HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 26, 19449-19455, June 30, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/19449    most recent M001278200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Andersson, J. Articles by Sjöberg, B.-M. Search for Related Content PubMed PubMed Citation Articles by Andersson, J. Articles by Sjöberg, B.-M. Social Bookmarking                      What's this?

Cysteines Involved in Radical Generation and Catalysis of Class III Anaerobic Ribonucleotide Reductase

A PROTEIN ENGINEERING STUDY OF BACTERIOPHAGE T4 NrdD^*

Jessica Andersson, MariAnn Westman, Margareta Sahlin, and Britt-Marie Sjöberg

From the Department of Molecular Biology, Stockholm University, SE-10691 Stockholm, Sweden

Received for publication, February 15, 2000, and in revised form, March 22, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Class III ribonucleotide reductase (RNR) is an anaerobic glycyl radical enzyme that catalyzes the reduction of ribonucleotides to deoxyribonucleotides. We have investigated the importance in the reaction mechanism of nine conserved cysteine residues in class III RNR from bacteriophage T4. By using site-directed mutagenesis, we show that two of the cysteines, Cys-79 and Cys-290, are directly involved in the reaction mechanism. Based on the positioning of these two residues in the active site region of the known three-dimensional structure of the phage T4 enzyme, and their structural equivalence to two cysteine residues in the active site region of the aerobic class I RNR, we suggest that Cys-290 participates in the reaction mechanism by forming a transient thiyl radical and that Cys-79 participates in the actual reduction of the substrate. Our results provide strong experimental evidence for a similar radical-based reaction mechanism in all classes of RNR but also identify important differences between class III RNR and the other classes of RNR as regards the reduction per se. We also identify a cluster of four cysteines (Cys-543, Cys-546, Cys-561, and Cys-564) in the C-terminal part of the class III enzyme, which are essential for formation of the glycyl radical. These cysteines make up a CX₂C-CX₂C motif in the vicinity of the stable radical at Gly-580. We propose that the four cysteines are involved in radical transfer between Gly-580 and the cofactor S-adenosylmethionine of the activating NrdG enzyme needed for glycyl radical generation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Ribonucleotide reductase (RNR)¹ catalyzes the reduction of ribonucleotides to their corresponding deoxyribonucleotides. As this is the only way for de novo synthesis of building blocks for DNA, RNR is an essential constituent of all living cells. At least three different classes of RNRs can be distinguished based on their polypeptide composition and cofactor requirements (1-3). Several prokaryotic organisms encode more than one class of RNR and in some cases also more than one representative from the same class.

All previously characterized class I and II RNRs operate via a radical based mechanism, considered to involve one cysteine residue that forms a transient thiyl radical during catalysis and a cysteine pair that provides the reducing electrons (4-6). However, the two classes differ in the way they acquire the thiyl radical. The resting class I RNR harbors a stable tyrosyl radical close to a diferric-oxo center within one of its components. The tyrosyl radical interacts with the active site in the other component via a long range radical transfer pathway. Class II RNRs require the cofactor adenosylcobalamin that by homolytic cleavage acts as a thiyl radical generator. Class III RNRs represent a third variant and harbors a stable glycyl radical within the so-called NrdD component that also harbors the active site region. An activating component, the 4Fe-4S NrdG protein, has the ability to generate the glycyl radical in NrdD via cleavage of the cofactor AdoMet (7-9). Class III RNR is an anaerobic enzyme, and the glycyl radical is extremely sensitive to oxygen. In the recently deduced three-dimensional structure of a class III enzyme from bacteriophage T4, only two cysteine residues were encountered in the active site region (10).

The radical generation procedure of class III RNRs is similar to that of pyruvate formate-lyase (PFL), a key enzyme in anaerobic glucose metabolism (11). An activating iron-sulfur protein cleaves AdoMet to generate a glycyl radical in PFL (12, 13). Two vicinal cysteines, Cys-418 and Cys-419, are crucial for catalysis, and both are capable of forming a transient thiyl radical (14-16). Recently, other proteins have also been suggested to harbor a glycyl radical (17, 18), making glycyl radical enzymes a distinct group of proteins (19).

Despite the lack of significant overall amino acid sequence similarities between class III RNRs on the one hand and class I and II RNRs on the other hand, or between class III RNRs and PFL enzymes, the catalytic cores of T4 class III RNR, protein R1 of class I RNR from Escherichia coli, and E. coli PFL are strikingly similar (10, 16, 19, 20). Interestingly, Cys-290 in the class III enzyme is in a position corresponding to the active site residues Cys-439 in the R1 structure (Fig. 1) and Cys-419 in the PFL structure. Cys-439 in R1 is the residue proposed to form the transient thiyl radical during catalysis of RNR. Additionally, Cys-79 from the class III enzyme is in a corresponding position to Cys-225 in R1, the residue in class I RNR that together with Cys-462 forms the redox active cysteine pair mentioned above (Fig. 1). Cys-462 in the class I structure has no cysteine counterpart in the class III RNR. These structural similarities suggest that Cys-290 in class III RNR could have the same proposed function as Cys-439 in the class I reaction mechanism, but it also suggests that the reaction mechanisms are different as regards Cys-79 in class III RNR and the redox active Cys-225-Cys-462 pair in the class I enzyme.

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Comparisons of the active sites of class III and class I RNR. A, active site region of bacteriophage T4 NrdD with GDP substrate modeled from active site region of E. coli protein R1 (10). Conserved residues Cys-79, Cys-290, and Asn-311 have been indicated. B, active site region of E. coli protein R1 cocrystallized with GDP (43). Conserved residues Cys-225, Cys-439, and Cys-462 are indicated.

We have earlier used site-directed mutagenesis to identify Gly-580 in phage T4 class III RNR as the site of the stable glycyl radical (21). In this work we have used the same approach to identify cysteine residues involved in the reaction mechanism of the T4 class III enzyme. Alignment of available class III RNR sequences identified five invariant, one highly conserved, and three moderately conserved cysteine residues (cf. Table I). We show that six of these cysteine residues are essential for class III RNR function. Cys-79 and Cys-290 are directly involved in catalysis, whereas a cluster of four cysteine residues (Cys-543, Cys-546, Cys-561, and Cys-564) in the C-terminal part of the enzyme is involved in the radical generation mechanism. We propose that Cys-290 forms a transient thiyl radical corresponding to Cys-439 in the class I RNR, which suggests that the initiation of the reaction for all three RNR classes is similar. In addition, we propose that the C-terminal cysteine cluster is involved in the radical transfer between AdoMet and the stable radical position at Gly-580 in the T4 enzyme.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Oligonucleotides used for site-directed mutagenesis and sequencing were synthesized by Scandinavian Gene Synthesis AB, Sweden. Restriction enzymes were from Amersham Pharmacia Biotech, New England Biolabs, or Roche Molecular Biochemicals. Chemicals were from Sigma, Saveen, and Amersham Pharmacia Biotech.

Bacterial Strains-- E. coli CJ236 (dut-1, ung-1, thi-1, relA1/pCJ105), E. coli MV1190 ( (lac-proAB), thi, supE, (srl-recA)306::Tn10/F'traD36, proAB, lacI^qZM15), E. coli JM109(DE3) endA1, recA1, gyrA96, thi, hsdR17, (r_K, mK⁺), relA1, supE44, (lac-proAB), [F', traD36, proAB, lacI^qZM15], (DE3), E. coli C1-a, a wild-type, prototrophic strain.

Plasmids-- Plasmids pET29T4nrdD and pET21aT4nrdG were described previously (21). Plasmid pEE1010 containing the gene for ferredoxin (flavodoxin) NADP⁺ reductase (22) was a kind gift from Vera Bianchi and Elisabeth Haggård-Ljungquist. The expression plasmid pDH1 for flavodoxin (23) was a kind gift from Peter Reichard.

Oligonucleotide-directed Mutagenesis-- All mutants were constructed with the Kunkel method (24, 25) using the Mutagene phagemid in vitro mutagenesis kit (version 2) as described previously (21). The mutagenic oligonucleotides also contained the insertion or deletion of a restriction enzyme cleavage site to facilitate the screening for mutants. The primers (and within parentheses the screening restriction enzyme) used for each mutation were as follows: C79S, 5'-d(TACTAAACAGCTATTAGTAAATG)-3' (AluI); C260S, 5'-d(TTTTGCTTGCGGACTCTAGAGCA)-3' (HhaI); C290S, 5'-d(AGAAACTACGGAACCCATCGGA)-3' (NlaIV); C453S, 5'-d(CTGTATCGAGCTTAGAGAAGCGATA)-3' (DdeI); C543S, 5'-d(CCACATGTAAAGCTTTTATCTACTG)-3' (AluI); C546S, 5'-d(GGGTACTTCCAGATGTAAAACAT)-3' (AflIII); C561S, 5'-d(CACAAATAGAAGAAACAAATCCG)-3' (MboII); C564S, 5'-CAGTTTCTCCAGAAATAGAACAA-3' (BpmI); C579S, 5'-CCAAATAACCAGATGTTCTTCTT-3' (AflIII). The mismatching bases are shown in bold, and the changed codon is underlined. All mutant plasmids were confirmed by sequencing using an ABI Prism Cycle sequencing kit from Perkin-Elmer. The analyzing gels were run by Katarina Gell at CMB, Karolinska Institute, Sweden.

Oxygen-dependent Cleavage-- JM109(DE3) strains containing wild-type or mutant pET29T4nrdD (kanamycin^R) plasmids and the wild-type pET21aT4nrdG plasmid (ampicillin^R) were grown aerobically in LB medium supplemented with 35 µg/ml kanamycin and 100 µg/ml carbenicillin. The cultures were grown until the A₆₄₀ reached 0.5 and were then induced with isopropyl-1-thio--D-galactopyranoside (IPTG; final concentration 1 mM). Samples of 0.5 absorbance units were withdrawn immediately before induction and after 3 h of induction. The samples were analyzed by SDS-PAGE on a 12.7% gel and stained with Coomassie Brilliant Blue. The gel was then densitometrically scanned, and the bands were quantified using the software ImageQuant from Molecular Dynamics. In Fig. 2, below the lanes, the extent of truncation is shown as percent NrdD', which is the amount of NrdD' divided by the sum of NrdD and NrdD'.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Oxygen-dependent cleavage of NrdD. Wild-type or mutant NrdD were coexpressed together with NrdG. Samples were withdrawn and analyzed on a 12.7% SDS-PAGE gel. Uninduced shows a wild-type NrdD sample before induction with IPTG; this is the background of the E. coli cells. The remaining sample lanes show NrdD wild-type and the mutants (as indicated in the figure) after 3 h of induction. LMW is a low molecular weight marker (Amersham Pharmacia Biotech); the molecular weight of the marker proteins is indicated in kDa. NrdD and the truncated form, called NrdD', are indicated by arrows. The bands were quantified after densitometric scanning, and the background from lane 1 was subtracted. At the bottom of the lanes, the percentage of cleavage is shown (% NrdD'), which is the ratio of NrdD' divided by the sum of NrdD and NrdD'.

Anaerobic Expression, Anaerobic Crude Extract Preparation, EPR Sample Preparation, and Activity Assays-- For anaerobic growth and anaerobic crude extract preparations of wild-type or mutant NrdD, and of wild-type NrdG, the procedure described in Ref. 21 was followed. Anaerobic activity assays and EPR measurements in anaerobic crude extracts are also described in Ref. 21.

Aerobic Expression and Purification of NrdD-- Wild-type and mutant proteins were prepared as described previously (26). In brief, the NrdD protein was expressed aerobically using IPTG induction. After harvesting, the frozen cells were disintegrated in an X-press (from BIOX), extracted in 20 mM Tris-HCl, pH 8.0, 1 mM DTT, and soluble proteins were batch-purified by streptomycin sulfate (1% final concentration) and ammonium sulfate (40% final concentration) precipitations. Further purification was achieved using hydrophobic interaction chromatography with butyl-Sepharose 4 Fast Flow media from Amersham Pharmacia Biotech.

Aerobic Expression and Crude Extract Preparation of NrdG-- The JM109(DE3) strain containing the plasmid pET21aT4nrdG was grown in Luria Broth medium, pH 7.0, supplemented with 50 µg/ml carbenicillin, and deionized water was used. The growth was done in a 10-liter container using a Microferm® fermentor from New Brunswick Scientific Co. The culture was grown at 30 °C, and when the cells reached an A₆₄₀ of 0.6 they were induced with IPTG (final concentration 0.5 mM), and the temperature was lowered (typically to 14-19 °C) to avoid formation of inclusion bodies. After 20 h of induction (A₆₄₀ = 1.0-1.4) the cells were harvested by centrifugation, and the pellet was frozen at 80 °C. The frozen pellet was pressed 3-5 times in an X-press and extracted with 100 mM Tris-HCl, pH 7.6, 5 mM DTT in a blender. Nucleic acids were removed by precipitation with streptomycin sulfate to a final concentration of 1%. The proteins were then precipitated with 40% ammonium sulfate and desalted over a NAP column containing Sephadex® G-25 medium from Amersham Pharmacia Biotech. The NrdG content was typically ~10% of the total protein fraction, as estimated from SDS-PAGE analyses. The desalted crude extract preparation was later used for EPR samples and enzymatic activity assays.

Aerobic Expression and Purification of Flavodoxin Reductase-- The protocol from Ref. 22 was followed with some minor changes. The C1-a strain containing the plasmid pEE1010 was grown to stationary phase for expression of flavodoxin reductase, and the cells were then lysed with lysozyme treatment and freeze-thawing. A Superose-12 column was used to purify the 29-kDa protein, and its flavin-related absorption spectrum was found identical to that in Ref. 22.

Aerobic Expression and Purification of Flavodoxin-- The protocol from D. Hoover was followed. The plasmid pDH1 containing the gene for flavodoxin was grown in the C1-a strain, and induction was made with IPTG; the cells were lysozyme-treated and freeze-thawed 3 times; neutralized FMN was added, 1-2 mg/liter culture. A DEAE column was used to separate the flavodoxin protein as confirmed by SDS-PAGE, its absorption spectrum, and the A₂₇₄/A₄₆₇ ratio.

Anaerobic Activation and EPR Analyses of Aerobically Purified Components-- Purified wild-type or mutant NrdD, crude extract preparation of NrdG, and a mixture containing all the activation components were flushed separately with argon at 4 °C for 40 min in order to remove oxygen. They were then transferred to a Forma Scientific anaerobic glove box where they were mixed and incubated anaerobically for 45 min at room temperature. The final concentrations were 37.5 µM pure NrdD, NrdG-containing crude extract (40 mg/ml total protein, ~75 µM NrdG), 30 mM Tris-HCl, pH 8.0, 30 mM KCl, 5 mM sodium formate, 5 mM DTT, 1.25 mM NADPH, 18 µM flavodoxin reductase, 6 µM flavodoxin, and 0.5 mM AdoMet, and the total volume was 160 µl. The samples were then transferred to EPR tubes, sealed, taken out of the anaerobic box, and frozen in liquid nitrogen.

X-band EPR measurements were performed at 77 K on a Bruker ESP300 spectrometer using a cold finger Dewar filled with liquid nitrogen. Double integrals of the EPR signals were determined using the Bruker software, and the radical content was calculated from the spin concentration comparing with a Cu²⁺-EDTA standard and a known glycyl radical standard.

Anaerobic Activation and Enzymatic Activity Assays of Aerobically Purified Components-- Purified wild-type or mutant NrdD, crude extract preparation of NrdG, an activation mixture containing all the necessary components for generating the glycyl radical, and a substrate mixture were flushed separately with argon at 4 °C for 40 min to remove oxygen before transferring them to the anaerobic box. NrdD, NrdG and the activation mixture were then mixed and incubated at room temperature for 10 min. The incubation mixture contained 0.12 µM pure NrdD, NrdG-containing crude extract (10 mg/ml total protein,~20 µM NrdG), 30 mM Tris-HCl, pH 8.0, 30 mM KCl, 5 mM DTT, 1.25 mM NADPH, 3 µM flavodoxin reductase, 2 µM flavodoxin, and 0.5 mM AdoMet, and the total volume was 25 µl. After 10 min, 25 µl of substrate mixture was added giving final concentrations of 30 mM Tris-HCl, pH 8.0, 30 mM KCl, 5 mM sodium formate, 5 mM DTT, 20 mM MgCl₂, 1 mM dATP, and 5 mM [³H]CTP. Incubation with substrate was stopped after 20 min by addition of 500 µl of 1 M perchloric acid. A carrier, 50 µl of dCMP, 5 mg/ml, was added as internal standard to follow recovery during the work-up procedure. After dephosphorylation to the monophosphate level by boiling, separation of [³H]dCMP and [³H]CMP was performed over a Dowex-50 column, and isocratic elution was made with 0.2 M acetic acid. The formed ³H-labeled product was then quantified in a scintillation counter.

Metal Analyses-- The analyses were made by ICP-AES, plasma emission spectrometry, or ICP-SMS, plasma mass spectrometry by SGAB Analytica, Luleå, Sweden. One additional purification step (anionic mono-Q chromatography) was added prior to analyses. Protein samples contained 85-90% NrdD as judged from SDS-PAGE analysis and Coomassie Blue staining.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Site-directed Mutagenesis-- To test our hypothesis that cysteine residues are involved in the reaction mechanism of the class III RNRs, we used site-directed mutagenesis to change conserved cysteines in the T4 NrdD protein. In all, nine cysteines (cf. Table I) were separately mutated to serine in order to abolish the redox function of the SH group while maintaining the size and the H-bonding capacity of the side chain. Five of these residues (Cys-79, Cys-290, Cys-543, Cys-561, and Cys-564) are invariant, and one (Cys-546) is highly conserved (Table I). Four of the C-terminal cysteines make up a CX₂C-CX₂C motif, below denoted the C-terminal cluster. Cys-260 and Cys-453 are only conserved in 12 of the 19 available NrdD sequences, as is Cys-579, adjacent to the site of the stable radical at Gly-580 (Table I). During the expression and purification all mutant Cys  Ser NrdD proteins behaved like wild type.

Oxygen-dependent Cleavage of Cys  Ser Mutant NrdD Proteins-- The class III anaerobic RNRs have been shown to undergo truncation at the site of the glycyl radical when the radical-containing enzyme is exposed to oxygen (27). A facile in vivo assay to detect formation of the glycyl radical is to monitor this truncation of NrdD when coexpressed aerobically with NrdG (21). This assay was used to monitor whether a glycyl radical is formed or not in the Cys  Ser mutants. Samples from the coexpressions were analyzed on a SDS-PAGE gel (Fig. 2). The full-length NrdD protein migrates approximately according to its theoretical molecular mass of 68 kDa (21). The truncated form, denoted NrdD', migrates slightly faster than the full-length NrdD, in good agreement with the anticipated truncation at Gly-580 ~3 kDa from the C terminus of the phage T4 NrdD.

The oxygen-dependent cleavage can clearly be seen for the mutants C79S, C260S, C290S, C453S, and C579S (Fig. 2). As a negative control, the G580A mutant is also shown; oxygen-dependent cleavage is not possible in this mutant since no radical can be formed at Ala-580 (21). To estimate the extent of truncation, the protein gel was densitometrically scanned, and the protein bands were quantified. Samples harvested before induction with IPTG contained an unrelated protein band that comigrates with truncated NrdD' protein (Fig. 2, Uninduced). This unrelated protein band was subtracted from the NrdD' protein band. As can be seen, the truncation occurs in approximately 60% of the polypeptide chains of wild-type NrdD. The extent of truncation agrees with the amount of radical per NrdD dimer (~0.55; cf. Table III) and has earlier been observed for NrdD from phage T4 and from E. coli (21, 28, 29). The mutants C79S, C260S, C290S, and C453S were truncated to 51-62%, and the C579S mutant was truncated to 39% (Fig. 2). This result clearly establishes that a glycyl radical could be formed in these five mutants. Fig. 2 also shows that no truncation occurs in the four C-terminal mutants C543S, C546S, C561S, and C564S suggesting that no glycyl radical could be formed in these mutants.

EPR Measurements and Enzymatic Activity Assays on Anaerobic Crude Extracts-- Anaerobic crude extracts from the four mutants C79S, C260S, C290S, and C453S were combined with anaerobically produced NrdG-containing extracts and incubated with AdoMet to promote generation of the glycyl radical and were then used for EPR measurements and enzymatic activity assays. All four mutant proteins displayed a doublet EPR signal with the characteristics of the glycyl radical of wild-type NrdD (Fig. 3A). The relative radical contents compared with wild-type NrdD are shown in Table II. As can be seen, the mutants C79S, C260S, and C453S have radical contents comparable to wild type; the mutant C260S has an even higher radical content than wild-type NrdD. The mutant C290S has roughly half the radical content of wild-type NrdD.

View larger version (34K): [in this window] [in a new window]   Fig. 3.   EPR spectra of NrdD:NrdG mixtures. A, anaerobic crude extracts of wild-type or mutant NrdD were mixed with anaerobic crude extract of NrdG and incubated anaerobically together with AdoMet for 45 min. Samples for EPR measurement were transferred to EPR tubes, sealed with a rubber top, and frozen in liquid nitrogen. EPR conditions were as follows: microwave power, 10 microwatts; modulation amplitude, 0.16 mT; receiver gain, 1 × 106; 4 scans. The arrow indicates a g value of 2.004. B, mixtures of purified NrdD and crude extract of NrdG. Aerobically expressed purified wild-type or mutant NrdD was mixed with aerobically expressed crude extract of NrdG and activated anaerobically. Samples for EPR measurement were transferred to EPR tubes, sealed with a rubber top, and frozen in liquid nitrogen. EPR conditions were as follows: microwave power, 8.5 microwatts; modulation amplitude, 0.16 mT; receiver gain, 5 × 105; 1 scan. The arrow indicates a g value of 2.004.

Interestingly, the enzymatic activity assays showed that C79S and C290S had completely lost enzymatic activity (Table II). Given the fact that a glycyl radical was formed in these mutants (Fig. 3A), the two cysteine residues must be important for a step occurring after the generation of the glycyl radical, plausibly the reaction mechanism. This is the first experimental evidence for the importance of Cys-79 and Cys-290 in the reaction mechanism of the class III RNRs.

The mutants C260S and C453S had enzymatic activities comparable to wild type showing that these mutants are not involved in the reaction mechanism. The C260S mutant has an even higher activity than wild-type NrdD, which was also reflected in a higher radical content.

EPR Measurements and Enzymatic Activity Assays on Purified NrdD Proteins-- In an attempt to obtain a system with pure proteins and a distinct activation procedure, we set out to express and purify NrdD and NrdG separately. Unfortunately, the overexpressed NrdG protein had a severe tendency to form inclusion bodies, and the resulting purified NrdG protein was highly unstable during assay conditions. We therefore resorted to an activation procedure using purified NrdD, crude extract of NrdG, and purified flavodoxin system (flavodoxin and flavodoxin reductase), see "Experimental Procedures." All protein fractions were prepared aerobically and then made anaerobic by argon flushing before the activation procedure. In this set of assays, the NrdD concentration was ~300-fold higher in the EPR measurements as compared with the enzymatic activity measurements, and the NrdG:NrdD ratios were 2 and 100-150, respectively. Accordingly, we used activation times of 45 min for EPR experiments and 10 min for activity measurements.

Activity assays of the four mutants C79S, C260S, C290S, and C453S confirmed the former results that C79S and C290S lack enzymatic activity and that C260S and C453S have activities comparable to wild-type NrdD (Table III). Again, C260S had slightly higher enzymatic activity than wild-type NrdD (Tables II and III). The C453S mutant had 68% of wild-type activity compared with 75% for the anaerobic crude extracts (Tables II and III).

For the mutants C79S and C260S, the EPR measurements gave the same results as for the anaerobic crude extracts (Fig. 3A) showing radical contents comparable to wild-type NrdD (Fig. 3B). The relative radical content of the C453S mutant was much lower than in the anaerobic crude extracts (Tables II and III). However, as this mutant has almost full enzymatic activity (Table III), it is plausible that these EPR conditions are suboptimal compared with the results from the anaerobic crude extracts.

Unexpectedly, no glycyl radical could be detected by EPR in the purified C290S sample (Fig. 3B). This particular experiment could thus not conclusively distinguish whether the lack of enzymatic activity was due to a failure in forming the glycyl radical or a cause of the mutation itself. However, since the two previous assays (Figs. 2 and 3A) showed glycyl radical formation in C290S, it indicates that the glycyl radical in the purified mutant C290S protein is not as stable as in wild-type NrdD.

Also the mutant C579S seemed to be less stable in assays with purified protein as compared with crude extract assays (data not shown). Even though this mutant had no detectable radical, its low but significant enzyme activity (Table III) suggests that some glycyl radical was formed (cf. Fig. 2). The detection limit of our EPR assay is about 0.5 µM glycyl radical, corresponding to ~2% of the wild-type radical content. As for the C290S mutant, C579S could form a glycyl radical as judged from the oxygen-dependent cleavage assay (Fig. 2).

The C-terminal Cysteine Cluster-- The cysteines of the C-terminal cluster Cys-543, Cys-546, Cys-561, and Cys-564 had no detectable glycyl radical by EPR and no enzymatic activity (Table III). The lack of glycyl radical formation confirmed the results from the oxygen-dependent cleavage assay (Fig. 2). The lack of enzymatic activity is logical since a glycyl radical is required for enzymatic activity.

One obvious explanation for the inability to form a glycyl radical would be that the complex between NrdD and NrdG was unable to form in these four mutant proteins. Generation of the glycyl radical in the NrdD subunit is promoted by the FeS center of the NrdG subunit (7, 9, 30). We therefore coexpressed each NrdD mutant with wild-type NrdG and purified the crude extracts over butyl-Sepharose columns. Similar conditions were used for purification of the holoenzyme complex NrdD(G580A)-NrdG(wild-type) (10). Fig. 4 shows that all four mutants, C543S, C546S, C561S, and C564S, were able to bind the NrdG subunit and that the relative yields of NrdD and NrdG components were comparable to that seen for the NrdD(G580A)-NrdG(wild-type) complex (21). This rules out the possibility that the lack of glycyl radical in any of these Cys  Ser mutants is due to impaired binding of the NrdG subunit and suggests that the three-dimensional structures of these mutants are unchanged.

View larger version (60K): [in this window] [in a new window]   Fig. 4.   SDS-PAGE analysis of cochromatographing NrdD and NrdG components. The samples contain C543S (lane 1), C546S (lane 2), C561S (lane 3), C5645 (lane 4), and wild-type NrdD (lane 5), in all cases mixed with excess NrdG. The migration of the NrdD and NrdG polypeptides is indicated. M denotes a marker protein mixture, molecular masses of 94, 67, 43, 30, 20.1 and 14.4 kDa.

Metal Analyses-- To investigate further the role of the C-terminal cysteine cluster, we performed a full transition metal analysis of wild-type NrdD and of an aliquot of the complex NrdD(G580A)-NrdG(wild-type) used for crystallizations. In parallel, NrdD(G580A) and the sample buffer were used as controls. None of the protein samples had any significant amounts of metal ions associated to them (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study we provide the first biochemical evidence that multiple cysteine residues are required for the generation of the stable glycyl radical and for enzymatic activity in the class III RNR from phage T4. Multiple cysteine residues have earlier been shown to be involved in the reaction mechanism of class I and II RNRs (4-6).

The reaction mechanisms of class I and II RNRs are very similar and are based on radical chemistry. Three active site cysteines are responsible for nucleotide reduction, and two C-terminal cysteines interact with a physiological protein reductant (1). The two systems differ mainly in the way the radical chemistry is initiated. Implications for a common reaction mechanism for all RNR classes stem from reactions with 2'-azido substrate analogues. 2'-Azido-CDP is a mechanism-based inhibitor that irreversibly destroys the tyrosyl radical of class I RNR and forms a new nitrogen-centered radical in a half-turnover reaction (31-33). The nitrogen-centered radical has also been observed in incubations of 2'-azido-CTP and a class II RNR (34). In this sense, the reaction mechanism of class III RNR is similar to that of class I and II RNRs. Incubations of E. coli class III RNR with 2'-azido-CTP led to destruction of the glycyl radical (35), and similar results have been obtained for class III RNR from phage T4.² On the other hand, differences in reaction mechanisms between class III RNRs and class I/II RNRs were expected, as class III RNRs use formate as reductant and are independent of protein-derived reductants, like "redoxins" (36). The recently solved three-dimensional structure of T4 NrdD corroborated this as only two conserved cysteines were localized to the active site of the class III RNR (10).

We have investigated the importance of several highly conserved cysteine residues in the reaction mechanism of the class III RNRs by site-directed mutagenesis. In one set of experiments we show that engineering of Cys-79 and Cys-290 renders catalytically inert protein, whereas engineering of Cys-260 and Cys-453 does not. The invariant residues Cys-79 and Cys-290 are localized to the active site region of T4 NrdD. By using the photoaffinity labeling technique, Olcott et al. (26) independently identified Cys-290 of T4 NrdD as part of a nucleotide-binding site. Cys-260 and Cys-453, on the other hand, are not absolutely conserved among class III RNRs (cf. Table I) and are located 21 and 15 Å from the active site region of T4 NrdD (10).

The identification of only two catalytically essential cysteines in the active site region of T4 NrdD confirms that the reaction mechanism of class III RNR is not identical to that of class I and II RNRs. However, extensive three-dimensional similarities between class I and III RNRs suggest that Cys-290 of T4 NrdD is homologous to Cys-439 of E. coli R1 (10, 20). It has therefore been proposed that Cys-290, whose sulfur atom is 4.0 Å from the 3' carbon atom of the modeled substrate, forms a transient thiyl radical that initiates catalysis by abstracting the 3'-H atom of the substrate (10), a suggestion that is fully corroborated by our results. Interestingly, Cys-290 of T4 NrdD and Cys-439 of R1 could also be aligned to Cys-419 of PFL, one of the cysteines shown to form a transient thiyl radical in the PFL system (14-16). Recent results for the class I RNR indicate involvement of a conserved carboxylate side chain in the steps following the abstraction of the 3'-hydrogen of the substrate (37-39). No direct counterpart of this residue is found in the class III RNR, instead formate may take this role.

Comparison of the class I and III structures shows that Cys-79 in T4 NrdD has a homologous position to Cys-225 in E. coli R1 (10, 20). The sulfur atom of Cys-79 is positioned 4.1 Å from the 2'-OH of the modeled substrate (10). In the class I reaction mechanism, Cys-225 and the conserved Cys-462 are the immediate reductants of the 2'-position. However, there is no cysteine counterpart in T4 NrdD of Cys-462 in E. coli class I RNR. Instead, one formate molecule is oxidized to CO₂ per molecule of reduced substrate in the class III mechanism (36). Based on these observations, we propose that Cys-79, together with formate, participates in later steps of the reaction mechanism and at some stage forms a transient thiyl radical. Formate is a small molecule that could easily access the active site. Most likely, it participates in the reaction mechanism as a carbon dioxide radical.

In another experiment, we engineered Cys-579, next to the site of the stable radical at Gly-580 in T4 NrdD. Cys-579 is a highly conserved residue, present in 12 out of 19 class III sequences (Table I). The C579S mutant had a low enzymatic activity (7% of wild-type activity; Table III) and approximately two-thirds of the wild-type glycyl radical content (Fig. 2). The corresponding C680S mutation in the class III RNR from E. coli was shown to have 14% of the wild-type activity and 20% of the wild-type radical content (29). We have thus confirmed that the cysteine residue immediately preceding the glycyl radical residue in class III RNRs is not directly involved in the radical formation or the reaction mechanism. However, its proximity to the glycyl radical may make it particularly susceptible to mutational changes. Cys-579 could play a structural role during generation and stabilization of the glycyl radical or when the Gly-580^·  Cys-290^· transition occurs.

In a third set of experiments, we have shown that the four C-terminal cysteines (Cys-543, Cys-546, Cys-561, and Cys564), which make up the cluster motif CX₂C-CX₂C, are essential for formation of the glycyl radical (Fig. 2). The cysteine cluster is reminiscent of a metal-binding motif, e.g. an iron-cysteine center or a zinc finger. We were, however, unable to identify any metal ions associated with the protein. One possibility is that bound metal ions were lost during purification of the protein. Even though the distance between the C-terminal cysteines and the glycyl radical is long,³ the C-terminal cysteine cluster could have a structurally important role in radical generation. Initially, it was suggested that the NrdD component of E. coli class III RNR contained an FeS cluster (40), but this metal site was later shown to be located in the NrdG subunit (30). The functional NrdG metal site is now known to be a Fe₄S₄⁺/Fe₄S₄²⁺ redox center that is easily interconverted to redox-inactive Fe₃S₄^+/0 and Fe₂S₂²⁺ forms (9, 41, 42). AdoMet interacts with the FeS center in the NrdG subunit, whereby a 5'-deoxyadenosyl radical is transiently formed (7). Subsequent radical transfer between the deoxyadenosyl radical and the glycine at the radical site in NrdD generates the stable glycyl radical. Considering that the Cys  Ser mutants of the C-terminal cluster behave like wild-type NrdD in NrdD-NrdG complex formation and that the activating NrdG is a wild-type enzyme, we suggest that the C-terminal cysteine cluster is directly involved in formation of the glycyl radical, e.g. by participating in radical transfer between the deoxyadenosyl radical and Gly-580. However, at this stage we cannot exclude that the C-terminal cysteine cluster forms a structurally important fold for the radical transfer. The structure of the C-terminal part of T4 NrdD, including the cysteine cluster, was not deducible in the initial electron density map (10), indicating that this part of NrdD may be rather flexible.

To summarize, we have shown that Cys-79 and Cys-290 are essential components of the reaction mechanism of the class III RNR from phage T4, and we propose that Cys-290 forms a transient thiyl radical that initiates the reaction mechanism and that Cys-79 mediates the formate-dependent reduction of the substrate. We have also shown that the four cysteines of the C-terminal cluster, Cys-543, Cys-546, Cys-561, and Cys-564, are required for formation of the glycyl radical. We propose that these residues, which are in a flexible part of the NrdD protein, are involved in radical transfer between AdoMet and Gly-580.

	ACKNOWLEDGEMENTS

We thank Bobby Arash for help with construction of the C-terminal mutants and Maria Lindström for help with expressing them. We also thank Sabrina Bodevin and Ulla-Maja Petersen for scientific discussions and suggestions about this manuscript.

	FOOTNOTES

^* This study was supported by grants from the Swedish Cancer Foundation and the Swedish Foundation for Strategic Research.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Tel.: 46-8-164150; Fax: 46-8-152350; E-mail: bitte@molbio.su.se.

Published, JBC Papers in Press, March 27, 2000, DOI 10.1074/jbc.M001278200

² J. Andersson, M. Sahlin, and B.-M. Sjöberg, unpublished observations.

³ D. Logan, personal communication.

	ABBREVIATIONS

The abbreviations used are: RNR, ribonucleotide reductase; AdoMet, S-adenosylmethionine; DTT, dithiothreitol; EPR, electron paramagnetic resonance; IPTG, isopropyl-1-thio--D-galactopyranoside; NrdD', truncated form of NrdD; PFL, pyruvate formate-lyase; PAGE, polyacrylamide gel electrophoresis.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. T. Logan, E. Mulliez, K.-M. Larsson, S. Bodevin, M. Atta, P. E. Garnaud, B.-M. Sjoberg, and M. Fontecave Bioinorganic Chemistry Special Feature: A metal-binding site in the catalytic subunit of anaerobic ribonucleotide reductase PNAS, April 1, 2003; 100(7): 3826 - 3831. [Abstract] [Full Text] [PDF]

				D. Padovani, E. Mulliez, and M. Fontecave Activation of Class III Ribonucleotide Reductase by Thioredoxin J. Biol. Chem., March 23, 2001; 276(13): 9587 - 9589. [Abstract] [Full Text] [PDF]

				J. Andersson, S. Bodevin, M. Westman, M. Sahlin, and B.-M. Sjoberg Two Active Site Asparagines Are Essential for the Reaction Mechanism of the Class III Anaerobic Ribonucleotide Reductase from Bacteriophage T4 J. Biol. Chem., October 26, 2001; 276(44): 40457 - 40463. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/19449    most recent M001278200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Andersson, J. Articles by Sjöberg, B.-M. Search for Related Content PubMed PubMed Citation Articles by Andersson, J. Articles by Sjöberg, B.-M. Social Bookmarking                      What's this?