Autolytic enzyme system of Streptococcus faecalis. V. Nature of the autolysin-cell wall complex and its relationship to properties of the autolytic enzyme of Streptococcus faecalis

Cell walls from exponential-phase cultures of Streptococcus faecalis ATCC 9790 contain an autolysin (a beta-N-acetylmuramide glycanhydrolase, E.C. 3.2.1.17) which has been isolated from trypsin-speeded wall autolysates. The autolysin, which was excluded from Bio-Gel P-60, was further fractionated by diethylaminoethyl (DEAE)-cellulose chromatography or filtration on Bio-Gel P-200. After DEAE-cellulose chromatography, which removed most of the wall polysaccharide, autolysin activity was extremely labile and was rapidly lost at -20 C, even in the presence of albumin. The P-60-excluded enzyme was rapidly bound by walls at both 37 C (50% bound in about 1 min) and 0 C (50% bound in less than 4 min). Wall-bound autolysin could not be removed by 1.0 m ammonium acetate (pH 6.9). Autolysin was also bound by walls that had been extracted with 10% trichloroacetic acid or treated with 0.01 n periodate, suggesting that the nonpeptidoglycan wall polymers are not important for binding. Wall-bound autolysin was more stable than the soluble enzyme to proteinase digestion, acetone (40%), 8 m urea (at 0 C), or to inactivation at 56 C. Two bacterial neutral proteinases (which do not hydrolyze ester bonds) activated latent wall-bound autolysin, suggesting that activation results from the cleavage of one or more peptide bonds. The group A streptococcal proteinase activated latent autolysin but differed from the other proteinases in that it did not inactivate soluble autolysin. The results suggest that the autolysin is not covalently linked to the wall. The high affinity of the walls for the autolysin appears to be responsible for the firm, not easily reversed binding.     

Extracellular cell wall lytic enzyme from Staphylococcus aureus: purification and partial characterization

An autolysin obtained from culture fluid of Staphylococcus aureus strain 8507 was purified 3,000-fold. One milligram of this preparation (S-5DL) will solubilize 12 mg of cell wall in 1 hr. The major activity is N-acetylmuramyl-l-alanine amidase. Recovery of lytic activity in the purified preparation was repeatably only 20% of the starting level. This suggests that other cell wall lytic enzymes may be present in the starting material. The S-5DL enzyme has been compared to freeze-thaw extracted enzyme (AFZ). Both enzymes precipitate in 0.01 m KPO(4) (pH 6.0) and dissolve in 0.1 to 0.7 m NaCl. Fifty per cent of the AFZ activity and 66% of the S-5DL activity bind rapidly to cell walls of S. aureus at 0 C in the presence of magnesium ion. None of the AFZ activity and 66% of the S-5DL activity bind to cell walls at 0 C in the absence of magnesium ion. The cell walls of nine different strains of S. aureus were compared for level of native autolysin activity. These same walls after inactivation of the native autolysin were tested for susceptibility to the S-5DL enzyme.     

Release of autolytic enzyme from Streptococcus, faecium cell walls by treatment with dilute alkali.

The autolytic enzyme (endo-beta-1,4-N-acetylmuramoylhydrolase) of Streptococcus faecium (S. faecalis ATCC 9790) was released in a soluble form from insoluble cell wall-autolytic enzyme complexes by treatment with dilute NaOH at 0 degree C. Treatment of cell wall-enzyme complexes, obtained from either exponential- or stationary-phase cells, with 0.008 to 0.01 N NaOH gave maximum yields of autolytic enzyme activity. At a fixed concentration of NaOH, the yield of autolysin increased with increasing wall densities and was accompanied by the release of methylpentose and phosphorus in amounts proportional to the autolysin. Since extraction of wall-enzyme complexes with 4.5 M LiCl at 0 degree C also removed methylpentose and phosphorus, release of enzyme with NaOH did not appear to result from hydrolysis of covalent linkages. The autolytic enzyme activity released from intact cells, or cell walls, was predominantly in the later (proteinase activable) form which could be activated by trypsin or a proteinase present in commerical bovine plasma albumin.     

New centrifugation technique for isolating enzymes from large cell structures: isolation and characterization of two Bacillus subtilis autolysins

Bacillus subtilis cell walls can be centrifuged through a linear gradient of 0 to 2 m LiCl and 10 to 25% sucrose so that different autolysins are removed by different salt concentrations and banded in separate positions as the walls pass through the gradient. Using this technique we have found that B. subtilis cell walls are isolated with two autolytic enzymes attached. One autolysin, a glycosidase, can be eluted from walls with 0.5 m LiCl, has a pH optimum between 5 and 8, is relatively heat-sensitive, and has a molecular weight of 60,000. The other autolysin, an alanine amidase, can be eluted from walls with 1.5 m LiCl, has a pH optimum around 8, is relatively heat-stable, has a molecular weight of 35,000, and is present in quantities ten times greater than the glycosidase.     

Rapid Methods for Extracting Autolysins from Bacillus subtilis

Two procedures are described for the extraction of autolysins from whole cells. One method uses 5 M LiCl at 4 C. The amount of enzyme obtained by this method is six times more than that obtained by autolysis of cell walls and fourteen times more than that obtained by extracting cell walls with LiCl. With the other method, cells are extracted with 2% Triton X-100. This is less efficient than the LiCl method but yields about one-half the amount of enzyme obtained by cell wall autolysis and about the same amount as obtained by extracting cell walls with salt. Both procedures yield autolysin with multiple pH optima. Autolysins can be extracted from several bacterial species by either the LiCl or the detergent method. The data suggest that these techniques have sufficient sensitivity to detect small differences in autolytic activity among mutants and various organisms and are also suitable for large-scale isolation of autolysin for purification and characterization studies.     

Protein-bound choline is released from the pneumococcal autolytic enzyme during adsorption of the enzyme to cell wall particles.

The inactive precursor form of the pneumococcal autolytic enzyme cloned in Escherichia coli was isolated by affinity chromatography on Sepharose-linked choline. The enzyme was recovered in an electrophoretically pure and activated form by elution from the affinity column with radioactive choline solution. When radioactive choline was used for elutions, the enzyme protein isolated contained protein-bound choline, at approximately 1 mol of choline per mol of enzyme protein, indicating the presence of a single choline recognition site. Radioactive choline remained bound to the enzyme protein during dialysis, precipitation by trichloroacetic acid or ammonium sulfate, and during gel filtration, but not during sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Incubation of the choline-labeled autolysin with pneumococcal cell walls at 0 degrees C resulted in the adsorption of the enzyme to the wall particles and a simultaneous release of free choline from the enzyme protein. It is suggested that the choline molecules that became bound to the enzyme protein during the activation of autolysin are expelled from the choline-binding site and replaced by choline residues from the wall teichoic acid as the autolysin molecules adsorb to their insoluble substrate before the onset of enzymatic wall hydrolysis.     

Sporangila autolysis in Clostridium perfringens type A

The autolytic system functioning in the release of mature spores and enterotoxin from sporangia of Clostridium prefringens was partially characterized. After sporangial autolysis in buffer, the supernatant fluid of the suspension contained autolysin active against purified sporangial walls. The autolysin was most active at pH 8 and 37°C, in the presence of Co²⁺ (0.3 · 10⁻³ M CoCl₂) and trypsin (48 μg/ml). Sodium dodecyl sulfate-treated sporangial walls further extracted with trichloroacetic acid to remove teichoic acid were a better enzyme substrate than walls treated only with sodium dodecyl sulfate. N-Acetylmuramyl-l-alanine amidase activity which released N-terminal alanine, and endopeptidase activity which hydrolysed the d-alanyl-glycine linkage liberating N-terminal glycine and C-terminal alanine, were both functional at pH 8. It is not known if one or two enzyme are involved. Autolysin appeared in cells as early as 2 h after inoculation into sporulation medium. Two asporogenic Stage 0 mutants grown in sporulation medium also produced autolysin identical in mode of action to that of the sporogenic wild type. Although the active cellular autolysin concentration subsequently decreased as cells sporilated, the walls of 8-h-old sporangia containing refractile heat-resistant spores were more susceptible to digestion by autolysin, than those of 2-, 4-, or 6-h-old cells grown in sporulation medium or of 4- or 14-h vegetative cells from growth medium. The results suggest that a progressive change may occur in the structure of the sporangial wall during spore morphogenesis, thus increasing its susceptibility to autolysis.     

Relationship between the Latent Form and the Active Form of the Autolytic Enzyme of Streptococcus faecalis

A 10-hr starvation of Streptococcus faecalis ATCC 9790 for the amino acids methionine and threonine results in cells which are resistant to autolysis and which contain greatly reduced quantities of both active and latent (proteinase activable) forms of the autolytic enzyme (an N-acetyl-muramide glycanhydrolase). Cell walls were isolated from cells harvested at various times during the recovery from such starvation and were assayed for active and latent forms of the autolysin. Within 10 min of recovery the latent enzyme began to increase. Only after 30 to 60 min did the active enzyme begin to increase; after a similar lag, the cells' proneness to lysis markedly increased. The intracellular localization of both forms of the autolysin was examined, using as an experimental tool the ability of added cell wall to bind autolysin. (14)C-lysine-labeled, inactivated cell walls were added to exponential-phase cells, which were then disrupted, and the mixed wall population was isolated. Measurement of the (14)C release during wall autolysis indicated that the active enzyme in the cells was not available for binding to the added (14)C-labeled walls and was therefore wall-bound in vivo. In contrast, up to 85% of latent autolysin activity was found to have been efficiently bound to the added (14)C walls. The results obtained suggest (i) cellular autolysis is a reflection of the level of active enzyme and not of latent enzyme, and (ii) autolysin is synthesized and mainly located in the cytoplasm as an inactive latent precursor (proenzyme) which is transported to sites on the cell wall associated with wall biosynthesis, where it becomes activated.     

Regulation of bacterial cell walls: correlation between autolytic activity and cell wall turnover in Staphylococcus aureus.

Cell wall turnover was examined in parent and mutant strains of Staphylococcus aureus. Peptidoglycan and teichoic acid were observed to undergo turnover in the wild-type strain during exponential growth; however, the rate of turnover did not decrease when the growth rate slowed, as the culture entered stationary phase. Isolated native cell walls and crude soluble autolytic enzyme were prepared from cells harvested during exponential and postexponential phases of growth. Native cell walls from both phases of growth autolyzed in buffer at identical rates; similarily, crude soluble enzyme from both preparations degraded radioactive cell walls at the same rate. Therefore, the activity of the autolysin in both exponential and postexponential cells was similar. The autolysis of whole cells of a mutant tar-1 was enhanced by 1.0 M NaCl. When 1.0 M NaCl was present under growing conditions, the rate of cell wall turnover was greatly increased. The presence of chloramphenicol, which inhibits whole-cell autolysis, also inhibited turnover. Analysis of the cell wall material recovered from spent medium revealed products consistent with the known mode of action of the endogenous autolysin. It is concluded that cell wall turnover in S. aureus is independent of the stage of culture growth but is dependent instead on the activity of the autolysin.     

Analysis of autolysins in temperature-sensitive morphological mutants of Bacillus subtilis.

The content and distribution of autolysin were measured in temperature-sensitive morphological mutants of Bacillus subtilis. Strains RUB1000 and RUB1012 grew as rods at 30 C. At 45 C the mutants contained disproportionately less teichoic acid than peptidoglycan and grew as irregular spheres. The amount of enzyme that could be extracted from rods was at least 31 times the amount extracted from spheres. The rate of autolysis of cell walls was 7- to 28-fold greater in rods than in spheres. The low activity found associated with the cell walls of spheres was not compensated for by larger amounts of autolytic activity in the cytoplasm. No activity was found in the growth medium at either temperature. The failure of the mutant cells to autolyze was due to low amidase activity and relatively resistant cell walls. Revertants of RUB1012 were isolated that had 13, 23, and 55% of the normal proportions of teichoic acid when grown at the nonpermissive temperature. Cell walls from the revertants were as sensitive to added amidase as the wild-type strain. None of the revertant strains regained the wild-type ability to produce more amidase at 45 C. However, the deficiency in autolysin observed with RUB1012 was partially restored in revertants containing higher proportions of teichoic acid.     

A mutant of Streptococcus pneumoniae that exhibits thermosensitive penicillin tolerance and the paradoxical effect

Mutants of Streptococcus pneumoniae that contain active autolysin and yet cannot be induced to lyse during treatment with penicillin (Lyt+Tol+ mutants) have been described. We have now shown that these mutants are temperature dependent (32 degrees C); at 37 degrees C these bacteria underwent penicillin-induced lysis. In addition, mutants at the lysis-permissive temperature showed the so-called 'paradoxical response' to penicillin. Temperature shift experiments indicated that the change from tolerant to lytic response or vice versa is a fast process. No differences were detected in autolysin specific activity or in the kinetics of inhibition of protein, peptidoglycan and teichoic acid syntheses in cells treated with penicillin at 32 and 37 degrees C. The results of genetic crosses indicated that the thermosensitivity of penicillin-induced autolysis in the Lyt+Tol+ mutants is not a property of the autolytic enzyme itself. The observations suggest that the thermosensitive process in the mutants represents either a step(s) in autolysin regulation or involves some difference in the structure of the cell walls produced at 32 degrees C versus 37 degrees C.     

Ultrastructural Studies on a Mutant of Bacillus subtilis Whose Growth is Inhibited Due to Insufficient Autolysin Production

The growth of Bacillus subtilis mutant betaA177 can be inhibited under special conditions in which not enough autolytic enzymes are produced for optimal growth. Electron microscopy studies show that during growth inhibition there is localized thickening of the cell wall at positions where cells bend. A model is proposed to explain this result. Rapid growth can be restored by adding lysozyme or a B. subtilis autolysin mixture to a growth-inhibited betaA177 culture. Such addition reduces the localized wall thickening and causes other changes in surface morphology which are described and discussed. Septum formation seems to be relatively less inhibited than cell elongation when lytic enzyme levels are reduced. Measurements were made demonstrating that walls at ends of cells are morphologically different from walls at sides of cells in cultures of betaA177 growing at 51 C.     

Autolytic Enzyme System of Streptococcus faecalis III. Localization of the Autolysin at the Sites of Cell Wall Synthesis

Cell walls from exponential-phase cultures of Streptococcus faecalis ATCC 9790 autolyzed in dilute buffers. Walls were isolated from cultures grown in the presence of (14)C-lysine for about 10 generations and then on (12)C-lysine for 0.1 to 0.8 of a generation (prelabeled). These walls released (14)C to the soluble fraction more slowly than they lost turbidity during the initial stages of autolysis. Walls isolated from cultures grown in the presence of (14)C-lysine for only the last 0.1 to 0.4 of a generation (postlabeled) released (14)C to the supernatant fluid more rapidly than they lost turbidity. Autolysin in both pre- and postlabeled walls was inactivated, and such walls were then incubated in the presence of unlabeled walls containing active autolysin. The inactivated walls lost their (14)C label only very slowly until autolysis of the unlabeled walls was virtually complete and release of soluble autolysin was expected. When this experiment was done in the presence of trypsin, a fourfold increase in the autolysis rate resulted, but the same pattern of (14)C release was observed. A parallel release of (14)C and loss of turbidity from pre- or postlabeled walls was observed upon trypsin "activation" and by addition of isolated soluble autolysin to inactivated walls. We conclude that the wall-bound autolysin acts first on the more recently synthesized portion of the wall. Trypsin appears to speed wall autolysis by activating additional latent autolysin in situ at sites in the older portion of the wall.     

Specific recognition of choline residues in the cell wall teichoic acid by the N-acetylmuramyl-L-alanine amidase of Pneumococcus.

Pneumococci growing on choline-containing medium are known to incorporate this amino alcohol into the wall teichoic acid and produce autolysin-sensitive cell walls. In contrast, bacteria grown on the choline analogue, ethanolamine, incorporate ethanolamine into the teichoic acid and synthesize cell walls that are resistant to the homologous autolysin. In this communication, we report experiments aimed at understanding the biochemical mechanism of this phenomenon. Ethanolamine-containing (autolysin-resistant) cell walls were methylated in vitro with methyl iodide. Under appropriate conditions, virtually all of the ethanolamine residues could be converted to choline. After methylation, the formerly autolysin-resistant walls could be quantitatively hydrolyzed by the pneumococcal autolysin. Methylated walls also recovered another property typical of cell walls isolated from choline-grown bacteria: they could induce the in vitro "conversion" of an inactive form of autolysin to the catalytically active form (Tomasz, A., and Westphal, M. (1971) Proc. Natl. Acad. Sci. U.S.A. 68, 2627-2630). The results suggest that the autolysin-catalyzed hydrolysis of amide bonds in the peptidoglycan requires an additional interaction between the enzyme protein and choline residues in the teichoric acid portion of the cell wall.     

Autolytic Enzyme System of Clostridium botulinum

Isolated cell walls of Clostridium botulinum type A strain 190L released an autolysin during autolysis of the cell walls. The autolysin was isolated from the cell walls, and partially purified 18.6-fold by ammonium sulfate precipitation, chromatography on DEAE-cellulose and gel filtration through Sephadex G-100. The purified preparation of the autolysin showed 2 major and 2 minor protein bands on Polyacrylamide gel electrophoresis. Some properties of the autolysin were examined using SDS-treated cell walls of the organisms as a substrate. The autolysin was active over a pH range of 6 to 8, with a maximum near pH 6.8. The lytic activity was stimulated by 10^?4 M each of Co⁺⁺, Mg⁺⁺ and Ca⁺⁺ in the order, whereas it was inhibited markedly by Cu⁺⁺. Mercaptoethanol (10^?4–10^?3 M) significantly activated the lytic action. Trypsin and nagarse (10 μg/ml) also stimulated the lytic activity. The lytic spectrum of the autolysin toward the SDS-treated cell walls obtained from various types of C. botulinum and C. perfringens indicated a relatively high specificity. After treatment with hot formamide the cell walls of C. botulinum increased in susceptibility to the autolysin.     

Mechanism of autolysis of Neisseria gonorrhoeae.

The major autolysin(s) of Neisseria gonorrhoeae was solubilized from envelopes by extraction with 2% Triton X-100 containing 0.5 M NaCl. Neither Triton X-100 nor NaCl alone could effectively release the autolysin(s). The major autolysin is N-acetylmuramyl-L-alanine amidase (EC 3.5.1.28). The pH optimum for this reaction was broad, ranging from 5.5 to 8.5. Optimal hydrolysis of peptidoglycan occurred in 2% Triton X-100 in 0.1 M KCl. Attempts to purify the autolysin were unsuccessful. A rapid assay for enzyme activity was developed using radioactive cell walls as a substrate ([3H]diaminopimelic acid).     

Autolytic Enzyme System of Streptococcus faecalis IV. Electron Microscopic Observations of Autolysin and Lysozyme Action

Cell walls (LOG walls) were isolated from cultures of Streptococcus faecalis ATCC 9790 in the exponential phase of growth. These walls were either allowed to undergo autolytic dissolution (in the presence or absence of trypsin) or wall autolysis was inactivated with sodium dodecylsulfate (SDS walls). Inactivated walls were treated either with lysozyme or with isolated, partially purified S. faecalis autolysin. During wall lysis, samples were removed, negatively stained with phosphotungstate, and examined in the electron microscope. Both lysozyme and isolated autolysin appeared to act over the entire surface of SDS walls. After partial dissolution, a fibrous network over the surface was revealed. Lysozyme digestion revealed the presence of prominent, highly-contrasted equatorial and subequatorial bands around the walls. After trichloroacetic acid extraction, the bands were seen less frequently and less distinctly in the partially lysozyme digested walls, suggesting that the bands contained nonpeptidoglycan polymers. In the absence of trypsin (which activates a latent form of the autolysin), autolysis of LOG walls appeared to start at the equatorial bands and to proceed back towards the apex of the coccus. Ribbons of wall material coming off the wide edge of the nearly hemispherical wall fragments were observed. Activation of latent autolysis resulted in lytic action over the entire wall surface. The results are consistent with the previously postulated location of active autolysin at the areas of new wall synthesis and the random location of latent autolysin in LOG walls.     

Some Properties of Two Autolytic-Defective Mutants of Streptococcus faecalis ATCC 9790

The isolation and some properties of two mutants of Streptococcus faecalis ATCC 9790 (S. faecium) which autolyze at a much slower rate than the wild type are described. Compared with the wild type, mutant E71 autolyzed more slowly, contained less active but more latent autolysin in the isolated wall fraction, and possessed a wall of very similar chemical composition and degree of cross-bridging. Ultrastructural studies of exponential phase cells showed that cells of E71 were on the average slightly longer and had slightly thickened walls compared to the wild type. Mutant E81 autolyzed much more slowly, grew exponentially in long chains (8 to 40 cells compared with mainly diplococci), contained much less active and latent autolysin in the wall, and possessed a wall of very similar chemical composition but with about twice the content of N-terminal groups. Mutant E81 walls were more susceptible to isolated autolysin but possessed an autolysin of the same specificity as the wild type. Ultrastructurally E81 cells were, on the average, significantly longer and had thicker walls than the wild type. Mutant E71 may be partially blocked at either transport of autolysin to the wall or in conversion of latent to active autolysin. The pleitropic effects noted in mutant E81 have been taken to suggest a possible membrane defect and to support the role of the autolysin in cell separation.     

Specificity of the autolysin of Streptococcus (Diplococcus) pneumoniae

A Streptococcus (Diplococcus) pneumoniae autolysin, partially purified from cellular autolysates, was optimally active at pH 7.0 and was stimulated by monovalent cations. Addition of autolysin to walls resulted in the appearance of only N-terminal l-alanine, whereas no glycosidase activity was observed. Walls which had been solubilized by autolysin were separated by gel filtration into a low-molecular-weight peptide containing amino acids in the same ratios found in intact walls and a high molecular fraction containing the amino acid-deficient peptidoglycan backbone. Thus, the major activity is an N-acetylmuramyl-l-alanine amidase. In addition, walls undergoing spontaneous lysis revealed no glycosidase activity but showed an increase in only N-terminal alanine. Autolysin, which was bound to walls in saline, was almost completely removed when walls were washed in distilled water, and all of the activity was recovered in the water wash fluid.     

Analysis of the autolysins of Bacillus subtilis 168 during vegetative growth and differentiation by using renaturing polyacrylamide gel electrophoresis.

The autolysins of Bacillus subtilis 168 were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis with substrate-containing gels. Four bands of vegetative autolytic activity of 90, 50, 34, and 30 kDa (bands A1 to A4) were detected in SDS and LiCl extracts and in native cell walls by using B. subtilis 168 vegetative cell walls as the substrate incorporated in the gel. The four enzyme activities showed different substrate specificities and sensitivities to various chemical treatments. The autolysin profile was not medium dependent and remained constant during vegetative growth. During sporulation, band A4 greatly increased in activity just prior to mother-cell lysis. No germination-associated changes in the profile were observed, although a soluble 41-kDa endospore-associated cortex-lytic enzyme was found. By using insertionally inactivated mutants, bands A1 and A2 were positively identified as the previously characterized 90-kDa glucosaminidase and 50-kDa amidase, respectively. The common filamentous phenotype of various regulatory mutants could not be correlated to specific changes in the autolysin profile.