Identification of a 5′-Deoxyadenosine Deaminase in Methanocaldococcus jannaschii and Its Possible Role in Recycling the Radical S-Adenosylmethionine Enzyme Reaction Product 5′-Deoxyadenosine

Abstract

We characterize here the MJ1541 gene product from Methanocaldococcus jannaschii, an enzyme that was annotated as a 5′-methylthioadenosine/S-adenosylhomocysteine deaminase (EC 3.5.4.31/3.5.4.28). The MJ1541 gene product catalyzes the conversion of 5′-deoxyadenosine to 5′-deoxyinosine as its major product but will also deaminate 5′-methylthioadenosine, S-adenosylhomocysteine, and adenosine to a small extent. On the basis of these findings, we are naming this new enzyme 5′-deoxyadenosine deaminase (DadD). The K_m for 5′-deoxyadenosine was found to be 14.0 ± 1.2 μM with a k_cat/K_m of 9.1 × 10⁹ M⁻¹ s⁻¹. Radical S-adenosylmethionine (SAM) enzymes account for nearly 2% of the M. jannaschii genome, where the major SAM derived products is 5′-deoxyadenosine. Since 5′-dA has been demonstrated to be an inhibitor of radical SAM enzymes; a pathway for removing this product must be present. We propose here that DadD is involved in the recycling of 5′-deoxyadenosine, whereupon the 5′-deoxyribose moiety of 5′-deoxyinosine is further metabolized to deoxyhexoses used for the biosynthesis of aromatic amino acids in methanogens.

INTRODUCTION

On the basis of sequence comparison, the MJ1541 gene product has been annotated as a 5′-methylthioadenosine/S-adenosylhomocysteine deaminase (EC 3.5.4.31/3.5.4.28) (MtaD) (1) and would thus be expected to catalyze the deamination of the adenine-containing nucleotide products of some S-adenosylmethionine (SAM) radical enzymes. Homologues of this enzyme from Thermotoga maritima, Pseudomonas aeruginosa, and Streptomyces flocculus (1–3) have each been enzymatically characterized using 5′-methylthioadenosine (MTA), S-adenosyl-l-homocysteine (SAH), and adenosine (Ad) as the substrates, but never 5′-deoxyadenosine (5′-dA). Here, we show that the MJ1541 gene encodes a metal-dependent 5′-deoxyadenosine deaminase (DadD) that specifically catalyzes the deamination of 5′-dA (Fig. 1).

FIG 1.

Enzymatic reactions catalyzed by the various enzymes discussed in the present study.

In a recent search of the Methanocaldococcus jannaschii genome for possible radical SAM enzymes, we found that 35 of the 1,811 protein-encoding genes (almost 2% of the genome) contain the canonical CX₃CX₂C motif indicating a possible radical SAM enzyme. Most of these gene products have no annotated function other than being identified as radical SAM enzymes. Radical SAM enzymes are found in many aspects of metabolism, such as the biosynthesis of cofactors, antibiotics, and in the repair of DNA (4). Since a major portion of radical SAM enzymes proceed via a mechanism that leads to the formation of 5′-dA (4), then in an organism with 35 radical SAM enzymes it seems likely that there would be a salvage pathway for the metabolism of the 5′-dA. The presence of a salvage pathway is important because it has been shown that both biotin synthase (BioB) and lipoyl synthase (LipA) are strongly inhibited by 5′-dA (4).

The importance of DadD in the salvage of this radical SAM enzyme reaction product is that in methanogens the MTA/SAH nucleosidase (EC 3.2.2.9), which normally metabolizes these compounds, is absent. The MTA/SAH nucleosidase works to remove adenine from the modified ribose sugar. Without the presence of this enzyme, methanogens instead perform an initial deamination step to salvage the by-products of the radical SAM reactions (Fig. 1).

MATERIALS AND METHODS

Chemicals.

5′-Deoxyadenosine, 2′-deoxyadenosine, 5′-methylthioadenosine, S-adenosylhomocysteine, and adenosine were obtained from Sigma-Aldrich.

Cloning, overexpression, and purification of M. jannaschii MJ1541 gene product in E. coli.

The MJ1541 gene (Swiss-Prot accession number Q58936) was amplified by PCR from genomic DNA using oligonucleotide primers MJ1541-Fwd (5′-GGTCATATGATATTGATAAAAAATG-3′) and MJ1541-Rev (5′-GCTGGATCCTTAGCTTCTCAAAATC-3′). PCR amplification was performed as described previously (5) using a 45°C annealing temperature. Purified PCR product was digested with NdeI and BamHI restriction enzymes and ligated into compatible sites in vector pT7-7. Sequence of the resulting plasmid, pMJ1541, was verified by DNA sequencing. pMJ1541 was transformed into Escherichia coli strain BL21-CodonPlus(DE3)-RIL cells (Stratagene). The transformed cells were grown in Luria-Bertani (LB) medium (200 ml) supplemented with 100 μg of ampicillin/ml at 37°C with shaking until they reached an optical density at 600 nm (OD₆₀₀) of 1.0. Recombinant protein production was induced by addition of lactose to a final concentration of 28 mM (5). After an additional 2 h of culture, the cells (200 ml) were harvested by centrifugation (4,000 × g, 5 min) and frozen at −20°C. Induction of the desired protein was confirmed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of total cellular proteins.

The frozen E. coli cell pellet containing the desired protein (∼0.4 g [wet weight]) was suspended in 3 ml of extraction buffer {50 mM N-[tris(hydroxymethyl)methyl]-2-aminoethanesulfonic acid (TES), pH 7.0; 10 mM MgCl₂; 20 mM dithiothreitol (DTT)} and lysed by sonication. The resulting expressed protein was found to remain soluble after heating the resulting cell extracts for 10 min at 70°C, followed by centrifugation (16,000 × g) for 10 min. This heating step allowed for the purification of the recombinant enzymes from the majority of E. coli proteins, which denature and precipitate under these conditions. The next step of purification was performed by anion-exchange chromatography of the 70°C soluble fractions on a MonoQ HR column (1 by 8 cm; Amersham Bioscience) using a linear gradient from 0 to 1 M NaCl in 25 mM Tris buffer (pH 7.5) over 55 min at a flow rate of 1 ml/min. Fractions of 1 ml were collected, and the desired protein was identified through SDS-PAGE analysis of the fractions. Protein concentrations was determined by Bradford analysis (6).

Recombinant expression of the MJ1541 gene product in the presence of Ni(II), Zn(II), Co(II), and Fe(II).

The metal containing protein was recombinantly produced by overexpression in E. coli grown in LB medium (200 ml), as described above, supplemented with either 1 mM NiCl₂, ZnCl₂, CoCl₂, or Fe(NH₄)₂(SO₄)₂. Expression under these conditions allowed for production of much larger amounts of soluble enzyme, as judged by the intensity of the protein band upon SDS-PAGE analysis of the cell extracts.

Site-directed mutagenesis.

The MJ1541 C294S, Y136R, E150R, and Y136R/E150R mutations were amplified by PCR from genomic DNA using the following oligonucleotide primers: MJ1541-C294S-Fwd, 5′-CTTAGGAACTGATGGAAGTGGAAGTAACAACAAC-3′; MJ1541-C294S-Rev, 5′-GTTGTTGTTACTTCCACTTCCATCAGTTCCTAAG-3′; MJ1541-Y136R-Fwd, 5′-GAGGGCAGTTTTAGCCCGCGGAATGATTGATTTATTTG-3′; MJ1541-Y136R-Rev, 5′-CAAATAAATCAATCATTCCGCGGGCTAAAACTGCCCTC-3′; MJ1541-E150R-Fwd, 5′-GAGGAGAGAAGGGAGAGACGGCTTAAAAATGCTGAGAAG-3′; MJ1541-E150R-Rev, 5′-CTTCTCAGCATTTTTAAGCCGTCTCTCCCTTCTCTCCTC-3′; MJ1541-Y136R/E150R-Fwd, using MJ1541-Y136R as the template and MJ1541-E150R-Fwd as the primer; or MJ1541-Y136R/E150R-Rev using MJ1541-Y136R as the template using MJ1541-E150R-Rev as the primer. Overexpression of the mutant proteins was done as described above.

MALDI-MS identification of SDS-PAGE protein bands.

The protein bands corresponding to the predicted molecular mass of the MJ1541 gene product were excised from the polyacrylamide gel, the cut gel bands were destained with 50 mM ammonium bicarbonate and water (liquid chromatography-mass spectrometry [LC-MS] grade) twice for 3 h. The gel pieces were then dehydrated in acetonitrile (LC-MS grade) for another 5 h. Finally, the acetonitrile was removed by centrifugation. Dehydrated gel pieces were rehydrated on ice using 5 μl of proteomics-grade porcine trypsin at 10 ng/μl in 25 mM ammonium bicarbonate. Once the gel pieces had rehydrated, the digests were incubated approximately 4 h at 37°C. Liquid was collected in the bottom of the tube by centrifugation at 10 × g for 5 min at room temperature. Digests were then incubated in a sonicating water bath for 15 min, and the liquid was again collected at the bottom of the tube by centrifugation. Approximately 1 μl of each digest was spotted onto a stainless steel matrix-assisted laser desorption ionization (MALDI) plate. Once dry the digests were overlaid with 1 μl of matrix solution. The matrix solution consisted of 4 mg of α-cyano-4-hydroxycinnamic acid/ml prepared in a 50:50 mixture of LC-MS-grade solvents acetonitrile and water (Spectrum Chemical) supplemented with 0.1% (vol/vol) trifluoroacetic acid and 10 mM ammonium chloride.

Once the matrix had dried the digests were analyzed by MALDI using an Applied Biosystems (AB) Sciex 4800 MALDI-TOF/TOF mass spectrometer. First, an MS spectrum was obtained for each digest for the m/z range of 800 to 4,000 using the MS reflector positive operating mode. The MS spectra were typically the average of 1,000 laser shots. A data dependent acquisition of the MS/MS spectra for the 10 most intense peaks was then performed utilizing the MS/MS 1-kV positive operating mode. Each MS/MS spectrum was typically the average of 3,500 laser shots. The MALDI software (4000 Series Explorer; AB Sciex) was then used to generate peak lists for searching on the public Mascot server (Matrix Science).

Measurement of native molecular weight of DadD.

The native molecular weight of DadD was determined by size exclusion chromatography as described previously using a Superose 12 HR column (7).

Metal analysis of metal expressed DadD.

Metal analysis of DadD was performed at the Virginia Tech Soil Testing Laboratory using inductively coupled plasma emission spectrophotometry. Instrumentation included a Spectro CirOS Vision made by Spectro Analytical Instruments equipped with a Crossflow nebulizer with a modified Scott spray chamber, nebulizer rate was 0.75 liter/min. A 50-mg/liter yttrium internal standard was introduced by peristaltic pump. Samples were analyzed for cobalt, copper, iron, manganese, nickel, and zinc.

Analysis of DadD activity and DadD mutants.

The enzymatic activity of DadD was determined by incubating 9 ng of the expressed DadD, Y136R, E150R, Y136R/E150R mutants, and each metal expressed DadD with 0.4 mM 5′-deoxyadenosine (5′-dA) in 50 mM Bis-Tris propane (pH 9.0) in a total volume of 100 μl for 10 min at 60°C. The enzymatic reaction was stopped by the addition 5 μl 2 M perchloric acid, followed by centrifugation at 16,000 × g for 5 min. The separated soluble material was then analyzed by high-pressure liquid chromatography (HPLC). To determine whether the activity of DadD is linear with time, a time course experiment was performed. In this experiment, the standard assay was used to measure DadD activity with incubation times of 5, 10, 15, 20, and 30 min.

The chromatographic analysis of 5′-dA and 5′-dI was performed on a Shimadzu HPLC System with a C18 reverse phase column (Varian Pursuit XRs, 250 by 4.6 mm, 5-μm particle size, operated at room temperature) equipped with photodiode array detection. The elution profile consisted of 5 min at 95% sodium acetate buffer (25 mM [pH 6.0], 0.02% NaN₃) and 5% methanol, followed by a linear gradient to 50% sodium acetate buffer–50% methanol over 25 min at 0.5 ml/min. Under these chromatographic conditions the following nucleotides eluted in the order: S-inosylhomocysteine (SIH) at 10.8 min, inosine at 12.3 min, S-adenosylhomocysteine (SAH) at 14.5 min, 5′-dI at 16 min, adenosine (Ad) at 17.2 min, 5′-dA at 20 min, methylthioinosine (MTI) at 22 min, and 5′-methylthioadenosine (MTA) at 26 min. Quantitation was based on absorbance at 260 nm for the adenosine- and 248 nm for the inosine-containing compounds using ε₂₆₀ = 14,900 M⁻¹ cm⁻¹ for adenosine and ε₂₅₀ = 12,300 M⁻¹ cm⁻¹ for inosine (8).

The chromatographic separation of 5′-dA and 5′-dI was also performed on the same Shimadzu HPLC System with a Phenomex Kinetex (C₁₈ reverse-phase column, 100 by 4.6 mm, 2.6-μm particle size) with an elution profile that consisted of 1 min at 98% sodium acetate buffer and 2% methanol, followed by a linear gradient to 70% sodium acetate buffer–30% methanol for 19 min and then another linear gradient to 20% sodium acetate buffer–80% methanol for 5 min, followed by an isocratic flow at 20% sodium acetate buffer–80% methanol for 5 min at 0.5 ml/min. Under these conditions, 5′-dI eluted at 12 min and 5′-dA eluted at 16 min. The quantitation was based on peak areas.

Temperature stability of DadD and mutants.

To determine the temperature stability of the enzymes, each was preincubated for 10 min at 25, 50, 60, and 70°C, and the remaining activity was assayed using the standard assay at 60°C.

Determination of pH optimum for the wild-type DadD.

The pH optimum was determined over the pH range 6.5 to 12.5 using three different buffer systems: 50 mM Bis-Tris propane (pH 6.5 to 9.0), 50 mM diethanolamine (pH 9.0 to 10.0), and 50 mM sodium phosphate (pH 11.0 to 12.5), using the standard assay with 0.2 mM 5′-dA.

Effect of added divalent metals on DadD activity.

To determine whether the activity of DadD was affected by the addition of metals, the enzyme was assayed in the presence of 0.05 mM ZnCl₂, MnCl₂, CoCl₂, NiCl₂, CuCl₂, or Fe(NH₄)₂(SO₄)₂·6H₂O added to the standard assay mixture as aqueous solutions.

Substrate specificity of DadD activity.

To establish the substrate specificity of the native DadD the enzyme and its mutants were assayed with different concentrations of 5′-dA, MTA, SAH, and Ad under the standard assay conditions.

Effect of DTT on DadD activity.

To determine whether DTT had an effect on the activity of the native DadD and C294S mutant the standard assay was performed with the addition of 0.02 M DTT to the assay. This concentration of DTT has been used previously with other metal-dependent DTT-dependent enzymes (7).

Effect of chelating agents on DadD activity.

To determine whether the presence of a chelating agent would have an effect on the native and nickel expressed DadD activity, 25 mM EDTA was added to the standard assay. The native and nickel-expressed DadD were also preincubated for 52 h at room temperature in the presence of 25 mM EDTA prior to the activity assay to establish whether the metal had been removed from the protein by monitoring the activity of the enzyme before and after the incubation. The nickel-expressed DadD was also dialyzed in buffer A [50 mM 2-(N-morpholino)ethanesulfonic acid sodium salt, 1 mM EDTA, 50 mM sodium sulfate, 25 mM dipicolinic acid] for 3 days at room temperature with stirring. The buffer was switched out for fresh buffer A after 24 and 48 h. Buffer B (same as dialysis buffer A but without dipicolinic acid) was exchanged for buffer A after 36 h. Fresh buffer B was exchanged for the old buffer three times at 12-h intervals to ensure the complete removal of dipicolinic acid from sample. The activity of the dialyzed enzyme was then checked according to the standard assay described above. This procedure was modified from that reported by Christopherson and coworkers (9).

Phylogenetic analysis of DadD and its homologues.

The evolutionary history was inferred using the neighbor-joining method (10). The optimal tree with a sum of branch length of 21.00517872 is shown (next to the branches). The evolutionary distances were computed using the Poisson correction method (11) and are in the units of the number of amino acid substitutions per site. The analysis involved 22 amino acid sequences. All positions containing gaps and missing data were eliminated. There were a total of 408 positions in the final data set. Evolutionary analyses were conducted in MEGA5 (12).

RESULTS

Cloning, overexpression, and purification of MJ1541 gene product in E. coli.

The MJ1541 derived protein was highly expressed in E. coli, as measured from the SDS-PAGE analysis of the total proteins in the E. coli cells after expression. MALDI analysis of the tryptic peptides recovered from this spot confirmed the presence of the desired protein. After sonication of the E. coli cells (3 ml) suspended in extraction buffer and centrifugation (14,000 × g) 5 min, SDS-PAGE analysis of the soluble material and the pellet demonstrated that most of the expressed MJ1541 protein was found in the soluble extract. Heating portions of the resulting crude soluble extract for 10 min at different temperatures indicated that DadD precipitated at temperatures above 70°C based on SDS-PAGE analysis. Thus, the first step in the purification of the native enzyme was heating of the sonicated cell extract to 70°C for 10 min prior to purification of the extract on MonoQ. DadD purified by this method showed one peak of activity eluting at 0.42 M NaCl. SDS-PAGE analysis showed the peak was >80% pure.

Recombinant expression of DadD in the presence of Ni(II), Zn(II), Fe(II), or Co(II).

In order to try to increase the amount of soluble protein expressed and alter the proteins metal content, the expression of the protein was conducted in LB medium to which either 1 mM Ni(II), Zn(II), Fe(II), or Co(II) had been added. On the basis of the measured enzymatic activity in the purified enzymes, the total DadD activity was found to be the same when expressed without additional metals or when expressed in the presence of Ni(II). The activity found in the enzymes expressed with added Zn(II) and added Fe(II) were 2-fold higher than when expressed with Ni(II) or without additional metal. The reason for the 2-fold higher activity remains unclear.

Each metal expression produced more soluble protein than with no added metal. This was confirmed by SDS-PAGE analysis of the protein extracts of the cells that showed much more soluble protein in the extract expressed in the presence of added metal. The protein from all of the separate expressions was then purified on MonoQ as described above for subsequent characterization. No cell growth was observed in the Co(II) expression.

Cloning, overexpression, and purification of MJ1541 mutant gene products in E. coli.

The MJ1541 mutants were highly expressed in E. coli as measured from the SDS-PAGE analysis of the total proteins in the E. coli cells after expression. After sonication of the cells suspended in extraction buffer and centrifugation (14,000 × g) 5 min, SDS-PAGE analysis of the pellet and soluble portion of the extract showed that most of the Y136R, E150R, and Y136R/E150R expressed proteins were in the soluble extract. Heating portions of the resulting crude soluble extracts for 10 min at different temperatures, followed by removal of the insoluble proteins by centrifugation, indicated that a small band in the soluble protein fraction, which could be the DadD mutants, precipitated at temperatures of >70°C. Thus, the first step in the purification was heating of the sonicated cell extract to 70°C for 20 min prior to separation of the extract on MonoQ. This separation produced one peak of DadD activity eluting at 0.44 M NaCl for the Y136R mutant, 0.46 M NaCl for the E150R mutant, and 0.46 M NaCl for the Y136R/E150R mutant.

Y136 and E150 were selected for site-directed mutagenesis because of the X-ray structure of the T. maritima enzyme (1) identified two arginines in the position of the Y136 and E150 of the M. jannaschii sequence. These residues were identified in the X-ray structure to be involved in binding the substrate, particularly in the stabilization of the carboxyl moiety of SAH. We wanted to use site-directed mutagenesis to determine whether by changing the Tyr and Glu to Arg we could improve DadD's ability to deaminate SAH by increasing DadD's ability to hydrolyze this substrate.

Measurement of native molecular mass of DadD.

The molecular mass of DadD was measured to be ∼230 kDa using size exclusion chromatography. This is consistent with the enzyme being a tetramer with a monomeric subunit mass of 57.5 kDa (data shown in Fig. S1 in the supplemental material).

Temperature stability of wild-type DadD and mutants.

The temperature stability of the DadD was found to be different for the purified wild-type and the DadD mutants. The wild-type and the C294S mutant strains were found to be stable after heating for 10 min at 60°C. The Y136R and E150R DadD mutants were found to be stable for only 10 min at 50°C, showing a significant loss of activity at 60°C (Table 1). The Y136R/E150R DadD double mutant showed the most significant change in temperature stability, only being stable at 25°C. The wild-type DadD, metal expressed, and C127S mutant were assayed at 60°C, the Y126R and E150R mutants were assayed at 50°C, and the double mutant was assayed at 25°C for the standard assay.

TABLE 1.

Temperature stability of purified wild-type and DadD mutant strains

DadD	Temp (°C)	% of retained activitya
WT	60	90
	70	34
Y136R	50	80
	60	40
E150R	50	96
	60	25
Y136R & E150R double mutant	25	100
	50	35

^aThat is, after heating for 10 min at the indicated temperature.

pH optimum of wild-type DadD.

The pH optimum of the DadD was found to be pH 9.0 in the Bis-Tris propane buffer (see Fig. S2 in the supplemental material). At a pH of >11.0, the enzyme no longer showed measurable activity due to denaturation of the enzyme. Denaturation of the enzyme was established by changing the pH of the MonoQ-purified fraction from 7.5 to 11 by the addition of 2.2 μl of 50 mM sodium phosphate (pH 12.5) to 13 μl of a solution of the purified DadD (0.6 μg). An assay done on the resulting sample at pH 9 showed no enzymatic activity, which lead us to conclude that after exposure to pH 11.0 DadD denatures.

Dependence DadD activity on added metals.

The activity of DadD was tested with several different divalent cations. The addition of metal to the assay had no effect on DadD activity, indicating that the catalytic metal(s) is already coordinated in the active site.

Effect of chelating agents on DadD activity.

The addition of EDTA to the standard assay stimulated the activity of DadD by 20%. This result is a similar to that seen when urease was incubated with EDTA (13) and indicates that the catalytic metal is not removed by EDTA. The stimulation could be a result of the removal of bound metal that is present on the enzyme that was causing inhibition (14). Dialysis of the wild-type DadD for 3 days against a dipicolinic acid containing buffer also failed to show any lowering enzyme activity, again indicating that chelation is not able to remove the bound metal in the active site.

DadD activity with alternate substrates.

DadD is annotated as a MTA/SAH deaminase so we tested its activity to deaminate 5′-dA, MTA, SAH, and Ad as the substrates using the purified enzyme and each mutant. The results of the kinetic analysis for each of these are summarized in Table 2. The results clearly showed that the preferred substrate is 5′-dA, where the catalytic efficiency of the enzyme for 5′-dA is 10³-fold higher than for MTA, and 10⁴-fold higher than with SAH or Ad (Table 2).

TABLE 2.

Kinetic parameters of DadD and mutants with 5′-dA, MTA, SAH, and Ad

DadD	Substrate	Mean Km (mM) ± SD	kcat (s−1)	kcat/Km (M−1 s−1)
WT	5′-dA	0.014 ± 0.0012	1.2 × 105	9.1 × 109
	MTA	0.11 ± 0.028	1.6 × 102	1.1 × 106
	SAH	1.1 ± 1.3	1.3 × 103	4.4 × 106
	Ad	0.15 ± 0.0047	1.1 × 102	7.5 × 105
Y136R	5′-dA	0.032 ± 0.033	3.3 × 102	1.7 × 107
	MTA	0.022 ± 0.0019	2.2 × 102	1.1 × 107
	SAH	0.16 ± 0.13	24.7	2.3 × 105
	Ad	0.072 ± 0.0018	1.4 × 102	2.9 × 106
E150R	5′-dA	0.021 ± 0.0061	2.1 × 102	9.8 × 107
	MTA	0.048 ± 0.0029	2.0 × 102	4.2 × 106
	SAH	3.9 ± 5.2	58.2	2.1 × 104
	Ad	0.15 ± 0.047	1.1 × 102	7.5 × 105
Y136R/E150R	5′-dA	0.05 ± 0.046	1.12	3.1 × 104
	MTA	0.073 ± 0.0054	1.39	1.9 × 104
	SAH	0.4 ± 0.1	2.41	5.3 × 103
	Ad	NMa	NM	NM

^aNM, not measurable.

Effect of DTT on the DadD activity.

The effect of DTT on the activity of DadD was tested by the addition of 0.02 M DTT in the standard assay. This experiment was performed as a result of there being four conserved cysteines in the amino acid sequences of most of the enzymes from the methanogens, and it was possible that their oxidation to disulfides during the isolation of the enzyme could have altered the activity. The presence of 0.02 M DTT had no effect on the enzymatic activity.

Phylogenetic analysis of DadD and its homologues.

Phylogenetic analysis of the DadD homologues showed they are a widely distributed among archaea and bacteria (Fig. 2) and are all annotated as MtaD enzymes. Figure 2 shows that there are three major branches. The M. jannaschii DadD is located in the lower branch (Fig. 2, bottom bracket) of the tree along with the Thermotoga maritima homolog, with a sequence identity of 38% to M. jannaschii. Only two other methanogens, M. infernus and M. fervens belong in the same branch. In the top branch (Fig. 2, top bracket) contains the P. aeruginosa MTAD enzyme that was found to be specific for MTA (15) and not SAH or Ad, the middle branch (Fig. 2, middle bracket) contains the S. griseus SAHD which was found to be specific for SAH and not MTA or Ad (2). This could represent three subfamilies that have various degrees of substrate specificity, but it is likely that all will catalyze the deamination of 5′-dA, MTA, SAH, and Ad to some degree of efficiency. The specificity of the enzymes may have evolved from the availability of 5′-dA, MTA, or SAH that is present in each subfamily of MTA/SAH deaminases. All of the homologs in the phylogenetic tree contain a sequence identity to M. jannaschii DadD ranging from 20 to 92%.

FIG 2.

Evolutionary relationships of 5′-methylthioadenosine/S-adenosylhomocysteine deaminases to M. jannaschii DadD. MtaD represents the homologs that do not have confirmed gene annotation; those confirmed are MTAD, SAHD, MtaD, and DadD. The organisms in the top bracket represent those that have the closest homology to MTAD and likely use MTA as their preferred substrate. The organisms in the middle bracket share homology with SAHD and likely use SAH as their preferred substrate. Those in the bottom bracket share homology with DadD of M. jannaschii (highlighted in the box), and we propose the use of 5′-dA as their preferred substrate. The accession numbers for the proteins used in the construction of the tree in order from top to bottom are as follows: B6YUF8, Q5JER0, O66851, A6UUG9, I9MDY4, O27549, Q4K8M5, A6V2Q5, G0PU20, M1Z1T6, Q9KC82, D3DYL9, Q8TYD4, Q2RJW1, B1BCH5, A4FW32, F2KTB3, Q8U0P7, D5VT65, Q9X034, Q58936, and C7P5A2.

DISCUSSION

5′-Deoxyadenosine deaminase (DadD) was originally annotated as a 5′-methylthioadenosine/S-adenylylhomocysteine deaminase (MtaD, EC 3.5.4.31/3.5.4.28) (16). At this time only the enzymes from Thermotoga maritima (1), Pseudomonas aeruginosa (3), and Streptomyces flocculus (17) have been studied. None of these enzymes were ever tested with 5′-deoxyadenosine (5′-dA) as the substrate. The enzyme from T. maritima has had its X-ray crystal structure determined and was shown to have a nickel ion bound to the enzyme at the catalytic site (PDB 1p1m), however, subsequent work done with excess ZnCl₂ present during the crystallization showed the protein to have Zn(II) in the active site during analysis of its structure and activity (1).

In most cases the presence of 5′-methylthioadenosine/S-adenosylhomocysteine nucleosidase (MtaN) is responsible for maintaining low levels of MTA so that the radical SAM enzymes can remain active (18). MtaN is an enzyme that is known to be involved in the recycling of MTA produced from SAM during the biosynthesis of the polyamines, spermidine, and spermine (19). MtaN works by catalyzing the first step of the methionine salvage pathway (20), where MTA is converted to 5′-methylthioribose-1-P. It could also just as well function where the MTA is deaminated into 5′-methylthioinosine before it undergoes phosphorolysis to 5′-methylthioribose-1-P. Here we propose that the same series of reactions are occurring, but the product is 5′-deoxyribose-1-P, from 5′-dI, which is converted into a precursor for aromatic amino acid biosynthesis instead of methionine.

Determination of DadD substrate specificity and amino acid residues involved in substrate binding.

Substrate specificity was determined by comparing the catalytic efficiency of the wild-type enzyme with each of the four substrates indicated in Table 2. The substrate binding was determined by incubating the wild-type and mutant DadD with various concentrations of the different substrates and measuring the catalytic efficiency of DadD for each substrate. Our results showed that 5′-dA was the preferred substrate for the enzyme, where the catalytic efficiency (k_cat/K_m) of the enzyme for 5′-dA was 10³-fold higher than for MTA and 10⁴ fold higher than with SAH or Ad (Table 2). These data indicate that the in vivo substrate is likely 5′-dA, the major product of many radical SAM enzymes (4).

The X-ray structure from the T. maritima MtaD gave us insight as to what residues may be important in binding and stabilizing the different substrates in the active site. The X-ray crystal structure showed that two arginines (Fig. 3, green) were involved in stabilizing the carboxylate moiety of SAH (1). Sequence alignments of the T. maritima MtaD and other homologous enzymes, including DadD, showed that these two arginines were not highly conserved, in DadD the residues in the position of the two arginines are completely different, this led us to postulate that these arginines maybe what imparts the T. maritima MtaD to have an increased ability to deaminate SAH in comparison with DadD. The K_m for SAH from T. maritima MtaD is 210 μM, which is lower than that of the M. jannaschii (1,100 ± 1,300 μM), indicating that the presence of the two arginines in T. maritima may play an important role in binding of SAH (Fig. 3). The K_m of the wild-type DadD for SAH is much higher than for the other substrates (Table 2), indicating that the loss of the two arginines in M. jannaschii DadD (Fig. 3, green) diminished its ability to deaminate SAH, because of DadD's inability to stabilize the carboxylate moiety of the SAH in the active site. The single mutants in which Y136 and E150 were replaced by arginine and the double mutant Y136R/E150R did not show an increase in the ability to bind or deaminate SAH (Table 2). Interestingly, the k_cat/K_m values for each of the single mutants of DadD showed that 5′-dA and MTA were the preferred substrates (Table 2). Surprisingly, the double mutant Y136R/E150R failed to have any activity toward Ad (Table 2).

FIG 3.

Sequence alignment of DadD from M. jannaschii with homologs from other methanogens, as well as with previously characterized enzymes. The residues are highlighted in the following way: a cyan background with black lettering indicates residues coordinating the metal, a blue background with white letters indicates residues interacting with the ribose moiety of the substrate, a pink background with black letters indicates residues interacting with the purine base, a purple background with white letters indicates residues involved in forming a hydrophobic pocket to bind the 5′-methylthio moiety of MTA, a gray background with black letters indicates residues interacting with the 5′-methylthio moiety of MTA, and the arginines residues interacting with the SAH carboxylate are highlighted by a green background with black letters and are underlined. M. jannaschii numbering is used.

Taken together, the data indicate that the two arginines stabilizing the carboxylate moiety of SAH are not the only residues involved, as previously believed for T. maritima (1). This is consistent with the data reported for the MtaD from P. aeruginosa, which showed no activity with SAH as a substrate, and it, too, lacks two arginines in the same position as the arginines from T. maritima (Fig. 3) (3). Determining the kinetic parameters of the Y136R, E150R, and Y136R/E150R mutants, the data indicate that the stabilization of the substrates, especially SAH, is much more complicated than previously suggested for the T. maritima MtaD based on X-ray analysis of the active site with SAH bound.

Characterization of the metal dependence of the M. jannaschii DadD.

The amidohydrolase superfamily is known to have a metal bound in the active site that is responsible for generating the nucleophilic water required for the deamination reaction (21). In MtaD a mononuclear metal center is likely coordinated by three histidines and an aspartate (Fig. 3, cyan) as is seen in all DadD homologs. The original crystal structure from T. maritima showed that the metal was nickel; however, upon further characterization of the T. maritima enzyme, it was cocrystallized with Zn(II) (1). The MTA deaminase (MTADA) from P. aeruginosa was also shown to use zinc for its catalytic metal (3). The catalytic metal in SAH deaminase from S. flocculus was not identified because the authors were unable to remove any metal through the use of several different chelating agents. From this work, they concluded that the SAH deaminase did not use a specific divalent cation (2). We were also unable to remove the metal from DadD or establish the nature of the catalytic metal (see the supplemental data).

Proposed route for reincorporation of 5′-dA into aromatic amino acids.

The conversion of the SAM radical product, 5′-dA, into the deoxyhexose used for aromatic amino acid biosynthesis would first proceed through the deamination of 5′-dA to 5′-deoxyinosine (5′-dI). Next, 5′-dI would undergo phosphorolysis to make hypoxanthine and 5′-deoxyribose-1-phosphate (5-dR-P) (Fig. 4). This step has been confirmed to be catalyzed by the MJ0060 gene product (R. H. White, unpublished data). The MJ0060 gene product is annotated as a 5′-methylthioadenosine phosphorylase (MTAP) which shares 47% sequence identity to the characterized MTAP from Pyrococcus furiosus. The P. furiosus MTAP was found to catalyze the phosphorolysis of MTA and Ad, but it could also catalyze the phosphorolysis of inosine with a k_cat/K_m of 9.74 × 10³ M⁻¹ s⁻¹ (22). Recently, a 5′-methylthioinosine phosphorylase (MTIP) from Pseudomonas aeruginosa has been identified and characterized (15). Interestingly, MTIP was found to be competitively inhibited by 5′-methylthioadenosine, the annotated substrate. The k_cat/K_m was found to be (1.8 ± 0.3) × 10⁶ M⁻¹ s⁻¹ for MTI with inosine being the only other substrate, but with 300-fold lower catalytic efficiency (15). However, the sequence identity between MTIP and the MJ0060 gene product is only 39%, lower than that of the MJ0060 gene product with MTAP (47%). We are currently studying the substrate specificity of the MJ0060 gene product.

FIG 4.

Proposed route for the incorporation 5′-dA into deoxyhexoses for the biosynthesis of aromatic amino acids in Methanocaldococcus jannaschii.

The canonical aromatic amino acid starts with the reaction of phosphoenolpyruvate (PEP) with erythrose 4-phosphate (E-4-P), a product the pentose phosphate pathway, to produce 3-deoxy-d-arabino-heptulosonate-7-phosphate catalyzed by 3-deoxy-d-arabino-heptulosonate 7-phosphate synthase (23). Not only are the genes encoding for these enzymes absent in archaeal genomes, so is the canonical pentose phosphate pathway (24). Without the pentose phosphate pathway, E-4-P is unable to be produced necessitating methanogens to have another method for the production of ribose-5-phosphate and/or E-4-P. The archaeal biosynthetic route for the production of 3-dehydroquinate (DHQ), the second intermediate in the shikimate pathway, is also different than the typical biosynthesis of DHQ in bacteria and eukaryotes. The archaeal pathway begins with the condensation of 6-deoxy-5-keto-fructose-1-phosphate (DKFP) and l-aspartate semialdehyde (25). A transaldolase reaction transfers the intact C₃ hydroxyacetone fragment (carbons 4, 5, and 6) to the l-aspartate semialdehyde, forming 2-amino-3,7-dideoxy-d-threo-hept-6-ulosonic acid. This amino sugar then undergoes an NAD-dependent oxidative deamination to produce 3,7-dideoxy-d-threo-hept-2,6-diulosonic acid, which cyclizes to DHQ (25). DKFP's biosynthesis was proposed to arise via transaldolase reaction between methylglyoxal and fructose-1,6-bis-P (26) with the methylglyoxal arising from lactaldehyde (27). We propose that the origin of the methylglyoxal and/or DKFP arises from the 5′-deoxyribose-1-P the product of the MJ0060 gene.

The hypoxanthine product could be incorporated into IMP by reaction with phosphoribosyl pyrophosphate catalyzed by a hypoxanthine phosphoribosyltransferase. The 5-deoxyribose 1-phosphate (5-dRP) product could be incorporated into the deoxyhexoses required for aromatic amino acid biosynthesis without coming from the traditional pentose pathway shunt. This could involve an isomerase (MJ0454 gene product) that is annotated to use the 5′-methylthio derivative of 5-dRP to produce 5-deoxyribulose-1-phosphate typically involved in the methionine salvage pathway (28, 29). This compound could be further metabolized by several yet to be identified enzymes to produce 6-deoxy-5-ketofructose, the required precursor for aromatic amino acid (Fig. 4). The identification of the enzymes and pathway for the biosynthesis of deoxyhexoses is currently being studied.

Conclusions.

In conclusion we have shown that the annotated 5′-methylthioadenosine/S-adenosyl-l-homocysteine deaminase MJ1541 gene product from M. jannaschii preferentially uses 5′-deoxyadenosine as its substrate compared to 5′-methylthioadenosine, S-adenosylhomocysteine, and adenosine. From this analysis we rename this enzyme 5′-deoxyadenosine deaminase (DadD). Here, we also suggest that other members of this family may be using 5′-deoxyadenosine as a preferred substrate as well, based on phylogenetic analysis with other homologs. Unfortunately, we were unable to fully identify the in vivo catalytic metal, but based on our data we propose it is either Ni(II) or Zn(II).