Dihydrofolate Reductase from Neisseria sp

ANTIMICROBIAL AGENTS

AND

CHEMOTHERAPY, Mar. 1979, p. 428-435

Vol. 15, No. 3

0066-4804/79/03-0428/08$02.00/0

Dihydrofolate Reductase from Neisseria sp. DEVRON R. AVERETT, 12* BARBARA ROTH,' JAMES J. BURCHALL,' AND DAVID P. BACCANARI' Wellcome Research Laboratories, Research Triangle Park, North Carolina 27709,' and Department of Bacteriology and Immunology, School of Medicine, University of North Carolina, Chapel Hill, North Carolina 275142 Received for publication 22 December 1978

gonorrhoeae. The genus Neisseria includes several species that are known to be pathogenic for man. N. gonorrhoeae and N. meningitidis are reported to the Center for Disease Control as agents of disease in approximately 1,000,000 and 1,400 cases per year, respectively (1), while Branhamella catarrhalis (fonnerly N. catarrhalis) has been linked to upper respiratory infections in compromised hosts (22, 23). The emergence of penicillin-resistant N. gonorrhoeae (24, 29) indicates that the range of therapeutic agents active against these organisms may be narrowed. Trimethoprim is an antibacterial agent of proven efficacy against many gram-negative or-

ganisms, and in combination with sulfamethoxazole it has been used to successfully treat gonorrhoea in a multidose regimen. However, the gonococcus is relatively nonsusceptible to trimethoprim alone. In Escherichia coli, trimethoprim inhibits the enzyme dihydrofolate reductase (5,6,7,8-tetrahydrofolate:oxidized nicotinamide adenine dinucleotide phosphate oxidoreductase; EC 1.5.1.3), resulting in an eventual cessation of thymidine, purine, and certain amino acid biosynthesis due to depletion of the intracellular tetrahydrofolate pool (6). E. coli growing in rich media is able to bypass trimethoprim inhibition 428

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

Members of the genus Neisseria are relatively nonsusceptible to trimethoprim, an inhibitor of dihydrofolate reductase. For example, the minimal inhibitory concentration (MIC) of trimethoprim for N. gonorrhoeae ranges from 2 to 70 ,ug/ ml, whereas the MIC for Escherichia coli is 0.2 ,ig/ml or less. In an effort to understand this difference, dihydrofolate reductase was partially purified from five Neisseria species and compared with the enzyme from E. coli. N. gonorrhoeae dihydrofolate reductase was similar to that from E. coli in molecular weight (18,000) and affinity for the substrates reduced nicotinamide adenine dinucleotide phosphate and dihydrofolate (Km = 13 and 8 ,uM, respectively). However, the gonococcal enzyme had a decreased affinity for trimethoprim, with an apparent K, of 45 x 10-9 M, some 30-fold greater than the E. coli value of 1.2 x 109 M. These enzymes also differed in their isoelectric points and pH activity profiles. Within the genus Neisseria, the dihydrofolate reductase isolated from N. meningitidis and N. lactamica resembled the N. gonorrhoeae enzyme, and only small differences were detected for the N. flavescens and Branhamella catarrhalis dihydrofolate reductases. These data indicate that the relatively poor affinity of trimethoprim for the dihydrofolate reductase from these organisms may be largely responsible for the relative nonsusceptibility of Neisseria sp. to trimethoprim. The contribution of other resistance mechanisms to the overall nonsusceptibility was assessed. Strains of N. gonorrhoeae with altered cell envelope permeability had MIC values less than twofold different from those of isogenic wild-type strains. Also, a direct relationship was observed between the affinity of trimethoprim analogs for gonococcal dihydrofolate reductase and the MIC of these compounds for the gonococcus. These observations suggest that the cell envelope of N. gonorrhoeae is not impermeable to trimethoprim. Changes in the amount of dihydrofolate reductase activity could cause alterations in the susceptibility of the gonococcus to trimethoprim, as demonstrated with N. gonorrhoeae strains selected for trimethoprim resistance after chemical mutagenesis. However, the level of dihydrofolate reductase activity in wild-type N. gonorrhoeae was similar to that of E. coli, indicating that the difference in the susceptibility of these organisms is not due to greater amounts of enzyme in N.

VOL. 15, 1979

NEISSERIA DIHYDROFOLATE REDUCTASE

MATERIALS AND METHODS Dihydrofolate was prepared by the method of Futterman (11) as modified by Blakely (5) and stored as a suspension in 5 mM HC1 at -70°C. Methotrexate was obtained from ICN Pharmaceuticals, and all other inhibitors were synthesized in the Organic Chemistry Department, Wellcome Research Laboratories. Dithiothreitol (DTT) was from Calbiochem, Ultrogel AcA54 was purchased from LKB, and reduced nicotinamide adenine dinucleotide phosphate (NADPH) was supplied by Sigma. Thiamine pyrophosphate was provided by P-L Biochemicals. N-Methyl-N'-nitro-N-nitrosoguanidine (NTG) and leucovorin were obtained from Aldrich. GC medium base, tryptic soy broth, and proteose peptone were obtained from Difco. Other chemicals were reagent grade. Bacterial strains. Bacterial strains used in this work and their sources are listed in Table 1. Cells were grown in an 8% C02 atmosphere on plates of GC medium base supplemented with Kellogg defined supplements (17). Broth medium (GCB broth) was of similar composition, but lacked agar and was supplemented with 9 mM NaHCO3. For storage, cells were suspended in trypic soy broth (3 g/100 ml) with added glycerol (25%, vol/vol), quick frozen, and kept at -70°C. Bacterial growth in broth. Inocula were prepared by washing cells from plates incubated for 18 to 24 h at 370C in 8% C02. Samples (400 ml) of prewarmed GCB broth in 1-liter flasks were inoculated to initial optical densities at 600 nm (OD6o) of at least 0.02 absorbance units. These were incubated at 370C under 8% C02 in a New Brunswick G-25 incubator at 150 rpm. Growth was monitored by following the change in absorbance at 600 nm. At the end of the logarithmic

phase of growth, cells were harvested by centrifugation at 15,000 x g for 10 min. Lysis. Cell pellets were frozen at -15°C overnight and thawed upon resuspension in 0.25% of the culture volume of 0.15 M NaCl-0.10 M potassium phosphate, pH 7.6-1 mM ethylenediaminetetraacetic acid-1 mM DTT (PBS-EDTA-DTT). The suspended cells were incubated at 00C for 5 min and centrifuged at 15,000 x g for 10 min. Alternatively, the unfrozen cell pellet was resuspended in PBS-EDTA-DTT and lysed by passage through a Aminco French pressure cell at 10,000 lb/in.2 The lysate was centrifuged at 15,000 x g. In either case, the supernatant fraction constituted the crude lysate. Enzyme assays. The standard assay for dihydrofolate reductase activity was performed in 0.1 M imidazolium chloride buffer, pH 6.4, with 60,uM NADPH and 40MM dihydrofolate. Mercaptoethanol was maintained at 12 mM. The final volume of the reaction mixture was 1.0 ml. Buffer and NADPH were prewarmed by a 2-min preincubation in a 300C water bath; after transfer to a Gilford model 240 spectrophotometer thermostated at 30°C, the cell extract was added, and the reaction was initiated by the addition of dihydrofolate. The decrease in absorbance at 340 nm was converted to units of activity as previously described (3). Kinetic determinations were performed under similar conditions, except that either the dihydrofolate or the NADPH concentration was varied as required and the varied substrate was added last. Enzyme preparation. Freeze-thaw lysates (1 to 5 ml) were loaded onto an Ultrogel AcA 54 gel permeation chromatography colunn (1.5 by 60 cm) equilibrated with 0.05 M KCl, 0.1 M potassium phosphate, pH 6.8, 1 mM EDTA, and 1 mM DTT at 5°C. After elution at 5 ml/h, the enzymatic activity and absorbance at 280 nm of each fraction were determined, and fractions containing dihydrofolate reductase activity were pooled and stored in aliquots at -15°C. In, determination. The concentration of inhibitor necessary to inhibit dihydrofolate reductase activity by 50% (h5o) was determined by a modification of the standard assay. Inhibitors were prepared to a known concentration in assay buffer, and varying amounts were added in place of an equal volume of assay buffer. After preincubation of buffer with inhibitor, enzyme and dihydrofolate were added, and the reaction was initiated with NADPH. Plots of the percentage of inhibition versus the logarithm of inhibitor concentration in the reaction mixture were used to estimate the I50.

Mutagenesis. N. gonorrhoeae strain F62 was mutagenized with NTG in 0.15 M NaCl-10 mM potassium phosphate, pH 6.0-4 mM MgCl2-0.5 mM spermine (M-buffer). These conditions have been shown to result in minimal autolysis of gonococci (9, 28). Cells were washed from plates of GCB into M-buffer to a final OD600 of 2.0. NTG was added to a final concentration of 9 ug/ml. Cells were incubated for 10 min at

370C,

diluted fivefold with M-buffer to reduce intra-

cellular NTG levels, and centrifuged at 15,000 x g for 10 min. This procedure resulted in a 50% decrease in viable cell counts compared with controls. CelLs were resuspended in GCB broth, and growth was allowed to proceed for two generations before plating on GCB agar with added trimethoprim.

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

by uptake of the required end products of biosynthesis (19). N. meningitidis lacks the enzymes required for utilization of exogenously supplied thymine and thymidine (16), suggesting that the nonsusceptibility of the Neisseria to trimethoprim is not due to the uptake of these tetrahydrofolate-sparing products. The present work was undertaken to determine whether the enzyme dihydrofolate reductase plays some role in the nonsusceptibility of these organisms to trimethoprim. Dihydrofolate reductase was shown to be the site of action of trimethoprim in the gonococcus. The enzyme was isolated free of contaminating activities, and several of its properties were determined. Strains of N. gonorrhoeae resistant to penicillin and other drugs have been examined to determine the extent of cross-resistance to trimethoprim. The similarity of dihydrofolate reductase from other Neisseria to that isolated from the gonococcus was assessed. The evidence presented in this work is consistent with the hypothesis that organisms in the genus Neisseria are refractory to inhibition by trimethoprim largely as a result of relatively weak binding of the drug to the dihydrofolate reductase found in these species.

429

430

ANTIMICROB. AGENTS CHEMOTHER.

AVERETT ET AL.

TABLE 1. Bacterial strains used in this work, their sources, their trimethoprim susceptibility, and the amount and susceptibility of dihydrofolate reductase from each Organism

Clinical isolate Clinical isolate Clinical isolate Clinical isolate Clinical isolate Clinical isolate Clinical isolate Clinical isolate Clinical isolate Clinical isolate

FA48

penA2 penB2 mtr

FA102

penA2

FA136

penA2 mtr

FA140

penA2 penB2 mtr

FA171

mtr

BR54

env penA2 penB2 mtr

BR84

env penA2 penB2 mtr

BR87

env penA2 penB2 mtr

A0078

Trimethoprim resistant

A0105

Trimethoprim resistant

AO110

Trimethoprim resistant

66 67 IPL

ppNGb ppNG

N. meningitidis

ppNG Group A, sulfa susceptible

Source and

MIC

reference (ad(p/mi) g/ml) reference

P. F. Sparling P. F. Sparling P. F. Sparling K. K. Holmes K. K. Holmes K. K. Holmes L. P. ElwelU L. P. Elwell L. P. Elwell P. F. Sparling; 27 P. F. Sparling; 27 P. F. Sparling; 27 P. F. Sparling; 27 P. F. Sparling; 27 P. F. Sparling; 27 P. F. Sparling; 26 P. F. Sparling; 26 P. F. Sparling; 26 NTG mutagenesis from F62 NTG mutagenesis from

F62 NTG mutagenesis from F62 L. P. Elweil; 10 L. P. Elwell; 10 L. P. Elwell; 10 L. P. Elwell; CDC 1894 ATCC 23970 ATCC 13120 ATCC 25238-1

Measured Sp act' anM (U/mg) (nM)

70

0.100 0.049 0.078 0.070 0.054 0.068 0.059 0.046 0.024 0.063

310 480 550 290 370 550 290 300 370 350

70

0.053

340

70

0.053

460

110

0.060

435

84

0.043

360

70

0.035

460

48

0.060

450

35

0.052

350

48

0.040

370

400

0.390

310

250

0.135

480

720

0.940

340

36 54 14 15

0.036 0.113

490 440

60 35 70 35 35 35 2 20

0.069

14 0.120 N. lactamica 50 0.039 N. flavescens 34 0.008c B. catarrhalis Avgd 46 ± 26 0.059 ± 0.024 398 ± 81 a Obtained at pH 6.4 in 0.10 M imidazolium chloride buffer, using freeze-thaw lysates of late-logarithmicphase cells. The specific activity of F62 lysed in a French pressure cell was 0.022 U/mg. ppNG, Penicillinase-producing N. gonorrhoeae. French pressure cell lysate. d Average values (± standard deviation) do not include the trimethoprim-resistant mutants or B. catarrhalis. b

MIC determination. Minimal inhibitory concentrations (MICs) were determined by using a broth dilution technique. Stock solutions were prepared by dissolving the compound in deionized water, with the addition of HCI to pH 1.5 to facilitate dissolution. The

solution was then neutralized with NaOH and filter sterilized with membrane filters (0.2-,um pore size; Nucleopore), and the concentration of the compound was determined spectrophotometrically. A solution of the compound in GCB broth was then prepared by the

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

Neisseria gonorrhoeae F62 DUR1004 zi NRL5797 NRL5340 NRL5618 S6273 S6255 S6219 FA19

Relevant characters

VOL. 15, 1979

NEISSERIA DIHYDROFOLATE REDUCTASE

431

=

TABLE 2. Susceptibility to trimethoprim of various species and the dihydrofolate reductase (FH2) from each' Species

Neisseria gonorrhoeae (F62)

MIC Sp act (yg/ml) (U/mg)

60

0.100

K. (AM)

pH profile

Assay pH

NADPH

Monophasic

6.4

13

7.0

7.2

Kg

I-O

Mol wt

FH2

(nM) (nM)

7.8

45

270 18,000

3.3

37

236

N. meningitidis 15 0.069 Monophasic 5 6.4 4.0 22 242 18,000 N. lactamica 14 0.120 Monophasic 7 6.4 6.4 24 174 18,000 N. flavescens 50 14 0.039 Monophasic 6.4 1.2 32 1,099 18,000 5.4 Branhamella catarrhalis 34 788 18,000 55 3 (.o%8b Monophasic 6.4 Escherichia coli 0.2 0.010 Biphasic 4.4 1.1 8.5 18,000 7.0 8.9 a MIC values were deterniined by serial dilution of trimethoprim in liquid media. Specific activities of crude lysates were corrected for background NADPH oxidase activity. The kinetic values were determined by a weighted regression procedure and were used to calculate the I5o value according to the equation I5so= Ki[(1 +

S)/K8]. b '

French pressure cell lysate. Lysed by sonic disruption.

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

remarkable in their trimethoprim susceptibility, having MIC values in the 14- to 50-,ug/ml range. The dihydrofolate reductase specific activities of the lysates of two of these strains (66 and 67) and the I50 values for trimethoprim were similar to those of the other clinical isolates. Strains (FA48 through BR87) that have altered susceptibility to a variety of antibiotics because of changes in envelope permeability have been characterized (13, 26, 27). In this isogenic set, strains FA102 through FA171 were more resistant than the parent strain FA19 as a result of gene transfer from FA48. For example, FA140, containing the penA2, penB2, and mtr loci, had MIC values elevated 120-fold for penicillin, 16-fold for erythromycin, and 4-fold for tetracycline, chloramphenicol, and rifampin (27). None of these strains had an MIC of trimethoprim elevated more than 60% (Table 1). Strains BR54, BR84, and BR87 carry the env RESULTS locus, which was shown to be a phenotypic supTrimethoprim susceptibility. Table 1 pre- pressor of the resistance noted above. For exsents the trimethoprim MIC values, I50 values, ample, BR84 was 16-fold more susceptible to and dihydrofolate reductase specific activities penicillin, 30-fold more susceptible to erythroobtained for a variety of strains of N. gonor- mycin, and 4-fold more susceptible to tetracyrhoeae. The MIC of trimethoprim for the first cline and chloramphenicol than is FA140 (26). 10 strains, terned "clinical isolates," ranged The trimethoprim MIC was decreased by no from 2 to 70 ,ug/ml, with an average of 40 ,Lg/ml. more than 2.5-fold in these strains (Table 1). These strains all had similar specific activities Both the specific activity and the trimethoprim and similar Io values for trimethoprim. The data I50 values for these strains (FA48 through BR87) are in good agreement with other reports (4, 12) were similar to the values obtained with the in which MIC values of 12 to 100 ,g/ml were clinical isolates. recorded. Other strains of N. gonorrhoeae that In contrast to these loci, mutations leading to have altered antibiotic susceptibility were ex- increases in the amount of dihydrofolate reducamined to determine whether any variation in tase can cause larger increases in the MIC of trimethoprim susceptibility would be observed. trimethoprnm (A0078, A0105, and A0110; Table The three strains nonsusceptible to penicillin by 1). In the case of strain A0110, a 9.5-fold increase virtue of plasmid-coded production of penicillin- in specific activity resulted in a 12-fold increase ase have been recently described (10). These in the MIC of trimethoprim when compared penicillinase-producing N. gonorrhoeae are un- with F62. The binding of trimethoprim to the addition of 1 volume of threefold-concentrated GCB to 2 volumes of the stock solution. Dilutions were then prepared with GCB broth as diluent. Inocula were prepared by washing cells from plates incubated overnight (18 to 24 h) into prewarmed GCB broth. The optical density of the cell suspension was adjusted to 0.333 absorbance units (3 x 10' colony-forming units/mi), and 0.1 ml of the suspension added to 1.0 ml of prewarnmed drug dilutions. Decreasing the inoculum by 10-fold delayed growth in all tubes. Inoculated tubes were incubated for 18 h at 37°C under an 8% C02 atmosphere. The growth in each tube was measured by absorbance at 600 nm or judged visually. End points were taken to be an absorbance of less than 0.1 (Klett 10) after 18 h of growth and corresponded to approximately 90% inhibition of growth. Miscellaneous. Protein concentration was determined by the method of Lowry et al. (20). All absorbance measurements were performed with a Gilford model 240 recording spectrophotometer.

432

ANTIMICROB. AGENTS CHEMOTHER.

AVERETT ET AL.

0.20 z

0.15

P: 0.10

FG. 1. Activity of N. gonorrhoeae F62 dNhydrofolate reductase at various pH values. Buffers used are 0.100 M Tris chloride (0), 0.100 M potassium phosphate (x), 0.100 M imidazolium chloride (0), 0.100 M histidine chloride (0), 0.100 M acetate Tris (V), and 0.100 M succinate Tris (A). Rates plotted are corrected for non-enzymatic changes in absorbance at 340 nm.

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

enzyme from these strains was not altered, as determined by 1so values. Enzyme purification. The dihydrofolate reductase from strain F62 had an I5o value representative of those determined for other N. gonorrhoeae strains (Table 1). On the basis of yield of enzyme, specific activity of crude lysates, and favorable growth characteristics, F62 was chosen for further study. This strain has the added advantage of being one of the more widely studied strains of N. gonorrhoeae. Lysates of F62 presented different elution profiles upon gel permeation chromatography depending upon the method of lysis. Lysates prepared by the freeze-thaw procedure had much less absorbance in the void-volume region than did French pressure cell lysates (data not shown). This difference in elution profiles was consistent with the roughly fourfold-greater specific activity of freeze-thaw lysates (Table 1). Both lysis techniques released the same total amount of dihydrofolate reductase. However, the French pressure cell lysates had markedly elevated NADPH oxidase activities compared with freeze-thaw lysates. The decrease in absorbance at 340 nm resulting from oxidation of NADPH by this contaminating enzyme activity interfered with the assay of dihydrofolate reductase in lysates prepared by using the French pressure cell. It is apparent that the purification of low-molecular-weight material from N. gonorrhoeae may be enhanced by the choice of lysis technique. Freeze-thaw lysis is a simple procedure that may be generally useful for purifying small proteins and metabolites from the gonococcus. The combination offreeze-thaw lysis and gel permeation chromatography result in a 25fold purification over French press lysates. The elution position of N. gonorrhoeae dihydrofolate reductase was not altered by the technique used for cell lysis, and calibration of the column provided an estimate of 18,000 for the molecular weight of this enzyme. Enzyme properties. The activity of the purified enzyme between pH 4 and 9 is shown in Fig. 1. The data indicate a single pH optimum near pH 5.2 for the gonococcal dihydrofolate reductase. Also, greater activity in the neutral pH range was obtained with imidazolium chloride buffer than with the other buffers. An apparent discrepancy existed when gonococcal dihydrofolate reductase activity around pH 6 was examined in different buffers, since activity decreased with increasing pH in acetate tris(hydroxymethyl)aniinomethane (Tris) and succinate Tris, but activity increased with increasing pH in 0.1 M histidine chloride solutions. This anomaly was found to result from a dependence of the pH activity profile upon the con-

centration of histidine. As shown in Fig. 2, the peak activity increased and shifted to more acidic pH values as the histidine concentration was decreased; extrapolation to zero histidine concentration indicated a pH optimum near pH 5.2, consistent with data obtained in other buffers. A similar relation was observed in imidazolium chloride buffers (data not shown). This observation allowed the selection of assay conditions that provide optimal enzyme activity at pH values close to neutrality. Under standard assay conditions, the pH optimum was 6.4. At this pH, acid-mediated background absorbance changes due to degradation or precipitation of substrates were not significant. Trimethoprim typically inhibits bacterial dihydrofolate reductases by competing with the substrate dihydrofolate. Lineweaver-Burke plots of inhibition of N. gonorrhoeae F62 dihydrofolate reductase by trimethoprim showed that in this case also trimethoprim is competitive with dihydrofolate. Similar results were obtained at both pH 6.4 and 7.0. The kinetic constants obtained from such analyses are presented in Table 2, with data for E. coli RT500 dihydrofolate reductase as a point of reference. The Km values were similar for the enzymes from both sources, but the K, values for trimethoprim differed 30fold. This significantly lower affinity of trimethoprim for N. gonorrhoeae dihydrofolate reductase was sufficient to account for much of the difference in MIC observed with these organisms. There was no significant difference in trimethoprim binding to gonococcal dihydrofolate reductase at either pH value examined. Other Neisseria species. Although N. gon-

NEISSERIA DIHYDROFOLATE REDUCTASE

VOL. 15, 1979 0.25

E C. I-

0.20 -

z 0.15I

0.10 .

(2 z

0 05 1

lo

4.5

5.2

5.8

6.4

7.0

pH

FIG. 2. Activity of N. gonorrhoeae F62 dihA vdrofolate reductase at various pH values in differer t con centrations of histidine chloride buffer.

orrhoeae is the most commonly reported pathogen in the genus Neisseria, other specii es are known to cause serious disease (1, 22, 23, 30). It was therefore of interest to examine the dihydrofolate reductase from each of several other representatives of this genus. The data (ITable 2) were obtained by use of procedures deve Xloped for N. gonorrhoeae F62. The various slpecies examined are listed in Table 2 according to the reported homology of their deoxyriboniucleic acid (DNA) with the DNA of N. gonorr)hoeae (18). These species each have a dihydrofolate 7 reductase that is closely similar to that fri AT gonorrhoeae and that may be differenitiated from the E. coli enzyme on the basis of ti he pH activity profile and decreased affinity fc)r trimethoprim. The species that have least hlomology with the gonococcus (N. flavescens a nd B. catarrhalis) are also those that appear to differ from the gonococcus in the parameters (examined in this study. The major enzymatic 4difference observed within the genus is the relaitively low Km for dihydrofolate of the N. flavewscens enzyme. The only other difference observted in this study was the relative resistance of I3. catarrhalis to lysis by the freeze-thaw techrmque. Passage through a French pressure cell w,as required to release the enzymatic activity; tbie low specific activity of B. catarrhalis crude lI ysates probably reflects the lack of selective pirotein release resulting from use of this procedur.e. Inhibitors. Differences in substituents c)n the benzyl ring of 2,4-diamino-5-benzylpyrimiidines have been shown to alter both the affiniity of these compounds for E. coli dihydrofolatte reductase and their in vitro antibacterial activity

(6, 25). Therefore, several trimethoprim analogs were examined to determine whether their affinity for N. gonorrhoeae dihydrofolate reductase would be sensitive to substitution on the benzyl moiety. Figure 3 presents the measured I.% values of these compounds plotted against their MIC values for strain F62. Clearly substitution affected the affinity of these compounds for gonococcal dihydrofolate reductase, as is shown by the 10-fold range of 150 values. The fact that a linear relation existed between the Iho and MIC of these compounds indicates that these substitutions do not markedly alter the penetration of these compounds into N. gonorrhoeae. On the other hand, the folate analog methotrexate, which bound very tightly to N. gonorrhoeae dihydrofolate reductase, did not inhibit growth of N. gonorrhoeae, even at concentrations above 1 mM. In addition, leucovorin at 0.5 ,iM did not alter the MIC of trimethoprim for the gonococcus, suggesting that these pteridines may not penetrate the gonococcal cell envelope effectively (data not shown). Therefore, inhibitors that bind tightly to N. gonorrhoeae dihydrofolate reductase will be efficacious against the organism only if differential cell permeability is not a factor. DISCUSSION Large differences in the affinity of trimethoprim for dihydrofolate reductases from mammalian and bacterial sources are well known and responsible for the selective action of the drug (6, 14). Among bacteria, such differences are less well documented. In this study, we have deter300

Soo mic (uM)

200

100 . 111

100

200

300 1lo (nM)

400

500

600

FIG. 3. Relationship between the affinities (h5o val-

ues) of various 2,4-diamino-5-benzylpyrimidines for N. gonorrhoeae F62 dihydrofolate reductase and the MICs of these compounds against this organism. The compounds are: (I) 2,4-diamino-5[3,4'-dimethoxy5'-bromobenzyl]pyrimidine; (II) 2,4-diamino-5-[3bromo-4'-methoxybenzyl]pyrimidine; aIII) 2,4-diamino-5-[3',4',5'-trimethoxybenzyl]pyrimidine (trimethoprim); and (IV) 2,4-diamino-5-[3',5'-dimethox-

ybenzyl]pyrimidine.

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

N

433

434

AVERETT ET AL.

few fold (Tables 1 and 2). Thus, the difference in enzyme levels alone is probably not of sufficient magnitude to account for the difference in susceptibility between E. coli and N. gonorrhoeae. We have shown that the dihydrofolate reductase from N. gonorrhoeae differs significantly from this enzyme isolated from E. coli. The gonococcal enzyme has a monophasic pH activity profile, in contrast to that of E. coli (2, 3). Most interestingly, the gonococcal dihydrofolate reductase has an apparent Ki for trimethoprim of 45 nM, contrasting sharply with the E. coli Ki of 1.2 nM. Also, the gonococcal enzyme appears to have a turnover number severalfold greater than that of the enzyme from E. coli (D. Averett, unpublished data) and an isoelectric point 1.5 pH units higher than that of E. coli dihydrofolate reductase (D. R. Averett and D. P. Baccanari, Abstr. Annu. Meet. Am. Soc. Microbiol. 1978, A2, p.1), indicating that structural differences exist to account for the differences in inhibitor binding. Other Neisseria species that are known to cause disease have also been examined to determine whether all members of this genus have a similar dihydrofolate reductase. It is apparent from the data presented in Table 2 that although small differences may exist between individual members of the genus, all these species are generally similar in the properties of their respective dihydrofolate reductases. All have relatively poor affinities for trimethoprim, Km values in the low-micromolar range, and monophasic pH profiles. Therefore, there appears to be some justification for the belief that effective gonococcal inhibitors based on the tdmethoprim model will be efficacious against at least some other members of the genus Neisseria. The information presented here resulting from a study of a single enzyme indicates a taxonomic relationship among the Neisseria that is consistent with data obtained from other systems such as DNA-DNA homology (18) and interspecific transformation (7, 15). The calculated 150 values (Table 2) suggest that N. meningitidis and N. lactamica are both similar to N. gonorrhoeae. N. flavescens and B. catarrhalis have 150 values severalfold greater than those of the other three Neisseria species, indicating that these species may be less closely related. Also, B. catarrhalis was not susceptible to lysis by the freeze-thaw method used for the other species, providing the possibility of a difference in the cell envelope. It would appear, then, that B. catarrhalis is least similar to the gonococcus, followed by N. flavescens, N. lactamica, and N. meningitidis.

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

mined that dihydrofolate reductase from members of the genus Neisseria binds trimethoprim 20- to 50-fold less tightly than does the E. coli enzyme, as measured by the apparent Ki. This difference in affmity is of sufficient magnitude to account for most of the difference in the susceptibility of these organisms to this antibacterial agent. We have also shown that substitutions to the trimethoprim molecule can affect the affinity of the gonococcal enzyme for these ligands, and that there is a regular relation between the 150 of a compound and its MIC for N. gonorrhoeae F62. The correlation between Iso and MIC provides direct evidence that more effective inhibitors of the gonococcal dihydrofolate reductase will result in superior inhibition of N. gonorrhoeae cellular growth. Differential permeability of the gonococcal cell envelope to these compounds appears not to be a significant factor in deternmning the relative efficacy against the gonococcus ofthe 2,4-diamino-5-benzyl-pyrimidines examined here. In addition, mutations that alter the gonococcal cell envelope to a degree sufficient to cause a 128-fold increase in the MIC of penicillin result in only a 2-fold increase in the MIC of trimethoprim. Thus, susceptibility of the gonococcus to trimethoprim might not be greatly affected by the prevalence of these envelope mutations in nature (8, 21). In order to conclude that the relative nonsusceptibility of the gonococcus to trimethoprIm is a result of a decreased affinity of trimethoprim for the gonococcal dihydrofolate reductase, it is necessary to demonstrate that this enzyme is the site of action of the drug. Evidence exists that dihydrofolate reductase is the target of trimethoprim inhibition in E. coli (14). We have shown the site of action in N. gonorrhoeae in two ways. First, the relation between the 150 of a compound for dihydrofolate reductase and the MIC of that compound against the intact organism provides evidence that this enzyme is the target of trimethoprim inhibition in the gonococcus. The second line of evidence leading to the conclusion that trimethoprim acts in the organism as it does against the isolated enzyme is observed in trimethoprim-resistant mutants. As one would expect for the case of a target enzyme, greater amounts of dihydrofolate reductase activity correlate with increased resistance to the drug. This observation raises the possibility that differences in the level of this enzyme in N. gonorrhoeae and E. coli could account for the relative nonsusceptibility of the gonococcus to trmethoprim. Comparison of the specific activity of French press lysates of N. gonorrhoeae F62 with lysates of E. coli shows that the levels of dihydrofolate reductase do not differ more than a

ANTIMICROB. AGENTS CHEMOTHER.

VOL. 15, 1979

NEISSERIA DIHYDROFOLATE REDUCTASE

ACKNOWLEDGMENTS We are particularly indebted to P. F. Sparling and to R. Twarog for helpful discussions.

orrhoeae. J. Clin. Microbiol. 4:71-81. 16. Jyssum, S. 1971. Utilization of thymine, thymidine and TMP by Neisseria meningitidis. Acta Pathol. Microbiol. Scand. Sect. B 79:778-788. 17. Kellogg, D. S., Jr., W. L. Peacock, Jr., W. E. Deacon, L. Brown, and C. I. Pirkle. 1963. Neisseria gonorrhoeae. I. Virulence genetically linked to clonal variation. J. Bacteriol. 85:1274-1279. 18. Kingsbury, D. T. 1967. Deoxyribonucleic acid homologies among species of the genus Neisseria. J. Bacteriol. 94:870-874. 19. Koch, A., and J. Burchall. 1971. Reversal of the antimicrobial activity of trimethoprim by thymidine in commercially prepared media. Appl. Microbiol. 22:812-817. 20. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 21. Maier, T. W., P. Warner, L Zubryzycki, and M. Chila. 1977. Identification of drug resistance loci in various clinical isolates of Neisseria gonorrhoeae. Antimicrob. Agents Chemother. 12:444-446. 22. Ninane, G., J. Joly, and M. Kraytman. 1978. Bronchopulmonary infection due to Branhamella catarrhalis: 11 cases assessed by transtracheal puncture. Br. Med. J. 1:276-278. 23. Percival, A., J. E. Corkill, J. Rowlands, and R. B. Sykes. 1977. Pathogenicity of and ,B-lactamase production by Branhamella (Neisseria) catarrhalis. Lancet i:1175. 24. Perine, P. L, C. Thornsberry, W. Schalla, J. Biddle, M. S. Siegel, K.-H. Wong, and S. E. Thompson. 1977. Evidence for two distinct types of penicillinaseproducing Neisseria gonorrhoeae. Lancet i:993-995. 25. Roth, B., E. A. Falco, G. H. Hitchings, and S. R. M. Bushby. 1962. 5-Benzyl-2,4-diaminopyrimidines as antibacterial agents. I. Synthesis and antibacterial activity in vitro. J. Med. Pharm. Chem. 5:1103-1123. 26. Sarubbi, F. A., Jr., P. F. Sparling, E. Blackman, and E. Lewis. 1975. Loss of low-level antibiotic resistance in Neisseria gonorrhoeae due to env mutations. J. Bacteriol. 124:750-756. 27. Sparling, P. F., F. A. Sarubbi, Jr., and E. Blackman. 1975. Inheritance of low level resistance to penicillin, tetracycline, and chloramphenicol in Neisseria gonorrhoeae. J. Bacteriol. 124:740-749. 28. Wegener, W. S., B. H. Hebeler, and S. A. Morse. 1977. Cell envelope of Neisseria gonorrhoeae: relationship between autolysis in buffer and the hydrolysis of peptidoglycan. Infect. Immun. 18:210-219. 29. Wilkinson, A. E. 1977. The sensitivity of gonococci to penicillin. J. Antimicrob. Chemother. 3:197-198. 30. Wilson, H. D., and T. L. Overman. 1976. Septicemia due to Neisseria lactamica. J. Clin. Microbiol. 4: 214-215.

Downloaded from http://aac.asm.org/ on March 20, 2016 by guest

LITERATURE CITED 1. Anonymous. 1978. Morbid. Mortal. Weekly Rep. 27:2. 2. Baccanari, D. P., D. Averett, C. Briggs, and J. Burchall. 1977. Escherichia coli dihydrofolate reductase: isolation and characterization of two isoenzymes. Biochemistry 16:3566-3571. 3. Baccanari, D., A. Phillips, S. Smith, D. Sinski, and J. Burchall. 1976. Purification and properties of Escherichia coli dihydrofolate reductase. Biochemistry 14: 5267-5273. 4. Bach, M. C., M. Finland, 0. Gold, and C. Wilcox. 1973. Susceptibility of recently isolated pathogenic bacteria to trimethoprim and sulfamethoxazole separately and combined. J. Infect. Dis. 128(Suppl.):S508-S533. 5. Blakely, R. L. 1960. Crystalline dihydropteroylglutamic acid. Nature (London) 188:231. 6. Burchall, J. J. 1974. Trimethoprim and pyrimethamine, p. 305-320. In J. W. Corcoran and F. E. Hahn (ed.), Antibiotics, vol. 3. Springer-Verlag, Berlin, Heidelberg, New York. 7. Catlin, B. W., and L S. Cunningham. 1961. Transforming activities and base contents of deoxyribonucleate preparations from various Neisseria. J. Gen. Microbiol. 26:303-312. 8. Eisenstein, B. I., and P. F. Sparling. 1978. Mutations to increased antibiotic sensitivity in naturally-occurring gonococci. Nature (London) 271:242-244. 9. Elmros, T., L. G. Burman, and G. D. Bloom. 1976. Autolysis of Neisseria gonorrhoeae. J. Bacteriol. 126: 969-976. 10. Elwell, L P., M. Roberts, L. W. Mayer, and S. Falkow. 1977. Plasmid-mediated beta-lactamase production in Neisseria gonorrhoeae. Antimicrob. Agents Chemother. 11:528-533. 11. Futterman, S. 1957. Enzymatic reduction of folic acid and dihydrofolic acid to tetrahydrofolic acid. J. Biol. Chem. 228:1031-1038. 12. Garrod, L P., and P. M. Waterworth. 1968. Action of three drug combinations on gonococci. Br. J. Vener. Dis. 44:75-79. 13. Guymon, L. F., and P. F. Sparling. 1975. Altered crystal violet permeability and lytic behavior in antibiotic-resistant and -sensitive mutants of Neisseria gonorrhoeae. J. Bacteriol. 124:757-763. 14. Hitchings, G. H., and J. J. Burchall. 1965. Inhibition of folate biosynthesis and function as a basis for chemotherapy. Adv. Enzymol. 27:417-468. 15. Janik, A., E. Juni, and G. A. Heym. 1976. Genetic transformation as a tool for detection of Neisseria gon-

435