J van Heijenoort

Ampd, Essential for Both Beta-Lactamase Regulation and Cell Wall Recycling, Is a Novel Cytosolic N-Acetylmuramyl-L-Alanine Amidase

In enterobacteria, the ampD gene encodes a cytosolic protein which acts as a negative regulator of β‐lactamase expression. It is shown here that the AmpD protein is a novel N‐acetylmuramyl‐L‐alanine amidase (E.C.3.5.1.28) participating in the intracellular recycling of peptido‐glycan fragments. Surprisingly, AmpD exhibits an exclusive specificity for substrates containing anhydro muramic acid. This anhydro bond is mainly found in the peptidoglycan degradation products formed by the periplasmic lytic transglycosylases and thus might behave as a‘recycling tag’allowing the enzyme to distinguish these fragments from the newly synthesized peptidoglycan precursors. The AmpD substrate (or substrates) which accumulates in the absence of the corresponding enzymatic activity acts as an intracellular positive effector for β‐lactamase expression and might represent an element of a communication network between the chromosome and the cell wall peptidoglycan.

Effects of Moenomycin on Escherichia coli

The antibiotic moenomycin is a valuable biochemical tool for studying the metabolism of peptidoglycan and the autolytic system in Escherichia coli, since as a specific inhibitor of peptidoglycan polymerases it can efficiently promote cell lysis. In liquid media the bacteriolytic effect on E. coli K12 was dependent on the concentration of moenomycin, on growth phase and on growth rate. Before lysis cells underwent major morphological alterations. In sucrose-containing medium complete transformation to osmotically sensitive spheroplasts was easily achieved by addition of moenomycin. The minimum inhibitory concentration of the antibiotic varied with the strain of E. coli and was highly dependent on the growth medium. A tritiated derivative of moenomycin, [3H]decahydromoenomycin A, was prepared and found to have the same inhibiting efficiency. Its binding to E. coli membranes and membrane proteins was investigated. The absence of irreversible binding suggested that moenomycin might be a competitive inhibitor of the peptidoglycan polymerases. Spontaneous moenomycin resistant variants were isolated at a frequency of about 10(-9).

Multilayered Distribution of Peptidoglycan in the Periplasmic Space of Escherichia coli

When a staining technique using phosphotungstic acid (PTA) in 10% (w/v) chromic acid was applied to cells of Escherichia coli, the periplasmic space was seen as a dark 15-nm-thick layer of uniform appearance and constant width. Our observations are consistent with peptidoglycan being the main material stained. Isolated sacculi as well as purified peptidoglycan (protein free) were also stained by the same procedure, the thickness of the peptidoglycan being 8.8 +/- 1.8 and 6.6 +/- 1.5 nm, respectively. The increased thickness of the PTA-stained layer in stationary phase cells correlated well with the increased thickness of isolated sacculi or purified peptidoglycan and with the increased amount of peptidoglycan in such cells. Thickness measurements on isolated peptidoglycan were compatible with a two to three layer structure for material from exponential phase cells and with a four to five layer structure for that from stationary phase cells. Furthermore, the results indicated an uneven distribution of peptidoglycan material in the periplasmic space, the peptidoglycan spanning the space from the inner to the outer membrane.

Peptidoglycan biosynthesis in Escherichia coli: variations in the metabolism of alanine and D-alanyl-D-alanine

The in vivo functioning of the alanine/D-alanyl-D-alanine pathway of Escherichia coli was investigated by determining precursor pool levels and specific enzyme activities under various growth conditions. Cells grown on D- or L-alanine showed several remarkable features compared with cells grown on other carbon sources: 10-fold higher values of the D-alanyl-D-alanine and the UDP-MurNAc-pentapeptide pools, a 240-fold increase of the alanine racemase activity, and the absence of bacteriolysis after treatment with D-cycloserine at high concentrations (50 micrograms ml-1). In cells grown on glucose, D-cycloserine (1 micrograms ml-1) led to depletion of the D-alanyl-D-alanine pool and to lysis, which was efficiently antagonized by chloramphenicol. A threefold increase of the dipeptide pool was observed when cells were treated with chloramphenicol alone. The alanine racemase activity was lowest in glucose-grown cells and the D-alanine:D-alanine ligase and D-alanyl-D-alanine-adding activities were the same whatever the carbon source. Molecular masses of 53-56 kDa and 56-60 kDa were estimated for the partially purified inducible alanine racemase and D-alanine:D-alanine ligase respectively.