Accumulation of trospectomycin by strains of Salmonella typhimurium , Escherichia coli and Haemophilus influenzae

Trospectomycin, unlike aminoglycosidic aminocyclitols, is accumulated by a nonsaturable, energy-independent, diffusional process in Salmonella typhimurium, Escherichia coli and Haemophilus influenzae. A deep rough mutant of S typhimurium was more susceptible and accumulated the drug faster, and F porin deficient mutants of E coli were more resistant than parental strains. Trospectomycin likely uses both porin and nonporin pathways to cross the outer membrane. An E coli strain effectively accumulated the drug anaerobically, explaining its anaerobic activity. An H influenzae strain accumulated trospectomycin at concentrations below those for which detectable uptake could be observed with E coli or S typhimurium strains, consistent with greater activity in Haemophilus species.

membrane.An E coli strain effectively accumulated the drug anaerobically, explaining its anaerobic activity.An H injluenzae strain accumulated trospectomycin at concentrations below those for which detectable uptake could be observed with E coli or S typhimurium strains, consistent with greater activity in Haemophilus species.Can J Infect Dis 1990;1(2):51-56 Key Words: Aminocyclitols, Trospectomycin, Uptake Neisseriagonorrheae, Proteus species, Bacteroides species, Clostridium species (including C difficile) and Chlamydia trachomatis.The two drugs show comparable activity against the enterobacteriaceae and generally exhibit cross-resistance (1).
Trospectomycin, spectinomycin and aminoglycosides are aminocyclitols which differ in several respects.One is the much greater antibacterial activity of aminoglycosidic aminocyclitols on aerobic bacteria in general versus anaerobic organisms, whereas the spectinomycins do not show this pattern of activity (1,2).The greater activity of aminoglycosidic aminocyclitols on aerobic bacteria is due to their more effective accumulation by bacteria carrying out electron transport which develop an electrical potential greater than the threshold level needed for aminoglycoside uptake (3,4).Both classes of antibiotics inhibit protein synthesis, although they differ in that the spectinomycins do not cause misreading, while the aminoglycosides do (5).This is not a sufficient reason to account for the difference in activity on aerobic and anaerobic bacteria as protein synthesis proceeds in both circumstances.Earlier studies reported that mutants of Escherichia coli defective in aspects of electron transport showed increased resistance to aminoglycosides but not to spectinomycin (6).This finding suggests that the uptake process for the spectinomycins might be different from that for the aminoglycosides.The authors have studied the nature of the transport process for trospectomycin as a representative aminocyclitol to determine if it is significantly different from that described for aminoglycoside aminocyclitols and whether or not it could account for the reported differences in activity on anaerobic bacteria.They also examined mutations affecting porin or lipopolysaccharide structure of salmonella or E coli to obtain information on the pathways used by trospectomycin to cross the outer membrane.These studies on trospectomycin uptake under a variety of conditions show that the uptake of this compound is by simple diffusion, consistent with its activity in anaerobic bacteria.

Bacterial strains:
The strains used in this study were obtained from KE Sanderson, University of Calgary (S typhimurium LT2 SL3770, S typhimurium LT2 SL3769 and  3 H)-trospectomycin were confirmed to be identical in bioactivity by identical zone sizes produced by 30.5 11g of each material spotted to 6.5 mm paper discs and incubated overnight with a lawn of E coli SA1306.Labelled and unlabelled compounds produced identical patterns on thin layer chromatography with CEL300-25 cellulose plates using n-butanol:water:acetic acid in ratios of 12:5:3 and iodine vapour as the developing reagent.Carbonyl cyanide m-chlorophenylhydrazone (CCCP) was supplied by the Sigma Chemical Co (StLouis , Missouri) .Media and MICs: Media used were brain heart infusion broth from Difco Laboratories (Detroit, Michigan) and nutrient broth from BBL Microbiology Systems (Cockeyville, Maryland).Minimum inhibitory concentrations (MICs), were determined in brain heart infusion broth, nutrient broth and nutrient broth with 10 mM magnesium chloride using an inoculum of 10 5 cells/mL and incubating tubes without shaking at 35°C for 16 h.All studies on H injluenzae used supplemented brain heart infusion broth containing CVA enrichment ( 10 mL/L) and hemin (10 llg/mL).Uptake studies: The basal medium used for uptake studies was nutrient broth.Cells were grown at 37°C overnight, a 1:10 dilution subcultured into fresh media and the incubation continued in 50 mL volumes until the A6oo reading reached 0.4.Cells were collected by centrifugation and resuspended at a concentration of 30 A6oo units/mL.The suspension was incubated in a block heater at the temperature to be used for uptake for 7 mins with shaking, and thereafter the appropriate amount of trospectomycin was added using a range of one part labelled drug to nine parts unlabelled, and one part labelled to 19 parts unlabelled, for final concentrations of 641J.g/mL or greater.For lower concentrations the ratio was between 1:3 and 1:6.
At the various times specified, 80 IJ.L samples were added to 2 mL of the same broth as rapidly as possible and filtered through Whatman GF /F microfibre glass filters, and a further 2 mL of broth was pi petted over the area of the filter occupied by the filtered cells.Filters were dried and counted in Beckman Ready Safe liquid scintillation cocktail in a Beckman LS6800 scintillation counter.Uptake is reported as ng trospectomycin/mg dry weight of cells.For uptakes using dead cells the cells were held at 70°C for 30 mins before concentration to 30 A6oo units/mL.No survivors were detected after this treatment using a dilution series capable of detecting 10 bacteria/mL.Uptake of H injluenzae was performed using supplemented brain heart infusion broth.Anaerobic uptake was carried out withE coli SA1306 grown at 37°C using an atmosphere of95% nitrogen and 5% carbon.Cells were collected by centrifugation at 4°C and resuspended in nutrient broth through which 100% nitrogen gas was bubbled.
Samples were taken as above under an atmosphere of 100% nitrogen.Other methods: All strains were confirmed to carry their lipopolysaccharide or porin phenotype using the methods for analysis of lipopolysac-

RESULTS
The patterns of trospectomycin accumulation at 37 or 43°C by S typhimurium, E coli and H influenzae were linear with time for at least 20 to 25 mins.More extensive studies using S typhimurium LT2 strain SL3770 showed that the rate of accumulation increased with the tempera- ture used for uptake determination, although the rate levelled off after 15 mins when uptake was done at 50°C (Figure 1).If it is assumed that all of the drug accumulated after zero time was intracellular, and that there is 1.6 )..lL cell water/mg cells (8), the intracellular concentrations at 37, 43 and 50°C were 180, 360 and 560 )..lg/mL cell water, respectively, or about 1.4, 2.8 and 4.4 times the extracellular concentration.Heat-killed cells showed excellent uptake at 37°C with the initial uptake rate faster than in viable cells at 37 or 43°C, demonstrating that an energy source was not necessary for accumulation.This was confirmed by the failure of CCCP to reduce uptake in S typhimurium LT2 SL3770 or SL1102 (Figure 1).Accumulation of trospectomycin by S typhimurium LT2 strains 3770 or 1102 at different drug concentrations gave a pattern of nonsaturability when the data were plotted by the Lineweaver-Burk method (Figure 2).Other tested circumstances where uptake was also not saturable were with EcoliSA1306 at 43°C and HinfluenzaeRd at 37°C (data not shown).

Accumulation of trospectomycin
Comparison of uptake by smooth S typhimurium SL3770 which possesses repeating osomatic side chain subunits of lipopolysaccharide (Ra chemotype) compared to the deep rough mutant S typhimurium SL1102 (Re chemotype), .::.::.showed that there was a small but consistently greater uptake by th e rou gh strain (Figures 1,2).This difference correlated with increased susceptibility to trospectomycin as demonstrated by a consistent twofold lower MIC value in three media for the deep rough mutant (Table 1).Strain 3769 with the intermed iate Rdl chemotype of lipopolysaccharide had a twofold lower MIC under two of the three testing conditions.Uptake and susceptibility of strains of E coli with differences in quantity of the F. C or both outer membrane porins were examined.Strain JF694 deficient in both porins and strain JF703 deficient in the F porin were more resistant by fourfold in nutrient broth with or without added magnesium.How-  ever, no appreciable consistent difference in uptake among the four strains could be detected (Figure 3).The inclusion of 10 mM magnesium chloride in nutrient broth increased MICs of both E coli and S typh imurium strains fourfold.S typhimurium LT2 strain SL1102 showed somewhat greater uptake in nutrient broth than in the same medium with 10 mM magnesium chloride, but no difference was detected with strain S L3 770 (Figure 2).The authors also examined uptake in E coli AB2495 and its rpsE-ribosomally spectinomycin-resistant derivative K.L2 52 and in E coli SA1306 and E coli SA1306 containing aminoglycoside nucleotidyl transferase (3") capable of inactivating spectinomycin (9).Both resistant isolates conferred resistance raising the MICs by four-and eightfold, respectively, but neither caused any difference in uptake of trospectomycin.
Trospectomycin has antibacterial activity against anaerobic bacteria and facultative bacteria grown anaerobically ( 1,1 0) and thus it was of interest to demonstrate that uptake occurs anaerobically.Figure 4 shows that E coli SA1306 accumulates the drug under anaerobic growth and uptake conditions at a similar rate to other E coli strains (Figure 3).H injluenzae strains have been reported more susceptible to trospectomycin than E coli. Figure 5 shows that uptake can be demonstrated in H injluenzae Rd at 37°C using 6 or 12 1-1-g/mL, whereas no uptake wa s detectable in E coli under these conditions at either concentration.

DISCUSSION
The present studies show that the aminocyclitol trospectomycin is accumulated by the bacteria examined through a process different from aminoglycosides.The latter require electron transport and a threshold electrical potential to be taken up (3).but trospectomycin undergoes simple energy-independent diffusion into the bacterial cells.This mode of accumulation is consistent with the observation that trospectomycin has equivalent activity under anaerobic and aerobic conditions.The present studies show that trospectomycin is effectively accumulated byE coli anaerobically and aerobically.There appears to be a modest cellular concentration of the drug over that in the extracellular medium.Such concentration is likely due to a substantial fraction bound either at the cell surface in spite of washing the cells, or intracellularly at target sites.However the ribosomal site modified in the rpsE-mutant does not contribute a significant amount to total cell binding, as no difference in total accumulation was observed compared to the parental strain.These results were supported by the fact that there was no change in accumulation by a strain adenylating spectinomycin, which also interferes with ribosomal binding (9).The increase in rate of uptake with temperature is characteristic of a diffusional process.Examination of uptake at 43°C was selected for many ofthe authors' studies because reproducibility was higher than at 37°C where the relative contribution of background zero time binding was greater.
Heat-killed dead cells took up trospectomycin faster than living cells.One explanation for this Accumulation of trospectomycin could have been inactivation of an active efflux system which has been described for other antibiotics (11,12).However. the use of the protonophore CCCP did not produce a consistent increase in accumulation which would have been expected for an energy-driven efflux process.The most probable explanation for the increased rate of accumulation at 37°C in heat-killed cells is that the outer membrane was modified, perhaps by release of lipopolysaccharide, thus enhancing uptake.A deep rough mutant of salmonella was more susceptible than a smooth strain, and the rate of uptake was also somewhat faster, supporting either a modest barrier function or an interaction with the polysaccharide portion of the lipopolysaccharide.The outer membrane has been shown to function as a barrier to relatively hydrophobic antimicrobial agents (13).Deficiency ofF porin in an isogenic set of E coli strains also increased resistance in nutrient broth but not in brain heart infusion broth.These observations on the effects of lipopolysaccharide and porins indicate that both porin and nonporin routes of cell entry through the outer membranes of two enteric bacteria may be used so that mutations in either do not produce major effects on susceptibility or uptake.
Magnesium chloride reduced the activity of trospectomycin for the salmonella or E coli strains tested.Although this compound reduced the rate of uptake in the rough salmonella strain it had little or no effect in the smooth strain SL3770 nor in several E coli strains examined (data not shown).These results are consistent with the view that magnesium antagonizes uptake to a small extent through an effect deep in lipopolysaccharides.However, they also suggest that some of the magnesium effect is not being exerted at the level of uptake but possibly at the ribosomal site of action.
Accumulation of trospectomycin by H injluenzae at concentrations much lower than those at which accumulation by E coli could be demonstrated indicates that one of the factors accounting for the better activity of this compound in the former bacterium is better penetration.The present studies do not exclude inhibition of ribosomal function at lower concentrations as an explanation for the enhanced activity alone or, more likely, in association with better uptake.H injluenzae has a much rougher type of lipopolysaccharide which could account in part for its better uptake (14).