MECHANISM OF ACTION OF PENICILLIN Maer JOSHUA LEDERBERG Department of Genetics, University of Wisconsin, Madison, Wisconsin Reprinted from JourNAL or BAcTERIOLOGY Vol. 73, No. 1, p. 144, January, 1957 Printed in U.S.A. Reprinted from JouRNAL or BacTERIOLOGY Vol. 73, No. 1, p. 144, January, 1957 Printed in U.S.A. NOTE MECHANISM OF ACTION OF PENICILLIN! JOSHUA LEDERBERG Depariment of Genetics, University of Wisconsin, Madison, Wisconsin Received for publication October 1, 1956 “The mechanism whereby penicillin exerts its cytotoxic effect remains obscure” (Eagle and Saz, Ann. Rey. Microbiol., 9, 173, 1955) notwith- standing the universal use of this antibiotic in chemotherapy. However, there has been a con- cordance by many workers on the development of protoplasts or L-forms of bacteria (for review see Liebermeister and Kellenberger, Z. Natur- forsch., 11, 200, 1956). These observations sup- port the argument that penicillin inhibits cell-wall synthesis, and thereby provokes osmotic fragility in the excoriated bacteria (Cooper, Bacteriol. Revs., 20, 28, 1956). The argument may be illustrated with obser- vations on Escherichia colt strain K-12 (Leder- berg, Proc. Natl. Acad. Sei. U. 8., 42, 8M4, 1956 —where the point was not amplified). Cells actively growing in customary broth media will lyse after one to two hr exposure to penicillin. In a protective hypertonic medium, i. e., one supplemented with m/3 sucrose plus m/100 MgSO,, the treated cells do not lyse but instead they balloon into spherical “protoplasts.”” Direct microscopic observations showed a one-for-one conversion of rods into protoplasts. The proto- plast suspension is osmotically fragile and lyses when diluted into water or ordinary broth. In the protective medium, however, the protoplasts remain almost fully viable, and will revert to typical (colony-forming) rods when diluted in protective media Jacking penicillin. Therefore, the bactericidal effect of penicillin in ordinary media is sufficiently explained by the induced osmotic fragility. As non-growing cells are not killed by penicillin, new wall-formation, rather than the existing wall, is the probable target. Lower concentrations of penicillin provoke the 1 Paper No. 641 of the Department of Genetics. This work has been supported by a researeh grant (C-2157) from the National Cancer In- stitute, Public Health Service. formation of long filamentous forms and may have a bacteriostatic effect by virtue of the inhi- bition of cell division. In the production of proto- plasts, it was observed that dividing cells usually swelled first from the point of incipient separa- tion. This suggests that the division-septum is especially sensitive to penicillin. Filamentous forms would arise when septum-formation was blocked without impairment of synthesis of the outer wall. It remains to define the target in biochemical terms. The simplest speculation is that penicillin inhibits a specifie wall-building polymerase. The chemotherapeutic specificity would then follow from the unique makeup of bacterial cell walls. Park (J. Biol. Chem., 194, 877, 1952) reported the accumulation of uridinc-pyrophosphate de- Tivatives of amino-sugars and various amino acids in penicillin-treated staphylococci. These derivatives may represent the activated forms of the residues for their polymeric condensation, which accumulate owing to the block in this reaction.?, However, more remote influences on eell-wall formation cannot be precluded. At any rate, further studies of antibiotic effects must be conducted with protected protoplasts, rather than with lysed or lysing cells in which the ramification of secondary lesions is an inevitable complication. The viability of penicillin-treated cells in pro- tective media is further indicated by their pro- liferation in penicillin-containing agar (but not in broth) to form L-colonies. That is, the “proto- plast’”’ is equivalent to the initial stage, the large body, of the L-eycle of Klieneberger-Nobel, Dienes, and others. Certain bacteria, such as Proteus, form protoplasts which are unusually resistant to osmotic shock, and have therefore been more amenable to previous experimentation on L-forms, ? For a more complete statement and substantia- tion of the same proposal, see Park and Strominger, Science, (accepted for publication), 1956. 144