Turnover of murein in oligopeptide permease-negative Escherichia coli cells appeared to be minimal or nonexistent. In one strain in which it was possible to measure turnover during the first generation of chase, it was found that the rate of turnover was constant throughout a chase of three generations. This result suggests that an "inside-to-outside" mode of growth of the sacculus does not occur in E. coli. Turnover, though minimal, was significantly higher from cells labeled uniformly than from cells labeled only in the lateral wall, suggesting that a significant portion of the observed turnover is related to cell separation. Actually, turnover only appeared to be minimal in opp mutant strains. Tripeptides were being released by turnover at a rate of about 509% per generation and then were efficiently recycled. This suggests that in addition to opp, a low-affinity uptake system for tripeptide derived from the sacculus may exist.For many years, it was thought that the Escherichia coli cell wall consisted of a single layer of murein and that during growth there was little or no release of wall material. Both of these assumptions have recently been questioned. Since 1984, several lines of evidence have suggested that the murein sacculus of exponentially growing E. coli cells might be multilayered, with a thickness of up to three layers. Electron microscopic observations based on new embedding techniques and freeze substitution revealed a periplasmic space filled with an electron-dense layer of uniform thickness which was several times thicker than the peptidoglycan layer observed in previous electron micrographs (9). The authors proposed that the peptidoglycan formed a gel consisting of a cross-linked outer shell with additional lesscross-linked material attached to the inner surface. Consistent with this view was the observation that staining of thin sections of E. coli by the Rambourg technique (phosphotungstic acid in 10% chromic acid) revealed a similar 15-nmthick layer in the periplasm apparently consisting primarily of peptidoglycan (13) Holtje and Glauner have reported the presence of trimers in muramidase digests of E. coli sacculi and have suggested that this is another indication that multiple layers may exist, since a trimer cannot be present in a monolayer. These authors went on to suggest that a multilayered wall, should one exist in E. coli, ought to be formed by the "inside-tooutside" mechanism that is known to occur in gram-positive bacilli (10).In the case of Bacillus subtilis, after the murein is labeled for a short pulse and then chased, there is a lag of about one generation before any radioactivity is released into the medium. After the first generation, about 50% of the radioactivity of the cylindrical wall is released per generation (15). This is interpreted to indicate that peptidoglycan is initially laid down next to the cytoplasmic membrane and eventually reaches the external surface, where it is degraded by autolytic enzymes. Since the principal enzymes used to polymerize and cross-link...

We examined the predacious gram-negative bacterium Bdellovibrio bacteriovorous 109J and free-living strains 109J-A1 and 109J-KA1 derived therefrom for penicillin-binding proteins (PBPs). We compared their PBPs with those of the host bacterium, Escherichia coli, and with those of a facultatively predacious bdellovibrio, B. stolpii UKi2, grown axenically. The multiple PBPs of the 109J strains and of UKi2 differed from each other and from those of E. coli, which suggests that screening for PBPs may be a convenient way to determine to what extent the bdellovibrios may represent a diverse group of organisms. A method for labeling furazlocillin and cefaperizone with iodine-125 is also described.We have examined the predacious gram-negative bacterium Bdellovibrio bacteriovorous 109J and free-living strains 109J-A1 and 109J-KA1 derived therefrom for penicillinbinding proteins (PBPs). We have compared their PBPs with those of the host bacterium, Escherichia coli, and with those of a facultatively predacious bdellovibrio, B. stolpii UKi2, grown axenically (2).B. bacteriovorous 109J was grown on stationary E. coli cells in dilute nutrient broth essentially as described by Shilo and Bruff (3). To 500 ml of E. coli RV cells grown to stationary phase in dilute nutrient broth were added 2 x 109 B. bacteriovorous 109J cells (multiplicity of infection, 1:200). The culture was aerated for 16 to 18 h, by which time lysis of the host bacteria was essentially complete. Five milliliters of 0.5% sodium glutamate was added to maintain the bdellovibrios. The bdellovibrios were isolated from the remaining host cells by differential centrifugation. Cells and cell debris were removed by centrifugation at 1,000 x g for 5 min, and the bdellovibrios were recovered from the supernatant by centrifugation at 10,000 x g for 20 min. The vibrios were washed and suspended in 0.5 ml of 10 mM phosphate buffer, pH 7.0. B. bacteriovorous 109J-Al and 109J-KA1 and B. stolpii UKi2 were grown in 1% peptone-0.3% yeast extract at 30°C with aeration and concentrated and washed as described above. The cell pellets of the 109J strains were mustard, whereas the UKi2 cells were mauve. The cells were ruptured by prolonged sonication (5 min at 0°C) and the envelopes were collected by centrifugation at 100,000 x g for 30 min.The cell envelopes were labeled with either tritiated benzyl penicillin or furazlocillin iodinated with 1251. lodination of furazlocillin was carried out as follows.