Mutants constitutive for the novel outer membrane protein Ic (e or E) contained a recently discovered binding protein for sn-glycerol-3-phosphate. The corresponding parental strains missing the outer membrane protein Ic (e, E) were negative or strongly reduced in the synthesis of the binding protein. In addition, strains that were previously isolated as mutants constitutive for the sn-glycerol-3-phosphate transport system (ugp+ mutants) and that produced the novel periplasmic proteins GP1 to GP4 also synthesized a new outer membrane protein with the same electrophoretic mobility on sodium dodecyl sulfate-polyacrylamide gels as protein Ic. Screening of different ugp+ mutants revealed the existence of three types in respect to the four novel periplasmic proteins GP1, -2, -3, and -4: (i) one containing all four proteins; (ii) one containing only proteins GP1, -2, and -3; (iii) one containing only proteins GP1, -2, and -4. In confirmation of the data presented in the accompanying paper by Tommassen and Lugtenberg (J. Bacteriol. 143:151-157, 1980), we found that purified GP1 is identical to alkaline phosphatase, whereas purified GP3 has binding activity of inorganic phosphate and is identical to the phosphate-binding protein. Moreover, growth conditions that lead in a wild-type strain to the derepression of alkaline phosphatase synthesis also derepressed the synthesis of the sn-glycerol-3-phosphate-binding protein as well as the corresponding transport system. Thus, the new sn-glycerol-3-phosphate transport system is part of the alkaline phosphatase regulatory system.Recently, we found a new transport system for sn-glycerol-3-phosphate (G3P) in mutants (ugp+ mutants) that arose as G3P+ suppressors in strains carrying a defective transport system for G3P coded for by gipT (3) at 48 min on the Escherichia coli chromosome (4). These ugp+ mutants mapped outside gipT and synthesized the new periplasmic proteins (GP1, -2, and -3). One of these proteins (GP2) was identified as a high-affinity binding protein for G3P (2).Along other lines, we were interested in the pore-forning activity of osmotic shock fluid in black lipid films due to soluble outer membrane proteins Ia and lb (6). In this respect, we were interested whether or not the newly discovered outer membrane protein Ic (18) also exhibited pore-forming activity when shock fluids of this mutant were used in the black lipid pore assay. Although this was indeed the case (R. Benz and U. Henning, manuscript in preparation) we noticed that the periplasmic proteins of the Iccarrying strain exhibited a polyacrylamide gel pattern very similar to those obtained from shock fluids of our G3P+ ugp+ mutants. 142

The ugp -dependent transport system for sn -glycerol-3-phosphate has been characterized. The system is induced under conditions of phosphate starvation and in mutants that are constitutive for the pho regulon. The system does not operate in membrane vesicles and is highly sensitive toward osmotic shock. The participation of a periplasmic binding protein in the transport process can be deduced from the isolation of transport mutants that lack the binding protein. As with other binding protein-dependent transport systems, this protein appears to be necessary but not sufficient for transport activity. The isolation of mutants has become possible by selection for resistance against the toxic analog 3,4-dihydroxybutyl-1-phosphonate that is transported by the system. sn -Glycerol-3-phosphate transported via ugp cannot be used as the sole carbon source. Strains have been constructed that lack alkaline phosphatase and glycerol kinase. In addition, they are constitutive for the glp regulon and contain high levels of glycerol-3-phosphate dehydrogenase. Despite the fact that these strains exhibit high ugp -dependent transport activity for sn -glycerol-3-phosphate they are unable to grow on it as a sole source of carbon. However, when cells are grown on an alternate carbon source, 14 C label from [ 14 C] sn -glycerol-3-phosphate appears in phospholipids as well as in trichloroacetic acid-precipitable material. The incorporation of 14 C label is strongly reduced when sn -glycerol-3-phosphate is the only carbon source. In the presence of an alternate carbon source, this inhibition is relieved, and sn -glycerol-3-phosphate transported by ugp can be used as the sole source of phosphate.

The ability of tetracycline to pass through phospholipid bilayers by diffusion was investigated. Liposomes did not retain enclosed tetracycline. Accumulation of tetracycline was observed with liposomes containing entrapped Tet repressor protein. These results indicate that the drug can pass through lipid bilayers. The antibiotic was also shown to bind to liposomes and isolated phospholipids.Incubation of Escherichia coli cells in medium containing tetracycline leads to uptake and accumulation of the antibiotic by these cells (1,4,5). To date no involvement of proteins in these processes has been demonstrated (5, 6). However, only a fraction of the antibiotic accumulation appears to be mediated by an active process: the treatment of cells with uncouplers reduces accumulation but does not abolish it. Active uptake of tetracycline has been shown to be driven by the proton motive force of the cell (5), but the basis for its accumulation by de-energized cells has so far eluded explanation. The present study was initiated to investigate the interaction of tetracycline with dissolved phospholipids and with phospholipid vesicles in the absence of membrane proteins.To test whether tetracycline can pass through phospholipid bilayers via passive diffusion, the permeability properties of liposomes to tetracycline were investigated. Two series of experiments were carried out. (i) The release of [3H]tetracycline enclosed in liposomes was determined by dialysis of liposomes against tetracycline-free buffer. Liposomes which contained radioactive rubidium, which is known not to pass through the lipid bilayer (8), served as a control. Tetracycline concentration in liposomes decreased gradually, whereas the concentration of liposome-enclosed 86Rb remained constant (after the first 30 min) over a period of 27 h (compare Fig. 1A and B). The rapid initial decrease seen with 86Rb is due to diffusion of 86Rb not enclosed in liposomes. The kinetics of diffusion of tetracycline and 86Rb from the dialysis tubes (without liposomes) is also shown in Fig. 1A and B, respectively. The data suggest that tetracycline present in the internal aqueous volume of the liposomes diffused passively through the phospholipid bilayers, whereas 86Rb was retained. The retention during the initial phase of dialysis of (nonencapsulated) tetracycline in the presence of liposomes (Fig. 1A) (3) (without alkali treatment) in the presence and absence of 0.5% sodium dodecyl sulfate. Protein could not be detected when the assay was carried out in the absence of the detergent. Equilibrium dialysis of Tet repressor protein encapsulated in liposomes (see below) in the presence of 100 ,ug each of trypsin and pronase per ml (Boehringer Mannheim Corp.) did not affect the results, confirming the absence of Tet repressor protein on the outside of liposomes. Equilibrium dialysis was performed with small dialysis tubes (VisKing; diameter, 1 cm; length, 2 cm). Either 2 ml of a liposome suspension (4.9 ,ul internal aqueous volume) containing 3 jig of Tet repressor protein per ml...

Experiments measuring the initial uptake of commercial (3H) tetracycline exhibit two distinct kinetic phases: a rapid phase followed by a slow phase. (3H) tetracycline purified by chromatography on a Dowex 50WX2 column exhibited only monophasic rapid uptake when tested with susceptible Escherichia coli cells. Cyanide inhibited the uptake of purified (3H) tetracycline only partially while transport of proline and maltose was entirely abolished. Energy independent accumulation of tetracycline may be accounted for by binding to cellular constituents. Uptake of tetracycline--as measured by inhibition of beta-galactosidase synthesis--was strongly affected by a shift in temperature from 37 degrees C to 21 degrees C while carrier-mediated transport systems revealed only minor reductions. Taken together with the non-saturability of tetracycline uptake and the evidence for diffusion of tetracycline through phospholipid bilayers [Argast and Beck (1984) Antimicrob Agents Chemother 26:263-265] these data support the hypothesis that tetracycline enters the cytoplasm by diffusion.

Two genes (tetC and tetD) were identified and located on transposon Tn10 between gene tetA and insertion sequence IS10R. Genes tetC and tetD encode proteins of apparent subunit molecular weights of 23,000 and 18,000, respectively. The TetD protein was found to be membrane associated. Tetracycline resistance levels promoted by transposon Tn10 were found to be unaffected in Escherichia coli K-12 when mutants lacking tetC or tetC and tetD were tested. The nucleotide sequence of genes tetC and tetD is reported in the accompanying article (K. Schollmeier and W. Hillen, J. Bacteriol. 160:499-503, 1984).