Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Cation Hexaammines Are Selective and Potent Inhibitors of the CorA Magnesium Transport System

Cation Hexaammines Are Selective and Potent Inhibitors of the CorA Magnesium Transport System THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 275, No. 22, Issue of June 2, pp. 16767–16773, 2000 © 2000 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Cation Hexaammines Are Selective and Potent Inhibitors of the CorA Magnesium Transport System* Received for publication, February 23, 2000, and in revised form, March 27, 2000 Published, JBC Papers in Press, March 27, 2000, DOI 10.1074/jbc.M001507200 Lisa M. Kucharski‡, Wil J. Lubbe, and Michael E. Maguire From the Department of Pharmacology, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106-4965 Cation hexaammines and related compounds are can be through direct, inner sphere, coordination of a ligand to the metal ion or through an indirect, outer sphere, association chemically stable analogs of the hydrated form of cat- ions, particularly Mg . We tested the ability of several with the bound water shell of the cation. In an outer sphere of these compounds to inhibit transport by the CorA or interaction, the relatively rigid geometry imposed by Mg MgtB Mg transport systems or the PhoQ receptor ki- essentially holds a water molecule in place within the active nase for Mg in Salmonella typhimurium. Cobalt(III)-, site of an enzyme. The outer sphere form of Mg -mediated ruthenium(II)-, and ruthenium(III)-hexaammines were catalysis has been demonstrated in exonuclease III, ribonucle- potent inhibitors of CorA-mediated influx. Cobalt(III)- ase H, and hairpin ribozymes (4 –7). and ruthenium(III)chloropentaammines were slightly Thus it is of interest to determine whether biochemical proc- less potent inhibitors of CorA. The compounds inhibited 21 esses involving Mg are of an inner sphere or outer sphere uptake by the bacterial S. typhimurium CorA and by the type. Cation hexaammines, such as cobalt(III)hexaammine archaeal Methanococcus jannaschii CorA, which bear (Co(III)Hex), are useful analogs of hydrated cations. The co- only 12% identity in the extracellular periplasmic do- valently and stably bound ammine groups mimic the size and main. Cation hexaammines also inhibited growth of S. shape of the hydration shell. These substitutionally inert typhimurium strains dependent on CorA for Mg up- 21 hexaammine compounds provide a means to differentiate be- take but not of isogenic strains carrying a second Mg tween inner and outer sphere Mg interactions (3, 8). If an uptake system. In contrast, hexacyano-cobaltate(III) enzyme is activated or inhibited by a substitutionally inert and ruthenate(II)- and nickel(II)hexaammine had little metal complex, an inner sphere interaction is not likely. There effect on uptake. The inhibition by the cation hexaam- are a variety of such compounds with variations in net charge mines was selective for CorA because none of the com- and number of hydrogen bond donors. pounds had any effect on transport by the MgtB P-type 21 21 This laboratory initially cloned and characterized the Sal- ATPase Mg transporter or the PhoQ Mg receptor kinase. These results demonstrate that cation hexaam- monella typhimurium CorA Mg transporter (9 –11). CorA is mines are potent and highly selective inhibitors of the virtually ubiquitous in the bacteria and archaea and forms CorA Mg transport system and further indicate that their constitutive Mg influx system (12). The topology of the initial interaction of the CorA transporter is with a CorA consists of a large N-terminal periplasmic domain and fully hydrated Mg cation. three transmembrane segments at the C terminus (13). We have also shown that even distant homologs of CorA function as Mg transporters. The CorA homolog of the Archeon Meth- The chemistry of Mg is unique among the biologically anococcus jannaschii has only about 16% overall sequence relevant cations. Compared with calcium, potassium, and so- identity to the S. typhimurium CorA. Expression of the M. dium, Mg possesses the largest hydrated radius, the smallest jannaschii CorA in a Mg transport-deficient strain of S. ionic radius, and the greatest charge density (1). Its volume typhimurium elicits Mg uptake with properties virtually change from hydrated to atomic cation is almost 400-fold, identical to that of the Salmonella transporter (14). 21 1 1 whereas Ca ,Na , and K are only 25-, 25-, and 4-fold, The mechanism of cation transport through CorA appears respectively. Waters of hydration are bound 3– 4 orders of mag- novel. In contrast to virtually all known cation transporters nitude more tightly to Mg than to other common biological and channels, CorA does not utilize electrostatic interactions to cations. The geometry of the hydration shell is octahedral with move the highly charge dense Mg cation across the mem- all six coordinate bond angles at angles of 90 6 6 . This con- brane. The transmembrane domain of S. typhimurium CorA trasts with the far more flexible geometry of the Ca cation, contains only a single negatively charged residue. Its mutation which can tolerate bond angles of 90 6 35 (1–3). to alanine has no effect on transport. Site-directed mutagenesis Study of Mg -dependent enzymes has revealed two types of of amino acids in the third transmembrane domain revealed interactions between Mg , enzymes, and ligands. Interaction the importance of three specific residues that line a single face of an a-helix. Likewise, mutagenesis of the second transmem- brane segment indicated that a single face of its a-helix is * This work was supported by National Institutes of Health Grant important (15, 16). Despite the lack of electrostatic interactions GM39447 (to M.E.M.) and a summer undergraduate fellowship from between Mg and the membrane domain of CorA, the large the American Society of Pharmacology and Experimental Therapeutics periplasmic domain is highly charged. It presumably functions, (to W. J. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby at least in part, to interact with hydrated Mg outside the cell marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ To whom correspondence should be addressed: Dept. of Pharmacol- The abbreviations used are: Co(III)Hex, cobalt(III)hexaammine; Ru(II) ogy, School of Medicine, Case Western Reserve University, 10900 Eu- Hex, ruthenium(II)hexaammine; Ru(III)Hex, ruthenium(III)hexaammine; clid Ave., Cleveland, OH 44106-4965. Tel.: 216-368-6187; Fax: 216-368- Ni(II)Hex, nickel(II)hexaammine; Co(III)Pent, chloropentaamminecobalt 3395; E-mail: [email protected]. (III)chloride; Ru(III)Pent, chloropentaammineruthenium(III)dichloride. This paper is available on line at http://www.jbc.org 16767 This is an Open Access article under the CC BY license. 16768 Cation Hexaammines Are Inhibitors of CorA Transport System supplemented N-minimal medium to an A between 0.050 and and provide an initial binding site for subsequent interaction 600 nm 0.080. These cultures were supplemented with Mg or Co(III)Hex as with the membrane domain. indicated and grown at 37 °C. At each time point, a 250-ml aliquot was We therefore tested cation hexaammines and related com- taken for luminescence and optical density readings. Luciferase activity pounds for their ability to inhibit Mg transport via CorA. was measured in triplicate for each sample by adding 10 ml of cells to 1 Both S. typhimurium and M. jannaschii CorA are potently ml of luciferase assay buffer, vortexing for 5 s, and reading the sample inhibited by cobalt(III)hexaammine and other similar analogs. in a luminometer for 10 s. Luminometer readings were normalized to cell number. In contrast, these same compounds did not inhibit other pro- Atomic Absorption—MM1927, MM1685, and MM281 were grown teins in S. typhimurium known to interact selectively with 21 overnight in LB broth plus 100 mM Mg and antibiotics. The cells were Mg . These results suggest that the initial binding site for washed twice with supplemented N-minimal medium containing no Mg in the CorA periplasmic domain binds a fully hydrated 21 Mg . The cells were resuspended in supplemented N-minimal medium Mg ion. plus Mg , Co(III)Hex, or cobalt chloride as indicated, to an A of 600 nm 1.5. At each time point an aliquot of 1.2 ml of cells was taken and MATERIALS AND METHODS overlaid on 300 ml of a 2:1 mixture of dibutyl and dioctyl phthalate. The 63 21 Plasmids and Strains—CorA-dependent Ni uptake and growth samples were spun in a microcentrifuge for 30 s, the supernatant was were assayed using strain MM1927 (16) for S. typhimurium experi- carefully aspirated, and the pellet was resuspended in 300 mlof1.0 N ments and MM1685 (14) for M. jannaschii experiments. MM1927 was nitric acid. The samples were then batch sonicated for 1 min, spun in a constructed by transforming the CorA plasmid, pMAS29, into MM281, microcentrifuge to pellet cellular debris, and the supernatant was aMg transport-deficient strain. MM1685 was constructed by trans- transferred to a new tube. Cobalt content of the supernatant was forming the M. jannaschii CorA-containing plasmid, pMETJA, into measured by atomic absorption as described previously (22). MM281. Growth inhibition was also assayed in MM199, a strain in Growth Inhibition—MM1927 and MM1685 were grown overnight in which corA has been inactivated by insertion of a transposon but that LB broth plus 100 mM Mg and antibiotics. The cells were washed still retains functional alleles for the MgtA and MgtB Mg transport- twice with Mg -free N-minimal medium and then diluted to a concen- 63 21 ers (9). MgtB-dependent Ni uptake was assayed using strain tration of 5.6 3 10 cells/ml in supplemented N-minimal medium plus MM1111, which is MM281 transformed with the mgtB-containing plas- 50 mM Mg . Cells (90 ml) and 10 ml of the appropriate concentration of mid pSPH39 (10). PhoPQ-dependent transcription of the mgtCB operon each inhibitor were aliquoted into each well of a 96-well plate. The plate was assayed using strain MM1106, which contains pTTC-Lux, a plas- lid was sterilely swabbed with NeverFog® (North American Oil Com- mid with a mgtCB-luciferase reporter fusion (17). pany, Atlanta, GA) anti-fog shield to protect against condensation. The Reagents and Buffers—Cobalt(III)-, ruthenium(II)-, and ruthenium- A of the cells in each well was measured every 30 min for 16 h in 650 nm TM (III)hexaammine chlorides were purchased from Aldrich. Nickel- a Molecular Devices THERMOmax microplate reader (16). MM199 (II)hexaammine, cobalt(III)- and ruthenium(III)chloropentaammine, po- was grown overnight in LB broth plus 100 mM Mg and 50 mg/ml tassium hexacyanocobaltate(III), and potassium hexacyanoruthenate(II) kanamycin. The cells were washed twice with Mg -free supplemented chlorides were obtained from Alfa Chemicals (Ward Hill, MA). N-minimal medium and then resuspended to an A of 0.1 in 5 ml of 600 nm Ni(III)hexaammine and Co(II)hexaammine were not tested, the former supplemented N-minimal medium containing 50 mM Mg plus either because it is not available and the latter because it is not stable in 100 mM Co(III)Hex, Ru(II)Hex or no inhibitor. The A of the cul- 600 nm aqueous solution at physiological pH. Co(III)Hex, Ru(II)Hex, Ru(III)- tures was determined after overnight growth at 37 °C. Statistical anal- Hex, and Ni(II)Hex stock solutions were stored at 220 °C. All other yses using unpaired t tests were calculated using GraphPad Prism solutions were made fresh for each use. All other reagents were ob- software (version 3.0). tained from Sigma, unless otherwise specified. Cultures were grown with shaking at 37 °C, with appropriate antibiotics, in Luria-Bertani RESULTS (LB) broth (18) or supplemented N-minimal medium, which is N-min- Cation Hexaammines Inhibit the CorA Mg Transporter— imal salts plus 0.4% (w/v) glucose, 0.1% (w/v) casamino acids, and 1 mM 21 63 21 leucine (15, 19). Ampicillin was used at 60 mg/ml, chloramphenicol at Mg uptake via CorA is studied using Ni as a substitute 28 21 170 mg/ml, and kanamycin at 50 mg/ml (20). Magnesium was added, because Mg is no longer available (21). We used this assay where indicated, as MgSO . Luciferase assay buffer contains 50 mM to examine the effect of various cation hexaammine compounds sodium phosphate buffer, pH 7.5, with 0.01% dodecyl aldehyde. on CorA-mediated uptake in S. typhimurium. Co(III)Hex, 63 21 Transport— Ni uptake was measured as described previously (9, 63 21 Ru(II)Hex, and Ru(III)Hex were potent inhibitors of Ni 21). MM1111 was grown overnight in LB broth containing 100 mM uptake (Fig. 1). All three were more potent inhibitors than the Mg . Cells were diluted 1:50 in supplemented N-minimal medium 21 21 21 21 containing 1 mM Mg and antibiotics. After 8 h growth, cells were actual substrates of CorA: Mg ,Ni , and Co . Co(III)Hex collected by centrifugation at 1,000 3 g for 15 min and washed twice in had an IC of 1 mM (Table I), which at the 200 mM Ni an equal volume of Mg -free N-minimal medium. These cells were concentration used, corresponds to a K of 500 nM. In contrast, resuspended to a final A of 0.1 in supplemented N-minimal me- 63 21 600 nm Ni(II)Hex did not inhibit Ni uptake at concentrations up to dium containing 10 mM Mg and antibiotics and incubated overnight at 30 mM (Fig. 2). 37 °C. MM1927 and MM1685 were grown overnight in LB broth con- The archeon M. jannaschii contains a CorA homolog with taining 100 mM Mg and appropriate antibiotics. After overnight growth, all strains were washed twice with Mg -free N-minimal me- about 38% overall sequence similarity (16% identity) to the S. dium and resuspended to an A between 1.0 and 2.0 in Mg -free 600 nm typhimurium CorA. The identity in the periplasmic domain of supplemented N-minimal medium. CorA-containing cells were washed approximately 235 residues is significantly lower at 32% over- with ice-cold buffer and kept on ice prior to the assay. MgtB-containing all similarity and 12% identity (12). We have previously shown cells could not be washed with cold media or stored on ice without 63 21 that this homolog is a true Mg transporter, functional in S. significant loss of transport activity (9). Transport of Ni was as- typhimurium with virtually identical properties as the native sayed at 37 °C by adding 0.1 ml of cells to 0.9 ml of supplemented 63 21 21 N-minimal medium containing 0.1–1.5 mCi of Ni and 200 mM Ni S. typhimurium transporter (14). In concert with those results, for assay of CorA or 20 mM Ni for assay of MgtB, plus the indicated cation hexaammines also inhibited uptake mediated by the M. concentration of inhibitor. These concentrations of Ni approximate jannaschii CorA. The rank order of potency (Co(III)Hex , the K for the respective transporters (9). Uptake was stopped after 5 Ru(II)Hex 5 Ru(III)Hex) was identical for both CorA transport- min for CorA or 30 min for MgtB by addition of 5 ml of ice-cold transport ers. As with the S. typhimurium CorA, Ni(II)Hex also did not wash buffer (N-minimal salts containing 5 mM Mg and1mM EDTA). Samples were filtered through BA85 (0.45 mm) nitrocellulose filters inhibit the M. jannaschii CorA transporter. The IC for each (Schleicher & Schuell) and washed once with 5 ml of transport wash hexaammine (Table I) did not differ significantly between the 63 21 buffer. Ni activity was determined by scintillation counting with an two CorA homologs (p . 0.1), although there was a slight efficiency of 85%. difference in Co affinity between the two CorA transporters Luciferase Assays—MM1106, a strain containing a mgtCB promoter- (p , 0.0001). These data indicate that the overall structure of luciferase reporter construct (17), was grown overnight at 37 °C in LB the cation binding site for the two transporters is functionally broth plus 100 mM Mg and antibiotics. The cells were washed three times with N-minimal medium containing no Mg and resuspended in identical. Cation Hexaammines Are Inhibitors of CorA Transport System 16769 FIG.1. Inhibition of S. typhimurium 63 21 CorA and MgtB Ni uptake. Trans- port was assayed as described under “Ma- terials and Methods” using a final Ni concentration of 200 mM for CorA (*, Œ, , r, F, f) and 20 mM for MgtB (ƒ, e, E, M). For each inhibitor, uptake at each concen- tration was normalized to the maximal uptake with no inhibitor present. The dose-response curves shown for CorA are the average of four independent experi- ments. Error bars are not shown for clar- ity but were within 6 4% at each inhibitor concentration. The dose-response curves shown for MgtB are the average of two independent experiments. Errors were within 6 5% at each inhibitor concentra- tion. Inhibitors are represented as: Mg (*), Co (Œ), Co(III)Hex (), Ru(II)Hex (ƒ), Ru(III)Hex (r), and Ni(II)Hex (e). Total cation uptake for CorA and MgtB 21 21 was about 1.0 and 0.2 nmol Ni min A , respectively. 600 nm TABLE I 63 21 Cation inhibition of Ni uptake by CorA 63 21 Uptake of Ni was assayed as described under “Materials and Methods” and the legends of Figs. 1–3. The concentration that inhibits 50% of 63 21 21 Ni uptake was determined from dose-response curves and is reported here as the IC . All assays were done at a final Ni concentration of 200 mM. IC Inhibitor S. typhimurium M. jannaschii mM Mg 10 10 Co 20 60 Cobalt(III)hexaammine 1 0.5 Ruthenium(II)hexaammine 3 3 Ruthenium(III)hexaammine 3 3 Nickel(II)hexaammine .30 .30 Chloropentaamminecobalt(III)chloride 90 100 Chloropentaammineruthenium(III)chloride 10 20 Hexacyanocobaltate(III) .1000 .1000 Hexacyanoruthenate(II) .1000 .1000 FIG.2. Inhibition of M. jannaschii 63 21 CorA Ni uptake. Transport was as- sayed as described under “Materials and Methods” using a final Ni concentration of 200 mM using MM1685, a Mg -trans- port deficient strain containing the M. jannaschii CorA gene on a plasmid. For each inhibitor, uptake at each concentra- tion was normalized to maximal uptake with no inhibitor present. The dose-re- sponse curves shown represent the aver- age of two independent experiments. Er- ror bars are not shown for clarity but were within 6 5% at each inhibitor concentra- tion. Inhibitors are represented as: Mg (*), Co (Œ), Co(III)Hex (), Ru(II)Hex (r), Ru(III)Hex (F), and Ni(II)Hex (f). Total cation uptake for M. jannaschii 21 21 CorA was about 0.15 nmol of Ni min A . 600 nm Cation Chloropentaammines Are Less Potent Inhibitors than strongly electronegative chlorine moiety might be expected to Hexaammines—Co(III)Pent and Ru(III)Pent differ from the alter the strength of hydrogen bonds formed by the ammine hexaammine compounds by the substitution of a covalently groups. Comparison of the interaction of cation hexaammines bound chlorine for one ammine group. The introduction of the or chloropentaammines with the transporter could reflect the 16770 Cation Hexaammines Are Inhibitors of CorA Transport System FIG.3. Chloropentaammines versus hexaammines as inhibitors of CorA 63 21 Ni uptake. The transport assay was performed as described under “Materials and Methods” using a final Ni concen- tration of 200 mM. The dose-response curves shown are the average of two in- dependent experiments. Error bars, not shown for clarity, were within 6 5% for each inhibitor concentration. Total cation uptake for CorA was about 1.0 Ni 21 21 min A . 600 nm importance of hydrogen bond based outer sphere interactions. tate(III), and hexacyanoruthenate(II) did not inhibit growth at For both S. typhimurium and M. jannaschii CorA transporters, mM. In contrast, 30 mM Co(III)Hex or concentrations of 100 Co(III)Pent and Ru(III)Pent were 5–10-fold less potent inhibi- Ru(III)Hex markedly inhibited growth of strain MM1927, tors than the corresponding hexaammine derivatives (Fig. 3). which is completely dependent on CorA for Mg uptake. In The IC values (Table I) for Co(III)Hex versus Co(III)Pent contrast, neither Co(III)Hex nor Ru(III)Hex inhibited the growth of strains dependent on the MgtA or MgtB transporters differed significantly (p , 0.0001), as did those of Ru(III)Hex versus Ru(III)Pent (p , 0.05). The potential importance of (data not shown). Further, the ability of Co(III)Hex or Ru(III)- outer sphere interactions involving hydrogen bond donors in Hex to inhibit CorA is not toxic per se, because neither com- CorA-mediated transport is further supported by the observa- pound inhibited growth of wild type S. typhimurium, which tion that potassium hexacyanocobaltate(III) and potassium expresses the MgtA and MgtB Mg transport systems in ad- 63 21 hexacyanoruthenate(II) derivatives could not inhibit Ni dition to CorA. uptake (Table I and data not shown). Potency of transport inhibition roughly corresponded with Growth Inhibition by Cation Hexaammines—We assayed the potency for growth inhibition because 100 mM Co(III)- and ability of the various cation derivatives to inhibit the growth of Ru(III)chloropentaammines were required to achieve the same a S. typhimurium strain in which CorA was the only functional growth inhibition as 30 mM Co(III)Hex or Ru(III)Hex (Fig. 4). Mg Similar growth inhibition was seen with an S. typhimurium transporter. This phenotypic assay provides a direct test 21 21 of the ability of the compounds to inhibit Mg uptake as strain dependent on the M. jannaschii CorA for Mg uptake opposed to inhibition of Ni uptake via CorA. It also provides (data not shown). an indication of the effectiveness of that inhibition in terms of The ability of the cation hexaammines to inhibit both trans- altering Mg port and growth could reflect an ability of the compounds to homeostasis. The ability of the cation derivatives to inhibit growth of a CorA-dependent strain of S. typhimurium inhibit transport sufficiently to deprive the cells of Mg ,orit correlated completely with the ability of the compounds to could reflect actual transport and intracellular accumulation of 63 21 inhibit Ni uptake via CorA. Ni(II)Hex, hexacyanocobal- the hexaammines leading to subsequent toxicity because of Cation Hexaammines Are Inhibitors of CorA Transport System 16771 FIG.4. Growth inhibition of S. ty- phimurium CorA-dependent strains. Growth of an S. typhimurium CorA-de- pendent strain was measured as de- scribed under “Materials and Methods.” A, growth inhibition by Ni(II)Hex, hexacyanoruthenate(II), and hexacyano- cobaltate(III) compared with addition of water. B, growth inhibition by hexaam- mine and chloropentaammine compounds at 30 and 100 mM. The control (water) curve is not shown for clarity but was identical to that in A. The data shown are a single experiment representative of three independent experiments. aMg transport-deficient strain, with both CoCl or Co(III)- TABLE II Cobalt content of cells after incubation with Cor Co(III)Hex Hex (Table II). The cobalt content of MM281 incubated with either CoCl or Co(III)Hex presumably reflects a residual Cobalt content 6 S.D. 2 Strain CorA expressed 21b amount of cation bound to the membrane or trapped in the Co Co(III)Hex periplasmic space. nmol/A 600nm In the presence of CoCl , the cobalt content of cells express- MM281 CorA 6.23 6 1.55 10.05 6 2.59 ing either the S. typhimurium or M. jannaschii CorA trans- MM1927 S. typhimurium CorA 19.33 6 4.72 8.32 6 2.71 porter was markedly greater than that of MM281 (p , 0.0001), MM1685 M. jannaschii CorA 24.42 6 1.70 9.59 6 3.64 reflecting previous data showing that Co influx can be me- Cobalt content was measured by atomic absorption as described diated by CorA. In contrast, the cobalt content of cells incu- under “Materials and Methods.” The results shown are the averages of three independent experiments. bated with Co(III)Hex did not differ from the negative controls Cells were grown in either 100 mM CoCl or 100 mM Co(III)Hex for 10 (p . 0.2). Thus we conclude that the cation hexa- and chloro- min. pentaammines inhibit growth by preventing Mg uptake suf- ficiently to interfere with cellular metabolism but are not them- interference with Mg selves transported into the cytosol. -dependent reactions. To determine Hexaammines Do Not Inhibit the MgtB Mg Transporter— which mechanism was most likely operative, we used atomic absorption spectroscopy to measure intracellular cobalt content MgtB is a P-type ATPase that mediates Mg influx in S. of cells incubated with Co(III)Hex or with CoCl as a positive typhimurium and has no sequence similarity to CorA (23). Its mechanism of Mg transport is therefore presumably differ- control (Table II). Negative controls included incubation of strains with Mg (data not shown) and incubation of MM281, ent than that of CorA, which is not dependent on ATP. No 16772 Cation Hexaammines Are Inhibitors of CorA Transport System FIG.5. Co(III)Hex does not repress PhoPQ-dependent transcription of the mgtCB operon. The luciferase activ- ity of strain MM1106, which contains a mgtCB-luxAB promoter fusion, was meas- ured as described under “Materials and Methods.” Culture aliquots were assayed in triplicate, and the averages of two in- dependent experiments are shown. inhibitory effect of cation hexaammine on MgtB-mediated The initial CorA cation binding site also has the property 21 21 Mg transport was seen (Fig. 1), indicating that Mg inter- that it does not discriminate by charge. Trivalent cations acts differently with the two transporters. (Co(III)Hex, Ru(III)Hex, Co(III)Pent, and Ru(III)Pent) as well 21 21 21 PhoPQ-mediated Transcription of the mgtCB Operon Is Not as divalent cations (Ru(II)Hex, Mg ,Ni , and Co ) are Affected by Co(III)Hex—PhoQ is the receptor or sensor kinase capable in interaction with CorA. of the PhoPQ two-component regulatory system (24, 25). The Selectivity for inhibition of CorA by analogs of hydrated physiological ligand for PhoQ is extracellular Mg (26) Tran- Mg is indicated by the failure of the Ni(II)Hex and hexacya- scription of the mgtCB operon is regulated by the PhoPQ; as nocobaltate(III) and hexacyanoruthenate(II) cations to inhibit 21 21 extracellular Mg decreases, Mg dissociates from PhoQ, CorA-mediated transport. The inability of the cyano com- resulting in the activation of the PhoP transcription factor. pounds can be reasonably attributed to their inability to form Cells grown in medium containing Mg were washed and the same type of hydrogen bonds as the ammine derivatives. resuspended in media containing either 30 mM Co(III)Hex, 10 Because Ni is a substrate for CorA, it seems surprising 21 21 mM Mg , both Co(III)Hex and Mg , or water. Transcription that Ni(II)Hex cannot inhibit uptake. This presumably reflects was activated in the cells resuspended in Co(III)Hex or water some difference in the properties of Ni(II)Hex versus the inhib- but was repressed in the presence of Mg (Fig. 5). Thus, itory cation hexaammines. Of the compounds tested here, ad- Co(III)Hex, at a concentration that can almost completely in- equate structural data appear to be available only for Ni(II)- hibit CorA-mediated Mg transport, could not repress mgtCB Hex, Co(III)Hex, and Co(III)Pent. All are stable at pH 7 in transcription via PhoPQ nor could Co(III)Hex block Mg re- aqueous solution, at least over the time frame of these experi- pression of transcription. We conclude that the Mg binding ments. For example, the half-life of Co(III)Hex under such site on PhoQ does not interact with the cation hexaammine. conditions is about 6 months (28, 29). The bond angle between the metal and the nitrogen of the ammines for all three com- DISCUSSION pounds is 90.0 6 0.2° (30 –34), indicating that the overall struc- Mg -dependent enzymes have been demonstrated to bind ture of the analogs is identical. Although the charge of the 21 21 Mg either through direct coordination to the Mg ion or cations is obviously different for Ni(II)Hex versus Co(III)Hex through an indirect interaction with the hydration shell. Cat- and Co(III)Pent, this is not a sufficient explanation for the ion hexaammines are analogs of hydrated cations that can be 21 failure of Ni(II)Hex to inhibit CorA because, as already noted used to study these Mg -ligand interactions (3, 8, 27). We 21 above, the CorA binding site does not distinguish as to charge. studied the effect of cation hexaammines on the Mg trans- The only significant difference between Ni(II)Hex and the two porters, CorA and MgtB, as well as the Mg binding sensor Co(III) compounds appears to be a small difference in their kinase PhoQ. Cation hexaammines were selective and potent size. The Co-N bond length is 1.97Å for both Co(III)Hex and inhibitors of CorA-mediated Mg transport but were not ef- Co(III)Pent while the Ni-N bond length is somewhat greater at fective Mg analogs for the MgtB and PhoQ proteins. 2.13 Å (30, 31). The N-H bond length is identical for all three. The most straightforward interpretation of the ability of This translates into a difference in diameter of about 0.3 Å and cation hexaammines to inhibit Mg uptake via CorA is that a difference in cation volume of approximately 20% (59.2– 60.7 the initial binding site for Mg on CorA binds a hydrated 3 3 Å for the Co(III) compounds versus 72.5 Å for Ni(II)Hex). If Mg ion. In contrast, their inability to inhibit MgtB or PhoQ this interpretation is correct (and sufficient), it suggests that indicates that these latter proteins bind a Mg ion that is at the cation binding site on CorA is between 4.9 and 5.2 Å in least partially stripped of its hydration shell. The selectivity of diameter. the cation hexaammines for CorA is further demonstrated by Growth of CorA-dependent strains was inhibited by cation their ability to inhibit the CorA transporter of M. jannaschii hexaammines. Cation hexaammines did not inhibit the growth despite low sequence homology. This latter result also strongly suggests that the structure of the periplasmic domains of these of a CorA knock-out strain, demonstrating that the inhibition is two CorA transporters is similar, although they share only a mediated by an interaction between CorA and the hexaam- 12% sequence identity in the extracellular periplasmic domain mines and that inhibition of CorA is sufficient to inhibit growth 21 21 that presumably is the initial binding site for Mg . if no other Mg transporters are present. The growth inhibi- Cation Hexaammines Are Inhibitors of CorA Transport System 16773 3. Cowan, J. A. (1995) The Biological Chemistry of Magnesium, VCH Publishers, tion is most likely due to competitive inhibition of Mg uptake New York, NY not through toxic effects of the cation hexaammines them- 4. Huang, H.-W., and Cowan, J. A. (1994) Eur. J. Biochem. 219, 253–260 5. Black, C. B., and Cowan, J. A. (1997) Eur. J. Biochem. 243, 684 – 689 selves. This conclusion is supported by the observation that 21 6. Hampel, A., and Cowan, J. A. (1997) Chem. Biol. 4, 513–517 although Co(III)Hex effectively inhibits CorA-mediated Mg 7. Uchiyama, Y., Iwai, S., Ueno, Y., Ikehara, M., and Ohtsuka, E. (1994) J. Bio- transport, cobalt is not apparently transported into the cell. In chem. (Tokyo) 116, 1322–1329 8. Cowan, J. A. (1998) Chem. Rev. 98, 1067–1087 turn, this is indirect evidence that there is a second step in- 9. Snavely, M. D., Florer, J. B., Miller, C. G., and Maguire, M. E. (1989) J. volved in cation transport, that is, at least the partial removal Bacteriol. 171, 4761– 4766 of the hydration shell. This is logical because if the hydration 10. Hmiel, S. P., Snavely, M. D., Florer, J. B., Maguire, M. E., and Miller, C. G. (1989) J. Bacteriol. 171, 4742– 4751 shell were not removed during passage through the membrane, 11. Hmiel, S. P., Snavely, M. D., Miller, C. G., and Maguire, M. E. (1986) J. it would imply that the pore or channel of CorA was 5Å in Bacteriol. 168, 1444 –1450 12. Kehres, D., Lawyer, C. H., and Maguire, M. E. (1998) Microb. Comp. Genomics diameter. Unfortunately, it is not possible to test how much of 43, 151–169 the hydration shell must be removed using partially substi- 13. Smith, R. L., Banks, J. L., Snavely, M. D., and Maguire, M. E. (1993) J. Biol. tuted cation ammines, e.g. a tetraammine-diaqua analog, be- Chem. 268, 14071–14080 14. Smith, R. L., Gottlieb, E., Kucharski, L. M., and Maguire, M. E. (1998) J. cause such compounds are unstable in aqueous solutions and Bacteriol. 180, 2788 –2791 hydrolyze readily (34, 35). 15. Smith, R. L., Szegedy, M. A., Kacharski, L. M., Walker, C., Wiet, R. M., Finally, CorA homologs have been found in nearly all bacte- Redpath, A., Kaczmarek, M. L., and Maguire, M. E. (1998) J. Biol. Chem. 273, 28663–28669 rial and archaeal genomes (12). This suggests that cation 16. Szegedy, M. A., and Maguire, M. E. (1999) J. Biol. Chem. 274, 36973–36979 hexaammines could inhibit growth of most organisms in these 17. Tao, T., Grulich, P. F., Kucharski, L. M., Smith, R. L., and Maguire, M. E. (1998) Microbiology 144, 655– 664 kingdoms. However, because many species have secondary 21 18. Miller, J. H. (1992) A Short Course in Bacterial Genetics, Cold Spring Harbor Mg transporters, cation hexaammines alone are not likely to Laboratory, Cold Spring Harbor, NY be bacteriocidal agents. Nonetheless, modified cation chloro- 19. Nelson, D. L., and Kennedy, E. P. (1971) J. Biol. Chem. 246, 3042–3049 20. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A pentaammines could have potential as antibiotics. Cation chlo- Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, ropentaammines of Co(III) and Ru(III) were potent inhibitors NY of CorA. The chlorine therefore had no major detrimental effect 21. Grubbs, R. D., Snavely, M. D., Hmiel, S. P., and Maguire, M. E. (1989) Methods Enzymol. 173, 546 –563 on binding, which implies that the binding site does not inter- 22. Moncrief, M. B. C., and Maguire, M. E. (1998) Infect. Immun. 66, 3802–3809 act with all six ligands bonded to the cation. It might therefore 23. Snavely, M. D., Miller, C. G., and Maguire, M. E. (1991) J. Biol. Chem. 266, 815– 823 be possible to replace the chlorine with a bacteriocidal moiety 24. Groisman, E. A., Chiao, E., Lipps, C. J., and Heffron, F. (1989) Proc. Natl. coupled via a spacer chain. Such a molecule might be selec- Acad. Sci. U. S. A. 86, 7077–7081 tively directed to the bacterial membrane because of the affin- 25. Soncini, F. C., Garcia, V. E., Solomon, F., and Groisman, E. A. (1996) J. Bacteriol. 178, 5092–5099 ity of cation pentaammine derivatives for the CorA transporter, 26. Vescovi, E. G., Ayala, Y. M., Di Cera, E., and Groisman, E. A. (1997) J. Biol. allowing selective delivery of the bacteriocidal component. Chem. 272, 1440 –1443 Thus, in addition to being useful tools for studying Mg 27. Cowan, J. A. (1993) J. Inorg. Biochem. 49, 171–175 inter- 28. Llewellyn, D. R., O’Connor, C. J., and Odell, A. L. (1964) J. Chem. Soc. 1964, actions, cation hexaammine compounds could potentially be developed as broad spectrum antibiotics based on their speci- 29. Rutenberg, A. C., and Drury, J. S. (1963) Inorg. Chem. 2, 219 30. Meek, D. W., and Ibers, J. A. (1970) Inorg. Chem. 9, 465– 470 ficity for CorA. 31. Hummel, H.-U., and Beiler, F. (1988) Z. Anorg. Allg. Chem. 565, 147–153 32. Iwata, M., and Saito, Y. (1973) Acta Crystallogr. Sect. B Struct. Sci. 29, Acknowledgments—We thank Dr. A. Romani for assistance with the 822– 832 atomic absorption measurements and Dr. J. Cowan for advice and for 33. Messmer, G. G., and Amma, E. L. (1968) Acta Crystallogr. Sect. B Struct. Sci. sharing expertise about cation hexaammines. 24, 417– 422 34. Basolo, F., and Pearson, R. G. (1967) in Mechanisms of Inorganic Reactions: A REFERENCES Study of Metal Complexes in Solution, pp. 158 –170, John Wiley & Sons, 1. Martin, R. B. (1990) Metal Ions Biol. 26, 1–13 Inc., New York 2. Smith, R. L., and Maguire, M. E. (1998) Mol. Microbiol. 28, 217–226 35. Hendry, P., and Ludi, A. (1990) Adv. Inorg. Chem. 35, 117–198 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Cation Hexaammines Are Selective and Potent Inhibitors of the CorA Magnesium Transport System

Kucharski, Lisa M.; Lubbe, Wil J.; Maguire, Michael E.
Journal of Biological Chemistry – Jun 1, 2000
Download PDF
7 pages

Loading next page...
 
/lp/unpaywall/cation-hexaammines-are-selective-and-potent-inhibitors-of-the-cora-eBf8R09LnV

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m001507200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 275, No. 22, Issue of June 2, pp. 16767–16773, 2000 © 2000 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Cation Hexaammines Are Selective and Potent Inhibitors of the CorA Magnesium Transport System* Received for publication, February 23, 2000, and in revised form, March 27, 2000 Published, JBC Papers in Press, March 27, 2000, DOI 10.1074/jbc.M001507200 Lisa M. Kucharski‡, Wil J. Lubbe, and Michael E. Maguire From the Department of Pharmacology, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106-4965 Cation hexaammines and related compounds are can be through direct, inner sphere, coordination of a ligand to the metal ion or through an indirect, outer sphere, association chemically stable analogs of the hydrated form of cat- ions, particularly Mg . We tested the ability of several with the bound water shell of the cation. In an outer sphere of these compounds to inhibit transport by the CorA or interaction, the relatively rigid geometry imposed by Mg MgtB Mg transport systems or the PhoQ receptor ki- essentially holds a water molecule in place within the active nase for Mg in Salmonella typhimurium. Cobalt(III)-, site of an enzyme. The outer sphere form of Mg -mediated ruthenium(II)-, and ruthenium(III)-hexaammines were catalysis has been demonstrated in exonuclease III, ribonucle- potent inhibitors of CorA-mediated influx. Cobalt(III)- ase H, and hairpin ribozymes (4 –7). and ruthenium(III)chloropentaammines were slightly Thus it is of interest to determine whether biochemical proc- less potent inhibitors of CorA. The compounds inhibited 21 esses involving Mg are of an inner sphere or outer sphere uptake by the bacterial S. typhimurium CorA and by the type. Cation hexaammines, such as cobalt(III)hexaammine archaeal Methanococcus jannaschii CorA, which bear (Co(III)Hex), are useful analogs of hydrated cations. The co- only 12% identity in the extracellular periplasmic do- valently and stably bound ammine groups mimic the size and main. Cation hexaammines also inhibited growth of S. shape of the hydration shell. These substitutionally inert typhimurium strains dependent on CorA for Mg up- 21 hexaammine compounds provide a means to differentiate be- take but not of isogenic strains carrying a second Mg tween inner and outer sphere Mg interactions (3, 8). If an uptake system. In contrast, hexacyano-cobaltate(III) enzyme is activated or inhibited by a substitutionally inert and ruthenate(II)- and nickel(II)hexaammine had little metal complex, an inner sphere interaction is not likely. There effect on uptake. The inhibition by the cation hexaam- are a variety of such compounds with variations in net charge mines was selective for CorA because none of the com- and number of hydrogen bond donors. pounds had any effect on transport by the MgtB P-type 21 21 This laboratory initially cloned and characterized the Sal- ATPase Mg transporter or the PhoQ Mg receptor kinase. These results demonstrate that cation hexaam- monella typhimurium CorA Mg transporter (9 –11). CorA is mines are potent and highly selective inhibitors of the virtually ubiquitous in the bacteria and archaea and forms CorA Mg transport system and further indicate that their constitutive Mg influx system (12). The topology of the initial interaction of the CorA transporter is with a CorA consists of a large N-terminal periplasmic domain and fully hydrated Mg cation. three transmembrane segments at the C terminus (13). We have also shown that even distant homologs of CorA function as Mg transporters. The CorA homolog of the Archeon Meth- The chemistry of Mg is unique among the biologically anococcus jannaschii has only about 16% overall sequence relevant cations. Compared with calcium, potassium, and so- identity to the S. typhimurium CorA. Expression of the M. dium, Mg possesses the largest hydrated radius, the smallest jannaschii CorA in a Mg transport-deficient strain of S. ionic radius, and the greatest charge density (1). Its volume typhimurium elicits Mg uptake with properties virtually change from hydrated to atomic cation is almost 400-fold, identical to that of the Salmonella transporter (14). 21 1 1 whereas Ca ,Na , and K are only 25-, 25-, and 4-fold, The mechanism of cation transport through CorA appears respectively. Waters of hydration are bound 3– 4 orders of mag- novel. In contrast to virtually all known cation transporters nitude more tightly to Mg than to other common biological and channels, CorA does not utilize electrostatic interactions to cations. The geometry of the hydration shell is octahedral with move the highly charge dense Mg cation across the mem- all six coordinate bond angles at angles of 90 6 6 . This con- brane. The transmembrane domain of S. typhimurium CorA trasts with the far more flexible geometry of the Ca cation, contains only a single negatively charged residue. Its mutation which can tolerate bond angles of 90 6 35 (1–3). to alanine has no effect on transport. Site-directed mutagenesis Study of Mg -dependent enzymes has revealed two types of of amino acids in the third transmembrane domain revealed interactions between Mg , enzymes, and ligands. Interaction the importance of three specific residues that line a single face of an a-helix. Likewise, mutagenesis of the second transmem- brane segment indicated that a single face of its a-helix is * This work was supported by National Institutes of Health Grant important (15, 16). Despite the lack of electrostatic interactions GM39447 (to M.E.M.) and a summer undergraduate fellowship from between Mg and the membrane domain of CorA, the large the American Society of Pharmacology and Experimental Therapeutics periplasmic domain is highly charged. It presumably functions, (to W. J. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby at least in part, to interact with hydrated Mg outside the cell marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ To whom correspondence should be addressed: Dept. of Pharmacol- The abbreviations used are: Co(III)Hex, cobalt(III)hexaammine; Ru(II) ogy, School of Medicine, Case Western Reserve University, 10900 Eu- Hex, ruthenium(II)hexaammine; Ru(III)Hex, ruthenium(III)hexaammine; clid Ave., Cleveland, OH 44106-4965. Tel.: 216-368-6187; Fax: 216-368- Ni(II)Hex, nickel(II)hexaammine; Co(III)Pent, chloropentaamminecobalt 3395; E-mail: [email protected]. (III)chloride; Ru(III)Pent, chloropentaammineruthenium(III)dichloride. This paper is available on line at http://www.jbc.org 16767 This is an Open Access article under the CC BY license. 16768 Cation Hexaammines Are Inhibitors of CorA Transport System supplemented N-minimal medium to an A between 0.050 and and provide an initial binding site for subsequent interaction 600 nm 0.080. These cultures were supplemented with Mg or Co(III)Hex as with the membrane domain. indicated and grown at 37 °C. At each time point, a 250-ml aliquot was We therefore tested cation hexaammines and related com- taken for luminescence and optical density readings. Luciferase activity pounds for their ability to inhibit Mg transport via CorA. was measured in triplicate for each sample by adding 10 ml of cells to 1 Both S. typhimurium and M. jannaschii CorA are potently ml of luciferase assay buffer, vortexing for 5 s, and reading the sample inhibited by cobalt(III)hexaammine and other similar analogs. in a luminometer for 10 s. Luminometer readings were normalized to cell number. In contrast, these same compounds did not inhibit other pro- Atomic Absorption—MM1927, MM1685, and MM281 were grown teins in S. typhimurium known to interact selectively with 21 overnight in LB broth plus 100 mM Mg and antibiotics. The cells were Mg . These results suggest that the initial binding site for washed twice with supplemented N-minimal medium containing no Mg in the CorA periplasmic domain binds a fully hydrated 21 Mg . The cells were resuspended in supplemented N-minimal medium Mg ion. plus Mg , Co(III)Hex, or cobalt chloride as indicated, to an A of 600 nm 1.5. At each time point an aliquot of 1.2 ml of cells was taken and MATERIALS AND METHODS overlaid on 300 ml of a 2:1 mixture of dibutyl and dioctyl phthalate. The 63 21 Plasmids and Strains—CorA-dependent Ni uptake and growth samples were spun in a microcentrifuge for 30 s, the supernatant was were assayed using strain MM1927 (16) for S. typhimurium experi- carefully aspirated, and the pellet was resuspended in 300 mlof1.0 N ments and MM1685 (14) for M. jannaschii experiments. MM1927 was nitric acid. The samples were then batch sonicated for 1 min, spun in a constructed by transforming the CorA plasmid, pMAS29, into MM281, microcentrifuge to pellet cellular debris, and the supernatant was aMg transport-deficient strain. MM1685 was constructed by trans- transferred to a new tube. Cobalt content of the supernatant was forming the M. jannaschii CorA-containing plasmid, pMETJA, into measured by atomic absorption as described previously (22). MM281. Growth inhibition was also assayed in MM199, a strain in Growth Inhibition—MM1927 and MM1685 were grown overnight in which corA has been inactivated by insertion of a transposon but that LB broth plus 100 mM Mg and antibiotics. The cells were washed still retains functional alleles for the MgtA and MgtB Mg transport- twice with Mg -free N-minimal medium and then diluted to a concen- 63 21 ers (9). MgtB-dependent Ni uptake was assayed using strain tration of 5.6 3 10 cells/ml in supplemented N-minimal medium plus MM1111, which is MM281 transformed with the mgtB-containing plas- 50 mM Mg . Cells (90 ml) and 10 ml of the appropriate concentration of mid pSPH39 (10). PhoPQ-dependent transcription of the mgtCB operon each inhibitor were aliquoted into each well of a 96-well plate. The plate was assayed using strain MM1106, which contains pTTC-Lux, a plas- lid was sterilely swabbed with NeverFog® (North American Oil Com- mid with a mgtCB-luciferase reporter fusion (17). pany, Atlanta, GA) anti-fog shield to protect against condensation. The Reagents and Buffers—Cobalt(III)-, ruthenium(II)-, and ruthenium- A of the cells in each well was measured every 30 min for 16 h in 650 nm TM (III)hexaammine chlorides were purchased from Aldrich. Nickel- a Molecular Devices THERMOmax microplate reader (16). MM199 (II)hexaammine, cobalt(III)- and ruthenium(III)chloropentaammine, po- was grown overnight in LB broth plus 100 mM Mg and 50 mg/ml tassium hexacyanocobaltate(III), and potassium hexacyanoruthenate(II) kanamycin. The cells were washed twice with Mg -free supplemented chlorides were obtained from Alfa Chemicals (Ward Hill, MA). N-minimal medium and then resuspended to an A of 0.1 in 5 ml of 600 nm Ni(III)hexaammine and Co(II)hexaammine were not tested, the former supplemented N-minimal medium containing 50 mM Mg plus either because it is not available and the latter because it is not stable in 100 mM Co(III)Hex, Ru(II)Hex or no inhibitor. The A of the cul- 600 nm aqueous solution at physiological pH. Co(III)Hex, Ru(II)Hex, Ru(III)- tures was determined after overnight growth at 37 °C. Statistical anal- Hex, and Ni(II)Hex stock solutions were stored at 220 °C. All other yses using unpaired t tests were calculated using GraphPad Prism solutions were made fresh for each use. All other reagents were ob- software (version 3.0). tained from Sigma, unless otherwise specified. Cultures were grown with shaking at 37 °C, with appropriate antibiotics, in Luria-Bertani RESULTS (LB) broth (18) or supplemented N-minimal medium, which is N-min- Cation Hexaammines Inhibit the CorA Mg Transporter— imal salts plus 0.4% (w/v) glucose, 0.1% (w/v) casamino acids, and 1 mM 21 63 21 leucine (15, 19). Ampicillin was used at 60 mg/ml, chloramphenicol at Mg uptake via CorA is studied using Ni as a substitute 28 21 170 mg/ml, and kanamycin at 50 mg/ml (20). Magnesium was added, because Mg is no longer available (21). We used this assay where indicated, as MgSO . Luciferase assay buffer contains 50 mM to examine the effect of various cation hexaammine compounds sodium phosphate buffer, pH 7.5, with 0.01% dodecyl aldehyde. on CorA-mediated uptake in S. typhimurium. Co(III)Hex, 63 21 Transport— Ni uptake was measured as described previously (9, 63 21 Ru(II)Hex, and Ru(III)Hex were potent inhibitors of Ni 21). MM1111 was grown overnight in LB broth containing 100 mM uptake (Fig. 1). All three were more potent inhibitors than the Mg . Cells were diluted 1:50 in supplemented N-minimal medium 21 21 21 21 containing 1 mM Mg and antibiotics. After 8 h growth, cells were actual substrates of CorA: Mg ,Ni , and Co . Co(III)Hex collected by centrifugation at 1,000 3 g for 15 min and washed twice in had an IC of 1 mM (Table I), which at the 200 mM Ni an equal volume of Mg -free N-minimal medium. These cells were concentration used, corresponds to a K of 500 nM. In contrast, resuspended to a final A of 0.1 in supplemented N-minimal me- 63 21 600 nm Ni(II)Hex did not inhibit Ni uptake at concentrations up to dium containing 10 mM Mg and antibiotics and incubated overnight at 30 mM (Fig. 2). 37 °C. MM1927 and MM1685 were grown overnight in LB broth con- The archeon M. jannaschii contains a CorA homolog with taining 100 mM Mg and appropriate antibiotics. After overnight growth, all strains were washed twice with Mg -free N-minimal me- about 38% overall sequence similarity (16% identity) to the S. dium and resuspended to an A between 1.0 and 2.0 in Mg -free 600 nm typhimurium CorA. The identity in the periplasmic domain of supplemented N-minimal medium. CorA-containing cells were washed approximately 235 residues is significantly lower at 32% over- with ice-cold buffer and kept on ice prior to the assay. MgtB-containing all similarity and 12% identity (12). We have previously shown cells could not be washed with cold media or stored on ice without 63 21 that this homolog is a true Mg transporter, functional in S. significant loss of transport activity (9). Transport of Ni was as- typhimurium with virtually identical properties as the native sayed at 37 °C by adding 0.1 ml of cells to 0.9 ml of supplemented 63 21 21 N-minimal medium containing 0.1–1.5 mCi of Ni and 200 mM Ni S. typhimurium transporter (14). In concert with those results, for assay of CorA or 20 mM Ni for assay of MgtB, plus the indicated cation hexaammines also inhibited uptake mediated by the M. concentration of inhibitor. These concentrations of Ni approximate jannaschii CorA. The rank order of potency (Co(III)Hex , the K for the respective transporters (9). Uptake was stopped after 5 Ru(II)Hex 5 Ru(III)Hex) was identical for both CorA transport- min for CorA or 30 min for MgtB by addition of 5 ml of ice-cold transport ers. As with the S. typhimurium CorA, Ni(II)Hex also did not wash buffer (N-minimal salts containing 5 mM Mg and1mM EDTA). Samples were filtered through BA85 (0.45 mm) nitrocellulose filters inhibit the M. jannaschii CorA transporter. The IC for each (Schleicher & Schuell) and washed once with 5 ml of transport wash hexaammine (Table I) did not differ significantly between the 63 21 buffer. Ni activity was determined by scintillation counting with an two CorA homologs (p . 0.1), although there was a slight efficiency of 85%. difference in Co affinity between the two CorA transporters Luciferase Assays—MM1106, a strain containing a mgtCB promoter- (p , 0.0001). These data indicate that the overall structure of luciferase reporter construct (17), was grown overnight at 37 °C in LB the cation binding site for the two transporters is functionally broth plus 100 mM Mg and antibiotics. The cells were washed three times with N-minimal medium containing no Mg and resuspended in identical. Cation Hexaammines Are Inhibitors of CorA Transport System 16769 FIG.1. Inhibition of S. typhimurium 63 21 CorA and MgtB Ni uptake. Trans- port was assayed as described under “Ma- terials and Methods” using a final Ni concentration of 200 mM for CorA (*, Œ, , r, F, f) and 20 mM for MgtB (ƒ, e, E, M). For each inhibitor, uptake at each concen- tration was normalized to the maximal uptake with no inhibitor present. The dose-response curves shown for CorA are the average of four independent experi- ments. Error bars are not shown for clar- ity but were within 6 4% at each inhibitor concentration. The dose-response curves shown for MgtB are the average of two independent experiments. Errors were within 6 5% at each inhibitor concentra- tion. Inhibitors are represented as: Mg (*), Co (Œ), Co(III)Hex (), Ru(II)Hex (ƒ), Ru(III)Hex (r), and Ni(II)Hex (e). Total cation uptake for CorA and MgtB 21 21 was about 1.0 and 0.2 nmol Ni min A , respectively. 600 nm TABLE I 63 21 Cation inhibition of Ni uptake by CorA 63 21 Uptake of Ni was assayed as described under “Materials and Methods” and the legends of Figs. 1–3. The concentration that inhibits 50% of 63 21 21 Ni uptake was determined from dose-response curves and is reported here as the IC . All assays were done at a final Ni concentration of 200 mM. IC Inhibitor S. typhimurium M. jannaschii mM Mg 10 10 Co 20 60 Cobalt(III)hexaammine 1 0.5 Ruthenium(II)hexaammine 3 3 Ruthenium(III)hexaammine 3 3 Nickel(II)hexaammine .30 .30 Chloropentaamminecobalt(III)chloride 90 100 Chloropentaammineruthenium(III)chloride 10 20 Hexacyanocobaltate(III) .1000 .1000 Hexacyanoruthenate(II) .1000 .1000 FIG.2. Inhibition of M. jannaschii 63 21 CorA Ni uptake. Transport was as- sayed as described under “Materials and Methods” using a final Ni concentration of 200 mM using MM1685, a Mg -trans- port deficient strain containing the M. jannaschii CorA gene on a plasmid. For each inhibitor, uptake at each concentra- tion was normalized to maximal uptake with no inhibitor present. The dose-re- sponse curves shown represent the aver- age of two independent experiments. Er- ror bars are not shown for clarity but were within 6 5% at each inhibitor concentra- tion. Inhibitors are represented as: Mg (*), Co (Œ), Co(III)Hex (), Ru(II)Hex (r), Ru(III)Hex (F), and Ni(II)Hex (f). Total cation uptake for M. jannaschii 21 21 CorA was about 0.15 nmol of Ni min A . 600 nm Cation Chloropentaammines Are Less Potent Inhibitors than strongly electronegative chlorine moiety might be expected to Hexaammines—Co(III)Pent and Ru(III)Pent differ from the alter the strength of hydrogen bonds formed by the ammine hexaammine compounds by the substitution of a covalently groups. Comparison of the interaction of cation hexaammines bound chlorine for one ammine group. The introduction of the or chloropentaammines with the transporter could reflect the 16770 Cation Hexaammines Are Inhibitors of CorA Transport System FIG.3. Chloropentaammines versus hexaammines as inhibitors of CorA 63 21 Ni uptake. The transport assay was performed as described under “Materials and Methods” using a final Ni concen- tration of 200 mM. The dose-response curves shown are the average of two in- dependent experiments. Error bars, not shown for clarity, were within 6 5% for each inhibitor concentration. Total cation uptake for CorA was about 1.0 Ni 21 21 min A . 600 nm importance of hydrogen bond based outer sphere interactions. tate(III), and hexacyanoruthenate(II) did not inhibit growth at For both S. typhimurium and M. jannaschii CorA transporters, mM. In contrast, 30 mM Co(III)Hex or concentrations of 100 Co(III)Pent and Ru(III)Pent were 5–10-fold less potent inhibi- Ru(III)Hex markedly inhibited growth of strain MM1927, tors than the corresponding hexaammine derivatives (Fig. 3). which is completely dependent on CorA for Mg uptake. In The IC values (Table I) for Co(III)Hex versus Co(III)Pent contrast, neither Co(III)Hex nor Ru(III)Hex inhibited the growth of strains dependent on the MgtA or MgtB transporters differed significantly (p , 0.0001), as did those of Ru(III)Hex versus Ru(III)Pent (p , 0.05). The potential importance of (data not shown). Further, the ability of Co(III)Hex or Ru(III)- outer sphere interactions involving hydrogen bond donors in Hex to inhibit CorA is not toxic per se, because neither com- CorA-mediated transport is further supported by the observa- pound inhibited growth of wild type S. typhimurium, which tion that potassium hexacyanocobaltate(III) and potassium expresses the MgtA and MgtB Mg transport systems in ad- 63 21 hexacyanoruthenate(II) derivatives could not inhibit Ni dition to CorA. uptake (Table I and data not shown). Potency of transport inhibition roughly corresponded with Growth Inhibition by Cation Hexaammines—We assayed the potency for growth inhibition because 100 mM Co(III)- and ability of the various cation derivatives to inhibit the growth of Ru(III)chloropentaammines were required to achieve the same a S. typhimurium strain in which CorA was the only functional growth inhibition as 30 mM Co(III)Hex or Ru(III)Hex (Fig. 4). Mg Similar growth inhibition was seen with an S. typhimurium transporter. This phenotypic assay provides a direct test 21 21 of the ability of the compounds to inhibit Mg uptake as strain dependent on the M. jannaschii CorA for Mg uptake opposed to inhibition of Ni uptake via CorA. It also provides (data not shown). an indication of the effectiveness of that inhibition in terms of The ability of the cation hexaammines to inhibit both trans- altering Mg port and growth could reflect an ability of the compounds to homeostasis. The ability of the cation derivatives to inhibit growth of a CorA-dependent strain of S. typhimurium inhibit transport sufficiently to deprive the cells of Mg ,orit correlated completely with the ability of the compounds to could reflect actual transport and intracellular accumulation of 63 21 inhibit Ni uptake via CorA. Ni(II)Hex, hexacyanocobal- the hexaammines leading to subsequent toxicity because of Cation Hexaammines Are Inhibitors of CorA Transport System 16771 FIG.4. Growth inhibition of S. ty- phimurium CorA-dependent strains. Growth of an S. typhimurium CorA-de- pendent strain was measured as de- scribed under “Materials and Methods.” A, growth inhibition by Ni(II)Hex, hexacyanoruthenate(II), and hexacyano- cobaltate(III) compared with addition of water. B, growth inhibition by hexaam- mine and chloropentaammine compounds at 30 and 100 mM. The control (water) curve is not shown for clarity but was identical to that in A. The data shown are a single experiment representative of three independent experiments. aMg transport-deficient strain, with both CoCl or Co(III)- TABLE II Cobalt content of cells after incubation with Cor Co(III)Hex Hex (Table II). The cobalt content of MM281 incubated with either CoCl or Co(III)Hex presumably reflects a residual Cobalt content 6 S.D. 2 Strain CorA expressed 21b amount of cation bound to the membrane or trapped in the Co Co(III)Hex periplasmic space. nmol/A 600nm In the presence of CoCl , the cobalt content of cells express- MM281 CorA 6.23 6 1.55 10.05 6 2.59 ing either the S. typhimurium or M. jannaschii CorA trans- MM1927 S. typhimurium CorA 19.33 6 4.72 8.32 6 2.71 porter was markedly greater than that of MM281 (p , 0.0001), MM1685 M. jannaschii CorA 24.42 6 1.70 9.59 6 3.64 reflecting previous data showing that Co influx can be me- Cobalt content was measured by atomic absorption as described diated by CorA. In contrast, the cobalt content of cells incu- under “Materials and Methods.” The results shown are the averages of three independent experiments. bated with Co(III)Hex did not differ from the negative controls Cells were grown in either 100 mM CoCl or 100 mM Co(III)Hex for 10 (p . 0.2). Thus we conclude that the cation hexa- and chloro- min. pentaammines inhibit growth by preventing Mg uptake suf- ficiently to interfere with cellular metabolism but are not them- interference with Mg selves transported into the cytosol. -dependent reactions. To determine Hexaammines Do Not Inhibit the MgtB Mg Transporter— which mechanism was most likely operative, we used atomic absorption spectroscopy to measure intracellular cobalt content MgtB is a P-type ATPase that mediates Mg influx in S. of cells incubated with Co(III)Hex or with CoCl as a positive typhimurium and has no sequence similarity to CorA (23). Its mechanism of Mg transport is therefore presumably differ- control (Table II). Negative controls included incubation of strains with Mg (data not shown) and incubation of MM281, ent than that of CorA, which is not dependent on ATP. No 16772 Cation Hexaammines Are Inhibitors of CorA Transport System FIG.5. Co(III)Hex does not repress PhoPQ-dependent transcription of the mgtCB operon. The luciferase activ- ity of strain MM1106, which contains a mgtCB-luxAB promoter fusion, was meas- ured as described under “Materials and Methods.” Culture aliquots were assayed in triplicate, and the averages of two in- dependent experiments are shown. inhibitory effect of cation hexaammine on MgtB-mediated The initial CorA cation binding site also has the property 21 21 Mg transport was seen (Fig. 1), indicating that Mg inter- that it does not discriminate by charge. Trivalent cations acts differently with the two transporters. (Co(III)Hex, Ru(III)Hex, Co(III)Pent, and Ru(III)Pent) as well 21 21 21 PhoPQ-mediated Transcription of the mgtCB Operon Is Not as divalent cations (Ru(II)Hex, Mg ,Ni , and Co ) are Affected by Co(III)Hex—PhoQ is the receptor or sensor kinase capable in interaction with CorA. of the PhoPQ two-component regulatory system (24, 25). The Selectivity for inhibition of CorA by analogs of hydrated physiological ligand for PhoQ is extracellular Mg (26) Tran- Mg is indicated by the failure of the Ni(II)Hex and hexacya- scription of the mgtCB operon is regulated by the PhoPQ; as nocobaltate(III) and hexacyanoruthenate(II) cations to inhibit 21 21 extracellular Mg decreases, Mg dissociates from PhoQ, CorA-mediated transport. The inability of the cyano com- resulting in the activation of the PhoP transcription factor. pounds can be reasonably attributed to their inability to form Cells grown in medium containing Mg were washed and the same type of hydrogen bonds as the ammine derivatives. resuspended in media containing either 30 mM Co(III)Hex, 10 Because Ni is a substrate for CorA, it seems surprising 21 21 mM Mg , both Co(III)Hex and Mg , or water. Transcription that Ni(II)Hex cannot inhibit uptake. This presumably reflects was activated in the cells resuspended in Co(III)Hex or water some difference in the properties of Ni(II)Hex versus the inhib- but was repressed in the presence of Mg (Fig. 5). Thus, itory cation hexaammines. Of the compounds tested here, ad- Co(III)Hex, at a concentration that can almost completely in- equate structural data appear to be available only for Ni(II)- hibit CorA-mediated Mg transport, could not repress mgtCB Hex, Co(III)Hex, and Co(III)Pent. All are stable at pH 7 in transcription via PhoPQ nor could Co(III)Hex block Mg re- aqueous solution, at least over the time frame of these experi- pression of transcription. We conclude that the Mg binding ments. For example, the half-life of Co(III)Hex under such site on PhoQ does not interact with the cation hexaammine. conditions is about 6 months (28, 29). The bond angle between the metal and the nitrogen of the ammines for all three com- DISCUSSION pounds is 90.0 6 0.2° (30 –34), indicating that the overall struc- Mg -dependent enzymes have been demonstrated to bind ture of the analogs is identical. Although the charge of the 21 21 Mg either through direct coordination to the Mg ion or cations is obviously different for Ni(II)Hex versus Co(III)Hex through an indirect interaction with the hydration shell. Cat- and Co(III)Pent, this is not a sufficient explanation for the ion hexaammines are analogs of hydrated cations that can be 21 failure of Ni(II)Hex to inhibit CorA because, as already noted used to study these Mg -ligand interactions (3, 8, 27). We 21 above, the CorA binding site does not distinguish as to charge. studied the effect of cation hexaammines on the Mg trans- The only significant difference between Ni(II)Hex and the two porters, CorA and MgtB, as well as the Mg binding sensor Co(III) compounds appears to be a small difference in their kinase PhoQ. Cation hexaammines were selective and potent size. The Co-N bond length is 1.97Å for both Co(III)Hex and inhibitors of CorA-mediated Mg transport but were not ef- Co(III)Pent while the Ni-N bond length is somewhat greater at fective Mg analogs for the MgtB and PhoQ proteins. 2.13 Å (30, 31). The N-H bond length is identical for all three. The most straightforward interpretation of the ability of This translates into a difference in diameter of about 0.3 Å and cation hexaammines to inhibit Mg uptake via CorA is that a difference in cation volume of approximately 20% (59.2– 60.7 the initial binding site for Mg on CorA binds a hydrated 3 3 Å for the Co(III) compounds versus 72.5 Å for Ni(II)Hex). If Mg ion. In contrast, their inability to inhibit MgtB or PhoQ this interpretation is correct (and sufficient), it suggests that indicates that these latter proteins bind a Mg ion that is at the cation binding site on CorA is between 4.9 and 5.2 Å in least partially stripped of its hydration shell. The selectivity of diameter. the cation hexaammines for CorA is further demonstrated by Growth of CorA-dependent strains was inhibited by cation their ability to inhibit the CorA transporter of M. jannaschii hexaammines. Cation hexaammines did not inhibit the growth despite low sequence homology. This latter result also strongly suggests that the structure of the periplasmic domains of these of a CorA knock-out strain, demonstrating that the inhibition is two CorA transporters is similar, although they share only a mediated by an interaction between CorA and the hexaam- 12% sequence identity in the extracellular periplasmic domain mines and that inhibition of CorA is sufficient to inhibit growth 21 21 that presumably is the initial binding site for Mg . if no other Mg transporters are present. The growth inhibi- Cation Hexaammines Are Inhibitors of CorA Transport System 16773 3. Cowan, J. A. (1995) The Biological Chemistry of Magnesium, VCH Publishers, tion is most likely due to competitive inhibition of Mg uptake New York, NY not through toxic effects of the cation hexaammines them- 4. Huang, H.-W., and Cowan, J. A. (1994) Eur. J. Biochem. 219, 253–260 5. Black, C. B., and Cowan, J. A. (1997) Eur. J. Biochem. 243, 684 – 689 selves. This conclusion is supported by the observation that 21 6. Hampel, A., and Cowan, J. A. (1997) Chem. Biol. 4, 513–517 although Co(III)Hex effectively inhibits CorA-mediated Mg 7. Uchiyama, Y., Iwai, S., Ueno, Y., Ikehara, M., and Ohtsuka, E. (1994) J. Bio- transport, cobalt is not apparently transported into the cell. In chem. (Tokyo) 116, 1322–1329 8. Cowan, J. A. (1998) Chem. Rev. 98, 1067–1087 turn, this is indirect evidence that there is a second step in- 9. Snavely, M. D., Florer, J. B., Miller, C. G., and Maguire, M. E. (1989) J. volved in cation transport, that is, at least the partial removal Bacteriol. 171, 4761– 4766 of the hydration shell. This is logical because if the hydration 10. Hmiel, S. P., Snavely, M. D., Florer, J. B., Maguire, M. E., and Miller, C. G. (1989) J. Bacteriol. 171, 4742– 4751 shell were not removed during passage through the membrane, 11. Hmiel, S. P., Snavely, M. D., Miller, C. G., and Maguire, M. E. (1986) J. it would imply that the pore or channel of CorA was 5Å in Bacteriol. 168, 1444 –1450 12. Kehres, D., Lawyer, C. H., and Maguire, M. E. (1998) Microb. Comp. Genomics diameter. Unfortunately, it is not possible to test how much of 43, 151–169 the hydration shell must be removed using partially substi- 13. Smith, R. L., Banks, J. L., Snavely, M. D., and Maguire, M. E. (1993) J. Biol. tuted cation ammines, e.g. a tetraammine-diaqua analog, be- Chem. 268, 14071–14080 14. Smith, R. L., Gottlieb, E., Kucharski, L. M., and Maguire, M. E. (1998) J. cause such compounds are unstable in aqueous solutions and Bacteriol. 180, 2788 –2791 hydrolyze readily (34, 35). 15. Smith, R. L., Szegedy, M. A., Kacharski, L. M., Walker, C., Wiet, R. M., Finally, CorA homologs have been found in nearly all bacte- Redpath, A., Kaczmarek, M. L., and Maguire, M. E. (1998) J. Biol. Chem. 273, 28663–28669 rial and archaeal genomes (12). This suggests that cation 16. Szegedy, M. A., and Maguire, M. E. (1999) J. Biol. Chem. 274, 36973–36979 hexaammines could inhibit growth of most organisms in these 17. Tao, T., Grulich, P. F., Kucharski, L. M., Smith, R. L., and Maguire, M. E. (1998) Microbiology 144, 655– 664 kingdoms. However, because many species have secondary 21 18. Miller, J. H. (1992) A Short Course in Bacterial Genetics, Cold Spring Harbor Mg transporters, cation hexaammines alone are not likely to Laboratory, Cold Spring Harbor, NY be bacteriocidal agents. Nonetheless, modified cation chloro- 19. Nelson, D. L., and Kennedy, E. P. (1971) J. Biol. Chem. 246, 3042–3049 20. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A pentaammines could have potential as antibiotics. Cation chlo- Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, ropentaammines of Co(III) and Ru(III) were potent inhibitors NY of CorA. The chlorine therefore had no major detrimental effect 21. Grubbs, R. D., Snavely, M. D., Hmiel, S. P., and Maguire, M. E. (1989) Methods Enzymol. 173, 546 –563 on binding, which implies that the binding site does not inter- 22. Moncrief, M. B. C., and Maguire, M. E. (1998) Infect. Immun. 66, 3802–3809 act with all six ligands bonded to the cation. It might therefore 23. Snavely, M. D., Miller, C. G., and Maguire, M. E. (1991) J. Biol. Chem. 266, 815– 823 be possible to replace the chlorine with a bacteriocidal moiety 24. Groisman, E. A., Chiao, E., Lipps, C. J., and Heffron, F. (1989) Proc. Natl. coupled via a spacer chain. Such a molecule might be selec- Acad. Sci. U. S. A. 86, 7077–7081 tively directed to the bacterial membrane because of the affin- 25. Soncini, F. C., Garcia, V. E., Solomon, F., and Groisman, E. A. (1996) J. Bacteriol. 178, 5092–5099 ity of cation pentaammine derivatives for the CorA transporter, 26. Vescovi, E. G., Ayala, Y. M., Di Cera, E., and Groisman, E. A. (1997) J. Biol. allowing selective delivery of the bacteriocidal component. Chem. 272, 1440 –1443 Thus, in addition to being useful tools for studying Mg 27. Cowan, J. A. (1993) J. Inorg. Biochem. 49, 171–175 inter- 28. Llewellyn, D. R., O’Connor, C. J., and Odell, A. L. (1964) J. Chem. Soc. 1964, actions, cation hexaammine compounds could potentially be developed as broad spectrum antibiotics based on their speci- 29. Rutenberg, A. C., and Drury, J. S. (1963) Inorg. Chem. 2, 219 30. Meek, D. W., and Ibers, J. A. (1970) Inorg. Chem. 9, 465– 470 ficity for CorA. 31. Hummel, H.-U., and Beiler, F. (1988) Z. Anorg. Allg. Chem. 565, 147–153 32. Iwata, M., and Saito, Y. (1973) Acta Crystallogr. Sect. B Struct. Sci. 29, Acknowledgments—We thank Dr. A. Romani for assistance with the 822– 832 atomic absorption measurements and Dr. J. Cowan for advice and for 33. Messmer, G. G., and Amma, E. L. (1968) Acta Crystallogr. Sect. B Struct. Sci. sharing expertise about cation hexaammines. 24, 417– 422 34. Basolo, F., and Pearson, R. G. (1967) in Mechanisms of Inorganic Reactions: A REFERENCES Study of Metal Complexes in Solution, pp. 158 –170, John Wiley & Sons, 1. Martin, R. B. (1990) Metal Ions Biol. 26, 1–13 Inc., New York 2. Smith, R. L., and Maguire, M. E. (1998) Mol. Microbiol. 28, 217–226 35. Hendry, P., and Ludi, A. (1990) Adv. Inorg. Chem. 35, 117–198

Journal

Journal of Biological Chemistry – Unpaywall

Published: Jun 1, 2000

There are no references for this article.