Sep 10, 1990 - The galE mutants of S. typhimurium, G30 ... Effects of DNP on uptake and utilization ofgalactose in galE ..... Bray, D., and P. W. Robbins. 1967.
Vol. 173, No. 10
JOURNAL OF BACTERIOLOGY, May 1991, p. 3128-3133
0021-9193/91/103128-06$02.00/0 Copyright © 1991, American Society for Microbiology
Energy Dependence of O-Antigen Synthesis in Salmonella typhimurium PAMELA A. MARINO, BARBARA C. McGRATH, AND M. J. OSBORN* Department of Microbiology, University of Connecticut Health Center, Farmington, Connecticut 06030 Received 10 September 1990/Accepted 8 March 1991
The uncoupler 2,4-dinitrophenol prevents in vivo synthesis of 0 antigen in Salmonella typhimurium by inhibiting the first reaction of the pathway, formation of galactosyl-pyrophosphoryl-undecaprenol. Inhibition was observed only in intact cells; dinitrophenol had no effect on activity of the synthase enzyme in isolated membrane fractions. In vivo inhibition could not be explained by changes in intracellular nucleotide pools or a shift in the equilibrium of the reaction and appeared to be specific for the first step in the pathway. Neither the subsequent mannosyl transferase, which catalyzes formation of the trisaccharide-lipid intermediate, mannosyl-rhamnosyl-galactosyl-pyrophosphoryl-undecaprenol, nor 0-antigen polymerase was inhibited. In addition, incorporation of galactose into core lipopolysaccharide was only modestly inhibited under conditions in which 0-antigen synthesis was abolished. The results suggest that maintenance of proton motive force is required for access of substrate, UDP-galactose and/or undecaprenyl phosphate, to the active site of the galactosyl-pyrophosphoryl-undecaprenol synthase enzyme.
The 0 antigen of Salmonella typhimurium is synthesized by a multienzyme system of the cytoplasmic (inner) membrane (10). The tetrasaccharide repeating unit of the 0-antigen chain is synthesized and polymerized via a series of oligo- and polysaccharide intermediates linked to the membrane-bound lipid coenzyme, undecaprenyl phosphate. The polymer chain is then transferred en bloc to core lipopolysaccharide, and the completed 0-antigenic lipopolysaccharide is finally translocated to the outer membrane. The final steps of 0-antigen assembly, polymerization and attachment to lipopolysaccharide, take place at the periplasmic face of the inner membrane (7, 8). The transmembrane topology of the glycosyl transferases responsible for synthesis of the tetrasaccharide-lipid repeating unit has not been established, but the available evidence suggests that the oligosaccharide is assembled at the cytoplasmic face of the membrane, where enzyme active sites would have direct access to cytosolic nucleotide sugar pools (2). Earlier studies have shown that maintenance of both proton motive force and intracellular ATP pools is required for translocation of newly synthesized lipopolysaccharide from the periplasmic face of the inner membrane to the outer membrane (5). Here, we show that proton motive force is also necessary in vivo for synthesis of 0 antigen. We were initially interested in testing the idea that transposition of the undecaprenol-linked tetrasaccharide unit from the presumptive site of synthesis at the cytoplasmic face of the inner membrane to the periplasmically oriented active site of the 0-polymerase enzyme might be energy dependent. The results gave no support for this hypothesis but unexpectedly revealed that in vivo synthesis of the first intermediate of the biosynthetic pathway is specifically dependent on maintenance of proton motive force. Evidence for energy-dependent transposition of the core lipopolysaccharide from the cytoplasmic to the periplasmic face of the inner membrane is presented by McGrath and Osborn (6).
*
MATERIALS AND METHODS Bacteria and growth conditions. The galE mutants of S. typhimurium, G30 and G30A (heptoseless deep rough, core deficient), have been described previously (11), as have the kdsAl(Ts) strains BCM1 (6, 7) and BCM2 (7). Cultures were grown at 37°C (30°C for BCM1 and BCM2) in PPBE medium (11) with vigorous aeration unless otherwise indicated. Materials. Radionuclides were purchased from New England Nuclear or Amersham. Bacterial alkaline phosphatase was from Worthington Biochemicals. All other chemicals were reagent grade or better and were obtained from standard sources. Analysis of intracellular products of galactose pulse. Strain G30A was pulsed with [3H]galactose for 5 min in the presence or absence of 1 mM 2,4-dinitrophenol (DNP) as described in the legend to Fig. 1. Cells were collected by centrifugation at 2°C and washed twice with cold 0.1 M phosphate buffer, pH 7.0, with or without 1 mM DNP. Intracellular products of galactose uptake were extracted with hot 70% ethanol and analyzed by paper electrophoresis as previously described (12). Determination of average degree of polymerization of 0 antigen. The method of Kent and Osborn (4) was employed. Strain G30A was labeled with [3H]galactose plus or minus 1 mM DNP for 5 min, and 0 polymer was extracted with cold 5% trichloroacetic acid (TCA) and isolated by gel filtration on a Sephadex G-50 column as described in the footnote to Table 2. Dephosphorylation, reduction with NaBH4, and acid hydrolysis were carried out as described previously (4). [3H]Galactitol and [3H]galactose were separated by paper chromatography (1) and counted. Determination of radioactivity. A Beckman LS2800 liquid scintillation counter was employed. Aqueous samples were counted in Liquiscint, and nonaqueous samples were counted in Filtron-X (both from National Diagnostics). RESULTS Effect of DNP on galactose uptake and phosphorylation. The major route of galactose transport in S. typhimurium is
Corresponding author. 3128
VOL. 173,
1991
ENERGY DEPENDENCE OF O-ANTIGEN SYNTHESIS IN VIVO
TABLE 1. Effect of DNP on uptake and utilization of galactosea Strain
G30
G30A
Concn of DNP
Uptake of [3H]galactose
(mM)
Total (nmol/ 7 x 108 cells)
% of control
0 0.5 1.0 0 0.5
5.77 4.22 3.72
100 73.1
1.26
1.0
0.96
100 90.4 76.2
1.14
64.4
Intracellular soluble radioactivity ered as UDP recovgalactose (%)
3129
TABLE 2. Effect of DNP on synthesis of polymeric 0 antigena during
Time and condition
dncurtiong 11 incubatlon (pulse)
for incubation (chase) 2
Zero time 10 min minus DNP Zero time 10 min plus DNP 10 min minus DNP
Amount of TCA-extractable
radioactivity Total
~~~~~~~(dpm[10'])
% recovered as
polymer
72.7 76.5 72.6 73.6
Minus DNP
73.6
a Cultures of G30A (5 ml) were incubated at 32°C plus or minus 1 mM DNP for 1 min before addition of [3H]galactose (50 ,M, 50 ,Ci/,umol). After 5 min, cultures were centrifuged for 5 min at 500 x g at 2°C and resuspended for chase in fresh medium containing 50 mM nonradioactive galactose with or without DNP. After 10 min at 32°C, cold TCA was added to 5% and cells were chilled, collected by centrifugation, and washed twice with cold 5% TCA. O-Antigen intermediates were released from undecaprenol linkage by extraction with cold TCA (4) and were fractionated by filtration through a 47-cm Sephadex G-50 column. Elution was with 50 mM NH4HCO3.
a Cells (5 x 108/ml) were incubated in PPBE medium with aeration for 2 min plus or minus DNP before addition of [3HIlgalactose (50 ,uM, 50 ,uCi/,mol). Aliquots were taken over a 5-min period and rapidly filtered (Millipore type HA, 0.45 ,), washed with 10 ml of medium, dried, and counted. At 5 min, 6-ml samples were taken for extraction with 70% ethanol and determination of intracellular UDP-[3Hlgalactose, as described in Materials and Methods.
an ATP-dependent system which uses periplasmic galactosebinding protein and is not directly dependent on mainte-
nance of proton motive force (13). This suggested that
possible requirements for proton motive force in 0-antigen synthesis might be probed by pulse-labeling galE mutants with galactose in the presence of an uncoupler such as DNP. Effects of DNP on uptake and utilization of galactose in galE strains are summarized in Table 1. The lipopolysaccharide mutant, G30A (galE, deep rough), lacks the core saccharide structure and incorporates galactose only into lipid-linked intermediates of 0-antigen synthesis. In all cases, galactose uptake was greater than 60% of that of control during a 5-min labeling period at concentrations of DNP up to 1 mM. In addition, the fraction of intracellular radioactivity recovered as UDP-galactose was unaffected even at the higher concentration of uncoupler. We therefore concluded that intracellular concentrations of UDP-galactose adequate to support pulse-labeling of 0 antigen and core lipopolysaccharide were
maintained under these conditions for at least 5 min. We have previously shown (5) that membrane potential is abolished within 30 s at 1 mM DNP, while ATP pools remain normal for several minutes. Effect of DNP on biosynthesis of 0 antigen in vivo. We were initially interested in testing the possibility that the undecaprenol-linked tetrasaccharide unit of 0 antigen is transposed from the cytoplasmic face to the periplasmic face of the inner membrane and that this transmembrane flip-flop is driven by proton motive force. Since polymerization takes place on the periplasmic side, this hypothesis predicts that formation of polymeric 0 antigen should be sensitive to uncoupler. Accordingly, strain G30A was pulse-labeled with galactose for 5 min in the presence or absence of 1 mM DNP, and 0-antigen-related saccharides were released from undecaprenol-linked intermediates by extraction of cells with cold 5% TCA. Extracts were fractionated by filtration on a Sephadex G-50 column in order to separate 0-specific polymer from low-molecular-weight species. As shown in Table 2, formation of polymer was profoundly inhibited in the presence of DNP; only 3.8% of the cell-associated radioactivity was recovered in the excluded fraction compared with 53% in the control. The inhibition was rapidly reversed upon removal of DNP. At the end of the 5-min pulse, cells were recovered by centrifugation, suspended in fresh medium containing nonradioactive galactose, and chased for 10 min with or without DNP (Table 2). Following chase in the absence of DNP, polymeric products accounted for about 90% of the total radioactivity whether or not the uncoupler
Plus DNP
20.5 26.3 7.2 8.6 18.3
53 93 3.5 2.7 88
had been present during pulse; thus, precursors accumulated during the pulse period could be utilized efficiently for 0-polymer formation during subsequent chase. Inhibition by DNP affected only the total amount of polymer formed, not the average degree of polymerization. Polymer fractions isolated by gel filtration were treated with alkaline phosphatase to dephosphorylate galactose-1-PO4 reducing termini and reduced with NaBH4. Radioactivity in galactitol and galactose was determined after acid hydrolysis and paper chromatography. Ratios of [3H]galactitol to [3H]galactose for control and DNP-treated cells gave average degrees of polymerization of 12 and 13, respectively. This result was consistent with the hypothesis that inhibition of polymerization by DNP was indirect, limiting flow of the tetrasaccharide monomer to the polymerase rather than the activity of the enzyme per se. Effect of DNP on synthesis of undecaprenol-linked oligosaccharide intermediates. If, as postulated, the effect of uncoupler was to prevent delivery of undecaprenol-linked monomer to polymerase enzyme on the trans side of the membrane, inhibition of polymerization should be accompanied by accumulation of unpolymerized oligosaccharide intermediate(s). This prediction proved incorrect. Incorporation of [3H]galactose into CHCl3-methanol-soluble oligosaccharide-lipid intermediates was virtually abolished in the presence of 1 mM DNP (Fig. 1). This result was confirmed by analysis of the water-soluble products of galactose pulse after extraction of possible 0-specific intermediates and mild acid hydrolysis to release free saccharide chains. No detectable accumulation of 0-related products was observed (data not shown). This observation strongly suggested that DNP acts to inhibit synthesis of the undecaprenol-linked monomer unit and not its utilization by 0 polymerase. Further, since galactosyl-pyrophosphoryl-undecaprenol (galactosyl-PP-Und) is the first intermediate to be formed, the finding pinpointed the effective site of DNP inhibition as the first reaction in the pathway, catalyzed by galactosyl-pyrophosphoryl-undecaprenol synthase: UDP-galactose + P-Und
