Homogeneous adenine deaminases (EC 3.5.4.2) from the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe and a putative ADA (adenosine deaminase; EC 3.5.4.4) from Arabidopsis thaliana were obtained for the first time as purified recombinant proteins by molecular cloning of the corresponding genes and their overexpression in Escherichia coli. The enzymes showed comparable molecular properties with well-known mammalian ADAs, but exhibited much lower kcat values. Adenine was the most favoured substrate for the yeast enzymes, whereas the plant enzyme showed only very low activities with either adenine, adenosine, AMP or ATP. Interestingly, the yeast enzymes also hydrolysed N6-substituted adenines from cytokinins, a group of plant hormones, cleaving them to inosine and the corresponding side chain amine. The hydrolytic cleavage of synthetic cytokinin 2,6-di-substituted analogues that are used in cancer therapy, such as olomoucine, roscovitine and bohemine, was subsequently shown for a reference sample of human ADA1. ADA1, however, showed a different reaction mechanism to that of the yeast enzymes, hydrolysing the compounds to an adenine derivative and a side chain alcohol. The reaction products were identified using reference compounds on HPLC coupled to UV and Q-TOF (quadrupole–time-of-flight) detectors.

The ADA1 activity may constitute the debenzylation metabolic route already described for bohemine and, as a consequence, it may compromise the physiological or therapeutic effects of exogenously applied cytokinin derivatives.

INTRODUCTION

The aminohydrolases adenine deaminase (ADE; EC 3.5.4.2), which catalyses the irreversible deamination of adenine to hypoxanthine, and ADA (adenosine deaminase; EC 3.5.4.4), which catalyses the irreversible deamination of adenosine to inosine, are enzymes that are responsible for the metabolic salvage of purines. Several subclasses of these enzymes have been already described and, with the knowledge of the full genome sequences of many organisms, it is possible to identify encoding genes and group the enzymes according to primary structure [1,2]. Fungal adenine deaminases share a relatively high sequence homology with both prokaryotic and eukaryotic ADAs that constitute a known family of α/β barrel enzymes [3]. Prokaryotic adenine deaminases form a specific group that differs structurally and evolutionarily from the fungal adenine deaminases.

The natural substrates of these enzymes are adenine and adenosine respectively, but the enzymes can also hydrolyse 6-chloro-substituted derivatives. ADA can also convert other purine compounds, such as 2-amino-6-chloropurine riboside [4], 6-methoxypurine riboside [5], 6-methylaminopurine ribonucleoside [6], as well as deoxyribose derivatives, AMP, ADP, ATP [7] and cAMP [8]. Metal ions in specific concentrations are, in some cases, essential for activity [9,10].

Proteins that exhibit ADA activity include not only ADA, but also ADA regulatory proteins (only in prokaryotes) with different conserved structural domains (EC 2.1.1.63) and ADAT (tRNA-specific ADA). ADAT was the first prokaryotic RNA-editing enzyme to be identified in Escherichia coli [11]. ADAR (ADA acting on RNA) has the ability to deaminate adenosines in long double-stranded RNA and convert them into inosines. These enzymes are commonly found in animals, but are not known in other organisms [12].

Two different isoenzymes of ADA, ADA1 and ADA2, are found in higher eukaryotes [13] and are encoded by different genes [14]. In humans, almost all ADA activity is attributed to the single-chain Zn2+-binding protein ADA1, whereas ADA2 is found in negligible quantities in serum and may be produced by monocytes [15]. ADA1 is expressed in all human tissues, with its activity levels being relatively high in the thymus and duodenum (approx. 10 nkat/mg), whereas the activity of ADA1 is more than 500-fold lower in liver [16].

Adenine and adenosine deaminases are constitutive components of purine metabolism and their impairment may cause serious disorders [17]. In humans, ADA deficiency has been linked to severe combined immunodeficiency and as a result ADA was approved for the first gene therapy trial [17]. Lack of ADA1 activity leads to an accumulation of dATP that causes inhibition of the activity of ribonucleotide diphosphate reductase, which is the enzyme that synthesizes the DNA and RNA required during lymphocyte proliferation. On the other hand, mutations leading to overexpression of ADA1 cause haemolytic anaemia. There is also some evidence that mutations on the other allele (ADA2) may lead to autism [18].

The crystal structure of ADA has been solved for the murine [19] and bovine [20] enzymes. The protein is folded as an eight-stranded parallel α/β barrel with a deep pocket at the β-barrel C-terminal end where the active site is formed from glutamic acid and aspartic acid residues and contains a Zn2+ ion.

There are only a few reports on ADAs in plants. Although the completed genome databases of Arabidopsis thaliana and other plants show the presence of putative encoding genes, the corresponding proteins have not been obtained so far nor has activity towards adenosine been demonstrated. The physiological role of these enzymes in plants is unclear, since no activity was detected in plant extracts. Some articles even speculate that ADA is not present at all in plant tissues [2123] or that the enzyme activity is too low for efficient adenosine recycling that might be instead controlled by adenosine kinase (EC 2.7.1.77) [24].

In the present paper, we describe the cloning and functional expression of three adenine and adenosine deaminase genes: AAH1 from Saccharomyces cerevisiae (encoding the protein Aah1p); SPBC1198.02 from Schizosaccharomyces pombe; and At4g04880 from A. thaliana. The genes from yeasts have already been cloned and annotated to encode adenine deaminases of the same subfamily as the enzyme from Aspergillus nidulans that was obtained as a recombinant protein and characterized [3]. The gene from A. thaliana was identified by a database search (annotated as an adenosine/AMP deaminase family protein). All three enzymes were obtained as recombinant proteins, purified to homogeneity and were shown to hydrolyse adenine compounds. Interestingly, the yeast enzymes also exhibited a low activity towards N6-substituted adenines that are commonly known as the plant hormones cytokinins. In plants, cytokinins act via specific plasma-membrane receptors within a two-component signalling system and regulate numerous developmental and physiological processes, including apical dominance, flower and fruit development, leaf senescence and seed germination [25]. They are often used as exogenous additives to control plant micropropagation. Some other cytokinin derivatives exhibit anticancer properties as a result of their ability to inhibit cyclin-dependent kinases [26] or are used in skin-protective cosmetics as they delay aging of human fibroblasts [27] and protect against oxidative damage of DNA [28]. In the present study, we found that a reference sample of human ADA1 displayed hydrolytic activity towards some of these compounds.

MATERIALS AND METHODS

Chemicals, vectors, enzymes and biological material

N6-(2-isopentenyl)adenine 9-glucoside, trans-zeatin {6-[(E)-4-hydroxy-3-methylbut-2-enylamino]purine}, trans-zeatin 9-riboside, trans-zeatin 9-glucoside, m-topolin [6-(3-hydroxybenzyl)-aminopurine], kinetin (6-furfurylaminopurine), olomoucine [6-benzylamino-2-(2-hydroxyethylamino)-9-methylpurine], olomoucine II {6-(2-hydroxybenzylamino)-(2R)-[1-(hydroxymethyl)propyl]amino-9-isopropylpurine}, roscovitine {6-benzylamino-2-[1-(hydroxymethyl)propyl]amino-9-isopropylpurine} and bohemine [6-benzylamino-2-(3-hydroxypropylamino)-9-isopropylpurine] were from OlChemIm (Olomouc, Czech Republic). N6-methyl-isopentenyladenine that had been synthesized according to a protocol published previously [29] was donated by Dr Kristin Bilyeu (Department of Plant Sciences, University of Missouri, Columbia, MO, U.S.A.). The cyclin-dependent kinase inhibitors bohemine, roscovitine, olomoucine and olomoucine II were further purified by HPLC to remove trace impurities on a System Gold HPLC (Beckman Coulter) equipped with a Symmetry C18 column (150×2.1 mm, 5 μm; Waters) in 15 mM ammonium formate (pH 4.0), with a linear gradient of methanol from 10–90% in 25 min followed by a 5 min isocratic elution at a flow rate of 0.3 ml/min at 25°C.

6-Amino-2-(3-hydroxypropylamino)-9-isopropylpurine and 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine, which were synthesized following a protocol published previously [30], were provided by Dr Libor Havlíček [Institute of Experimental Botany, Academy of Sciences of the Czech Republic (ASCR), Prague, Czech Republic].

PCR fragment-isolation kits, plasmid purification kits and Ni2+–NTA (Ni2+-nitrilotriacetate) agarose were from Qiagen. The protein assay kit used to assess protein concentration was from Bio-Rad. Advantage DNA polymerase used for cloning into the pDrive vector (Qiagen) was from Clontech and the blunt-end-generating proofreading polymerase Phusion™ used for cloning into pET100/D-TOPO and pET151/D-TOPO vectors (Invitrogen) was obtained from Finnzymes. ADA from human erythrocytes (ADA1) was a certified reference material (BCR-647) purchased from Sigma (catalytic activity 2.55 μkat/l [31]). All other chemicals were from Sigma.

Cloning of the yeast adenine deaminase genes

The S. cerevisiae strain used was 23344c (MATα ura3), which is derived from the wild-type strain S1278b [32]. S. pombe wild-type 972 h- was provided by Dr Marie Kopecká (Department of Biology, Faculty of Medicine, Masaryk University, Brno, Czech Republic). Fractions of genomic DNA from S. pombe and S. cerevisiae were isolated according to a protocol published previously [33].

The gene SPBC1198.02 from S. pombe (GeneID: 2540066, no introns) was amplified using Advantage DNA polymerase (annealing at 58°C, 30 cycles of PCR) and oligonucleotide primers 5′-GGAGCGATATTGTGGGTGAT-3′ (forward) and 5′-TTGATTAAGCTTGAACTTCCACAG-3′ (reverse), cloned into the pDrive vector by U/A cloning and replicated in chemically competent One Shot® TOP10 E. coli cells (Invitrogen) according to the manufacturer's protocol.

The gene was then cloned into the expression vectors pET100/D-TOPO (fusing a 6×His tag and an Xpress™ epitope to the N-terminus of the cloned gene product) and pET151/D-TOPO (fusing a 6×His tag and a V5 epitope to the N-terminus of the cloned gene product) using the forward primer 5′-CACCATGAGCAATCTACCTA-3′ (for both vectors), the reverse primers 5′-AGCTTGAACTTCCACAG-3′ (pET100/D-TOPO, with the gene-specific stop codon removed) or 5′-TTAAGCTTGAACTTCCACAG-3′ (pET151/D-TOPO, with the gene-specific stop codon maintained), and the blunt-end-generating proofreading polymerase Phusion™ (annealing at 55°C, 30 cycles of PCR). The insert region of the construct was fully sequenced to confirm the gene identity and the vectors were then used to transform competent cells BL21 Star™(DE3) One Shot® Chemically Competent E. coli cells (Invitrogen) according to the manufacturer's protocol.

The AAH1 gene from S. cerevisiae (GeneID: 855581, no introns) was cloned first into the pDrive vector using touchdown PCR [annealing at 64–58°C, 30 cycles of PCR, addition of 5% (v/v) Triton X-100] with the primers 5′-CGAAACTGCACTGAAATGGCGCA-3′ (forward) and 5′-CGTGAAATACAAAAGGTCCAGC-3′ (reverse) and then cloned into pET100/D-TOPO and pET151/D-TOPO [annealing at 56°C, 30 cycles of PCR, addition of 5% (v/v) Triton X-100] with the forward primer 5′-CACCATGGTTTCTGTGGAGT-3′ and the reverse primers 5′-ATGCGAATATTTAGTGACTACTT-3′ (pET100/D-TOPO) or 5′-CTAATGCGAATATTTAGTGACTACTTCGT-3′ (pET151/D-TOPO) by analogous procedures as used above for S. pombe.

Cloning of a putative ADA gene from A. thaliana

Total RNA was isolated from 1-week-old A. thaliana ecotype Col0 seedlings using TRIzol (Invitrogen) and reverse transcribed using RevertAid™ H Minus M-MuLV reverse transcriptase and an oligo(dT)18 primer (Fermentas).

From the cDNA, At4g04880 (GeneID: 825826, 10 introns) was amplified and cloned into the pDrive vector [annealing at 57°C, 30 cycles of PCR, addition of 10% (w/v) betain] with the primers 5′-ACAAAAAAAAAATGGAATGGATACAAT-3′ (forward) and 5′-ATTACAATCAATATTCGAGATGAATGTTAT-3′ (reverse), and then into pET100/D-TOPO and pET151/D-TOPO vectors (annealing at 60°C, 30 cycles of PCR) with the forward primer 5′-CACCATGGAATGGATACAATCACTG-3′ and the reverse primers 5′-AACGTGCTCTGGCGAGGC-3′ (pET100/D-TOPO) or 5′-CTAAACGTGCTCTGGCGAG-3′ (pET151/D-TOPO) by the same procedure as described for the yeast adenine deaminase genes.

Expression and purification of recombinant proteins

Protein expression in E. coli BL21 Star™(DE3) cells was induced by 0.2 mM IPTG (isopropyl β-D-thiogalactoside) using a cell culture of an attenuance (D) of 0.6 at 600 nm maintained at 18°C overnight in LB (Luria–Bertani) medium containing 1% glucose and 100 μg/ml of antibiotics (ampicillin for SPBC1198.02 and carbenicillin for Aah1p and At4g04880).

Recombinant proteins were purified on Ni2+–NTA agarose (Qiagen) following the manufacturer's protocol with modifications for purification of 6×His-tagged proteins from E. coli under native conditions. Glycerol was added as a stabilizing agent to all buffers used throughout the purification, i.e. lysis buffer [50 mM NaH2PO4 (pH 8.0), 300 mM NaCl, 10 mM imidazole, 0.5% Triton X-100 and 10% (v/v) glycerol], washing buffer (same composition as the lysis buffer without Triton X-100) and elution buffer [50 mM NaH2PO4 (pH 8.0), 300 mM NaCl, 300 mM imidazole and 20% (v/v) glycerol].

Cells overexpressing the studied enzymes were harvested by centrifugation at 5000 g for 10 min, resuspended in lysis buffer to a volume of 5% (for SPBC1198.02) or 2.5% (for Aah1p and At4g04880) of the original culture volume. The cells were placed in 1.5 ml Eppendorf tubes and disrupted by three cycles of repeated freezing in liquid nitrogen and thawing at 42°C.

Conditions for protein purification were optimized for 1.2 ml of cell extract per tube for each protein as follows: SPBC1198.02, binding with 0.3 ml of 50% (v/v) Ni2+–NTA slurry, elution with 0.15 ml of elution buffer; Aah1p and At4g04880, binding with 0.15 ml 50% (v/v) Ni2+–NTA slurry, elution with 0.075 ml of elution buffer. The recombinant proteins were concentrated using an Amicon Ultra-4 centrifugation device with a 10 kDa cut-off (Millipore) to a concentration of approx. 30 mg/ml and stored at 4°C. Protein purity was checked by SDS/PAGE (10% gels) in Tris/glycine running buffer [34]. Before separation, samples were heated to 95°C for 10 min in the presence of 1% SDS and 1% 2-mercaptoethanol. PageRuler™ unstained protein ladder (Fermentas) was used as a molecular-mass marker. Gels were stained for proteins with Coomassie Brilliant Blue R-250.

Enzyme activity assay

Interactions of the enzymes with substrates were monitored by absorption changes in the range of 200–500 nm using an Agilent (HP) 8345 diode array spectrophotometer (Agilent Technologies) in a temperature-controlled cell with magnetic stirring. Molar absorption coefficients of the substrates and products were determined experimentally from reference compounds and used for the calculations of product concentration as shown below for adenine and adenosine. The assay method is based on a protocol for adenine deaminase published previously [35].

Adenine deaminase activity was assayed as an increase in the concentration of hypoxanthine (calculated from its absorption at 240 nm, εHyp240=8850 M−1·cm−1), with 0.067 mM adenine as a substrate, in 0.2 M potassium phosphate buffer (pH 6.7) at 33°C. The reaction was started by the addition of substrate. Adenine itself has an absorption maximum at 260 nm (εAde265 =11050 M−1·cm−1) (where Ade is adenine) and thus it contributes to the total absorption at 240 nm (εAde240=5850 M−1·cm−1). The concentration of a product released from the enzymatic reaction was therefore calculated as [P]=ΔA240/(εP240−εS240), where ΔA240 is the linear increase in absorbance during a given time period, εP240 is the molar absorption coefficient of the product and εS240 is the molar absorption coefficient of the substrate, all at 240 nm.

The activity of ADA was determined as an increase in inosine concentration at 240 nm (εIno240=12930 M−1·cm−1, where Ino is inosine) using the same method as stated above [0.067 mM adenosine as substrate and 0.2 M potassium phosphate buffer (pH 7.0) at 37°C], with a correlation with adenosine absorption at 240 nm (εAdo240=5740 M−1·cm−1, where Ado is adenosine).

The activity in the presence of AMP/ATP was assayed as an increase in IMP (inosine monophosphate)/ITP (inosine triphosphate) concentrations at 240 nm (εIMP240=6670 M−1·cm−1 and εITP240=6060 M−1·cm−1) as stated above, using εS240 determined experimentally: AMP=3390 M−1·cm−1 and ATP =3080 M−1·cm−1.

The activity with cytokinins was assayed by determination of hypoxanthine production (or inosine for ribosides) by the same method as stated above with the following εS240 values determined experimentally: N6-isopentenyladenine =3920 M−1·cm−1, N6-isopentenyladenosine=3780 M−1·cm−1, N6-isopentenyladenine 9-glucoside=3840 M−1·cm−1, N6-methylisopentenyladenine=3080 M−1·cm−1, trans-zeatin =7900 M−1·cm−1, cis-zeatin=7680 M−1·cm−1, benzyladenine=5070 M−1·cm−1, m-topolin=2020 M−1·cm−1 and kinetin=4150 M−1·cm−1.

The activity of human ADA with the cyclin-dependent kinase inhibitors roscovitine (ε255=7660 M−1·cm−1) and bohemine (ε255=9000 M−1·cm−1) was assayed as an increase in the concentration of the debenzylated products measured at an absorbance of 255 nm, i.e. 6-amino-2-[(1-hydroxymethyl)-propyl]amino-9-isopropylpurine (ε255=8800 M−1·cm−1) and 6-amino-2-(3-hydroxypropylamino)-9-isopropylpurine (ε255 =9600 M−1·cm−1) respectively, using such calculations as stated above.

Since some cytokinin derivatives, especially N6-benzyl-substituted ones, show limited solubility in aqueous buffers, they were dissolved in DMSO prior to being added to the reaction mixture. The addition of 0.7% DMSO (final concentration) decreased the rate of the control reaction with adenine as a substrate to approx. 70%.

Identification of recombinant proteins by MALDI-TOF (matrix-assisted laser-desorption ionization–time-of-flight) MS

Protein samples were first resolved by SDS/PAGE (10% gels) followed by Coomassie Brilliant Blue R-250 staining. Protein bands were excised from the gel, cut into small pieces and placed into 0.65 ml microtubes (Eppendorf). In-gel digestion using 1 μM raffinose-modified trypsin [36] was performed without prior reduction/alkylation at 37°C for 12 h. MALDI probes were prepared using an MSP AnchorChip 600/96 microScout target (Bruker Daltonics) and an α-cyano-4-hydroxycinnamic acid matrix [37]. Measurements were performed in the reflectron mode for positive ions on a Microflex MALDI-TOF LRF20 mass spectrometer (Bruker Daltonics) equipped with a nitrogen laser (337 nm wavelength). MS were accumulated from 100–200 shots at a laser repetition rate of 10 Hz and the examined m/z range was 500–4000. The instrument was calibrated externally using a mixture of peptide standards (Bruker Daltonics). The acquired spectra were processed by FlexAnalysis 2.4 and Biotools 3.0 software (Bruker Daltonics). Protein identification was achieved using the online version of the program Mascot (Matrix Science), and searches were performed against non-redundant protein databases (MSDB and Swiss-Prot).

CD spectra

CD spectra were recorded at 22°C using a Jasco J-810 spectrometer (Jasco). Data were collected from 195–260 nm, at 100 nm/min, with a response time of 1 s and a bandwidth of 2 nm, using a 0.1-cm path length quartz cuvette containing the protein in 0.1 M potassium phosphate buffer (pH 7.0). Collected data were expressed in terms of the mean residue ellipticity ([θ]m.r.w.) using the equation: [θ]m.r.w.=[θ]obs.×Mw×100/(n×c×l), where [θ]obs. is the observed ellipticity in degrees, Mw is the protein molecular mass, n is number of residues, l is the cell path-length, c is the protein concentration and the factor 100 originates from the conversion of the molecular mass to mg/dmol.

Identification of reaction products by LC–MS

A hybrid Q-TOF (quadrupole–time-of-flight) micro mass spectrometer (Waters) was used for high-resolution identification and confirmation of enzyme-generated products. The reaction mixture for identification of the product of S. pombe adenine deaminase containing 178 nkat/ml enzyme and 0.1 mM substrate in 0.2 M potassium phosphate buffer (pH 6.7) was incubated at 33°C (incubation time of 5 h for adenine and 16 h for cytokinin as substrate). The reaction mixture of human ADA with cytokinins and cyclin-dependent kinase inhibitors contained 5.1 μkat/ml enzyme and 0.01 mM substrate in 0.2 M potassium phosphate buffer (pH 7.0) and was incubated at 37°C for 24 h. Standard samples of the substrates and products (the same concentrations as for the reaction mixtures) were first dissolved in DMSO and then in the corresponding reaction buffer (the final DMSO concentration was 1%) and aliquots (20 μl) were used for analysis.

Measurements were performed in connection with HPLC analysis on an Alliance separation module 2965 equipped with a photodiode array detector 2996 (Waters) using a reversed-phase Symmetry C18 column (150×2.1 mm, 5 μm; Waters) and a post-column splitting of 1:1 of the mobile phase between the UV detector and Q-TOF mass analyser [38]. Following the injection, elution was performed with a 25 min binary linear gradient of (A) 15 mM ammonium formate (pH 4.0) and (B) methanol [0 min, 2% (v/v) B; 2–25 min, 90% (v/v) B] at a flow rate of 0.25 ml/min and a column temperature of 30°C. Electrospray ionization in the positive-ion mode was performed using the following parameters: source block/desolvation temperature, 100°C/350°C; capillary/cone voltage, 2500/25 V; and spray/cone gas flow (N2), 50/500 litres/h. In the full-scan mode, data were acquired in the mass range 50–1000 m/z, with a cycle time of 33 ms, a scan time of 2 s and collision energy of 4 V. For exact mass-determination experiments, a lock spray was used for external calibration with a mixture of 0.1 M NaOH/10% (v/v) formic acid/acetonitrile [1:1:8 (by vol.)] as a reference. Accurate masses were calculated and used for determination of the elementary composition and structure of the analytes, with a fidelity of 5 p.p.m.

RESULTS

Cloning and expression of adenine and adenosine deaminase genes

Adenine and adenosine deaminases hydrolyse the N6-amino group of adenine/adenosine, but they may act also on other substrates, such as purines containing chlorine, methoxy or methylamine at position 6 [46]. In this context, the enzymatic degradation of N6-furfuryladenine (a cytokinin known as kinetin) by S. pombe [39] is particularly interesting, because yeast do not possess the enzyme cytokinin dehydrogenase (EC 1.5.99.12), which commonly metabolizes cytokinins in plants [40]. As S. pombe is known to contain adenine deaminase [3], we decided to test the hypothesis that adenine deaminases and possibly also ADAs may act on cytokinin bases and ribosides respectively. Since it is difficult to purify these enzymes their native sources, two yeast adenine deaminase genes and a homologous gene from A. thaliana, have been chosen for cloning and functional expression to obtain active recombinant enzymes.

Translated sequences of the studied yeast genes share 45.7% identity and 61.1% similarity at the amino-acid level as calculated by BioEdit software 7.0.5.3 using the BLOSUM 62 matrix [41], whereas when compared with the A. thaliana enzyme, they show only approx. 16% identity and 31% similarity at the amino-acid level. When comparing these protein sequences with mammalian ADAs from cow and human, they show approx. 20% identity and approx. 40% similarity at the amino acid level. Despite the low similarity, all important active-site residues deduced from the structure of bovine ADA [20] are conserved as shown in the protein sequence alignment in Figure 1(A). The corresponding A. thaliana gene product contains an adenosine/AMP conserved domain cd00443 (different from the cd01320 domain present in the yeast enzymes examined in the present study) that is also found in some prokaryotic and eukaryotic enzymes annotated as ADAs, adenosine/AMP deaminases or ADA-like proteins.

Protein sequence alignment of adenine and adenosine deaminases

Figure 1
Protein sequence alignment of adenine and adenosine deaminases

(A) Multiple protein sequence alignment by ClustalW (BioEdit software 7.0.5.3 [41]) of S. cerevisiae Aah1p (GeneID: 855581), S. pombe SPBC1198.02 (GeneID: 2540066), A. thaliana ADA (GeneID: 825826), Homo sapiens ADA1 (H. sapiens ADA1) (GeneID: 100) and Bos taurus ADA (B. taurus ADA) (GeneID: 280712). Black and grey shaded boxes indicate identical and similar amino acids present at that position in at least 70% of the input sequences respectively. Active-site residues that were identified in the structure of B. taurus ADA [20] are marked below the alignment as follows: histidyl ligands to Zn2+ (*) and substrate-binding residues (s). (B) Protein sequence alignment of S. pombe adenine deaminase SPBC1198.02 (GeneID: 2540066) and the cytokinin receptor CRE1/WOL/AHK4 (GeneID: 814714) from A. thaliana.

Figure 1
Protein sequence alignment of adenine and adenosine deaminases

(A) Multiple protein sequence alignment by ClustalW (BioEdit software 7.0.5.3 [41]) of S. cerevisiae Aah1p (GeneID: 855581), S. pombe SPBC1198.02 (GeneID: 2540066), A. thaliana ADA (GeneID: 825826), Homo sapiens ADA1 (H. sapiens ADA1) (GeneID: 100) and Bos taurus ADA (B. taurus ADA) (GeneID: 280712). Black and grey shaded boxes indicate identical and similar amino acids present at that position in at least 70% of the input sequences respectively. Active-site residues that were identified in the structure of B. taurus ADA [20] are marked below the alignment as follows: histidyl ligands to Zn2+ (*) and substrate-binding residues (s). (B) Protein sequence alignment of S. pombe adenine deaminase SPBC1198.02 (GeneID: 2540066) and the cytokinin receptor CRE1/WOL/AHK4 (GeneID: 814714) from A. thaliana.

Interestingly, there is a part of the amino-acid sequence of S. pombe adenine deaminase which shows a notable homology with the CHASE domain of the cytokinin receptor CRE1/WOL/AHK4 (cytokinin response 1/wooden leg/Arabidopsis histidine kinase 4). This receptor is responsible for cytokinin signal perception in A. thaliana and interacts with the hormone via the CHASE domain [42]. As identified by an NCBI BLAST search (blastp), the CHASE domain (amino-acid residues 198–411 of CRE1/WOL/AHK4) shows 25% identity and 41% similarity in a stretch of 120 residues (amino-acid residues 199–320) to amino-acid residues 153–275 of SPBC1198.02 (Figure 1B). This homology was shown for the enzyme from S. pombe, but not for the other studied proteins.

All three proteins of interest were obtained by cloning the corresponding genes into pET vectors, followed by IPTG-inducible expression in E. coli. At temperatures higher than 18°C, the recombinant protein was localized to inclusion bodies. As shown in Figure 2, homogeneous preparations of all three enzymes from E. coli cell extracts were obtained by single-step affinity purification on Ni2+–NTA agarose. The binding and elution conditions were optimized for each protein as described in the Materials and methods section. Although the specific activities of both expression products were the same, protein constructs carrying an additional 3′-translated overhanging sequence from the pET100/D-TOPO vector (see Table 1) fused to the C-terminus of the protein were much more resistant to proteolytic degradation in E. coli cell extracts than proteins lacking the extended C-terminal sequence and were therefore used throughout the present study.

SDS/PAGE of adenine and adenosine deaminases produced from pET100/D-TOPO clones

Figure 2
SDS/PAGE of adenine and adenosine deaminases produced from pET100/D-TOPO clones

SDS/PAGE was performed using a discontinuous buffer system according to Laemmli [34], and the protein bands were visualized by staining with Coomassie Brilliant Blue R-250. Lane 1, E. coli extract before induction (10 μg of protein); Lane 2, cytosolic fraction after induction of S. cerevisiae Aah1p (10 μg of protein); Lane 3, cytosolic fraction after induction of S. pombe SPBC1198.02 (10 μg of protein); Lane 4, cytosolic fraction after induction of A. thaliana ADA (9 μg of protein); Lane 5, purified S. cerevisiae Aah1p, estimated molecular mass of 46.0 kDa (2.2 μg); Lane 6, purified S. pombe SPBC1198.02, estimated molecular mass of 48.0 kDa (2 μg); Lane 7, purified A. thaliana ADA, estimated molecular mass of 46.5 kDa (1.7 μg); M, molecular-mass markers (kDa).

Figure 2
SDS/PAGE of adenine and adenosine deaminases produced from pET100/D-TOPO clones

SDS/PAGE was performed using a discontinuous buffer system according to Laemmli [34], and the protein bands were visualized by staining with Coomassie Brilliant Blue R-250. Lane 1, E. coli extract before induction (10 μg of protein); Lane 2, cytosolic fraction after induction of S. cerevisiae Aah1p (10 μg of protein); Lane 3, cytosolic fraction after induction of S. pombe SPBC1198.02 (10 μg of protein); Lane 4, cytosolic fraction after induction of A. thaliana ADA (9 μg of protein); Lane 5, purified S. cerevisiae Aah1p, estimated molecular mass of 46.0 kDa (2.2 μg); Lane 6, purified S. pombe SPBC1198.02, estimated molecular mass of 48.0 kDa (2 μg); Lane 7, purified A. thaliana ADA, estimated molecular mass of 46.5 kDa (1.7 μg); M, molecular-mass markers (kDa).

Table 1
Biochemical characterization of studied adenine and adenosine deaminases

The molecular masses (Da) were calculated from the amino-acid sequence using Bioedit 7.0.5.3 [41], see Figure 2 for SDS/PAGE of the recombinant proteins. The values for S. cerevisiae, S. pombe and A. thaliana proteins do not include the 6× His tag and Xpress™ epitope [(M)RGSHHHHHH-GMASMTGGQQMGRDLTDDDDKDHPET (3757 Da)] and the vector 3′-translated overhanging sequence [KGELNDPAANKARKEAELAAATAEQ (2589 Da)] from pET100/D-TOPO fused to the N- and C-termini of the recombinant protein respectively.

Enzyme
FeatureS. cerevisiae Aah1pS. pombe SPBC1198.02A. thaliana ADAHomo sapiens ADA1
Molecular mass 39633 41197 39947 40762 
pH optimum 7.0* 6.7* 6.7* 6–8 [44
Temperature optimum (°C) 30–37§ 33§ 30§ 37 [15
Specific activity (nkat/mg) 139 326 0.12 8970 [44
Km (μM) 55* 32* – 52 [44
kcat (s−15.5 13.4 0.005 366** 
kcat/Km 0.10 0.46 – 7.0 
Enzyme
FeatureS. cerevisiae Aah1pS. pombe SPBC1198.02A. thaliana ADAHomo sapiens ADA1
Molecular mass 39633 41197 39947 40762 
pH optimum 7.0* 6.7* 6.7* 6–8 [44
Temperature optimum (°C) 30–37§ 33§ 30§ 37 [15
Specific activity (nkat/mg) 139 326 0.12 8970 [44
Km (μM) 55* 32* – 52 [44
kcat (s−15.5 13.4 0.005 366** 
kcat/Km 0.10 0.46 – 7.0 
*

Measurements were performed in 0.2 M phosphate buffer, at optimum temperature.

Adenine was used as a substrate.

Adenosine was used as a substrate and measurements were performed in 0.2 M phosphate buffer (pH 7.0) at 37°C.

§

Measurements were performed in 0.2 M phosphate buffer, at optimum pH.

Measurements were performed with 0.067 mM substrate in 0.2 M phosphate buffer, at optimum pH and temperature.

Measurements were performed with 0.033 mM adenosine in 0.2 M phosphate buffer (pH 6.7) at 30°C.

**

Calculated from published specific activity [44] and calculated molecular mass.

Protein expression was evaluated by MALDI-TOF peptide mass fingerprinting (see Supplementary Figure S1 at http://www.bioscirep.org/bsr/028/bsr0280335add.htm) performed after SDS/PAGE and in-gel digestion. The recombinant proteins were unambiguously assigned to the UniProtKB/Swiss-Prot databases with accession numbers P53909 (S. cerevisiae Aah1p), Q9P6I7 (S. pombe SPBC1198.02) and Q8LPL7 (A. thaliana ADA). The corresponding sequence coverage values were 65, 40 and 69% respectively, probability-based MOWSE scores (Mascot search) were in the range 121–292 and the RMS (root mean square) error was below 50 p.p.m.

Far-UV CD spectra were used to assess the proper folding and secondary structure of studied recombinant proteins. Comparison of the measured spectra demonstrated that all tested aminohydrolase enzymes are correctly folded. All proteins showed spectra typical of a predominantly α-helix conformation, with two negative features at approx. 221 and 208 nm [43]. No significant differences between the spectra of all three aminohydrolases were found (see Supplementary Figure S2 at http://www.bioscirep.org/bsr/028/bsr0280335add.htm), which suggests that the secondary structure content of all tested enzymes is very similar.

Biochemical characterization of the recombinant proteins

Activity of purified proteins was stable for 1–2 weeks when stored at 4°C at concentrations above 10 mg/ml with the addition of 20% (v/v) glycerol. Prolonged storage in solution was possible at −20°C with 50% (v/v) glycerol and at −80°C with 10–20% (v/v) glycerol or as a freeze-dried powder. After dissolving the freeze-dried enzymes, however, the activity decreased significantly within 3–4 h. Thanks to the straightforward purification method, all further studies were performed with fresh enzyme preparations. In general, the activity was highest in potassium phosphate buffers, decreasing to approx. 75% in Tris/HCl buffers and to 30% in Mops buffers (results for S. pombe adenine deaminase). Potassium phosphate buffer was therefore used for all studies.

A comparison of general biochemical properties of the enzymes from S. cerevisiae, S. pombe and A. thaliana with those published for human ADA (ADA1 [44]) is shown in Table 1. Compared with human ADA1, the yeast enzymes have a similar pH optimum and Km value for the best substrate. Their kcat values, however, are much lower than that of the human enzyme. The kcat value of the A. thaliana protein is extremely low and it is not clear if the protein is really a deaminase enzyme.

Identification of reaction products by HPLC–MS

Reaction products of the studied enzymes were analysed and identified using standards on HPLC equipped with UV and MS detectors (Figure 3). Degradation of adenine (136 m/z) by S. pombe adenine deaminase showed the expected product hypoxanthine with a typical UV absorption maximum at 250 nm and 137 m/z. Hypoxanthine was also confirmed as a product of the degradations of the cytokinins isopentenyladenine (204 m/z) and m-topoline (242 m/z). The side chain cleavage product, supposedly an amine, was not detected in the chromatogram. Further experiments with a standard compound, benzylamine, showed that amines eluted near the dead volume and thus no identification was possible.

HPLC–MS analysis of reaction products of S. pombe adenine deaminase (A–C) and human ADA (D–F)

Figure 3
HPLC–MS analysis of reaction products of S. pombe adenine deaminase (A–C) and human ADA (D–F)

For (A–F), monitoring was performed at 250 nm and the retention times are indicated for the reaction products of S. pombe adenine deaminase (A–C) and human ADA (D–F). (A) Standard compounds hypoxanthine (2.62 min), adenine (3.24 min) and isopentenyladenine (22.18 min); (B) adenine degradation by S. pombe adenine deaminase producing hypoxanthine (2.60 min), insets show UV absorption and MS spectra, with the substrate adenine completely converted; (C) isopentenyladenine degradation by adenine deaminase producing hypoxanthine (2.60 min), insets show UV and MS spectra, the substrate isopentenyladenine was only partially converted (22.20 min); (D) separation of roscovitine (26.46 min) and (E) 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (20.35 min) standard compounds, insets show UV absorption and MS spectra; and (F) roscovitine degradation by human ADA producing 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (20.20 min); inset shows MS spectrum (see the Materials and methods section for details). AU, absorbance unit.

Figure 3
HPLC–MS analysis of reaction products of S. pombe adenine deaminase (A–C) and human ADA (D–F)

For (A–F), monitoring was performed at 250 nm and the retention times are indicated for the reaction products of S. pombe adenine deaminase (A–C) and human ADA (D–F). (A) Standard compounds hypoxanthine (2.62 min), adenine (3.24 min) and isopentenyladenine (22.18 min); (B) adenine degradation by S. pombe adenine deaminase producing hypoxanthine (2.60 min), insets show UV absorption and MS spectra, with the substrate adenine completely converted; (C) isopentenyladenine degradation by adenine deaminase producing hypoxanthine (2.60 min), insets show UV and MS spectra, the substrate isopentenyladenine was only partially converted (22.20 min); (D) separation of roscovitine (26.46 min) and (E) 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (20.35 min) standard compounds, insets show UV absorption and MS spectra; and (F) roscovitine degradation by human ADA producing 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (20.20 min); inset shows MS spectrum (see the Materials and methods section for details). AU, absorbance unit.

Analysis of the reaction mixture of human ADA with cytokinins did not result in detection of hypoxanthine or any other detectable products. However, when the reaction with a cyclin-dependent kinase inhibitor roscovitine (355 m/z) was analysed, an unexpected product was detected, having a molecular mass of 1 Da less than 6-hydroxy-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine which would have been the expected product if the reaction had proceeded as for the S. pombe adenine deaminase. Using accurate mass determination, the compound was identified as 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (265 m/z). Since it was possible to obtain the compound from the stock in our laboratory, the identification was further verified by comparing the retention time and UV spectrum (maxima at 222, 256 and 295 nm). Another cyclin-dependent kinase inhibitor, olomoucine II (371 m/z), which is identical to roscovitine at the purine part of the molecule, resulted in the same hydrolytic product. Analogous product identification using an available reference compound was performed for the reaction of bohemine (341 m/z) that was hydrolysed to 6-amino-2-(3-hydroxypropylamino)-9-isopropylpurine (251 m/z, UV spectrum maxima at 222, 255 and 296 nm). Hydrolysis of olomoucine (299 m/z) formed 6-amino-2-(2-hydroxyethylamino)-9-methylpurine (209 m/z) that was identified by accurate mass determination (unfortunately a reference compound was not available for the hydrolytic product of olomoucine). The expected side chain cleavage product benzylalcohol was also searched for with a standard compound, but as it eluted near the dead volume in the chromatogram, no clear identification was possible.

On the basis of the above experimental results, we propose reaction schemes for yeast adenine deaminase and human ADA as shown in Figure 4.

Reaction schemes of adenine and adenosine deaminases

Figure 4
Reaction schemes of adenine and adenosine deaminases

(A) Hydrolysis of adenine/adenosine to hypoxanthine/inosine by adenine and adenosine deaminases and (B) hydrolytic cleavage of kinetin by yeast adenine deaminase, both occurring at the endo-side of the substrate N-6 amino group. (C) Hydrolytic exo-side cleavage of roscovitine by human ADA (ADA1).

Figure 4
Reaction schemes of adenine and adenosine deaminases

(A) Hydrolysis of adenine/adenosine to hypoxanthine/inosine by adenine and adenosine deaminases and (B) hydrolytic cleavage of kinetin by yeast adenine deaminase, both occurring at the endo-side of the substrate N-6 amino group. (C) Hydrolytic exo-side cleavage of roscovitine by human ADA (ADA1).

Activity assays with adenine/adenosine, cytokinins and cyclin-dependent kinase inhibitors

Adenine and adenosine deaminase activity of purified enzyme preparations can be assayed simply, apart from TLC and radioactivity measurements [45], by monitoring UV spectra reflecting the changes in substrate or product hypoxanthine/inosine concentrations [3,20,35]. The latter option was preferable, since it could be presumably used also for substrates other than adenine/adenosine that do not release ammonia, but primary amines when metabolized by the enzyme.

UV absorption spectra of adenine show a maximum at 260 nm, which is close to the maximum for hypoxanthine at 250 nm (Figure 5C), but adenine shows far less absorption at 240 nm, where its conversion to hypoxanthine can be monitored (Figure 5A, [20]). The same method can be applied to adenosine/inosine, AMP/IMP and ATP/ITP that show similar shapes of absorption spectra with slightly different molar absorption coefficients (see the Materials and methods section). Ammonia, the other released product, shows negligible contribution to the total absorption [46].

Spectrophotometric assays of adenine deaminase activity

Figure 5
Spectrophotometric assays of adenine deaminase activity

Activity assay of S. pombe adenine deaminase. (A) Adenine hydrolysis accompanied by spectral changes as a result of the production of hypoxanthine. (B) N6-Isopentenyladenine hydrolysis and production of hypoxanthine; inset shows a detailed view of the spectrum region at approx. 240 nm. (C) UV spectra of 0.13 mM adenine (1), hypoxanthine (2) and N6-isopentenyladenine (3) standards. The maximum spectral difference at 240 nm is indicated, which was used for monitoring hypoxanthine production (see the Materials and methods section for details).

Figure 5
Spectrophotometric assays of adenine deaminase activity

Activity assay of S. pombe adenine deaminase. (A) Adenine hydrolysis accompanied by spectral changes as a result of the production of hypoxanthine. (B) N6-Isopentenyladenine hydrolysis and production of hypoxanthine; inset shows a detailed view of the spectrum region at approx. 240 nm. (C) UV spectra of 0.13 mM adenine (1), hypoxanthine (2) and N6-isopentenyladenine (3) standards. The maximum spectral difference at 240 nm is indicated, which was used for monitoring hypoxanthine production (see the Materials and methods section for details).

Deamination of cytokinins that have the main absorption band shifted to approx. 269 nm and release hypoxanthine or inosine can also be monitored at 240 nm (Figures 5B and 5C). The other respective products, N6-side chain amines, have negligible contributions to absorption at 240 nm.

Human ADA cleaves adenosine, producing spectral changes similar to those of the yeast adenine deaminases acting on adenine, and its activity can be assayed as the production of inosine at 240 nm (Figure 6A). The enzyme, however, does not cleave cytokinins at all. Interaction of human ADA with cyclin-dependent kinase inhibitors showed spectral changes that were completely different from those of adenine/adenosine reactions and the cytokinin reactions of the yeast enzymes. The nature of these changes was only understood fully after identification of the reaction products by HPLC coupled to a Q-TOF mass spectrometer as described above. As shown in Figure 6(B), the spectrum of hydrolysed roscovitine shows a significant increase in the region of 250–260 nm, reflecting the release of the hydrolytic product 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine with a characteristic absorption maximum at 255 nm (Figure 6C). On the basis of determining the molar absorption coefficient of the compound that was synthesized (see the Materials and methods section), it was possible to assay the rate of hydrolysis of roscovitine and olomoucine II that gives the same product. The same approach was used to determine the rate of enzymatic hydrolysis of bohemine to 6-amino-2-(3-hydroxypropylamino)-9-isopropylpurine. As the hydrolytic products of olomoucine, the other tested compound, were not available, it was not possible to assay the activity directly, but only to estimate its values using the molar absorption coefficient of the product of bohemine that is thought to show similar absorption.

Spectrophotometric assays of human ADA activity

Figure 6
Spectrophotometric assays of human ADA activity

Activity assay of H. sapiens adenosine deaminase ADA1. (A) Adenosine hydrolysis accompanied by spectral changes as a result of the production of inosine. (B) Spectral changes as a result of the cleavage of roscovitine. (C) UV spectra of 0.02 mM roscovitine (1) and 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (2) standards. The maximum spectral difference at 255 nm is indicated, which was used for monitoring product formation (see the Materials and methods section for details).

Figure 6
Spectrophotometric assays of human ADA activity

Activity assay of H. sapiens adenosine deaminase ADA1. (A) Adenosine hydrolysis accompanied by spectral changes as a result of the production of inosine. (B) Spectral changes as a result of the cleavage of roscovitine. (C) UV spectra of 0.02 mM roscovitine (1) and 6-amino-2-[(1-hydroxymethyl)propyl]amino-9-isopropylpurine (2) standards. The maximum spectral difference at 255 nm is indicated, which was used for monitoring product formation (see the Materials and methods section for details).

Reaction rates of adenine and adenosine deaminases with cytokinins and cytokinin derivatives

Of the three recombinant proteins prepared in this work, only the yeast enzymes were capable of hydrolytic cleavage of cytokinins. Cyclin-dependent kinase inhibitors derived from cytokinins, but not the cytokinins themselves, were also substrates of human ADA1. The relative reaction rates of all studied enzymes with various substrates are shown in Table 2. The yeast enzymes both showed substrate specificity that unambiguously confirms their classification as adenine deaminases, despite being annotated in databases as ADAs (e.g. entries P53909 and Q9P6I7 in UniProtKB/Swiss-Prot databases). Adenine was by far the best substrate, followed by AMP and ATP, whereas the cleavage of adenosine and cytokinins proceeded at rates of several percentage of that of adenine. The hydrolysis of the N6-secondary amino group of cytokinins produced a hypoxanthine/inosine molecule and an amine from the cytokinin side-chain as described above (Figure 4).

Table 2
Comparison of reaction rates of adenine and adenosine deaminases with various substrates

Reaction rates were assayed with 0.067 mM substrate by detection of product formation at 240 nm, except for the cleavage products of cyclindependent kinase inhibitors, which were were assayed at 255 nm (see the Materials and methods section) under the optimum conditions as stated in Table 1. All reaction mixtures contained 0.7% DMSO (final concentration), which was used to dissolve the substrate. n.d., not detected.

Relative reaction rate (%)
SubstratesS. cerevisiae Aah1pS. pombe SPBC1198.02A. thaliana ADAH. sapiens ADA1
Adenine 100 100 100* 9.9 
Adenosine 1.4 2.0 30* 100 
AMP 3.3 4.5 56* 9.1 
ATP 3.3 3.3 68* 8.8 
N6-Isopentenyladenine 2.1 2.9 n.d. n.d. 
N6-Isopentenyladenosine 1.2 1.9 n.d. n.d. 
N6-Isopentenyladenine 9-glucoside n.d. 1.1 n.d. n.d. 
N6-Methyl-isopentenyladenine n.d. n.d. n.d. n.d. 
cis-Zeatin 1.7 2.5 n.d. n.d. 
trans-Zeatin 1.8 1.6 n.d. n.d. 
trans-Zeatin riboside n.d. n.d. n.d. n.d. 
trans-Zeatin 9-glucoside n.d. n.d. n.d. n.d. 
Benzyladenine 0.7 3.8 n.d. n.d. 
Kinetin 2.6 2.8 n.d. n.d. 
m-Topolin 2.5 4.0 n.d. n.d. 
Cyclin-dependent kinase inhibitors     
 Bohemine n.d. n.d. n.d. 13.6 
 Roscovitine n.d. n.d. n.d. 6.8 
 Olomoucine n.d. n.d. n.d. 7.3 
 Olomoucine II n.d. n.d. n.d. 7.0 
Relative reaction rate (%)
SubstratesS. cerevisiae Aah1pS. pombe SPBC1198.02A. thaliana ADAH. sapiens ADA1
Adenine 100 100 100* 9.9 
Adenosine 1.4 2.0 30* 100 
AMP 3.3 4.5 56* 9.1 
ATP 3.3 3.3 68* 8.8 
N6-Isopentenyladenine 2.1 2.9 n.d. n.d. 
N6-Isopentenyladenosine 1.2 1.9 n.d. n.d. 
N6-Isopentenyladenine 9-glucoside n.d. 1.1 n.d. n.d. 
N6-Methyl-isopentenyladenine n.d. n.d. n.d. n.d. 
cis-Zeatin 1.7 2.5 n.d. n.d. 
trans-Zeatin 1.8 1.6 n.d. n.d. 
trans-Zeatin riboside n.d. n.d. n.d. n.d. 
trans-Zeatin 9-glucoside n.d. n.d. n.d. n.d. 
Benzyladenine 0.7 3.8 n.d. n.d. 
Kinetin 2.6 2.8 n.d. n.d. 
m-Topolin 2.5 4.0 n.d. n.d. 
Cyclin-dependent kinase inhibitors     
 Bohemine n.d. n.d. n.d. 13.6 
 Roscovitine n.d. n.d. n.d. 6.8 
 Olomoucine n.d. n.d. n.d. 7.3 
 Olomoucine II n.d. n.d. n.d. 7.0 
*

Measured with 0.033 mM substrate in 0.2 M phosphate buffer (pH 6.7) at 30°C.

Estimated using the molar absorption coefficient of inosine at 240 nm (spectral data for the product hypoxanthine 9-glucoside are not available).

Estimated using the molar absorption coefficient of the product of bohemine at 255 nm (spectral data for the specific product are not available).

The enzyme from A. thaliana, in addition to its low activity, did not show a clear preference for either adenine, adenosine, AMP or ATP, and did not cleave cytokinins at all.

Human ADA showed substrate specificity that clearly differed from the above enzymes. The highest activity was detected with adenosine, followed by 10-fold lower values obtained with adenine, AMP and ATP. Surprisingly, the enzyme did not cleave cytokinins, but showed reaction rates similar to adenine and ATP/AMP, with the derivatives of cytokinins bearing hydrophobic chains at C2 and C9 that are known as effective cyclin-dependent kinase inhibitors. As found by HPLC–MS analysis of the products, the N6-secondary amino group of these compounds was cleaved in an opposite manner (i.e. at its exo-side) than in the reaction of yeast adenine deaminase with cytokinins, leaving the primary amino group on the purine ring and assumedly releasing benzylalcohol derived from the N6 side chain (Figure 4).

DISCUSSION

In yeast, adenine deaminase is responsible for the metabolic salvage of purine compounds. The AAH1 gene in S. cerevisiae is strongly down-regulated when yeast cells enter quiescence under nutrient-limiting conditions. The down-regulation is maintained via proteasome- and SCF (Skp1/Cullin/F-box protein)-dependent degradation [47]. As shown in the present study, kcat values of the yeast enzymes are approx. 50-fold lower than that of human ADA1 [44], but probably still sufficient for metabolic function. The results obtained in the present study with the recombinant adenine deaminase from S. pombe are in agreement with results published previously which describe the enzymatic degradation of kinetin by this yeast [39]. This protein exhibits partial amino-acid sequence similarity to the CHASE domain of the cytokinin receptor, which may account for its ability to bind and hydrolyse not only kinetin, but also other cytokinins and cytokinin derivatives, albeit at lower rates than its natural substrate adenine (Table 2). It is not likely, however, that yeast cells adopt a specific metabolic pathway for cleaving cytokinins, since yeasts do not commonly interact with cytokinin-producing systems.

The protein from A. thaliana encoded by the gene At4g04880 shows very low levels of activity without an obvious substrate preference for adenine, adenosine, AMP or ATP. The protein may perform different functions from the metabolism of purine compounds, which in plants is believed to be controlled by adenosine kinase [24]; perhaps it may be involved in plant defence. Results from Zimmermann et al. [48] show that the expression of At4g04880 is strongly induced by biotic stress (probeset 255299_at, array name Agro_inf_rep_3). The gene expression is 10-fold upregulated in mature Arabidopsis siliques infected with Agrobacterium tumefaciens. As found in rice, genes coding for deoxycytidine deaminase (Os07g14150) and AMP deaminase (Os05g28180) are upregulated as well during an infection with the fungal pathogen Magnaporthe grisea and may be involved in RNA editing [49]. These proteins, however, do not share significant sequence homology with the product of the gene At4g04880, the former belonging to the cytidine deaminase family and the latter to the AMP deaminase family that includes the embryonic factor FAC1 (fetal Alz-50 clone 1) [50].

Some synthetic cytokinin derivatives exhibit anticancer properties as a result of their ability to inhibit cyclin-dependent kinases [26] and currently some are undergoing testing in clinical trials as drugs for treatment of some types of cancers and glomerulonephritis; others, such as kinetin, have an application in skin-protective cosmetics, as it delays aging of human fibroblasts [27] and protects against oxidative damage of DNA [28]. Our results show that, in humans, the high activity of ADA (ADA1) may attenuate the level of some of the active compounds and thus compromise their physiological or therapeutic effects. These compounds are hydrolysed (debenzylated) via a different reaction mechanism to the natural substrate adenosine and at a lower rate, but as a result of the very high turnover number of the enzyme, the rate may still be quite significant. Debenzylation of bohemine by mouse liver microsomes supplemented with NADPH has been described previously as one of the main biotransformation routes [30]. The different product specificity of yeast and human adenine and adenosine deaminases when acting on secondary amino groups may be similar to the case of polyamine oxidases. Plant and bacterial polyamine oxidases oxidize the carbon on the endo-side of the N4 nitrogen of spermidine and spermine, whereas animal polyamine oxidases and yeast (S. cerevisiae) spermine oxidase oxidize the carbon on the exo-side of the N4 nitrogen [51].

We suggest that a survey of potential biologically active compounds derived from adenine should include also an assessment of hydrolytic cleavage by ADA1 using the methods described in the present study. As none of the enzymes studied in the present paper was not capable of cleaving N-methylisopentenyladenine and the hydrolytic cleavage of a tertiary amino group is highly unlikely, we also suggest that additional methylation of these compounds at N6 may fully prevent the degradation by deaminases.

Abbreviations

     
  • ADA

    adenosine deaminase

  •  
  • ADAT

    tRNA-specific ADA

  •  
  • CRE1/WOL/AHK4

    cytokinin response 1/wooden leg/Arabidopsis histidine kinase 4

  •  
  • IMP

    inosine monophosphate

  •  
  • IPTG

    isopropyl β-D-thiogalactoside

  •  
  • ITP

    inosine triphosphate

  •  
  • MALDI-TOF

    matrix-assisted laser-desorption ionization–time-of-flight

  •  
  • Ni2+–NTA

    Ni2+–nitrilotriacetate

  •  
  • Q-TOF

    quadrupole–time-of-flight

We thank Radka Chaloupková (National Center for Biomedical Research, Faculty of Science, Masaryk University, Brno, Czech Republic) for measuring CD spectra.

FUNDING

This work was supported by the Ministry of Education, Youth and Sports [grant number MSM 6198959216]; and by the Czech Science Foundation [grant number 522/06/0022].

References

References
Maier
 
S. A.
Galellis
 
J. R.
McDermid
 
H. E.
 
Phylogenetic analysis reveals a novel protein family closely related to adenosine deaminase
J. Mol. Evol.
2005
, vol. 
61
 (pg. 
776
-
794
)
Pospíšilová
 
H.
Frébort
 
I.
 
Aminohydrolases acting on adenine, adenosine and their derivatives
Biomed. Pap. Med. Fac. Univ. Palacky Olomouc Czech Repub.
2007
, vol. 
151
 (pg. 
3
-
10
)
Ribard
 
C.
Rochet
 
M.
Labedan
 
B.
Daignan-Fornier
 
B.
Alzari
 
P.
Scazzocchio
 
C.
Oestreicher
 
N.
 
Sub-families of α/β barrel enzymes: A new adenine deaminase family
J. Mol. Biol.
2003
, vol. 
334
 (pg. 
1117
-
1131
)
Pickard
 
M. A.
 
Purification and some properties of the soluble and membrane-bound adenosine deaminases of Micrococcus sodonensis ATCC 11880 and their distribution within the family Micrococcacea
Can. J. Biochem.
1975
, vol. 
53
 (pg. 
344
-
353
)
Lupidi
 
G.
Marmocchi
 
F.
Cristalli
 
G.
 
Inhibition studies on membrane adenosine deaminase from human placenta
Biochem. Mol. Biol. Int.
1998
, vol. 
46
 (pg. 
1071
-
1080
)
Nygaard
 
P.
 
Hoffee
 
P. A.
Jones
 
M. E.
 
Adenosine deaminase from Escherichia coli
Methods in Enzymology: purine and pyrimidine nucleotide metabolism
1978
, vol. 
vol. 51
 
New York
Academic Press
(pg. 
508
-
512
)
Jun
 
H. K.
Kim
 
T. S.
Yeeh
 
Y.
 
Purification and characterization of an extracellular adenosine deaminase from Nocardioides sp. J-326TK
Biotechnol. Appl. Biochem.
1994
, vol. 
20
 (pg. 
265
-
277
)
Jun
 
H. K.
Kim
 
T. S.
Sakai
 
T.
 
Purification and characterization of extracellular adenosine deaminase from a Streptomyces sp
J. Ferment. Bioeng.
1991
, vol. 
71
 (pg. 
6
-
11
)
Nygaard
 
P.
Duckert
 
P.
Saxild
 
H. H.
 
Role of adenine deaminase in purine salvage and nitrogen metabolism and characterisation of the ade gene in Bacillus subtilis
J. Bacteriol.
1996
, vol. 
178
 (pg. 
846
-
853
)
Matsui
 
H.
Shimaoka
 
M.
Kawasaki
 
H.
Takenaka
 
Y.
Kurahashi
 
O.
 
Adenine deaminase activity of the yicP gene product of Escherichia coli
Biosci. Biotechnol. Biochem.
2001
, vol. 
65
 (pg. 
1112
-
1118
)
Wolf
 
J.
Gerber
 
A. P.
Keller
 
W.
 
tadA, an essential tRNA-specific adenosine deaminase from Escherichia coli
EMBO J.
2002
, vol. 
21
 (pg. 
3841
-
3851
)
Keegan
 
L. P.
Leroy
 
A.
Sproul
 
D.
O'Connell
 
M. A.
 
Adenosine deaminases acting on RNA (ADARs): RNA-editing enzymes
Genome Biol.
2004
, vol. 
5
 
209
Hirschhorn
 
R.
Ratech
 
H.
 
Ratazzi
 
M. C.
Scandalia
 
J. G.
Whitt
 
G. S.
 
Isozymes of adenosine deaminase
Current Topics in Biological and Medical Research
1980
, vol. 
vol. 1
 
New York
Alan R. Liss
(pg. 
132
-
157
)
Zavialov
 
A. V.
Engström
 
A.
 
Human ADA2 belongs to a new family of growth factors with adenosine deaminase activity
Biochem. J.
2005
, vol. 
391
 (pg. 
51
-
57
)
Van der Weyden
 
M. B.
Kelley
 
W. N.
 
Human adenosine deaminase. Distribution and properties
J. Biol. Chem.
1976
, vol. 
251
 (pg. 
5448
-
5456
)
Aronow
 
B. J.
Silbiger
 
R. N.
Dusing
 
M. R.
Stock
 
J. L.
Yager
 
K. L.
Potter
 
S. S.
Hutton
 
J. J.
Wiginton
 
D. A.
 
Functional analysis of the human adenosine deaminase gene thymic regulatory region and its ability to generate position-independent transgene expression
Mol. Cell. Biol.
1992
, vol. 
12
 (pg. 
4170
-
4185
)
Gaspar
 
H. B.
Bjorkegren
 
E.
Parsley
 
K.
Gilmour
 
K. C.
King
 
D.
Sinclair
 
J.
Zhang
 
F.
Giannakopoulos
 
A.
Adams
 
S.
Fairbanks
 
L. D.
, et al 
Successful reconstitution of immunity in ADA-SCID by stem cell gene therapy following cessation of PEG-ADA and use of mild preconditioning
Mol. Ther.
2006
, vol. 
14
 (pg. 
505
-
513
)
Persico
 
A. M.
Militerni
 
R.
Bravaccio
 
C.
Schneider
 
C.
Melmed
 
R.
Trillo
 
S.
Montecchi
 
F.
Palermo
 
M. T.
Pascucci
 
T.
Puglisi-Allegra
 
S.
, et al 
Adenosine deaminase alleles and autistic disorder: Case-control and family-based association studies
Am. J. Med. Genet.
2000
, vol. 
96
 (pg. 
784
-
790
)
Wilson
 
D. K.
Rudolph
 
F. B.
Quiocho
 
F. A.
 
Atomic structure of adenosine deaminase complexed with a transition-state analog: understanding catalysis and immunodeficiency mutations
Science
1991
, vol. 
252
 (pg. 
1278
-
1284
)
Kinoshita
 
T.
Nishio
 
N.
Nakanishi
 
I.
Sato
 
A.
Fujii
 
T.
 
Structure of bovine adenosine deaminase complexed with 6-hydroxy-1,6-dihydropurine riboside
Acta Crystallogr., Sect. D: Biol Crystallogr.
2003
, vol. 
59
 (pg. 
299
-
303
)
Yabuki
 
N.
Ashihara
 
H.
 
Catabolism of adenine nucleotides in suspension-cultured plant cells
Biochim. Biophys. Acta
1991
, vol. 
1073
 (pg. 
474
-
480
)
Katahira
 
R.
Ashihara
 
H.
 
Profiles of purine biosynthesis, salvage and degradation in disks of potato (Solanum tuberosum L.)
tubers. Planta
2006
, vol. 
225
 (pg. 
115
-
126
)
Guranowski
 
A.
 
Purine catabolism in plants
Plant Physiol.
1982
, vol. 
70
 (pg. 
344
-
349
)
Moffatt
 
B. A.
Stevens
 
Y. Y.
Allen
 
M. S.
Snider
 
J. D.
Pereira
 
L. A.
Todorova
 
M. I.
Summers
 
P. S.
Weretilnyk
 
E. A.
Martin-McCaffrey
 
L.
Wagner
 
C.
 
Adenosine kinase deficiency is associated with developmental abnormalities and reduced transmethylation
Plant Physiol.
2002
, vol. 
128
 (pg. 
812
-
821
)
Sakakibara
 
H.
 
Cytokinins: activity, biosynthesis and translocation
Annu. Rev. Plant Biol.
2006
, vol. 
57
 (pg. 
431
-
449
)
Doležal
 
K.
Popa
 
I.
Kryštof
 
V.
Spíchal
 
L.
Fojtíková
 
M.
Holub
 
J.
Lenobel
 
R.
Schmülling
 
T.
Strnad
 
M.
 
Preparation and biological activity of 6-benzylaminopurine derivatives in plants and human cancer cells
Bioorg. Med. Chem.
2006
, vol. 
14
 (pg. 
875
-
884
)
Rattan
 
S. I.
Clark
 
B. F.
 
Kinetin delays the onset of ageing characteristics in human fibroblasts
Biochem. Biophys. Res. Commun.
1994
, vol. 
201
 (pg. 
665
-
672
)
Olsen
 
A.
Siboska
 
G. E.
Clark
 
B. F.
Rattan
 
S. I.
 
N(6)-Furfuryladenine, kinetin, protects against Fenton reaction-mediated oxidative damage to DNA
Biochem. Biophys. Res. Commun.
1999
, vol. 
265
 (pg. 
499
-
502
)
Wang
 
J.
Letham
 
D. S.
 
Cytokinin oxidase – purification by affinity chromatography and activation by caffeic acid
Plant Sci.
1995
, vol. 
112
 (pg. 
161
-
166
)
Rypka
 
M.
Veselý
 
J.
Chmela
 
Z.
Riegrová
 
D.
červenková
 
K.
Havlíček
 
L.
Lemr
 
K.
Hanuš
 
J.
černý
 
B.
Lukeš
 
J.
Michalíková
 
K.
 
In vitro biotransformation of 2,6,9-trisubstituted purine-derived cyclin-dependent kinase inhibitors bohemine by mouse liver microsomes
Xenobiotica
2002
, vol. 
32
 (pg. 
1017
-
1031
)
Bota
 
A.
Gella
 
F. J.
Profilis
 
C.
Férard
 
G.
Hadjivassiliou
 
A. G.
Hørder
 
M.
Schiele
 
F.
Segura
 
R.
Canalias
 
F.
 
Production and certification of an enzyme reference material for adenosine deaminase 1 (BCR 647)
Clin. Chim. Acta
2001
, vol. 
306
 (pg. 
79
-
89
)
Bechet
 
J.
Greenson
 
M.
Wiame
 
J. M.
 
Mutations affecting the repressibility of arginine biosynthetic enzymes in Saccharomyces cerevisiae
Eur. J. Biochem.
1970
, vol. 
12
 (pg. 
31
-
39
)
Rose
 
M. D.
Winston
 
F.
Hieter
 
P.
 
Methods in Yeast Genetics: a laboratory course manual
1990
New York
Cold Spring Harbor Laboratory Press
Laemmli
 
U. K.
 
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
Nature
1970
, vol. 
227
 (pg. 
680
-
685
)
Kalckar
 
H. M.
 
Differential spectrophotometry of purine compounds by means of specific enzymes. II. Determination of adenine compounds
J. Biol. Chem.
1947
, vol. 
167
 (pg. 
445
-
459
)
Šebela
 
M.
Štosová
 
T.
Havliš
 
J.
Wielsch
 
N.
Thomas
 
H.
Zdráhal
 
Z.
Shevchenko
 
A.
 
Thermostable trypsin conjugates for high throughput proteomics: synthesis and performance evaluation
Proteomics
2006
, vol. 
6
 (pg. 
2959
-
2963
)
Thomas
 
H.
Havliš
 
J.
Peychl
 
J.
Shevchenko
 
A.
 
Dried-droplet probe preparation on AnchorChipTM targets for navigating the acquisition of matrix-assisted laser desorption/ionization time-of-flight spectra by fluorescence of matrix/analyte crystals
Rapid Commun. Mass Spectrom.
2004
, vol. 
18
 (pg. 
923
-
930
)
Novák
 
O.
Tarkowski
 
P.
Tarkowská
 
D.
Doležal
 
K.
Lenobel
 
R.
Strnad
 
M.
 
Quantitative analysis of cytokinins in plants by liquid chromatography/single quadrupole mass spectrometry
Anal. Chim. Acta
2003
, vol. 
480
 (pg. 
207
-
218
)
Guern
 
J.
Doree
 
M.
Leguay
 
J. J.
Heslot
 
H.
 
Enzymatic degradation of kinetin and some N6-substituted adenines in Schizosaccharomyces pombe
C.R. Hebd. Seances Acad. Sci., Ser. D, Sci. Nat.
1972
, vol. 
275
 (pg. 
377
-
380
)
Galuszka
 
P.
Spíchal
 
L.
Kopečný
 
D.
Tarkowski
 
P.
Frébortová
 
J.
Šebela
 
M.
Frébort
 
I.
 
Atta-ur-Rahman
 
 
Metabolism of plant hormones cytokinins and their function in signaling, cell differentiation and plant development
Studies in Natural Product Chemistry
2009
, vol. 
vol. 34
 
Amsterdam
Elsevier
(pg. 
203
-
264
)
Hall
 
T. A.
 
BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT
Nucleic Acids Symp. Ser.
1999
, vol. 
41
 (pg. 
95
-
98
)
Heyl
 
A.
Wulfetangle
 
K.
Pils
 
B.
Nielsen
 
N.
Romanov
 
G. A.
Schmülling
 
T.
 
Evolutionary proteomics identifies amino acids essential for ligand-binding of the cytokinin receptor CHASE domain
BMC Evol. Biol.
2007
, vol. 
7
 pg. 
62
 
Fasman
 
G. D.
 
Circular Dichroism and the Conformational Analysis of Biomolecules
1996
New York
Plenum Press
pg. 
740
 
Daddona
 
P. E.
Kelley
 
W. N.
 
Analysis of normal and mutant forms of human adenosine deaminase – a review
Mol. Cell. Biochem.
1980
, vol. 
29
 (pg. 
91
-
101
)
Deeley
 
M. C.
 
Adenine deaminase and adenine utilization in Saccharomyces cerevisiae
J. Bacteriol.
1992
, vol. 
174
 (pg. 
3102
-
3110
)
Grammaticakis
 
P.
 
Contribution a l'etude spectrale de quelques derives de l'isocyanate de phenyle. I. Absorption dans l'ultra-violet moyen des derives phenylcarbamyles de quelques amines, hydroxylamines, phenylhydrazines, cetimines et oximes
Bull. Soc. Chim. Fr.
1947
, vol. 
14
 (pg. 
664
-
674
)
Escusa
 
S.
Camblong
 
J.
Galan
 
J.-M.
Pinson
 
B.
Daignan-Fornier
 
B.
 
Proteasome- and SCF-dependent degradation of yeast adenine deaminase upon transition from proliferation to quiescence requires a new F-box protein named Saf1p
Mol. Microbiol.
2006
, vol. 
60
 (pg. 
1014
-
1025
)
Zimmermann
 
P.
Hirsch-Hoffmann
 
M.
Hennig
 
L.
Gruissem
 
W.
 
GENEVESTIGATOR: Arabidopsis microarray database and analysis toolbox
Plant Physiol.
2004
, vol. 
136
 (pg. 
2621
-
2632
)
Gowda
 
M.
Venu
 
R.-C.
Li
 
H.
Jantasuriyarat
 
C.
Chen
 
S.
Bellizzi
 
M.
Pampanwar
 
V.
Kim
 
H. R.
Dean
 
R. A.
Stahlberg
 
E.
, et al 
Magnaporthe grisea infection triggers RNA variation and antisense transcript expression in rice
Plant Physiol.
2007
, vol. 
144
 (pg. 
524
-
533
)
Xu
 
J.
Zhang
 
H.-Y.
Xie
 
C-H.
Xue
 
H.-W.
Dijkhuis
 
P.
Liu
 
C.-M.
 
EMBRYONIC FACTOR 1 encodes an AMP deaminase and is essential for the zygote to embryo transition in Arabidopsis
Plant J.
2005
, vol. 
42
 (pg. 
743
-
758
)
Tavladoraki
 
P.
Rossi
 
M. N.
Saccuti
 
G.
Perez-Amador
 
M. A.
Polticelli
 
F.
Angelini
 
R.
Federico
 
R.
 
Heterologous expression and biochemical characterization of a polyamine oxidase from Arabidopsis involved in polyamine back conversion
Plant Physiol.
2006
, vol. 
141
 (pg. 
1519
-
1532
)

Supplementary data