Citrate-Mg”’ Transport in Bacillus subtilis STUDIES WITH 2-FLUORO-L-ER YTHRO-CITRATE AS A SUBSTRATE*

Z-Fluoro-L-erythlo-[3,4,5, 6J4C4]citrate is actively transported in Bacillus subtilis cells by the citrate-Mg2+ transport system. The kinetic data for uptake of the analogue are very similar to those previously determined for citrate transport. Citrate and 2-fluoro-L-erythro-citrate mutually inhibit their uptake in a competitive manner. Only insignificant metabolism of the analogue was detected after uptake into cells induced for citrate-Mg*f transport. 2-Fluoro-L-erythro-citrate induces the citrate-Mg2f transport system with a delay of about one generation time relative to induction caused by citrate. The fluoro analogue probably causes indirect induction of citrate-Mg zf transport by inhibition of aconitase (EC 4.2.1.3) and subsequent accumulation of endogenous citrate. 2-Fluoro-L-erythro-citrate strongly inhibits aconitase in extracts of B. subti2is cells. B. subtilis mutants resistant toward growth inhibition by 2-fluoro-DL-erythro-citrate in ribose minimal medium were isolated and found to be defective in citrate-Mg2+ transport. These mutants can still grow, albeit very slowly, on plates containing citrate minimal medium. Preliminary results suggest that a second system exists in B. subtilis cells which transports citrate slowly and with lower affinity than the previously described citrate-Mg2f transport system.

6J4C4]citrate is actively transported in Bacillus subtilis cells by the citrate-Mg2+ transport system. The kinetic data for uptake of the analogue are very similar to those previously determined for citrate transport. Citrate and 2-fluoro-L-erythro-citrate mutually inhibit their uptake in a competitive manner.
Only insignificant metabolism of the analogue was detected after uptake into cells induced for citrate-Mg*f transport.
2-Fluoro-L-erythro-citrate induces the citrate-Mg2f transport system with a delay of about one generation time relative to induction caused by citrate.
The fluoro analogue probably causes indirect induction of citrate-Mg zf transport by inhibition of aconitase (EC 4.2.1.3) and subsequent accumulation of endogenous citrate.
B. subtilis mutants resistant toward growth inhibition by 2-fluoro-DL-erythro-citrate in ribose minimal medium were isolated and found to be defective in citrate-Mg2+ transport. These mutants can still grow, albeit very slowly, on plates containing citrate minimal medium. Preliminary results suggest that a second system exists in B. subtilis cells which transports citrate slowly and with lower affinity than the previously described citrate-Mg2f transport system.
When citrate is taken up by a bacterial cell it is quickly metabolized via the tricarboxylic acid cycle or the citrate fermentation pathway.
This fast metabolism of citrate has previously hindered investigations of bacterial citrate transport.
For accurate kinetic investigation of this transport system, experimental conditions were needed under which metabolism of the transported substrate is prevented after it is taken up by the bacterial cell. Previously we had reported that citrate was not catabolized in a mutant of Bacillus subtilis, defective in aconitase activity (EC 4.2.1.3) (1). We had found that the citrate fermentation pathway did not function in our B. subtilis strains and that the citrate synthase (EC 4.1.3.7) did not catalyze any significant degradation of citrate under in viva conditions (1). Then we wanted to study the inducibility of the citrate-Mg* transport * This work was supported by the Deutsche Forschungsgemeinschaft (SFB 74). system in B. subtilis cells which had structural chauges in the membrane composition.
The aconitase-less mutant could not be used for this purpose since it was endogenously induced for citrate uptake.
Induction studies of citrate transport could be carried out in wild type cells of B. subtilis but only in the presence of extensive metabolism of the substrate.
Thus, we looked for a citrate analogue which could be transported by the citrate-Mg2+ transport system without being metabolized.
We knew from our previous inhibitor studies that the hydroxy group in the 3 position and the carboxy group in the 6 position of the citrate molecule were structural prerequisites for a substrate of citrate-Mg2+ transport in B. subtilis (2). ii single substitution in any other position of the citrate molecule, however, would give rise to structural isomers, thereby posing the problem of purifying for transport studies a single analogue from a synthetic mixture of four optical isomers. This problem could be overcome by making use of the citrate synthase reaction which converted 2-fluoroacetyl-CoA and oxalacetate to 2-fluoro-L-erythro-citrate (3). This isomer was expected to be a strong inhibitor of aconitase in B. subtilis (4). We enzymatically synthesized 2-fluoro-L-erythro- [3,4,5, 6-14C4]citrate. In addition, the unlabeled compound was prepared according to Dummel and Kun (4) and used as reference and carrier material.
In this publication we describe the properties of this citrate analogue in kinetic studies of the citrate-Mg2+ transport system. Moreover we report on the isolation of B. subtilis mutants resistant toward uL-erythrofluorocitrate and on a characterization of some of them as defective in citrate-Mg2f transport activity. We measured the dependency of this reaction on citrate concentration in the presence of two different concentration of L-erythro-fluorocitrate. The activity of NADPH oxidase in the cell extract was also determined and used for correction of aconitase activity as previously reported (14).
Isolation of Mutants Resistant Toward 2-Fluoro-x-erythrocitrate-A culture (10 ml) of B. subtilis SB-26 growing exponentially in ribose minimal medium was mutagenized with 50 pg of N-methyl-N'-nitrosoguanidine per ml for 10 min at 43". After filtration and washing the cells were diluted 1: 15 in ribose minimal medium and grown overnight.
The cells were then washed four times with glutamate-free ribose minimal medium and aliquots of 5 X lo7 cells were spread on plates containing glutamate-free ribose minimal medium with 20 mM DL-erythro-fluorocitrate. After 4 to 5 days of incubation some colonies appeared which grew much faster than the slow growing background cells. Single colonies were picked and restreaked onto fresh plates containing DL-erythro-fluorocitrate in glutamate-free ribose minimal medium. The mutants were stored on tryptose blood agar (Difco) before being grown for citrate and malate transport assays.
As we shall see below this compound also inhibited aconitase in a crude extract from B. subtilis cells. In order to rule out extensive degradation of the analogue inside the cells, we incubated B. subtilis cells induced for citrate transport with 2-fluoro-n-eythro- [3,4,5, 6-14C4]citrate and analyzed the dissimilation products (1). Since we had previously found that a hot water treatment of 2-fluoro-n-erythro-[3,4,5, 6J4C4]citrate led to extensive destruction of this compound, we chose to extract the cells with a toluene-water mixture (15). This procedure did not change the chromatographic properties of radioactive Lqthro-fluorocitrate used as reference material. We analyzed the toluene water extract by electrophoresis and two different chromatographic runs on thin layer plates. Details are described under "Experimental Procedure." Fig. 1 shows an example of thin layer chromatography of the toluene-water extract. None of the systems showed significant catabolism of 2-fluoro-nerythro-[3,4,5, 6-l*C4]citrate.
More than 96% of the radioactivity applied on the plates was concentrated in the position of L-eythro-fluorocitrate.
To check the possibility that '*CO2 was released from 2-fluoro-n-erythro-[3,4,5,6J*CJcitrate during our transport assays, we incubated citrate-induced cells of B. subtilis SB-26 under the same conditions as during our transport assays together with the radioactive substrate in closed Warburg vessels. After 5 min we stopped the incubation by addition of perchloric acid from the side arm of the vessel. The released '*CO2 was trapped on a strip of filter paper (Whatman No. 1) moistened with hyamine hydroxide (Packard).
After the end of the reaction the paper was dried and counted in scintillation fluid (16). The experiment was run in parallel with [6J*C]citrate and 2-fluoro-n-erythro-[3,4,5,6J*CJcitrate as substrates.
As previously reported (17) almost all radioactivity was released as '*CO2 from [6-**C]citrate immediately after this compound was taken up by B. subtilis SB-26 due to metabolism via the aconitaseisocitrate dehydrogenase pathway.
This result further supports the conclusion that Lerythro-fluorocitrate is only insignificantly metabolized under our incubation conditions. Comparative Transport Studies with B. subtilis Cells Using The bacteria had been harvested -after growth in NYE medium nlus 5 mM citrate.
Thev were incubated for 20 min with 2-fluorb-L-erythro-[3,4,5,6-'"C4jcitrate and extracted with toluene-water, 5:95 (v/v) (15). Excess of salt was removed during electrophoresis of the radioactive extract on a cellulose plate (0.25 mm) in ammonium formate-formic acid, 0.1 M, pH 2.8. All of the radioactive material was pooled after the electrophoretic run, reapplied on a cellulose plate (0.25 mm), and chromatographed in the solvent system phenol-water-formic acid (13). After drying the plate was incubated with Kodak x-ray film for 4 weeks.  [1,5-i4C&itrate (A) were carried out for the indicated times as described under "Experimental Procedure." %Fluoro-z-erythro-citrate and Citrate as Substrates- Fig.  2 shows the time dependency of transport of 2-fluoro-n-erythro-citrate in B. subtilis SB-26. When [l ,5J*Cz]citrate was used as substrate for uptake studies in this strain we found for incubations of up to 1 min similar transport rates as for 2-fluoro-n-erythro-[3,4,5,6-i4C4]citrate.
The accumulated label from metabolized [I, 5-r*Ca]citrate is mainly trapped in the glutamate pool of the cells (1).
The data in Table I show that CCCP, an uncoupler of oxidative phosphorylation, blocked uptake of radioactivity of both 2-fluoro-n-erythro-citrate and citrate. Moreover, fully induced cells transport both substrates at similar rates.
Usually slightly more fluorocitrate than citrate was taken up by noninduced cells. The transport of n-ergthro-fluorocitrate showed the same dependency on Mg2+ ions as reported for citrate uptake (6). Using cells of B. subtilis SB-26 we determined a K, of 0.4 mM for the dependency of n-erythro-fluorocitrate transport on Mg2+ concen-2040  El, data plotted according to Lineweaver and Burk (18) to obtain the apparent Km and V,,,.
tration. This is in agreement with the value of 0.45 I& for citrate transport in B. sub&?is 60781 (6). of 105 f 20 clmoles per min per g dry weight. We had reported for citrate uptake in B. subtilis 60871 an apparent R, of 0.55 and a V,,, of 122 I.rmo1e.c per min per g dry weight of cells (6). Thus, the kinetic constants for citrate and L-eythro-fluoro~itrate uptake are very similar.
Another necessary criterion for establishing that L-ergthrofluoroeitrate and citrate are transported by the same system is mutual inhibition. Fig. 4A shows that citrate tra~po~ in B. subtilis 60871 is competitively inhibited by 2-fluoro-L-q&ocitrate with a & of 0.5 f 0.1 m;s/l. For the competitive inhibition of citrate transport by 2-Auoro-n-eythzo-citrate we found in the same mutant a Kd of 0.6 mM. The lack of specificity toward the L and D isomer of fluorocitrate recalls our earlier observation that both L-and n-malate inhibit citrate transport competitively (2). Transport of n-eytl&ro-Auorocitrate in B. subtilis 60781 is competitively inhibited by citrate (KC = 0.35 f 0.1 rnx) as illustrated in Fig. 4B and n-ewthro-fluoroeitrate (Ki = 0.35 mx). These results support the idea that n-q&o-fluorocitrate can be used as citrate analogue in kinetic investigations of the citrate-Mgz+ transport system in B. ~b~iZi~. The measurements of accumulation of label from [1,5-1CJcitrate during short time incubations also yielded results sufficiently accurate for estimation of citrate"Mg~ transport activity. We found for citrate transport in noninduced B. subtilis SB-26 grown in NYE medium an apparent Km of 1.5 mM and a V,, of 18 Mmoles per min per g dry weight. The corresponding data for transport of 2-fluoro-L-evthro-citrate were an apparent Km of 8.3 rnM and a V maE of 18 Kmoles per min per g dry weight (data not shown). The results of our transport measurements under noninduced conditions should be considered as rough estimations. Our measurements of the low maximal velocities showed a considerable variabilit,y in the presence of high substrate concentrations which was probably due to variable diffusion and binding of the substrate to the cells.

Induction of Citrate-Mgw
Transport 6y L-qthro- Fluoroci~   Fig, 5 illustrates an experiment to check whether L-erythro-fluorocitrate was able to induce the citra~-egg transport system in B. subtilis SB-26 cells growing in NYE medium. Induction caused by n-q&o-fluorocitrate took place only after a iiO-min delay relative to induction by citrate itself. The maximal activity induced by L-erythro-fluorocitrate was only two-thirds of the maximum transport activity which is induced by citrate. We conclude that r-erythro-fluorocitte probably did not directly induce citrate-Mg~ transport. It seems more likely that it caused inhibition of aconitase activity, which led to an internal accumulation of citrate. This might have acted as an endogenous inducer similar to the situationfoundin theaconitasemutant 60871 (1). Dummel and Kun (4) had reported that 2-fluoro-Lervthro-citrate blocked aconitase from rat liver mitochondria. We confirmed this inhibitory effect of the fluoro analogue on aconitase activity in a crude extract from B. subtilis SB-26.
After incubation of the extract (final protein concentration, 0.2 mg per ml) with 2 PM 2-fluoro-L-eqthro-citrate for 5 min at 25' prior to addition of citrate we didnot observeany NADPH formation.

C~r~~~~~~ of B. subtilis Muons
Resistant Toward DL-erythro-Fluoroeitrate-In pilot experiments we had observed that addition of 5 mM DL-erythro-fluorocitrate to glutamate-free ribose minimal medium extended the doubling time of B. ~bt~Z~s SB-26 cells from 3 hours to 13 hours, although growth never stopped complexly. This growth inhibition could be relieved by addition of 100 Mg of n-glutamate per ml of medium. This recalled the growth properties of the aconitase-less mutant B. subtilis 60871 which could grow in glucose minimal medium only if glu~rna~ was also present (1). Thus, it seemed likely that nn-eqthro-fluorocitrate retarded growth of B. subtilis SB-26 by inhibiting the aconitase activity.
The parent strain SB-26 transported citrate under these conditions 15. to 20.fold faster.
The mutants showed 25% less uptake of radioactivity from L-malate compared with the parent strain. Since malate is metabolized during transport, further characterization is needed to determine whether this decreased uptake is due to a pleiotropic effect in these mutants.
The mutants C-l and C-11 showed normal growth rates in glucose and malate minimal liquid medium and on agar plates. Surprisingly, they were still able to grow on plates containing citrate minimal medium although at a very much reduced rate compared to the parent strain SB-26. When B. subtilis SB-26 had formed colonies of 1.5 mm in diameter on solid citrate minimal medium, colonies of B. subtilis C-11 were barely visible. transferred the corresponding gene locus to another strain, B. subtilis SB-25, which is a histidine ausotroph.
Donor DNA was prepared from B. subtilis C-11 and used to transform the recipient SB-25 to histidine prototrophy.
Among the transformants we found a class of 1 to 2% which showed only very slow growth on citrate minimal agar plates. Three of them were grown under inducing conditions and assayed for citrate transport. They exhibited the same low uptake rate as B. subtilis C-11. The original recipient strain SB-25 showed a normal rate of citrate-Mg >+ transport after induction. From this growth difference we determined the frequency of reverting the mutant C-11 back to normal growth on plates containing citrate minimal medium.
We found a frequency of 10m7. Thus B. subtilis C-11 most likely harbors a single mutation which causes the defect in citrate-Mg2+ transport activity.

DISCUSSION
The effects of fluoro analogues 011 di-or tricarbosylate transport systems in B. subtilis and in mitochondria are completely different. Kun and associates had reported t,hat, 3-fluoro-Lerythro-malate and 2-fluoro-L-erythro-citrnte did not penetrate the inner mitochondrial membrane (19,20). 2-Fluoro-L-eruthromalatc, however, competitively inhibited the mitochondrial malate carrier and 2-fluoro-L-erythro-citrate irreversibly innctivatcd the tricarbosylate carrier in mitochondria at a wnw11tration of lop7 to lo-* M (20). The citrate transport, system in Aerobacter aerogenes was also competit,ively inhibited by fluorocitrate (probably a mixture of all isomers) but according to St,ein and Sachan the analogue was not taken up by these bacteria (21). In B. sub&is on the other hand, 3-fluoro-L-eythro-malate is a substrate for CXdicarboxylate transport (22), and 2-fluoro-rerythro-citrate is a substrate for the citrate-Mg* transport system. Although the fluoro analogues were accumulated in these bacteria 10. to 30.fold over t,he external concentration, no estensive metabolism of these compounds was detected. Thus, the fluoro analogues may replace malate and citrate in transport studies under conditions where fast metabolism of the transported substrate would influence interpretation of the results. Lawford and Williams (23) had blocked aconit'ase act,ivit,y with 2042 nn-fluorocitrate in two species of Pseudomonas in order to study transport of citrate without concomitant metabolism. They pointed out that m-fluorocitrate also inhibited the uptake of citrate under their experimental conditions. This complicated the interpretation of their transport kinetics. In B. subtilis we observed a strong inhibition by in-erythro-fluorocitrate of citrate catabolism via aconitase and isocitrate dehydrogenase.
We were able to isolate mutants resistant to this inhibition which proved to be defective in citrate transport activity.
At present we cannot say, however, whether these mutants are blocked in induction of citrate-Mg2+ transport or in a structural gene of a corresponding transport protein.
The fluorocitrate-resistant B. subtilis mutants could still grow on plates containing citrate minimal medium, albeit very slowly when compared to the parent strain. This suggested that B. subtilis cells can transport citrate under those conditions via another transport system. Probably this second system is expressed in cells grown in yeast extract medium (NYE medium) without addition of citrate. We tried to determine the apparent K, for uptake of citrate and of L-erythro-fluorocitrate under these conditions. Our results were somewhat variable due to the relatively high substrate concentrations necessary to saturate the second transport system. We estimated that both K, values were about 3-and 15fold higher, respectively, than the corresponding data determined for citrate-Mg*f transport in citrate-induced cells. Thus, it seems likely that this low affinity system which transported citrate relatively slowly is normally used by B. subtilis cells for transport of other tricarboxylates or possibly C-4 dicarboxylates. We had earlier reported that L-malate competitively inhibited citrate uptake in B. subtilis (2). In this context it may be added that according to a recent report three different transport systems for tricarboxylic acids could be induced in Salmonella typhimurium (24).