However, the pH optimum may reflect not only the characteristics of the enzyme but also the susceptibility of the chloroform-killed cells to the enzyme action. More relevant information on the pH optimum of the enzymatic activity itself may be obtained by using simpler substrates. Pyocin Rl has a close resemblance to some phage tails (32, 33) and may be considered as a defective lysogenic phage (32, 34), although conclusive evidence for this has not yet been presented. The pyocinogenic factors of pyocin Rl, R2, and R3 appear to be integrated into the bacterial chromosome at a corresponding position close to the trp-1 region (35-37). However, it is still questionable whether PRl-lysozyme is the product of a bacterial gene or is derived from genetic information of the pyocinogenic factor. In any case, more precise investigations of phage-tail-like bacteriocin and lytic enzyme systems from a genetic as well as a biochemical viewpoint are required.
/. Biochem.
Bacteriolytic Enzyme Induced from Pyocinogenic Pseudomonas aeruginosa Purification and Characterization of PRl-Lysozyme 1 Norimichi OCHI, Masao AZEGAMI, and Shin-ichi ISHH Department of Biochemistry, Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-ku, Sapporo, Hokkaido 060 Received for publication, August 31, 1977
A bacteriolytic enzyme, PRl-lysozyme, has been purified from the lysate of mitomycin Cinduced pyocinogenic Pseudomonas aeruginosa, by acrinol treatment, Amberlite CG-50 chromatography, ammonium sulfate fractionation, Sephadex G-100 gel filtration and two cycles of SP-Sephadex C-50 chromatography. Homogeneity of the preparation was demonstrated by three electrophoretic techniques. PRl-lysozyme is a basic protein (pi, 9.4) and consists of a single polypeptide chain having a molecular weight of 24,000. The amino acid composition of the protein was analyzed, and no cysteine residue was found among more than 210 amino acid residues. The optimum pH for enzymatic activity was 6.4 and the enzyme exhibited about 50 to 70 times greater specific activity than hen egg-white lysozyme when assayed with chloroform-killed P. aeruginosa as a substrate. By analyzing the products of enzymatic action on purified peptidoglycan of P. aeruginosa, the enzyme was identified as an ,/V-acetylmuramidase, i.e., the same classification as hen egg-white lysozyme. PRl-lysozyme did not show any activity towards intact cells of gram-positive and gram-negative bacteria tested. However, the enzyme was able to lyse chloroform-killed gram-negative and grampositive bacteria.
Although many bacteriophage lytic enzyme systems have been investigated (1-10), few studies have been reported on the lytic enzymes produced by bacteriocinogenic bacteria which produce phagetail-like bacteriocin (IT). The fact that pyocin1
This work was supported in part by a grant for
ogenic Pseudomonas aeruginosa .PI 5 can produce a bacteriolytic enzyme together with pyocin Rl is interesting in view of the evolutionary relationship with the bacteriophage lytic enzyme systems. It seemed valuable to clarify the physico-chemical and biological characteristics of PRl-lysozyme by using a purified enzyme preparation for comparison with
^UTuHuVonapt Abbreviations: GIcNAc, W-acetylglucosamine; MurNAc, W-acetylmuramic acid; A,pm, /n«0-diaminopimelic acid; BSA, bovine serum albumin; MES 2(A'-morpholino)ethanesulfonic acid; BES, N,N-b\s(2hydroxyethyl>2-aminoethanesulfonic acid; SDS, sodium dodecyl sulfate; BPB, bromphenol blue. Vol. 83, No. 3, 1978
727
organisms. The purposes of the work described in this article were first, to establish a method of purification, and second, to characterize the enzyme , . . . „ . . -. • 8 at 0.5 mA/gel (0.5x8 cm) until pyronin G marker dye li i entered the resolving gel, and at 1.5 mA/gel until the < ~ U marker reached 1 cm from the bottom of the gel. After * Hojf -50 / - 50 % electrophoresis, gels were fixed in 10% trichloroacetic acid, stained with Coomassie brilliant blue R-250 and destained in a mixture of methanol, acetic acid, and 1 water ( 1 : 1 : 8 , v/v/v). SDS-gel electrophoresis was ^t—•—• -0 0 m performed at 3 mA/gel (0.5 x 7 cm) until bromphenol 10 20 30 40 blue (BPB) entered the gel, and then at 6 mA/gel for Fraction Number Fig. 1. Second SP-Sephadex C-50 chromatography of 3.5 h. Before electrophoresis, a lyophilized sample was PRl-lysozyme. Fraction VI was adsorbed on an SP- denatured in 0.01 M sodium phosphate buffer, pH7.1, Sephadex C-50 column (1.3x7 cm) equilibrated with containing 1% SDS and 5% 2-mercaptoethanol in a 0.05 M potassium phosphate buffer, pH 6.0. Elution boiling-water bath for 5 min. Staining and destaining was carried out at 4°C with a linear gradient of pH were carried out as described above. Isoelectric focus(6.0 to 7.2) and concentration of potassium phosphate ing was carried out at 100 volts for 1 h and then at 150 (0.05 to 0.175 M). Theflowrate was 20 ml/h, and 3.0 ml volts for 1.5 h. After electrophoresis, the gels (0.5x8 fractions were collected. , Absorbance at 280 ran; cm) were fixed in 3.5% perchloric acid for 30 min at • , lytic units per ml; , phosphate concentration; 45°C, stained with Coomassie brilliant blue G-250 in the same solvent. x x x , specific activity (lytic u
Ml
-
f\ ur 1 t
E
ty(unitsi
Is
TABLE I. Summary of the purification of PRl-lysozyme. Fraction I II m IV V VI Vn
Lysate Acrinol supernatant Amberlite CG-50 ( N H J ^ 45-75% Sephadex G-100 SP-Sephadex C-50 SP-Sephadex C-50
Volume (liters)
Total
100 135 6.6 0.15 0.57 0.115 0.033
— — 2,185 383 128 34.5 28.4
Total units 12,090 12,950 10,670 6,600 4,990 2,730 2,150
Yield (%) — — 4.89 17.3 39.0 78.3 75.7
100
107 88.3 54.6 41.3 22.6 17.8
J. Biochem.
PURIFICATION AND CHARACTERIZATION OF PR1-LYSOZYME electrophoresis at p H 4 . 3 , electrophoresis in 0 . 1 % SDS and analytical gel isoelectric focusing in a p H gradient from 3 to 10. The results indicate that the preparation was electrophoretically homogeneous. //. Characterization of PRl-Lysozyme—Molecular weight and isoelectric point: By slab gel electrophoresis in 0.1 % SDS, the molecular weight of the enzyme was estimated to be 24,000, based on a calibration curve obtained by using five proteins of known molecular weight as standards. The purified enzyme preparation also exhibited homogeneity on gel filtration through Sephadex G-100 (2.2 x 90 cm column; 0.05 M potassium phosphate buffer, p H 6.0, containing 0.15 M NaCl). A slightly higher
731
molecular weight value of 26,000 to 27,000 was obtained by the method of Andrew (27). The isoelectric point of the enzyme was estimated to be 9.4 by analytical gel isoelectric focusing with 2 % Ampholine, p H range 3 to 11. The results indicate that PRl-lysozyme consists of a single polypeptide chain and is a basic protein, as is hen egg-white lysozyme or T4-lysozyme, even though the molecular weight is higher than those of the latter lysozymes. Amino acid composition of PRl-lysozyme: Table H shows the amino acid composition of PRl-lysozyme. The values listed are presented in terms of molar ratios on the basis of 34 mol of alanine residues per mol of the enzyme. From
TABLE II. Amino acid composition of PRl-lysozyme. Samples were hydrolyzed with 6 N H O in evacuated, sealed tubes at 110°C for 24, 48, and 86 h, and with 4 N methanesulfonic acid at 115°C for 24 h. Values listed are expressed in terms of molar ratios, on the basis of 34 alanine residues per molecule of the enzyme. Values for serine and threonine were obtained by extrapolation to zero time of hydrolysis, and the maximum values were taken for valtne and isoleucine. Residues/Molecule Amino acid
Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Half-cystine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Tryptophan
24 h
Hydrolysis for 48 h 86 h
18.4 7.24 6.64 29.7 9.96 19.0 34.0 7.98 1.24 0 5.90 24.5 3.82 4.96 1.83 2.13 21.8 —
19.3 7.88 6.48 29.3 10.7 19.9 34.0 8.27 1.15 0 6.09 25.4 3.99 5.03 1.99 2.34 22.6 —
18.6 7.18 5.80 29.0 10.1 19.4 34.0 8.73 1.14 0 6.28 24.1 4.42 4.95 1.83 2.33 21.4 —
Total * Values obtained after hydrolysis with 4 N methanesulfonic acid. Vol. 83, No. 3, 1978
24 h» 19.2 7.83 6.68 30.9 11.0 20.3 34.0 8.32 1.71 0 6.39 25.5 4.41 5.04 2.06 2.01 23.0 4.86
Nearest integer
T4-Lysozyme (24)
19 8 7 30 10-11 20 34 9 2 0 7 25-26 5 5 2 2
22 11 6 13 3 11 15 9 5 2 10 16 6 5 1 13 13 3
23 5 213-215
164
732
the sum of the integer values, the total number of amino acid residues per molecule of the enzyme is estimated to be 213 to 215. From these values, the molecular weight of the enzyme was calculated to be 23,400 to 23,600, which is compatible with the value 24,000 determined by gel electrophoresis rather than the value obtained by gel filtration. Table II also indicates that there was no cysteine residue among more than 210 amino acid residues, and the number of basic amino acid residues was 27, accounting for more than 12% of the total number of residues. Glucosamine, galactosamine, and other amino sugars were not detected. Effect of pH on enzyme activity: The activity of PRl-lysozyme at various pH values was measured in potassium phosphate buffer (pH 5.0 to 8.0), Tris-HCl buffer (pH 7.5-9.0), 2-4) glycosidic linkages between TV-acetylmuramic acid and iV-acetylglucosamine ia
100
SO
v ^
\
1 10 : *
5
:
\
\
A
I
30 40 50 60 Temperature (°C)
00
Fig. 5. Stability of the enzyme to heat. The enzyme (0.32 units/ml) was incubated in 0.05 M potassium phosphate buffer, pH 6.5 containing 0.01% BSA for: • , 5min; O, lOmin; A, 20min; A, 60min at the indicated temperatures. After incubation, samples were chilled to 0°C and the residual activity was measured. Each activity is expressed as a percentage of the activity of a control sample which had been incubated at 0°C for the same time.
20
40 60 80 100 120 Incubation Time(min)
Fig. 6. Mode of action of PRl-lysozyme. Peptidoglycan (1.5 mg) was hydrolyzed with PRl-lysozyme (0.14 fig) in 1.7 ml of 0.02 M sodium acetate buffer, pH 6.0, at 37°C. At fixed intervals, 50 fi\ samples were withdrawn, and assayed for the indicated characteristics. Reducing groups and free amino groups were determined by the method of Park and Johnson (18) and by ninhydrin colorimetry (19), respectively. • , Reducing groups; O, free amino groups.
TABLE 111. Specificity of PRl-lysozyme. Peptidoglycan (1.5 mg) was incubated with or without 0.2 fig of PRllysozyme for 3 h at 37°C in 1.7 ml of 0.02 M sodium acetate buffer, pH 6.0. Digestion with 10 fig of hen egg-white lysozyme was also carried out in the same manner. The reaction mixtures were incubated with 2 mg of sodium borohydride at about pH 10 for another 48 h at 37°C. After reduction, the samples were dried and hydrolyzed with 6 N HC1. The hydrolysates were analyzed with an automatic amino acid analyzer. Other experimental details are given in the text. /imol/mg of peptidoglycan (molar ratio1)
Compound
Without enzyme Muramic acid Muramicitol Glucosamine Glucosaminitol Glutamic acid Alanine Diaminopimelic acid » Glutamic acid is taken as 100. Vol. 83, No. 3, 1978
0.71 0.05 0.81 0.05 0.95 1.53 0.86
(75) (5.3) (85) (5.3) (100) (161) (91)
Hen egg-white lysozyme 0.11 0.69 0.75 0.06 0.89 1.45
(12) (76) (84) (6.7) (100) (163)
0.86
(97)
PRl-lysozyme 0.03 0.82
(3.4) (94)
0.78 0.05 0.87 1.49 0.89
(90) (5.7) (100) (170) (102)
734
N. OCHT, M. AZEGAMI, and S. ISHII strates (Table IV). Values listed are expressed in terms of specific activity (units/^jjo), and values obtained with hen egg-white lysozyme are also listed for comparison. Neither PRl-lysozyme nor hen egg-white lysozyme showed any activity toward intact cells of gram-negative bacteria, while both
the glycan chain of peptidoglycan, as illustrated in Fig. 7. Bacteriolytic spectrum of PRl-lysozyme: The bacteriolytic spectrum of PRl-lysozyme was investigated by using intact, EDTA-treated, and chloroform-killed cells of various bacteria as sub-
PRl-lysozyme, T2, T4 6
1
)
' "^"
G1 cNAc
Mu r NAc
T3, T5, T7
G1 cNAc —
Vi-III. HM7
D-Glu
I I
L-Ala MurNAc PRl-lysozyme, T 2 , T4-
Fig. 7. Sites of peptidoglycan attacked by PRl-lysozyme and by other phage lytic enzymes. A part of the structure of peptidoglycan of gram-negative bacteria is illustrated. TABLE IV. Bacteriolytic spectrum of PRl-lysozyme. Bacteria listed in the table were grown in Nutrient BrothNaCl medium (Nutrient Broth 15 g, NaCl 5 g/liter) and collected at the late log phase (about 2x10* cells/ml) by centrifugation. Intact, EDTA-treated, and chloroform-killed cells of each bacterium were prepared as substrates. Intact and EDTA-treated cells were prepared by two cycles of suspension and centrifugation in 0.05 M potassium phosphate buffer, pH 6.4, and 0.05 M potassium phosphate buffer, pH 6.4, containing 5 mM EDTA, respectively. Chloroform-killed cells were obtained as described by Black and Hogness (2). Activities of enzymes were 0.31 units/ml and 0.36 units/ml for PRl-lysozyme and hen egg-white lysozyme, respectively. Specific activity (units//* Hen egg-white lysozyme EDTAtreated cells
Chloroformkilled cells
0 0 0 0
0.32 0.35 0.23 0.21
1.66 1.82 1.75 1.65
0
0.35
1.47
2.59 2.38
2.49 2.26
2.64 2.35
Intact cells Gram-negative bacteria Escherichia coll C Escherichia coli K12W6 Pscudomonas aeruginosa PI 5 Pseudomonas aeruginosa PI 1 Klebsiella pneumoniae Salmonella typhimurium Aerobacter aerogerus Serratia marcescens Proteus vulgaris Gram-positive bacteria Micrococcus lysodeikticus Staphyhcoccus aureus Bacillus subtilis
PRl-lysozyme Intact cells
EDTAtreated cells
Chloroformkilled cells
0 0 0 0 0 0 0 0 0
5.36 6.90 6.82 5.36
68.2 79.3 69.6 64.0 58.5 71.0 61.2 50.1 54.3
0 0
0 0
5.64 4.04 3.83 4.38
32.0 20.2 13.9
/ . Biochem.
PURIFICATION AND CHARACTERIZATION OF PR1-LYSOZYME
735
the enzymes were able to lyse chloroform-killed the hypothesis that these enzymes recognize not cells of these bacteria. However, hen egg-white only the sites which they cleave, but also structures lysozyme lysed intact cells of gram-positive bac- near them. teria such as M. lysodeikticus and S. aureus. These Composition analyses of PRl-lysozyme findings indicate that there is a distinct difference showed the absence of cysteine and cystine residues. in affinity for bacterial cells between hen egg-white T2-(23), T4-lysozyme (24), lambda endolysin (25), lysozyme and PRl-lysozyme, though the two en- and N20F' lytic enzyme (4) also contain no disulfide zymes have the same specificity against the pep- bond in their molecules. The lack of disulfide tidoglycan of P. aeruginosa. bonds may be a common feature of lytic enzymes related to bacteriophages and bacteriocins. This would distinguish them from bacterial or fungal DISCUSSION lytic enzymes, which are known to have one disulPRl-lysozyme was purified to homogeneity from fide bond in the molecule (26, 27), and from the a mitomycin C-induced lysate of pyocinogenic enzymes of higher organisms, such as hen eggP. aeruginosa by a reasonably convenient proce- white lysozyme, which has four disulfide bonds dure. PRl-lysozyme was shown to be an N- (28). The lack of disulfide bonds may partly acetylmuramidase, as is hen egg-white lysozyme, account for the heat lability of PRl-lysozyme. since both enzymes hydrolyzed ./V-acetylmuramide Our preliminary experiments suggest that the NH,linkages only, leaving N-acetylglucosaminide link- terminus and COOH-terminus of the enzyme ages intact, with peptidoglycan of P. aeruginosa molecule were methionine and leucine, as deteras a substrate. However, PRl-lysozyme hydro- mined by the dansylation (29) and carboxypeptidase lyzed Af-acetylmuramide linkages almost com- Y methods (30), respectively. If this is the case, pletely in the peptidoglycan, while the extent of PRl-lysozyme has the same terminal amino acid hydrolysis by hen egg-white lysozyme only reached residues as T4-lysozyme (24). almost 90% even though the amount was increased It is of interest to compare the lysozymes of up to 50 times that of the former enzyme (Table higher organisms and lytic enzymes related to ni). This superiority of PRl-lysozyme over hen phages and phage-tail-like bacteriocins as regards egg-white lysozyme was also observed when the structural differences and biological significance. activities of the two enzymes were compared with Such differences may be a result of evolutionary chloroform-killed gram-negative bacteria as sub- selective pressures related to enzyme stability, or strates by turbidimetric measurement (Table IV). the rate of enzyme synthesis, or both. Rapid Presumably, the chloroform treatment removes the synthesis of lytic enzyme could be advantageous lipid components, at least in part, from the bac- to phage or bacteriocin reproduction, because lysis terial cell surface and makes the cell wall susceptible of the host cells must occur within a short time, to attack by lytic enzymes. However, it is note- whereas a high degree of stability might not be worthy that hen egg-white lysozyme is able to lyse significant. It is known that T2- and T4-lysozymes gram-positive bacteria such as M. lysodeikticus and act upon the host cell from inside the cell (31) and S. aureus. Thesefindingsmay reflect differences of this may also be the case with PRl-lysozyme, peptidoglycan structure, especially that of the since the only biological function for these enzymes peptide portion, or differences of components on may be to lyse host cells from inside to release the the cell surface, between gram-positive and gram- newly formed phages or bacteriocins. In the case negative bacteria. In addition, Arima et al. (22) of higher organisms, the opposite situation may have reported that P. aeruginosa phage 95-induced exist, i.e., high stability is a significant factor and lytic enzyme, LE95, is an L-alanyl-D-glutamic acid rapid formation is not particularly relevant. Lysoendopeptidase and does not hydrolyze various zymes of higher organism may have to act against small fragments of peptidoglycan even though they bacteria which have invaded the organism, from contain the peptide moiety attacked by LE95, outside the bacterial cell. suggesting that the presence of the glycan chain In the present experiments, the number of of the peptidoglycan is essential for enzymatic PRl-lysozyme molecules produced from a single activity. These observations may be explained by cell appeared to be about 4x10*. This number Vol. 83, No. 3, 1978
736
N. OCHI, M. AZEGAMI, and S. ISHH
9. Jastrzemski, K.B. (1975) Ada Biochim. Pol. 22, 297-304 10. Taylor, A. (1971) Nature New Biol. 234, 144-145 11. Kato, F., Ogata, S., & Hongo, M. (1976) Agric. Biol. Chem. 40, 1101-1105 12. Yanai, A., Kato, K., Beppu, T., & Arima, K. (1976) Agric. Biol. Chem. 40, 1505-1508 13. Ornstein, L. (1964) Ann. N.Y. Acad. Sci. 121, 321349 14. Davis, B.J. (1964) Ann. N.Y. Acad. Sci. 121, 404-^27 15. Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, 4406-4412 16. Wrigley, C.W. (1971) in Methods in Enzymology (Jakoby, W.B., ed.) Vol. 22, pp. 559-564, Academic Press, New York 17. Simpson, R.J., Neuberger, M.R., & Liu, T.Y. (1976) J. Biol. Chem. 251, 1936-1940 18. Park, J.T. & Johnson, M.J. (1949) / . Biol. Chem. 181, 149-151 19. Yemm, E.W. & Cocking, E.C. (1955) Analyst 80, 209-213 20. Ghuysen, J.M., Tipper, D.J., & Strominger, J.L. (1966) in Methods in Enzymology (Neufeld, E.F. & Ginsburg, V., eds.) Vol. 8, pp. 685-711, Academic Press, New York 21. Andrew, P. (1964) Biochem. J. 91, 222-233 22. Yanai, A., Kato, K., Beppu, T., & Arima, K. (1976) Biochem. Biophys. Res. Commun. 68, 1146-1152 23. Inouye, M. & Tsugita, A. (1968) / . Mol. Biol. 37, 213-223 24. Tsugita, A. & Inouye, M. (1968) J. Mol. Biol. 37, 201-212 25. Black, L.W. & Hogness, D.S. (1969) J. Biol. Chem. 244, 1976-1981 26. Shih, J.W. & Hash, J.H. (1971) J. Biol. Chem. 246, 994-1006 27. Jackson, R.L. & Wolfe, R.S. (1968) J. Biol. Chem. 243, 879-888 We wish to thank Miss M. Fujimoto for her cooperation 28. Imoto, T.. Johnson, L.N., North, A.C.T., Phillips, D.C., & Rupley, J.A. (1972) in The Enzymes (Boyer; in amino acid analyses, and Dr. R. Tirrell for his advice P.D., ed.) Vol. 7, pp. 665-868, Academic Press, during the preparation of this paper. New York 29. Hartley, B.S. (1970) Biochem. J. 119, 805-822 30. Hayashi, R., Moore, S., & Stein, W.H. (1973) REFERENCES /. Biol. Chem. 248, 2296-2302 1. Tsugita, A., Inouye, M., Terzaghi, E., & Streisinger, 31. Emrich, J. & Streisinger, G. (1968) Virology 36, G. (1968) J. Biol. Chem. 243, 391-397 387-391 2. Black, L.W. & Hogness, D.S. (1969) / . Biol. Chem. 32. Kageyama, M. (1964) / . Biochem. 55, 49-53 244,1968-1975 33. Jshii, S., Nishi, Y., & Egami, F. (1965) / . Mol. 3. Taylor, A. (1970) Biochem. Biophys. Res. Commim. Biol. 13,428-431 41, 16-24 34. Ito, S. & Kageyama, M. (1970) / . Gen. Appl. 4. Moo-Penn, W., Mowry, C , & Wiesmeyer, H. (1969) Microbiol. 16, 231-240 Biochim. Biophys. Ada 178, 330-346 35. Kageyama, M. (1974) / . Gen. Appl. Microbiol. 20, 5. Arima, K., Yanai, A., & Bcppu, T. (1976) Agric. 269-275 Biol. Chem. 40, 1301-1306 . 36. Kageyama, M. (1970) / . Gen. Appl. Microbiol. 16, •6. Inouye, M., Arnheim, N., & Sternglanz, R. (1973) 523-530 J. Biol. Chem. 248, 7247-7252 37. Kageyama, M. & Inagaki, A. (1974) / . Gen. Appl. Microbiol. 20, 257-267 "7. Hongo, M.,Tahara, Y., & Ogata, S. (1974) Agric. Biol. Chem. 38, 755-761 8. DeMartini, M., Halegoua, S., & Inouye, M. (1975) / . Virol. 16, 459-461
may be compared with about 100 to 200 pyocin Rl particles produced per induced cell under the same conditions. These numbers are compatible with those of a temperate phage, lambda, in which the number of lambda endolysin molecules per cell is about 6x10*, and that of matured phage particles is about 100 (2). The optimum pH for the activity of PR1lysozyme was found to be near 6.4 with chloroformkilled P. aeruginosa cells as a substrate. However, the pH optimum may reflect not only the characteristics of the enzyme but also the susceptibility of the chloroform-killed cells to the enzyme action. More relevant information on the pH optimum of the enzymatic activity itself may be obtained by using simpler substrates. Pyocin Rl has a close resemblance to some phage tails (32, 33) and may be considered as a defective lysogenic phage (32, 34), although conclusive evidence for this has not yet been presented. The pyocinogenic factors of pyocin Rl, R2, and R3 appear to be integrated into the bacterial chromosome at a corresponding position close to the trp-1 region (35-37). However, it is still questionable whether PRl-lysozyme is the product of a bacterial gene or is derived from genetic information of the pyocinogenic factor. In any case, more precise investigations of phage-tail-like bacteriocin and lytic enzyme systems from a genetic as well as a biochemical viewpoint are required.
/. Biochem.
