Isolation and purification of a Campylobacter upsaliensis autolysin.

Autolytic activity in the soluble and sediment fractions of sonicates of the spiral and the coccoid form of Campylobacter upsaliensis could not be demonstrated by native (nondenaturing) polyacrylamide gel electrophoresis (PAGE). Autolysins were detected, however, by using denaturing sodium dodecyl sulfate (SDS)-PAGE gels containing either purified Escherichia coli peptidoglycan or whole cells of Micrococcus luteus (Micrococcus lysodeikticus) as the turbid substrate, with subsequent renaturation by treatment with Triton X-100 buffer. In renaturing gels that contained Escherichia coli peptidoglycan, 14 putative autolytic bands ranging from 200 to 12 kDa were detected. In similar gels containing whole cells of M. luteus, only a single band appeared with a molecular mass of 34 kDa. This band corresponded to one of the bands present in the gels containing Escherichia coli peptidoglycan. This common autolysin was isolated by adsorbing it from Campylobacter upsaliensis soluble fractions onto M. luteus cells and then subjecting these cells to renaturing SDS-PAGE in gels containing Escherichia coli peptidoglycan. The 34-kDa autolysin differed from a single 51-kDa autolysin unique to the M. luteus cells, and when isolated from an SDS-PAGE gel, was pure when tested by isoelectric focusing. The N-terminal amino acid sequence analysis showed the first 15 amino acids of the 34-kDa autolysin to have 67% identity to a part of antigenic protein PEB4 of Campylobacter jejuni. The purified autolysin was used to immunize rabbits and the antibodies produced precipitated autolytic activity from cell lysates. The specificity of the antibodies was shown by Western blotting: only a single specific band occurred, with a molecular mass of 34 kDa, and thus it seems unlikely that the 34-kDa autolysin was derived from any of the other autolysins that were detected.     

Bacteriolytic effect of membrane vesicles from Pseudomonas aeruginosa on other bacteria including pathogens: conceptually new antibiotics.

Pseudomonas aeruginosa releases membrane vesicles (MVs) filled with periplasmic components during normal growth, and the quantity of these vesicles can be increased by brief exposure to gentamicin. Natural and gentamicin-induced membrane vesicles (n-MVs and g-MVs, respectively) are subtly different from one another, but both contain several important virulence factors, including hydrolytic enzyme factors (J. L. Kadurugamuwa and T. J. Beveridge, J. Bacteriol. 177:3998-4008, 1995). Peptidoglycan hydrolases (autolysins) were detected in both MV types, especially a periplasmic 26-kDa autolysin whose expression has been related to growth phase (Z. Li, A. J. Clarke, and T. J. Beveridge, J. Bacteriol. 178:2479-2488, 1996). g-MVs possessed slightly higher autolysin activity and, at the same time, small quantities of gentamicin. Both MV types hydrolyzed isolated gram-positive and gram-negative murein sacculi and were also capable of hydrolyzing several glycyl peptides. Because the MVs were bilayered, they readily fused with the outer membrane of gram-negative bacteria. They also adhered to the cell wall of gram-positive bacteria. g-MVs were more effective in lysing other bacteria because, in addition to the autolysins, they also contained small amounts of gentamicin. The bactericidal activity was 2.5 times the MIC of gentamicin, which demonstrates the synergistic effect of the antibiotic with the autolysins. n-MVs were capable of killing cultures of P. aeruginosa with permeability resistance against gentamicin, indicating that the fusion of n-MV to the outer membrane liberated autolysins into the periplasm, where they degraded the peptidoglycan and lysed the cells. g-MVs had even greater killing power since they liberated both gentamicin and autolysins into these resistant cells. These findings may help develop a conceptually new group of antibiotics designed to be effective against hard-to-kill bacteria.     

Initial characterization of two extracellular autolysins from Pseudomonas aeruginosa PAO1.

Two extracellular autolysins have been detected in the spent culture supernatants of Pseudomonas aeruginosa PAO1 by using renaturing polyacrylamide gel electrophoresis. The two autolysins were isolated from the culture supernatant by trichloroacetic acid precipitation and were shown to have apparent molecular masses of 26 and 29 kDa. The 26-kDa autolysin first appears during the early exponential phase of growth and then declines sharply, while the 29-kDa autolysin first appears in the late exponential phase of growth and continues well into the stationary phase. Fractionation of whole cells indicated that the 26-kDa enzyme was also localized within the periplasm, with a lesser amount of activity associated with the cytoplasmic membrane. The 29-kDa autolytic activity was distributed within the cell equally between the periplasm and the cytoplasmic membrane. The pH optima of the isolated 26- and 29-kDa autolysins are 6.0 and 5.0, respectively. Further evidence from both protease susceptibility and inhibition studies confirms that these two extracellular autolysins isolated from P. aeruginosa PAO1 are separate and distinct.     

Zymogram and Preliminary Characterization of Lactobacillus helveticus Autolysins

The autolysins of Lactobacillus helveticus ISLC5 were detected and partially characterized by renaturing sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis with substrate-containing gels (zymogram). By using lyophilized Micrococcus luteus cells or heated whole cells of L. helveticus ISLC5 (0.2% [wt/vol]) as a substrate, several lytic activities were detected in the whole-cell SDS extract of strain ISLC5 (i) one activity at 42.4 kDa, which was named autolysin A, and (ii) six other activities having very similar molecular weights (29.1, 29.6, 30, 30.8, 31.7, and 32.8 kDa), which were named autolysins B (B1 through B6, respectively). As regards the temporal distribution of the enzymes, autolysins A and B were detected in the cells harvested from the beginning of the exponential growth phase. Autolysin A appeared to be associated only with viable cells, whereas the autolysins B remained associated with the cell envelope several days after the complete loss of culture viability. When SDS-treated walls of L. helveticus ISLC5 were used as a substrate, a supplementary lytic activity appeared at 37.5 kDa; it was considered a peptidoglycan hydrolase, since it was not able to induce lysis of whole-cell substrate. The autolysins of 30 other strains of L. helveticus from various geographical origins were also analyzed by zymogram; all the activity profiles obtained were similar to that of strain ISLC5 in terms of the number of lytic bands and their apparent molecular weights. Only the relative intensities of the lytic bands corresponding to autolysins A and B were variable depending on the strains. This observation suggested that autolysins are highly conserved enzymes. A concentrated crude lysate of the virulent bacteriophage 832-B1 infecting L. helveticus was also analyzed by zymogram; one lytic activity with an apparent molecular weight of 31.7 kDa, very close to the weights of the autolysins B, was observed. Finally, the autolysins of L. helveticus ISLC5 were successfully extracted from whole cells by using a 1 M lithium chloride solution; they were partially purified by precipitation, selective resolubilization, and gel filtration chromatography, which led to a 20-fold increase in specific activity.     

A major autolysin of Pseudomonas aeruginosa: subcellular distribution, potential role in cell growth and division and secretion in surface membrane vesicles.

A 26-kDa murein hydrolase is the major autolysin of Pseudomonas aeruginosa PAO1, and its expression can be correlated with the growth and division of cells in both batch and synchronously growing cultures. In batch cultures, it is detected primarily during the mid-exponential growth phase, and in synchronous cultures, it is detected primarily during the cell elongation and division phases. Immunogold labeling of thin sections of P. aeruginosa using antibodies raised against the 26-kDa autolysin revealed that it is associated mainly with the cell envelope and in particular within the periplasm. It is also tightly bound to the peptidoglycan layer, since murein sacculi, isolated by boiling 4% sodium dodecyl sulfate treatment, could also be immunogold labeled. Since division is due to cell constriction in this P. aeruginosa strain (septa are rarely seen), we cannot comment on the autolysin's contribution to septation, although constriction sites were always heavily labeled. Some labeling was also found in the cytoplasm, and this was thought to be due to the de novo synthesis of the enzyme before translocation to the periplasm. Interestingly, the autolysin was also found to be associated with natural membrane vesicles which blebbed from the surface during cell growth; the enzyme is therefore part of the complex makeup of these membrane packages of secreted materials (J. L. Kadurugamuwa and T. J. Beveridge, J. Bacteriol. 177:3998-4008, 1995). The expression of these membrane vesicles was correlated with the expression of B-band lipopolysaccharide.     

Cell cycle parameters of Proteus mirabilis: interdependence of the biosynthetic cell cycle and the interdivision cycle.

We investigated the time periods of DNA replication, lateral cell wall extension, and septum formation within the cell cycle of Proteus mirabilis. Cells were cultivated under three different conditions, yielding interdivision times of approximately 55, 57, and 160 min, respectively. Synchrony was achieved by sucrose density gradient centrifugation. The time periods were estimated by division inhibition studies with cephalexin, mecillinam, and nalidixic acid. In addition, DNA replication was measured by thymidine incorporation, and murein biosynthesis was measured by incorporation of N-acetylglucosamine into sodium dodecyl sulfate-insoluble murein sacculi. At interdivision times of 55 to 57 min murein biosynthesis for reproduction of a unit cell lasted longer than the interdivision time itself, whereas DNA replication finished within 40 min. Surprisingly, inhibition of DNA replication by nalidixic acid did not inhibit the subsequent cell division but rather the one after that. Because P. mirabilis fails to express several reactions of the recA-dependent SOS functions known from Escherichia coli, the drug allowed us to determine which DNA replication period actually governed which cell division. Taken together, the results indicate that at an interdivision time of 55 to 57 min, the biosynthetic cell cycle of P. mirabilis lasts approximately 120 min. To achieve the observed interdivision time, it is necessary that two subsequent biosynthetic cell cycles be tightly interlocked. The implications of these findings for the regulation of the cell cycle are discussed.     

Murein biosynthesis in synchronized cells of Proteus mirabilis. Quantitative analysis of O-acetylated murein subunits and of chain terminators incorporated into the sacculus during the cell cycle

Cells of Proteus mirabilis, synchronized by sucrose density gradient centrifugation, were grown in complex medium containing radioactive N-acetylglucosamine. At various times, labelled murein sacculi were isolated and digested with endo-N,O-acetylmuramidase from Chalaropsis. The murein fragments thus obtained were separated into disaccharide peptides as the monomeric subunits and into peptide-cross-linked subunits by gel filtration. The subunits were further differentiated into O-acetylated and non-O-acetylated species, and into subunits containing anhydro-N-acetylmuramic acid which were glycan chain terminators in the native sacculi. Quantification of the subunit species gave the following results. At specific times during the cell cycle, murein subunits were lost from the polymer and a transient decrease in cross-linkage was observed. The overall degree of cross-linkage in mature murein, i.e. the ratio of peptide-cross-linked subunits versus uncross-linked subunits, was 1.15 as determined by regression analysis. Anhydro-N-acetylmuramic-acid-containing murein subunits representing glycan chain terminators were found either peptide-cross-linked or uncross-linked as monomers. Since these two subunit species were recovered in a defined ratio of 1.6, mature murein consisted of at least two different types of glycan chains. On average, each chain contained 15.4 murein subunits. About 60% of the murein subunits in mature murein were O-acetylated and showed a higher degree of cross-linkage than the non-O-acetylated portion. Finally, following the composition of the sacculus during the cell cycle revealed a complex precursor-product relationship between non-O-acetylated and O-acetylated subunits during murein maturation. The data allowed us to deduce several features of the assembly process of murein sacculi.     

Bactericidal effect of gentamicin-induced membrane vesicles derived from Pseudomonas aeruginosa PAO1 on gram-positive bacteria

Previous studies have shown that gentamicin-induced membrane vesicles (g-MVs) from Pseudomonas aeruginosa PAO1 possess both the antibiotic (gentamicin) and a potent peptidoglycan hydrolase (PGase; autolysin) that is effective in killing gram-negative pathogens. This present study evaluated the therapeutic potential of g-MVs against four gram-positive bacteria. Bactericidal assays and electron microscopy of thin sections revealed that Bacillus subtilis 168 and Staphylococcus aureus D2C were susceptible to killing mediated by g-MVs, Listeria monocytogenes ATCC 19113 was slightly susceptible, whereas Enterococcus hirae ATCC 9790 was unaffected. g-MVs were generally more effective against the bacteria than was soluble gentamicin, suggesting they could have more killing power than natural membrane vesicles containing no antibiotic. Electron microscopy and hydrophobic interaction chromatography showed that more membrane vesicles (MVs) initially attached to B. subtilis (hydrophilic) than to predominantly hydrophobic E. hirae, L. monocytogenes, and S. aureus. Zymograms containing murein sacculi as an enzyme substrate illustrated that all organisms except E. hirae were sensitive to the 26-kDa autolysin to varying degrees. Peptidoglycan O-acetylation did not influence susceptibility to MV-mediated lysis. Though not universally effective, the g-MV delivery system remains a promising therapeutic alternative for specific gram-positive infections.     

Production and characterization of a monoclonal antibody to the O-acetylated peptidoglycan of Proteus mirabilis.

A monoclonal antibody (PmPG5-3) specific for the O-acetylated peptidoglycan of Proteus mirabilis 19 was produced by an NS-1 myeloma cell line and purified from ascites fluid by a combination of ammonium sulfate precipitation and affinity chromatography. The monoclonal antibody (an immunoglobulin M) was characterized by a competition enzyme-linked immunosorbent assay to be equally specific for both insoluble and soluble O-acetylated peptidoglycan but weakly recognized chemically de-O-acetylated P. mirabilis peptidoglycan, the non-O-acetylated peptidoglycans from Escherichia coli and Bacillus subtilis, and the peptidoglycan monosaccharide precursors N-acetylglucosamine and N-acetylmuramic acid dipeptide. The monoclonal antibody did not react with D-alanine or lipopolysaccharide isolated from P. mirabilis. Based on this evidence, the binding epitope on the P. mirabilis peptidoglycan is predicted to be linear and to comprise the glycan backbone, including both the N- and O-acetyl moieties. Monoclonal antibody PmPG5-3 was used to localize the O acetylation of the P. mirabilis peptidoglycan by immunoelectron microscopy. Murein sacculi of P. mirabilis were heavily and randomly labelled with the immunogold, whereas very little labelling and no labelling were observed on the sacculi isolated from de-O-acetylated P. mirabilis and E. coli, respectively. Based on the apparent pattern of immunogold labelling, a physiological role for peptidoglycan O acetylation in P. mirabilis is proposed.     

Penicillin-binding proteins of listeria monocytogenes--a re-evaluation

Intact Listeria monocytogenes cells or membranes isolated from them were treated with [3H]penicillin to allow identification of the penicillin binding proteins (PBPs) located in the cytoplasmic membrane. In the former case the PBPs were released from the cells following disruption of the cell wall murein with Listeria monocytogenes bacteriophage lysin. The procedure described by Dougherty et al. (1996) for Escherichia coli, with some modifications, was used to evaluate the M(r)s of the individual PBPs and allowed direct quantitation of their copy number.     

Homology among bacterial catalase genes

Catalase activities in crude extracts of exponential and stationary phase cultures of various bacteria were visualized following gel electrophoresis for comparison with the enzymes from Escherichia coli. Citrobacter freundii, Edwardsiella tarda, Enterobacter aerogenes, Klebsiella pneumoniae, and Salmonella typhimurium exhibited patterns of catalase activity similar to E. coli, including bifunctional HPI-like bands and a monofunctional HPII-like band. Proteus mirabilis, Erwinia carotovora, and Serratia marcescens contained a single band of monofunctional catalase with a mobility intermediate between the HPI-like and HPII-like bands. The cloned genes for catalases HPI (katG) and HPII (katE) from E. coli were used as probes in Southern hybridization analyses for homologous sequences in genomic DNA of the same bacteria. katG was found to hybridize with fragments from C. freudii, Ent. aerogenes, Sal. typhimurium, and K. pneumoniae but not at all with Ed. tarda, P. mirabilis, S. marcesens, or Er. carotovora. katE hybridized with C. freundii and K. pneumoniae DNAs and not with the other bacterial DNAs.     

Thickness and elasticity of gram-negative murein sacculi measured by atomic force microscopy

Atomic force microscopy was used to measure the thickness of air-dried, collapsed murein sacculi from Escherichia coli K-12 and Pseudomonas aeruginosa PAO1. Air-dried sacculi from E. coli had a thickness of 3.0 nm, whereas those from P. aeruginosa were 1.5 nm thick. When rehydrated, the sacculi of both bacteria swelled to double their anhydrous thickness. Computer simulation of a section of a model single-layer peptidoglycan network in an aqueous solution with a Debye shielding length of 0.3 nm gave a mass distribution full width at half height of 2.4 nm, in essential agreement with these results. When E. coli sacculi were suspended over a narrow groove that had been etched into a silicon surface and the tip of the atomic force microscope used to depress and stretch the peptidoglycan, an elastic modulus of 2.5 x 10(7) N/m(2) was determined for hydrated sacculi; they were perfectly elastic, springing back to their original position when the tip was removed. Dried sacculi were more rigid with a modulus of 3 x 10(8) to 4 x 10(8) N/m(2) and at times could be broken by the atomic force microscope tip. Sacculi aligned over the groove with their long axis at right angles to the channel axis were more deformable than those with their long axis parallel to the groove axis, as would be expected if the peptidoglycan strands in the sacculus were oriented at right angles to the long cell axis of this gram-negative rod. Polar caps were not found to be more rigid structures but collapsed to the same thickness as the cylindrical portions of the sacculi. The elasticity of intact E. coli sacculi is such that, if the peptidoglycan strands are aligned in unison, the interstrand spacing should increase by 12% with every 1 atm increase in (turgor) pressure. Assuming an unstressed hydrated interstrand spacing of 1.3 nm (R. E. Burge, A. G. Fowler, and D. A. Reaveley, J. Mol. Biol. 117:927-953, 1977) and an internal turgor pressure of 3 to 5 atm (or 304 to 507 kPa) (A. L. Koch, Adv. Microbial Physiol. 24:301-366, 1983), the natural interstrand spacing in cells would be 1.6 to 2.0 nm. Clearly, if large macromolecules of a diameter greater than these spacings are secreted through this layer, the local ordering of the peptidoglycan must somehow be disrupted.     

Novel type of murein transglycosylase in Escherichia coli.

The purification and properties of a novel type of murein transglycosylase from Escherichia coli are described. The purified enzyme appears as a single band on sodium dodecyl sulfate-polyacrylamide gels and has an apparent molecular weight of approximately 65,000 as estimated by gel filtration and gel electrophoresis. It degrades pure murein sacculi from E. coli almost completely into low-molecular-weight products. The two prominent muropeptide fragments in the digest are the disaccharide-tripeptide N-acetylglucosamine-N-acetylmuramic acid-L-alanine-D-iso-glutamic acid-meso-diaminopimelic acid and the corresponding disaccharide-tetrapeptide N-acetylglucosamine-N-acetylmuramic acid-L-alanine-D-iso-glutamic acid-meso-diaminopimelic acid-D-alanine. The unique feature of these compounds is that the disaccharide has no reducing end group and that the muramic acid residue possesses an internal 1 leads to 6 anhydro linkage. The new lytic enzyme is designated as a murein: murein transglycosylase. Its possible role in the rearrangement of murein during cell growth and division is discussed.     

Serotypes and pyocin types of Pseudomonas aeruginosa isolated from natural waters

The susceptibility of strains of Enterobacter cloacae, Klebsiella pneumoniae, Serratia marcescens, Pseudomonas aeruginosa, Proteus mirabilis and Escheriehia coli to six aminoglycosides was tested in media of different osmolarity and ionic content. We observed that increasing osmolarity decreased susceptibility of these Gram-negative bacteria to all antibiotics used. On the other hand, raising of ionic strength increased the susceptibility to tobramycin, neomycin and gentarnicin in all bacteria tested.     

Expression of the lactose transposon Tn951 in Escherichia coli, Proteus and Pseudomonas

The control of beta-galactosidase specified by the lactose transposon Tn951 (inserted into RP1 to give pGC9114) has been studied in Escherichia coli K12, Proteus mirabilis, Pseudomonas aeruginosa and Pseudomonas putida; in the first two species comparison could be made with Flac. In E. coli K12, the Tn951 and chromosomally encoded enzymes showed marked qualitative differences in regulatio, the former giving a substantially lower maximum induced level and induction ratio. Several parameters were slightly affected by strain background. In P. mirabilis, beta-galactosidase control determined by both Flac (in accord with earlier work) and pGC9114 was markedly different from E. coli in that maximal induced levels were about an order of magnitude lower and the induction ratio was reduced to 3 to 5. In Ps. aeruginosa and Ps. putida, Tn951-specified lac expression was qualitatively similar to that in P. mirabilis. Possible reasons for anomalous expression in Proteus and Pseudomonas are discussed.     

The extraction of cell walls of Pseudomonas aeruginosa with aqueous phenol. The insoluble residue and material from the aqueous layers

1. The insoluble residue and material present in the aqueous layers resulting from treatment of cell walls of Pseudomonas aeruginosa with aqueous phenol were examined. 2. The products (fractions AqI and AqII) isolated from the aqueous layers from the first and second extractions respectively account for approx. 25% and 12% of the cell wall and consist of both lipopolysaccharide and muropeptide. 3. The lipid part of the lipopolysaccharide is qualitatively similar to the corresponding material (lipid A) from other Gram-negative organisms, as is the polysaccharide part. 4. The insoluble residue (fraction R) contains sacculi, which also occur in fraction AqII. On hydrolysis, the sacculi yield glucosamine, muramic acid, alanine, glutamic acid and 2,6-diaminopimelic acid, together with small amounts of lysine, and they are therefore similar to the murein sacculi of other Gram-negative organisms. Fraction R also contains substantial amounts of protein, which differs from that obtained from the phenol layer. 5. The possible association or aggregation of lipopolysaccharide, murein and murein sacculi is discussed.     

Metalloantibiotics: synthesis and antibacterial activity of cobalt(II), copper(II), nickel(II) and zinc(II) complexes of kefzol

Kefzol (kzl), a beta-lactam antibiotic, possesses various donor sites for interaction with transition metal(II) ions [Co(II), Cu(II), Ni(II) and Zn(II)] to form complexes of the type [M(kzl)2]Cl2 and [M(kzl)Cl], with molar ratio of metal: ligand (M:L) of 1:2 and 1:1 respectively. These complexes were prepared and characterized by physicochemical and spectroscopic methods. Their IR and NMR spectra suggest that kefzol potentially acts as a bidentate, tridentate as well as monoanionic tetradentate ligand. The complexes have been screened for antibacterial activity and results were compared with the activity of the uncomplexed antibiotic against Staphylococcus aureus, Klebsiella pneumoniae, Pseudomonas aeruginosa, Escherichia coli and Proteus mirabilis. The metal complexes were found to be more potent against one or more bacterial species than the uncomplexed kefzol.     

Fishmeal extract bile salt lactose agar--a differential medium for enteric bacteria

Fishmeal extract bile salt lactose agar (FEBLA), a new differential medium for enteric bacteria was developed and evaluated for its ability to grow and differentiate lactose fermenters (LF) from non-lactose fermenters (NLF) in comparison with MacConkeys agar. Performance of FEBLA was at par with the latter. On FEBLA medium, the contrast between LF and NLF colonies was pronounced and Klebsiella pneumoniae produced more mucoid colonies than on MacConkeys agar (Hi Media). Unlike MacConkeys agar, a 24 h culture of K. pneumoniae cells on FEBLA were longer and thicker with abundant capsular material around the bacilli. Escherichia coli produced long and thick cells but only after 48h. No change in cell morphology was evident with regard to Salmonella typhi, S. paratyphi A, Shigella flexneri, Pseudomonas aeruginosa, Proteus mirabilis, Proteus vulgaris, Citrobacter koseri and Acinetobacter baumannii. Performance of the medium was controlled using E. coli and S. flexneri. FEBLA is simple, cost effective and may be a suitable alternative in the preliminary identification of enteric bacteria.     

In vitro synthesis of cross-linked murein and its attachment to sacculi by PBP1A from Escherichia coli

The penicillin-binding protein (PBP) 1A is a major murein (peptidoglycan) synthase in Escherichia coli. The murein synthesis activity of PBP1A was studied in vitro with radioactive lipid II substrate. PBP1A produced murein glycan strands by transglycosylation and formed peptide cross-links by transpeptidation. Time course experiments revealed that PBP1A, unlike PBP1B, required the presence of polymerized glycan strands carrying monomeric peptides for cross-linking activity. PBP1A was capable of attaching nascent murein synthesized from radioactive lipid II to nonlabeled murein sacculi. The attachment of the new material occurred by transpeptidation reactions in which monomeric triand tetrapeptides in the sacculi were the acceptors.     

The two-step lysis system of pneumococcal bacteriophage EJ-1 is functional in Gram-negative bacteria: triggering of the major pneumococcal autolysin in Escherichia coli

The holin function Ejh of the pneumococcal bacteriophage EJ-1 has been characterized. It shows structural features similar to, and functionally complemented, the prototype member of the holin family. In Escherichia coli and Pseudomonas putida the Ejh product caused cellular death, and changes in cell morphology could be accounted for by lesions in the cytoplasmic membrane. Expression of ejh resulted in the inhibition of growth in a variety of phylogenetically distant bacterial genera, suggesting a broad spectrum of action. Concomitant expression of the ejh and ejl (encodes a lysin) genes led to lysis of E. coli and P. putida cells. Remarkably, the Ejl lysin was able to attack murein from bacteria lacking choline in their sacculi, which suggests that pneumococcal lysins have a broader substrate specificity than previously assumed. Furthermore, the Ejh holin was able to trigger activity of the major pneumococcal autolysin cloned and expressed in E. coli , and this raised new questions about the regulation of this model autolysin. A new function for holins in systems where the phage lysin is supposed to be associated with the membrane is proposed.