Transport of Divalent Cations with Tetracycline as Mediated by the Transposon TnlO-encoded Tetracycline Resistance Protein*

Tetracycline uptake into inverted membrane vesi- cles from TnlO-bearing Escherichiu coli cells required divalent cations. The degree of the stimulation of tet- racycline uptake by various divalent

Transport of Divalent Cations with Tetracycline as Mediated by the Transposon TnlO-encoded Tetracycline Resistance Protein* (Received for publication, October 11,1989)  It is clear that the s°Co2+ uptake was mediated by the tetracycline resistance protein, because the membrane vesicles from tetracycline-sensitive cells did not show the uptake of 6oCo2+ and tetracycline. The e°Co2+ uptake was inhibited in the presence of other divalent cations, without any significant effect on tetracycline uptake, indicating that these cations are also transported with tetracycline by the tetracycline resistance protein.
Among the mechanisms underlying bacterial resistance to antibiotics, the tetracycline resistance mediated by a plasmidencoded resistance gene is a unique system, because the resistance is based on the active efflux of the drug out of the cells (1,2). Similar mechanisms based on efflux are known for the bacterial resistance to some poisons and heavy metals (3)(4)(5)(6)(7)(8). Arsenate and arsenite efflux systems (8,9) coded by genes on plasmids are primary ATP-driven anion pumps (9, lo), whereas cadmium (3) and ethidium (5) are excreted by cation/proton antiporters coded by the resistance plasmids. Tetracycline is excreted by similar antiporters coupled with proton influx in resistant cells (2,11). Tetracycline molecules are mainly present as the protonated (neutral) or monoanionic form at physiological pH (12). However, the exchange between tetracycline and protons was an electrically neutral reaction (ll), suggesting the participation of a third ion(s) in tetracycline transport.
The active efflux of tetracycline from resistant cells requires the magnesium ion (2). However, it is not known whether this ion is transported with tetracycline or simply activating the transport protein. Tetracycline forms a chelate complex with * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.. a wide variety of divalent cations including the magnesium ion (13). Considering the high magnesium concentration in Escherichia coli cells (14), most of the tetracycline molecules accumulated in the cells should be present as a chelate complex. Therefore, it may be reasonable to assume that the chelate complex may play an important role(s) in the transport process.
In this study, we first tested a wide variety of chelateforming divalent cations as stimulators for TnlO-mediated tetracycline transport. These divalent cations could stimulate tetracycline transport, the degree of stimulation being correlated with their ability for chelate-formation.
Among the divalent cations tested, the cobalt ion was the most effective stimulator.
Furthermore, cobalt ion (@'Co'+) was taken up coupled with tetracycline transport. These results suggest that divalent cations were transported by tetracycline transport mechanism.  (17). Thus, we measured the concentration of the chelation complex using absorbance at 390 or 400 nm in 50 mM MOPS-KOH buffer (pH 7.5). The wavelengths used were 400 nm for the complexes with Co'+ and Ca*+, and 390 nm for those with Mg2f, Mn'+, and Cd". The molecular extinction coefficient used for Co*+, Mg2+, Mn'+, Cd'+, and Ca*+ were 9.6 x 103, 1.05 X lo*, 1.04 x 104, 7.9 X 103, and 8.2 X 103, respectively.

Requirement
of a Divalent Cation for Tetracycline

Uptake by Inverted Membrane
Vesicles-The magnesium ion is known to be required for tetracycline transport mediated by the tetracycline resistance protein (2). A wide variety of divalent cations including Mg2+ was examined as to their effects on tetracycline uptake into inverted membrane vesicles prepared from tetracycline-resistant cells (Fig. 1). All the divalent cations tested stimulated tetracycline uptake, Co*+ being the most effective among the cations tested. The initial rate of the uptake and the final concentration of tetracycline in the vesicles in the presence of Co2' were about 2.0-to 2.5-fold higher than those in the presence of Me. The degree of stimulation were the following order: Co'+ > Mn2+ > Mg2+ > Cd*+ > Ca'+. In the absence of a divalent cation, there was no significant uptake of tetracycline (Fig. l), suggesting that a divalent cation is essential for the tetracycline transport. Vesicles from sensitive cells did not show significant uptake of tetracycline in the presence or absence of these cations (data not shown).

Kinetics of the Tetracycline
Uptake-When the initial rate of the tetracycline uptake in the presence of a constant tetracycline concentration (10 pM) was plotted against cation concentrations, a saturation curve was observed with all cations tested ( Fig. 2A). The kinetic constants (K, and V,.,) of the tetracycline uptake with these cations were estimated from the initial uptake rate by means of three methods (Lineweaver-Burk plotting, Hanes-Woolf plotting, and Eadie plotting) (Table I). Except for Cd*+ ion, the three methods gave similar values for each cation. On the other hand, the kinetic constants for the uptake with Cd" varied significantly depending on the method used, suggesting that the uptake did not follow simple Michaelis-Menten type kinetics. Among the cations tested, Co*+ showed the smallest K, and the largest Vmsx values (Table I) (Table I) were in agreement with the decreasing order of the stimulating effect (Fig.  1). Excess tetracycline could not be used for this assay, because high concentration of tetracycline increased the noncarrier-mediated permeation of tetracycline across the membrane (data not shown). Thus, K, values for divalent cations estimated above may reflect the stability of metal-tetracycline complexes.
Next the initial rate of tetracycline uptake in the presence of an excess divalent cation (500 PM) was plotted against tetracycline concentrations by the method of Hanes-Woolf plotting (data not shown). In this case, K, values for tetracycline were approximately similar (20-50 pM) regardless of the cation species present (Table II). Thus, the affinity of the carrier protein to tetracycline or its metal-chelate complex was not affected by the cation species. On the other hand, the turnover rate of the transport carrier protein for tetracycline (V,..) was dependent on the cation species (Tables I and II). The initial rate of tetracycline uptake at a constant tetra-  Fig. 2A) was replotted against the divalent cation concentration to obtain Hill plots (Fig. 2, B and C). The Hill coefficients for Co'+, Mn*+, Mp"', and Ca*+ were estimated to be 1.1, 1.1, 1.0, and 0.98, respectively, indicating that the uptake of one tetracycline molecule requires one divalent cation.

Dissociation
Constants of the Chelate Complexes of Tetracycline with Divalent Cations-The tetracycline molecule has three dissociable protons (18). The major molecular species at pH 7.5 are the monovalent anion (TH-) and the neutral form (TH,). The anionic form of tetracycline easily forms a chelate complex with a wide variety of divalent cations (13,(19)(20)(21). The concentration of a chelate complex in 50 mM MOPS-KOH buffer (pH 7.5) was measured, and the dissociation constant (I&) of the metal-tetracycline complex was estimated from Scatchard plots (Table III). The Kd value for the complex differed widely depending on the cation species. The increasing order of the Kd values was in agreement with the decreasing order of the stimulation of the tetracycline uptake by these cations. It seems unlikely that the free tetracycline molecule is a substrate for the transport carrier protein, because a lower Kd value should result in a lower concentration of free tetracycline. The coincidence between the orders of the Kd and K, values for various cations (Table  I) indicates that the substrate of the carrier protein is a metalchelate complex of tetracycline, and the differences in the Km value is due to a difference in the stability of the metaltetracycline complex.
Cobalt Ion Uptake by Inverted Membrane Vesicles from Tetracycline-resistant Cells-The uptake of @'Co'+ by inverted membrane vesicles prepared from tetracycline resistance cells was measured in the presence and absence of an energy source (NADH) and tetracycline (Fig. 3A). In the absence of tetracycline, no significant uptake of 6oCo2+ was detected, although the reaction mixture was well supplied with NADH. When 10 pM tetracycline was present, a very little, but significant, FIG amount of 6oCo2+ was taken up even in the absence of NADH (possibly downhill transport of Co'+-tetracycline complex). This 6oCo2+ uptake was greatly stimulated by the addition of NADH. The 6oCo2+ uptake due to the addition of NADH was completely inhibited by the addition of CCCP. Therefore, it is clear that the 6oCo2+ was taken up with tetracycline driven by a proton motive force. 6oCo2+ uptake by vesicles prepared from tetracycline-sensitive cells was not observed (Fig. 3B).
These results indicate that the @Co2+ uptake was mediated by the tetracycline resistance gene.
Inhibition of Cobalt Ion Uptake by the Magnesium Ion-If s°Co2+ is taken up by inverted vesicles as a metal-chelate complex with tetracycline, the @'Co2+ uptake should be inhibited by the addition of other chelate-forming cations. When Mg2+ was added to the assay mixture, the uptake of s°Co2+ was competitively inhibited (Fig. 4) without concomitant inhibition of the tetracycline uptake (data not shown). 6oCo2+ uptake in the presence of Mg2+ was approximately proportional to the concentration of Co'+-tetracycline complex calculated from the dissociation constants of the Co2+-tetracycline and M$+-tetracycline complexes. These results indicated that the M%+-tetracycline complex was formed and taken up into vesicles.
The Effect of Ionophores on the Transport of Cobalt Ion with Tetracycline-Tetracycline uptake into inverted vesicles is inhibited by CCCP or nigericin, but valinomycin rather stimulates the uptake (11). YJo2+ uptake into inverted vesicles showed the same behavior; CCCP or nigericin inhibited the uptake, whereas valinomycin stimulated the uptake by a factor of about 1.4 in the presence of 0.1 M KC1 (data not shown), indicating the electrically neutral nature of the Co'+tetracycline transport driven by antiport with proton.
When tetracycline and Co'+ uptake was stopped by the addition of CCCP (or nigericin), tetracycline leaked out slowly from the vesicles without any significant leakage of Co'+ ( Fig.  5), possibly due to the binding of Co'+ at the inner surface of the vesicles. On the other hand, when gramicidin S was added to the reaction mixture, both Co'+ and tetracycline rapidly leaked out from the vesicles (Fig. 5), because gramicidin S competitively removes bound cations from a membrane (22) due to its polycationic nature. Gramicidin D, which is an ionophore similar to gramicidin S but has no polycationic nature, caused no such rapid efflux of Co'+ and tetracycline (data not shown). DISCUSSION The experiments described in the present paper showed that a divalent cation was transported with tetracycline by the TnlO tetracycline resistance protein, probably as a metalchelate complex. The K,,, values for divalent cations reflected the dissociation constants (Kd) for the metal-chelate complexes. The affinity of the transport carrier protein for the metal-chelate complex is independent of the cation species, because the K, values for tetracycline in the presence of excess cations was not affected by the cation species. On the other hand, the turnover rate of the transport carrier protein FIG. 6 of the cations used in this study shows the following increasing order: Co2+ < Mr?+ < Mg2' < Ca2+ (23), and this order is in agreement with the decreasing order of the V,,,,. values. The uptake of 6oCo2+ was about 2-fold greater than the [3H] tetracycline uptake under the same conditins. This is clearly different from the 1:l stoichiometry indicated by the Hill plot. This discrepancy may be due to the leakage of free tetracycline molecules, by diffusion, out of the vesicles through the lipid bilayer (22), as a result, the observed tetracycline uptake is underestimated. This was essentially confirmed by the fact that tetracycline actually leaked out by downhill diffusion from the vesicles without any significant leakage of Co2+ when the uptake was stopped by the addition of CCCP (Fig. 5), although the efflux rate was slower than that expected from the discrepancy probably due to cancelling the pH gradient by CCCP. The pH gradient is not only favorable to the TetAmediated uptake of the chelation complex but also favorable to the diffuse back out of free tetracyclines from the vesicles. The ApH-driven quick diffusion of tetracycline across lipid bilayer membranes was observed in liposomes made from E. coli phospholipids. 2 The reason why the carrier-mediated efflux of Co'+ with tetracycline from the vesicles did not occur when CCCP was added (Fig. 5B) may be as follows: the metal-chelate complexes taken up into vesicles possibly dissociate into cations and free tetracycline molecules and then cations bind to the inner surface of the vesicles, whereas free tetracycline diffuse out slowly from the vesicles through lipid bilayers (22). This assumption was supported by the observation that the rapid efflux of both the Co2+ and tetracycline from the vesicles occurred on the addition of gramicidin S (Fig. 5), which removes cations from the membrane due to the replacement of membrane-bound cations by the antibiotic (24), resulting in the reformation of metal-tetracycline complexes and then the rapid efflux of the complexes mediated by the transport carrier proteins.
In intact E. coli cells, energy-dependent accumulation of tetracycline is known to be driven by ApH but not A# (25) and inhibited by Mp2+. 3 Argast and Beck (26) claimed that tetracycline permeates into the cytoplasm by diffusion, and they presented a direct evidence supporting the tetracycline diffusion through phospholipid bilayers (24). We also confirmed that tetracycline was accumulated in liposomes driven by ApH? The inhibitory effect of M$+ on the uptake into the cells is probably due to the decrease in free tetracycline concentration in the medium. The resistant cells also have such a mechanism for tetracycline accumulation, although the accumulated tetracycline is actively excluded from the cells. We propose here a model for "the tetracycline transport cycle" in the resistant cells (Fig. 6). A tetracycline molecule penetrates into the cells through lipid bilayer of the cytoplasmic membrane as a neutral form (TH2). The diffused TH2 dissociates into TH-and proton in the cytoplasm, because intracellular pH is usually higher than that in the medium. The high intracellular concentration of M$' (l-4 mh@ may also facilitate the dissociation by the formation of the M$+tetracycline chelate complex, resulting in an intracellular accumulation of tetracycline. In the resistant cells, the chelate complex is then excluded as the substrate for the efflux system mediated by the tetracycline-resistant protein coupling with proton influx. One-to-one complex of monoanionic tetracycline and divalent cation should be a true substrate for the efflux system, because the exchange of this monocationic chelate complex for proton was electrically neutral, confirming our previous observation (11). This model was also consistent with our observation that the cells harboring TnlO showed higher level of tetracycline resistance (minimum inhibitory concentration, 200 pg/ml) at pH 6.0 than that at pH 7.0 (minimum inhibitory concentration, 120 pg/ml).
It is known that cobalt ion is taken up into E. coli cells via Mg2+ (14,27) or Cd" (28) transport systems. However, the tetracycline-dependent 6oCo2+ uptake into the inverted membrane vesicles is independent of these systems, because Co*+ transport mediated by these intrinsic systems should be carried out only from inside to the outside of the inverted vesicles as these vesicles have opposite orientation from cells. The inverted vesicles from tetracycline-sensitive cells actually did not show the uptake of 6oCo2+ dependent on respiration.
Although the origin of the tetracycline resistance gene on plasmids is not obvious, McMurry et al. (29) reported that a very weak tetracycline efflux was induced in the susceptible E. coli cells when the cells were grown in the nutrient rich medium. The tetracycline transport mediated by the resistance gene may be originated from such intrinsic efflux systems and such systems may be in general for exclusion of toxic compounds or waste products from the cells as Mg*+-chelate complexes.