Uptake of AMP, ADP, and ATP in Escherichia coli W

Transcription

1 Biosci. Biotechnol. Biochem., 75 (1), 7 12, 2011 Uptake of AMP, ADP, and ATP in Escherichia coli W Kimiko WATANABE, y Satsuki TOMIOKA, Kiyoko TANIMURA, Hisae OKU, and Koichiro ISOI School of Pharmacy and Pharmaceutical Sciences, Mukogawa Women s University, Kyubancho, Koshien, Nishinomiya, Hyogo , Japan Received January 26, 2010; Accepted October 9, 2010; Online Publication, January 7, 2011 [doi: /bbb ] The uptake activity ratio for AMP, ADP, and ATP in mutant (T-1) cells of Escherichia coli W, deficient in de novo purine biosynthesis at a point between IMP and 5-aminoimidazole-4-carboxiamide-1--D-ribofuranoside (AICAR), was 1:0.43:0.19. This ratio was approximately equal to the 5 0 -nucleotidase activity ratio in E. coli W cells. The order of inhibitory effect on [2-3 H]ADP uptake by T-1 cells was adenine > adenosine > AMP > ATP. About 2-fold more radioactive purine bases than purine nucleosides were detected in the cytoplasm after 5 min in an experiment with [8-14 C]AMP and T-1 cells. Uptake of [2-3 H]adenosine in T-1 cells was inhibited by inosine, but not in mutant (Ad-3) cells of E. coli W, which lacked adenosine deaminase and adenylosuccinate lyase. These experiments suggest that AMP, ADP, and ATP are converted mainly to adenine and hypoxanthine via adenosine and inosine before uptake into the cytoplasm by E. coli W cells. Key words: purine nucleotide; uptake; 5 0 -nucleotidase; purine-nucleoside phosphorylase; Escherichia coli Nucleotides are probably the most important molecules in living cells. In most microorganisms, purine nucleotides are synthesized via de novo pathways or from external purine compounds by salvage pathways. In Gram-negative bacteria, the outer membrane contains various protein channels formed by two major porins, OmpF and OmpC. These proteins allow for the passive diffusion of small hydrophilic molecules ( Dalton) across the outer membrane. Bocquet-Pages et al. have suggested that uptake of AMP and other nutrients requires porins OmpF and OmpC to cross the outer membrane. 1) Little is known about the uptake of AMP, ADP, and ATP in E. coli. Since the 5 0 -nucleotidase (EC ) of E. coli K-12 is located in the periplasmic space, 2) purine nucleotides are first dephosphorylated to produce the corresponding nucleoside in the perplasmic space. An usha knochout mutant did not grow with AMP as sole carbon source. 3) AMP was dephosphorylated by 5 0 -nucleotidase, and the adenosine formed was converted to adenine cleaved by a membrane-associated purine-nucleoside phosphorylase (EC ) prior to its uptake into the cytoplasm by E. coli K-12 cells. 4) On the other hand, Komatsu has reported three different adenosine uptake systems in isolated membrane vesicles of E. coli K-12: the first was strongly inhibited by adenine nucleosides, pyrimidine nucleosides, and showdomycin; the second was strongly inhibited by adenine nucleosides, guanine nucleosides, and inosine but weakly inhibited by pyrimidine nucleosides and showdomycin; and the third was strongly inhibited by adenine. 5) Many studies have indicated the existence of two nucleoside-transport systems on the cytoplasmic membrane, NupC and NupG. 6 10) NupC is responsible for the transport of all pyrimidine and purine nucleosides (except guanosine and deoxyguanosine), while NupG transports all purine and pyrimidine nucleosides tested to date. 6) It is likely that the first and second systems of Komatsu 5) correspond to NupC and NupG for intact adenosine, and that the third system resembles Hochstadt-Ozer s findings. 11) Hochstadt-Ozer reported that adenosine is taken up into the cytoplasm via cleavage to adenine by membrane purine nucleoside phosphorylase. We have reported that purine nucleoside phosphorylase was located in the periplasm in E. coli W. 12) Therefore, the adenosine formed from AMP, ADP, and ATP can be converted to adenine by purine nucleoside phosphorylase located in the periplasm in E. coli W, or be taken up into the cytoplasm by the NupG or the NupC protein in E. coli W. Rapid conversion of adenosine to inosine, hypoxanthine, and adenine by E. coli leads to difficulties in the accurate quantitative estimation of adenosine uptake. Since we found that the uptake of [8-14 C]adenosine was inhibited by inosine and hypoxanthine in mutant T-1 cells of E. coli W, which were deficient in de novo purine biosynthesis at a point between IMP and AICAR, but was not inhibited by either agent in mutant Ad-3 cells which lacked adenosine deaminase and adenylosuccinate lyase, we tried to clarify the uptake pathways of AMP, ADP, and ATP in E. coli W using mutant Ad-3 in this study. Materials and Methods Materials. [8-14 C]adenine, [8-14 C]adenosine, [2-3 H]adenosine, [U- 14 C]inosine, [8-14 C]hypoxanthine, [8-14 C]AMP, [2-3 H]ADP, [8-14 C]ADP, and [8-14 C]ATP were purchased obtained from Amersham Japan (Tokyo). Bacterial strains and culture conditions. Mutant strains were derived from E. coli W (ATCC9637). Mutations were performed with N-methyl-N 0 -nitro-n-nitrosoguanidine (100 mg/ml) by the method of y To whom correspondence should be addressed. Present address: New Industry Creation Hatchery Center (NICHe), Tohoku University, Aramaki-Aoba, Aoba-ku, Sendai, Miyagi , Japan; Tel: ; Fax: ; kimiko-watanabe@ niche.tohoku.ac.jp Abbreviations: AICAR, 5-aminoimidazole-4-carboxiamide-1--D-ribofuranoside; OM, outer membrane; CM, cytoplasmic membrane

2 8 K. WATANABE et al. Adelberg et al. 13) Ad-3 cells were isolated as a mutant deficient in adenylosuccinate lyase (EC ) and adenosine deaminase (EC ). 12) T-1 cells were isolated as a mutant requiring adenosine or inosine for growth. Mutant 8 9 cells were also isolated from the T-1 cells by selecting colonies requiring a purine base or purine nucleoside, but not AMP after mutational treatment. Cells were grown at 37 Cin minimal Davis-Mingioli Medium 14) with 0.1 mm adenine or 0.1 mm adenosine on a rotary shaker. Growth response of mutant strains. Growth experiments were conducted aerobically at 37 C in test tubes for 8 h. The ability to utilize purines for growth was determined by measuring bacterial growth in 5 ml of Davis-Mingioli medium with 0.1 mm or 1 mm purine bases, nucleosides, or nucleotides. Inoculations were made at cells/ml. Measurement of 5 0 -nucleotidase and 3 0 -nucleotidase activities. The fraction precipitating between 70% and 75% saturation of ammonium sulfate of the periplasmic fraction, prepared by the osmotic shock method as described by Neu and Heppe, 15) was used as the partially purified osmotic fluid of 5 0 -nucleotidase. The reaction mixtures for 5 0 -nucleotidase contained 0.1 M sodium acetate-acetic acid buffer (ph 6.0), 5 mm CoCl 2,16mM CaSO 4,2mM adenosine nucleotides, and cells or 0.5 mg of enzyme in a total volume of 1 ml, as described by Nossal and Heppel. 16) The reaction mixture for nucleotidase contained 1 mm sodium acetate buffer (ph 6.0), 5 mm MgSO 4,1mM CoCl 2,5mM 3 0 -AMP, and cells, as described by Anraku. 17) These mixtures were preincubated at 45 C for 10 min to inactivate the inhibitor for 5 0 -nucloetidase before the addition of nucleotides, and then incubated at 37 C for 5 60 min. After the reaction period, 0.1 ml of 6 N HClO 4 was added to stop the reaction, and the mixture was left in ice water for 10 min. The amount of adenosine formed was measured by HPLC following neutralization of perchloric acid with KOH and filtration to remove KClO 4. Adenosine was separated by reverse-phase paired ion chromatography on a mbondapack C18 column (30-cm 4-mm inside diameter, 10 mm particle size, Waters, Milford, M). The chromatography system consisted of a Waters model 510 solvent equipped with an UV detector at 254 nm. A gradient was run using H 2 O with 5 mm tetrabutylammonium phosphate, ph 7.5 (PIC Reagent A, Waters), solution A, and methanol with 5 mm tetrabutylammouium phosphate, solution B, at a flow rate of 1 ml/min. The gradient included 2 min of 5% solution B, an increase to 40% solution B over 7 min, and then holding for 2 min, and a decrease to 5% solution B over 7 min. The eluted peaks were monitored at 254 nm with a Some S-310A variable wavelength UV detector. The retention time for adenosine was 12.5 min. Assay for uptake. Uptake of purine compounds was measured by conducting membrane filtration assays as described by Kaback, 18) with some modifications. Since the depletion of ATP in cells has a specific stimulatory effect on the transport of purine nucleosides, 19) we used mutant strains that were deficient in de novo purine biosynthesis instead of E. coli W. Cells of the mutant strains were grown exponentially and suspended in 10 mm potassium phosphate buffer (ph 7.2). The reaction mixture (1 ml) contained a suspension of mutant T-1 or mutant Ad-3 cells in 50 mm potassium phosphate buffer (ph 7.2) with 0.4% glucose, 10 mm MgCl 2, and 50 mg of chloramphenicol. The reaction was initiated by the addition of the labeled base, nucleoside, or nucleotide (10 mm [8-14 C]adenine (370 GBq/mol), 10 mm [8-14 C]adenosine (185 GBq/mol), 10 mm [2-3 H]adenosine (92.5 GBq/mol), 10 mm [U- 14 C]adenosine (185 GBq/ mol), 10 mm [8-14 C]hypoxanthine (185 GBq/mol), 10 mm [8-14 C]AMP (185 GBq/mol), 10 mm [8-14 C]ADP (185 GBq/mol) 10 mm [2-3 H]ADP (92.5 GBq/mol), or 10 mm [8-14 C]ATP (185 GBq/mol)). After incubation at 37 C for 5 30 min, measured samples (0.2 ml) were removed directly on a 25 mm diameter cellulose nitrate filter with pores 0.45 mm in diameter lying on a stainless steel mesh support connected to suction. The filters were washed quickly 3 times with 6 ml of 10 mm potassium phosphate buffer (ph 7.2) warmed to 37 C. The washed filters and 10 ml of a dioxane-based scintillation fluid (ALX-2, Daiichi Kagaku, Tokyo) were put into bottles, and the bottles were used for the measurement of radioactivity with a scintillation counter, a Model LSC-700 (Aloka, Tokyo). Radioactivity was converted into moles after correction for counting efficiency. Analysis of radioactive purine compounds from [8-14 C]AMP in the cells. A reaction mixture (10 ml) containing a suspension of T-1 or Ad-3 cells in 10 mm potassium phosphate buffer (ph 7.2) with 0.8% glucose, 10 mm MgCl 2, and 50 mg of chloramphenicol was used as indicated in the assay for uptake. The reaction mixtures were preincubated for 5 min at 37 C. The reaction was initiated by the addition of 10 mm [8-14 C]AMP (185 GBq/mol). After incubation at 37 C for 5 min, the reaction mixtures (5 ml) were filtrated to collect the cells on cellulose nitrate filters and were washed quickly 3 times with 6 ml of 10 mm potassium phosphate buffer (ph 7.2) warmed to 37 C. The filters obtained were extracted with 1.5 ml of 3 N HClO 4 containing adenine, adenosine, AMP, ADP, ATP, inosine, and hypoxanthine (1 mm) as a carrier, for 30 min at room temperature, and were neutralized with 3 N KOH and then centrifuged to remove KClO 4 and bacterial cells, giving a final volume of 2 ml. A 40-mL sample of the solution was injected onto the HPLC column and eluted under the following conditions: Products of purine synthesis were analyzed by HPLC following neutralization of perchloric acid with KOH and filtration to remove KClO 4. Purines were separated by reverse-phase paired ion chromatography, as described for the measurement of 5 0 -nucleotidase activity. The retention times for hypoxanthine, inosine, adenine, adenosine, IMP, AMP, ADP, and ATP were 6.1, 10.3, 11.5, 12.5, 14.0, 15.5, 17.5, and 18.8 min respectively. Fractions were collected every 0.5 min, and then radioactivity was measured by liquid scintillation count using a dioxane-based scintillation cocktail. Results Growth response of mutant strains to purine compounds E. coli W cells and mutant T-1, 8 9 cells grew fully in Davis-Mingioli medium supplemented with 0.1 mm adenine, adenosine, inosine, hypoxanthine, guanine, or guanosine (Table 1). Ad-3 cells grew in the medium supplemented with adenosine, adenine, or AMP, but not inosine, hypoxanthine, guanine, or guanosine as purine source (Table 1). Although the growth of the T-1 and Ad-3 cells in the medium with 0.1 mm and 1 mm AMP reached 60 and 80% of that in the medium with 0.1 mm adenosine or adenine respectively, these cells did not grow in the medium with 1 mm ADP or 1 mm ATP (Fig. 1A, B). A mutant (8 9) isolated from T-1 cells was unable to utilize AMP for growth as sole source of a purine compound (Table 1). Table 1. Growth Response of E. coli W and T-1, 8-9, and Ad-3 Cells to Purine Compounds Growth Purine compound E. coli W T Ad-3 None þþþ 0.01 mm Adenine þþþ þ þ þ 0.1 mm Adenine þþþ þþþ þþþ þþþ 0.1 mm Adenosine þþþ þþþ þþþ þþþ 0.1 mm Hypoxanthine þþþ þþþ þþþ 0.1 mm Inosine þþþ þþþ þþþ 0.1 mm Guanine þþþ þþþ þþþ 0.1 mm Guanosine þþþ þþþ þþþ 0.1 mm AMP þþþ þ þ 1mM AMP þþþ þþ þþ 0.1 mm ADP þþþ 1mM ADP þþþ 0.1 mm ATP þþþ 1mM ATP þþþ þþþ, full growth; þþ, considerably visible growth; þ, barely visible growth;, no growth Each strain was grown in Davis-Mingioli medium with purine compound for 8 h and growth was measured.

3 Uptake of AMP, ADP, and ATP in E. coli 9 A 1.2 B 1.2 Optical density of medium (260 nm) Optical dencity of medium (260 nm) Time (h) Time (h) Fig. 1. Growth of Mutant T-1 (A) and Ad-3 (B) Cells on AMP, ADP, ATP, and Adenosine. Cells were incubated at 37 C in Davis-Mingioli medium supplemented with 0.1 mm adenosine ( ), 0.1 mm AMP ( ), 1 mm AMP ( ), 1 mm ADP ( ), or 1 mm ATP ( ). Table 2. Specific Activities of 5 0 -Nucleotidase for AMP in Intact Cells and Partially Purified Periplasmic Fluid (enzyme) of E. coli W, T-1, and 8 9 Cells Strain Intact cells 5 0 -Nucleotidase 3 0 -Nucleotidase (n moles/min/ cells) Enzyme 5 0 -Nucleotidase (mmoles/min/mg of protein) E. coli W 87 2:5 16 0:1 1:7 0:08 T :2 15 2:0 1:4 0: :7 0:1 20 1:6 ND ND, not detected Table 3. Specific Activities of 5 0 -Nucleotidase for AMP, ADP, and ATP in Intact Cells and Partially Purified Periplasmic Fluid of E. coli W Nucleotide Substrate specificity of 5 0 -nucleotidase Intact cells (mmoles/min/2: cells) Enzyme (mmoles/min/mg of protein) AMP 0:18 0:02 (1) 1:43 0:04 (1) ADP 0:073 0:003 (0.41) 0:62 0:03 (0.43) ATP 0:034 0:002 (0.19) 0:30 0:01 (0.21), The values in parentheses are related to AMP, which was set at 1. Data are the means for at least three independent experiments Nucletotidase and 3 0 -nucleotidase in E. coli W, mutant T-1, and 8 9 cells Although the 3 0 -nucleotidase activity levels of intact cells of mutant strains T-1 and 8 9 were almost equal to those of E. coli W cells, 5 0 -nucleotidase activity levels of intact 8 9 cells were about 2% of those of E. coli W and T-1 cells (Table 2). The 8 9 cells were deficient in 5 0 -nucleotidase. The specific activity ratio of 5 0 -nucleotidase for AMP, ADP, and ATP in both intact cells and partially purified osmotic shocked fluid of E. coli W was 1:0.4:0.2 (Table 3). The specific activity of 5 0 -nucleotidase for ADP and for ATP in these cells was lower than that for AMP in E. coli W (Table 3). Uptake of [8-14 C]AMP, [8-14 C]ADP and [8-14 C]ATP Uptake of punine nucleotides, purine nucleosides, and purine bases required the presence of 0.4% glucose, and was inhibited by the addition of 5 mm 2,4-dinitrophenol (data not shown). Table 4. Uptake of [8-14 C]AMP, [8-14 C]ADP, and [8-14 C]ATP by E. coli W, T-1, and 8 9 Cells Uptake Nucleotide T E. coli W (n moles /cells) AMP 0:28 0:04 (1) ND 0:24 0:01 (1) ADP 0:12 0:02 (0.43) ND 0:078 0:002 (0.33) ATP 0:054 0:006 (0.19) ND 0:031 0:002 (0.13), The values in parentheses are related to AMP, which was set at mm [8-14 C]AMP, 10 mm [8-14 C]ADP or 10 mm [8-14 C]ATP and cells were used in reaction mixtures (1 ml) with 0.4% glucose, 10 mm MgCl 2, and 50 mg of chloramphenicol. After 5 min, 0.2 ml of sample was used for measurements. Data are the means for at least three independent experiments. The uptake activity ratios for [8-14 C]AMP, [8-14 C]ADP, and [8-14 C]ATP by E. coli W and the mutant T-1 cells during the 5-min reaction were 1:0.33:0.13 and 1:0.43:0.19 respectively (Table 4). The quantities of AMP, ADP, and ATP taken up by the cells were greater for the mutant T-1 than for E. coli W. As Munch- Petersen and Pihl have reported, 19) it appears that the depletion of ATP in cells has a specific stimulatory effect on the transport of purine nucleosides. The uptake activity ratio for AMP, ADP, and ATP in mutant T-1 was the same as the 5 0 -nucletidase-specific activity ratio in the intact cells and the partially purified periplasmic fluid of E. coli W (Table 3). These results indicate that uptake of AMP, ADP, or ATP into the cells was dependent on the specific activity of 5 0 -nucleotidase. Therefore, it is thought that 8 9 cells defective in nucleotidase can not take up any AMP, ADP or ATP. This indicates that the concentration of adenosine formed from AMP by 5 0 -nucleotidase is enough for the growth of T-1 cells, while that formed from ADP or from ATP is too low to support growth. Inhibitory effects of purine compounds on the uptake of [2-3 H]ADP Adenine, adenosine, AMP, and ATP at 100 mm inhibited the uptake of 10 mm [2-3 H]ADP by 85, 79, 66 and 52% respectively after a 5-min reaction in mutant T-1 cells (Table 5). The order of the inhibitory effect on [2-3 H]ADP uptake by T-1 cells was adenine > adenosine > AMP > ATP. The inhibitory effect of adenine was greater than that of AMP or adenosine.

4 10 K. WATANABE et al. Table 5. Inhibitory Effects of Purines on the Uptake of [2-3 H]ADP Table 6. Distribution of Radioactivity from [8-14 C]AMP in the by Mutant T-1 Cells Mutant T-1 and Ad-3 Cells Purines added [2-3 H]ADP uptake Inhibitory effect (n moles/ cells) (%) None 2:9 0:2 0 Adenine 0:43 0:04 85 Adenosine 0:60 0:07 79 AMP 1:0 0:07 66 ATP 1:4 0: mm [2-3 H]ADP (185 Bq/mol) and cells were used in reaction mixtures (1 ml) in the presence of a 10-fold molar excess of adenine, adenosine, AMP or ATP by the mutant T-1 cells. After 15 min, 0.2 ml of each sample was measured. Data are the means for at least three independent experiments. If the labeled adenosine converted from [2-3 H]ADP by 5 0 -nucleotidase was transported to the cytoplasm via inosine converted by adenosine deaminase without conversion to adenine and hypoxanthine, the inhibitory effect of adenosine and of AMP on [2-3 H]ADP uptake might have been greater than that of adenine, but adenine had the greatest inhibitory effect of any of the purine compounds. Therefore, it is likely that purine nucleotides were converted to purine bases via purine nucelosides and then transported into the cytoplasm across the cytoplasmic membrane. Analysis of radioactive purine compounds in the cells after incubation with [8-14 C]AMP About 80% of the radioactivity taken up into the cytoplasm was detected as purine nucleotides after 5 min in the experiments with mutant T-1 and Ad-3 cells and [8-14 C]AMP (Table 6). About 2-fold more radioactive purine bases than purine nucleosides were detected in the cytoplasm after 5 min in the experiments with [8-14 C]AMP and T-1 and Ad-3 cells (Table 6). These results indicate that adenosine, which was formed from AMP, ADP, or ATP, to be converted not only to hypoxanthine by adenosine deaminase and then purine nucleoside via inosine, but also to adenine by purine nucleoside phosphorylase in the periplasm. It is generally believed that exogenous adenine, hypoxanthine, and guanine are converted to their corresponding nucleosides and then further phosphorylated to form AMP, IMP, and GMP respectively, as the main as well as the de novo pathway replenishing the purine nucleotide pool. Therefore, the adenosine and inosine detected might be formed from adenine and hypoxanthine respectively in the cytoplasm in T-1 cells. GMP, GDP, GTP, guanosine, and guanine were not detected, because we did not add these cold compounds as a carrier. Hence, less radioactivity was detected in the mutant T-1 cells than in the Ad-3 cells. Because the interconversion of purine nucleotides taken up into the cytoplasm depends on the level of the corresponding purine compounds, IMP was not detected in the Ad-3 cells (Table 6). The proposed uptake pathway, salvage pathways for purine compounds in E. coli W, and defective steps in mutant strains T-1, 8 9, and Ad-3 are shown in Fig. 2. Red and enclosed characters indicate radioactive compounds detected in the cytoplasm after 5 min in the experiments with [8-14 C]AMP and Ad-3 and T-1 cells respectively. Radioactivity cpm (%) Purines detected Mutant T-1 Mutant Ad-3 AMP (7.5) (7.6) ADP 2; (24.0) 1; (12.7) ATP 4; (47.2) 5; (59.3) Adenosine (1.8) (5.2) Adenine (3.6) 1; (12.4) IMP (3.3) 0 (0) Inosine 54 3 (0.06) 0 (0) Hypoxanthine (1.1) 0 (0) Total detected 8; (89.1) 8; (97.2) Counts injected 9; (100) 9; (100) onto HPLC %, percentage detected on the HPLC column The reaction was initiated by addition of 10 mm [8-14 C]AMP. After 5 min at 37 C, the reaction mixtures (5 ml) were filtrated, extracted and measured as described in Materials and Methods. Data are the means for at least three independent experiments. Norholm and Dandanell reported that NupC transported inosine, but its affinity for inosine was too low to support growth. 20) They also reported that NupG did not appear to support growth on adenosine as sole source of carbon even though adenosine was effectively taken up by NupG in an assay with purine nucleosides, and suggested that a high concentration of adenosine might inhibit growth. As Hill and Pittillo reported, 21) adenosine and the other compounds might block the de novo synthesis of purines and thiamine by means of a feedback mechanism (Fig. 2). Effects of heterologous purine compounds on the uptake of various purine compounds Unlabeled purine compounds at mm were used in the uptake of 10 mm labeled purine compounds. Uptake of [2-3 H]adenosine and of [U- 14 C]inosine was mutually and competitively inhibited in the mutant T-1 cells, but that of [2-3 H]adenosine was not inhibited by inosine in the mutant Ad-3 cells (Table 7a). Uptake of [2-3 H]adenosine and of [8-14 C]hypoxanthine was mutually and noncompetitively inhibited in the T-1 cells, but that of [2-3 H]adenosine was not inhibited by hypoxanthine in the Ad-3 cells (Table 7d). Uptake of [2-3 H]adenosine was inhibited by guanosine and guanine in the T-1 cells, but not in the Ad-3 cells (data not shown). Inhibition of adenosine uptake not only by inosine but also by hypoxanthine differed between the T-1 and the Ad-3 cells. Uptake of [8-14 C]adenine was not inhibited by hypoxanthine or inosine in either type of mutant cells, but that of [8-14 C]hypoxanthine was competitively inhibited by adenine in both. Uptake of [U- 14 C]inosine was noncompetitively inhibited by adenine in both types of mutant cells (Table 7b, c). Uptake of [8-14 C]adenine was not inhibited by hypoxanthine or inosine in either type of mutant cells, but that of [8-14 C]hypoxanthine was competitively inhibited by adenine in both types. Uptake of [8-14 C]adenine and [2-3 H]adenosine and of [U- 14 C]inosine and [8-14 C]hypoxanthine was mutually and competitively inhibited in both mutant cell lines (Table 7e, f). The apparent K m values for the uptake of adenosine, inosine, adenine, and hypoxanthine were 9.0, 18, 6.1,

5 Uptake of AMP, ADP, and ATP in E. coli 11 (O M) (Periplasm) (C M) (Cytoplasm) AMP ADP ATP AMP ADP ATP 1 Adenosine (Ad-3) 2 3 Inosine (8-9) Adenine 3 3 Adenosine 2 3 Adenine Inosine ATP ADP 7 6 IMP XMP GMP 8 (T-1) (8-9) L-Histidine AICAR 9 (Ad-3) Feed back inhibition AMP (Ad-3) Adenylosuccinate Succino- AICA Thiamine metabolisum Hypoxanthine Hypoxanthine GDP GTP de novo Fig. 2. Proposed Uptake Pathway, Salvage Pathways for AMP, ADP and ATP in E. coli W and Defective Steps in Mutant Strains T-1, 8 9, and Ad-3. Red and enclosed characters indicate radioactive compounds detected in the cytoplasm after 5 min in the experiments with [8-14 C]AMP and Ad-3 and T-1 cells respectively. Reactions 1, 2, and 3 involve 5 0 -nucleotidase, adenosine deaminase, and purine-nucleoside phosphorylase respectively. Reactions 4 and 7 involve adenine phosphoribosyltransferase and hypoxanthine phosphoribosyltransferase respectively. Reaction 5 involves adenosine kinase or phosphotransferase. Reaction 6 involves inosine kinase or phosphotransferase. Reactions 8 and 9 involve AMP deaminase and adenylosuccinate lyase respectively. The site of mutation is represented by oblique lines. Table 7. Inhibitory Effects of Heterologous Purine Compounds on the Uptake of Purine Compounds in the Mutant T-1 and Ad-3 Cells Uptake substrate Inhibitory Type of inhibition substrate T-1 Ad-3 a b c d e f [2-3 H]Adenosine Inosine Competitive [U- 14 C]Inosine Adenosine Competitive Competitive [8-14 C]Adenine Hypoxanthine [8-14 C]Hypoxanthine Adenine Competitive Competitive [8-14 C]Adenine Inosine [U- 14 C]Inosine Adenine Noncompetitive Noncompetitive [2-3 H]Adenosine Hypoxanthine Noncompetitive [8-14 C]Hypoxanthine Adenosine Noncompetitive Noncompetitive [2-3 H]Adenosine Adenine Competitive Competitive [8-14 C]Adenine Adenosine Competitive Competitive [U- 14 C]Inosine Hypoxanthine Competitive Competitive [8-14 C]Hypoxanthine Inosine Competitive Competitive, no inhibition The type of inhibition was determined from reciprocal plots. and 9.6 mm respectively in the T-1 cells. Although the K m value for the uptake of hypoxanthine was only about 1.6-fold larger than that of adenine, uptake of hypoxanthine was inhibited by adenine, and that of adenine was not inhibited by hypoxanthine in the T-1 and Ad-3 cells (Table 7). Discussion T-1 cells, deficient in de novo purine biosynthesis, utilize 1 mm AMP for growth, but not ADP or ATP (Table 1). Because the uptake of ADP and ATP by the T-1 cells was 43% and 19% that of AMP respectively (Table 4), it appears that the concentration of adenosine formed from AMP by 5 0 -nucleotidase was enough for the growth of T-1 cells, while that formed from ADP and from ATP was too low to support growth. Patching et al. have reported that uptake of [U- 14 C]uridine was inhibited by adenosine and inosine in E. coli BL21(DE3), containing the plasmid vector expressing NupG, but not by adenine. 10) Our results show that adenine had the greatest inhibitory effect on [2-3 H]ADP s uptake of any of the purine compounds in the T-1 cells (Table 5). Furthermore, uptake of [2-3 H]adenosine was inhibited by inosine in the T-1 cells, but was not inhibited in the Ad-3 cells (Table 7a). The inhibition of [2-3 H]adenosine uptake by inosine can be explained by the dilution of the radioactivity of [2-3 H]adenosine in T-1 cells, because adenosine was converted to inosine by the adenosine deaminase

6 12 K. WATANABE et al. activity in T-1 cells. Furthermore, the inhibition of adenosine in the uptake of [8-14 C]adenine can be explained by the dilution of the radioactivity of adenine in T-1 and Ad-3 cells, because adenosine may be converted to adenine by the purine nucleoside phosphorlase activity prior to its uptake into the cytoplasm. The uptake of [8-14 C]hypoxanthine or [U- 14 C]inosine was inhibited by adenine in T-1 and Ad-3 cells, but the uptake of [8-14 C]adenine was not inhibited by hypoxanthine or inosine (Table 7b, c). Hypoxanthine did not inhibited the uptake of [8-14 C]adenine, but adenine inhibited the uptake of [8-14 C]hypoxanthine in T-1 and Ad-3 cells (Table 7b). These results suggest that adenosine is taken up by the NupG protein into the cytoplasm, and that it is converted to adenine or hypoxanthine via inosine before uptake of it into the cytoplasm by E. coli W cells. These inhibition patterns may be explained by the existence of two specific transport proteins for purine bases formed from purine nucleosides spanning the cytoplasmic membrane, one responsible for uptake of adenine and the other for uptake of adenine and hypoxanthine. Roy-Burman and Visser have reported that uptake of hypoxanthine and guanine was mutually and competitively inhibitory, but was not inhibited by adenine in E. coli B. 22) We have no explanation for this difference between E. coli W and B. Hochstadt-Ozer and Stadtman reported that adenine is taken up by a translocation process involving adenine phosphoribosyltransferase, 23 25) but this is not supported by other evidence. 26,27) Burton found that E. coli PurP adenine permease was responsible for high-affinity adenine uptake, and there was only 50% inhibition of the adenine uptake system in the presence of a 250-fold excess of hypoxanthine. 26) Andersen et al. found that UraA protein (the uracil permease) was necessary for the uptake of uracil at low concentrations in E. coli K ) Kratza and Frillingos reported that YgfO and YicE functioned as specific, high-affinity transporters for xanthine in E. coli K-12, but displayed no detectable transport of uracil and hypoxanthine. 29) The nucleobase/ascorbate transporter (NAT) family includes UraA, PurP, YgfO, and YicE proteins. Our results suggest that the adenosine formed from AMP, ADP, and ATP is taken up into the cytoplasm via the conversion to adenine and hypoxanthine in E. coli W. Adenosine deaminase is induced by adenine and hypoxanthine, while adenosine and inosine induce the synthesis of purine nucleoside phosphorylase in E. coli. 30) Therefore, the adenosine formed from AMP, ADP, and ATP might be rapidly converted to adenine and hypoxanthine via adenosine and inosine by adenine deaminase and purine nucleoside phosphorylase in the periplasm and then taken up into the cytoplasm by the purine base-uptake systems in E. coli W cells. This adenosine uptake system might be the third system reported by Komatsu. 5) Further studies with a nupc or nupg mutant are needed to confirm the uptake model of purine nucleosides across the cytoplasmic membrane developed in this study. Acknowledgments We thank Dr. Yuzo Yoshida (Mukogawa Women s University), Dr. Tadayuki Imanaka (Ritsumeikan University), Dr. Youhei Yamagata (Tokyo University of Agriculture and Technology) and Dr. Masafumi Goto (Tohoku University) for helpful discussion. References 1) Bocquet-Pages C, Lazdunski C, and Lazdunski A, Eur. J. Biochem., 118, (1981). 2) Neu HC, J. Biol. Chem., 242, (1969). 3) Kakehi M, Usuda Y, Tabira Y, and Sugimoto S, J. Mol. Microbiol. Biotechnol., 13, (2007). 4) Yagil E and Beacham IR, J. Bacteriol., 121, (1975). 5) Komatsu Y, Biochim. Biophys. Acta, 330, (1973). 6) Mygind B and Munch-Petersen A, Eur. J. Biochem., 59, (1975). 7) Zhang Y, Craig JE, and Gallagher MP, J. Bactriol., 174, (1992). 8) Craig JE, Zhang Y, and Gallagher MP, Mol. Microbiol., 11, (1994). 9) Xie H, Patching SG, Gallagher MP, Litherland GJ, Brough AR, Venter H, Yao SYM, Ng AML, Young JD, Herbert RB, Henderson PJF, and Baldwin SA, Mol. Membr. Biol., 21, (2004). 10) Patching SG, Baldwin SA, Baldwin AD, Young JD, Gallagher MP, Henderson PJ, and Herbert RB, Org. Biomol. Chem., 7, (2005). 11) Hochstadt-Ozer J, J. Biol. Chem., 247, (1972). 12) Watanabe K, Fukumoto H, and Isoi K, Biochem. Biophys. Res. Commun., 134, (1986). 13) Adelberg EA, Mandel M, and Ching Chen GC, Biochem. Biophys. Res. Commun., 18, (1965). 14) Davis BD and Mingioli ES, J. Bacteriol., 60, (1950). 15) Neu HC and Heppel LA, J. Biol. Chem., 240, (1965). 16) Nossal NG and Heppel LA, J. Biol. Chem., 241, (1966). 17) Anraku Y, J. Biol. Chem., 239, (1964). 18) Kaback HR, J. Biol. Chem., 243, (1968). 19) Munch-Petersen A and Pihl NJ, Proc. Natl. Acad. Sci. USA, 77, (1980). 20) Norholm MHH and Dandanell G, J. Bacteriol., 183, (2001). 21) Hill DL and Pittillo RF, Antimicrob. Agents Chemother., 4, (1973). 22) Roy-Burman S and Visser DW, J. Biol. Chem., 250, (1975). 23) Hochstadt-Ozer J and Stadtman ER, J. Biol. Chem., 246, (1971). 24) Hochstadt-Ozer J and Stadtman ER, J. Biol. Chem., 246, (1971). 25) Hochstadt-Ozer J and Stadtman ER, J. Biol. Chem., 246, (1971). 26) Burton K, Biochem. J., 168, (1977). 27) Burton K, J. Gen. Microbiol., 129, (1983). 28) Andersen PS, Frees D, Fast R, and Mygind B, J. Bacteriol., 177, (1995). 29) Karatza P and Frillingos S, Mol. Membr. Biol., 22, (2005). 30) Neuhard J and Nygaard P, Escherichia coli and Salmonella typhimuriu Cellular and Molecular Biology Vol. 1, eds. Ingraham JL, Low KB, Neidhardt FC, Magasanik B, Schaechter M, and Umbarger HE, American Society for Microbiology, Washington DC, pp (1987).