No abstract

The D-alanyl-D-alanine-adding enzyme encoded by the murF gene catalyzes the ATP-dependent formation of UDP-N-acetylmuramyl-L-gamma-D-Glu-meso-diaminopimelyl-D-Ala-D-Ala (UDP-MurNAc-tripeptide). MurF has been cloned from Escherichia coli and expressed as a glutathione S-transferase (GST) fusion using the tac promoter-based pGEX-KT vector. From induced, broken cell preparations, highly active fusion was recovered and purified in one step by affinity chromatography. The purified fusion protein was strongly inhibited by substrate UDPMurNAc-tripeptide, a response unaltered by changes in assay pH or by cleavage from the fusion partner. However, this effect was suppressed by the addition of 0.5 M NaCl. Initial velocity and dead-end inhibitor studies with the fusion enzyme were most consistent with a sequential ordered kinetic mechanism for the forward reaction in which ATP binds to free enzyme, followed by tripeptide and D-Ala-D-Ala in sequence prior to product release. Reported homologies between the MurF protein and the three preceding steps of cytoplasmic murein biosynthesis, MurC, -D, and -E, [Ikeda et al. (1990) J. Gen. Appl. Microbiol. 36, 179-187], raise the prospect that all of these enzymes will be found to proceed via this mechanism.

We used the molecular beam epitaxial growth of CuCI on CaF2(I I I ) to determine if' scaling theory provides insight into the kinetic mechanisms of heteroepitaxy. %e measured quantitative surface topographs of several films representing the island nucleation, growth, and coalescence regimes of film growth with an atomic force microscope, and found that the static scaling exponent of all the surfaces was a 0.84 0.05. This a value is closer to theoretical predictions in which surface diAusion is the dominant smoothening mechanism than to those involving evaporation and recondensation.The classification scheme used to understand film growth modes over most of the last three decades was proposed by Bauer [I]. This paradigm recognized three different processes that have been named after some of their earliest investigators: Frank-van der Merwe (FM) for monolayer by monolayer growth [2], Volmer-Weber (VW) for initial film nucleation by 3D crystallite growth [3], and Stranski-Krastanov (SK) for formation of an initial uniform layer followed by 3D crystallite growth [4].All of these models are based on thermodynamic considerations, and have been discussed in detail previously [5,6]. The validity of the paradigm depends on the attainment of local equilibrium on the growing surface, which requires that the mass transport processes parallel to the macroscopic surface be fast compared to the flux of arriving species. In modern technological applications, the drive toward lower substrate temperatures and higher growth rates pushes practical growth of materials by vapor phase processes away from the 'idealized thermodynamic models toward a nonequilibrium or kinetic limit.Theoretical considerations showed that the thermodynamic models had to be modified to account for kinetic limitations [7]. The concept of scaling was introduced to the field by Family and Vicsek [8] to provide a framework for understanding the self-afine (or fractal-like [9]) topologies of the nonequilibrium surfaces. Most recently, a group of continuum models based on the competition between roughening of a surface caused by the stochastic arrival of depositing species and smoothening resulting from surface diffusion and other transport processes [10] have been proposed. Three of these models predict different surface topologies: Kardar-Parisi-Zhang (KPZ) [11], Wolf-Villain (WV) [12], and Villain [13] and Lai-Das Sarma (VLS) [14]. However, these models are implicitly valid only for homodeposition processes. The purpose of this paper is to see if some of the insights gained from these kinetic growth theories can be applied in understanding heteroepitaxy as well.According to scaling theory, the discreteness of the de-positing material is the main cause for the growing surface to become self-a%ne; the interface width g, i.e., the standard deviation of the surface height H, can be expressed in the form [8,15] & (t) =L 'f(t/L'), (I a) which reduces to z(t) =L2 for small L with t =const, and to &t'. (t ) = t ' as L =~, (lb) (Ic)where L is the length scale ...