(Received for publication, December 3, 1996, and in revised form, February 3, 1997)
From the Biotechnological Institute, Koglevej 2, DK-2970 Hørsholm, Denmark
Hexose oxidase from Chondrus crispus catalyzes the oxidation of a variety of mono- and disaccharides including D-glucose, D-galactose, maltose, and lactose. The enzyme has previously been partially purified and was reported to be a highly glycosylated, copper-containing protein with a relative molecular mass of approximately 130,000 (Sullivan, J. D., and Ikawa, M. (1973) Biochim. Biophys. Acta 309, 11-22). We report here the purification to homogeneity of hexose oxidase from C. crispus. The purified enzyme was cleaved with cyanogen bromide and endoproteinase Lys-C and the peptide fragments were subjected to amino acid sequence analysis. Oligonucleotides were designed on the basis of the peptide sequences and a cDNA clone encoding C. crispus hexose oxidase was obtained using polymerase chain reaction on reverse transcribed cDNA. The nucleotide sequence of the hexose oxidase cDNA contained an open reading frame of 546 amino acid residues with a predicted relative molecular mass of 61,898. No significant sequence similarity was found between hexose oxidase and other protein sequences available in data bases. Expression of the hexose oxidase cDNA in Pichia pastoris as an active enzyme confirmed the identity of the DNA sequence. Native hexose oxidase from C. crispus was characterized and compared with purified, recombinant enzyme.
It has previously been reported that the red algae Iridophycus flaccidum and Chondrus crispus contain the enzyme hexose oxidase (D-hexose:O2 oxidoreductase, EC 1.1.3.5) (1, 2). This enzyme catalyzes the oxidation of mono- and disaccharides to their corresponding lactones, with concomitant reduction of molecular oxygen to hydrogen peroxide. Hexose oxidase is able to oxidize a variety of substrates including D-glucose, D-galactose, maltose, cellobiose, and lactose. The wide substrate specificity clearly distinguishes this enzyme from glucose oxidase (EC 1.1.3.4), which is highly specific for D-glucose. Hexose oxidase from C. crispus has been partially purified and characterized by Sullivan and Ikawa (2) and was shown to be a copper-containing glycoprotein with a relative molecular mass (Mr) of ~130,000.
The functionally related enzyme, glucose oxidase, has been investigated
comprehensively (3, 4), and is now widely used in analytical
biochemistry and in the food industry. The applications of glucose
oxidase include (i) conversion of glucose to glucono--lactone, which
is used as an acidifying agent in cheese production, (ii) removal of
glucose from foods, (iii) production of hydrogen peroxide, and (iv)
removal of molecular oxygen from foods and pharmaceuticals (5). In
contrast to glucose oxidase, relatively little is known about the
structure, enzymology, and applicability of hexose oxidase. The wide
substrate specificity of hexose oxidase provides a greater applicability, but investigations of the enzyme have met with difficulties due to the low amounts of enzyme present in red algae. We
have therefore initiated work to develop a process for the production
of recombinant hexose oxidase, which will facilitate further
characterization and industrial use of the enzyme. We describe here the
purification and molecular cloning of hexose oxidase from C. crispus, and we present the cDNA sequence of the enzyme. Our
results show that the hexose oxidase gene of C. crispus encodes a polypeptide of 546 amino acids with a predicted
Mr of 61,898. Furthermore, we show that
recombinant hexose oxidase can be produced in the methylotropic yeast,
Pichia pastoris, and we discuss the similarities and
differences between the previously published data and those of the
native and recombinant hexose oxidase reported here.
Escherichia coli strain
DH5 (Life Technologies) was used for routine molecular biological
work. Heterologous expression experiments were conducted in P. pastoris strain KM71 (his4,
aox1::ARG4) (Research Corporation
Technologies, Inc., Tucson, AZ). The cloning vector pT7Blue for
E. coli was from Novagen, and the P. pastoris
expression vector pPIC3 was from Research Corporation Technologies.
The assay procedure was essentially as described by Sullivan and Ikawa (2). One enzyme unit is defined as that amount of enzyme which catalyzes the production of 1 nmol of H2O2/min at 25 °C and at a substrate concentration of 0.05 M.
Purification of Hexose Oxidase from C. crispusThe red
seaweed C. crispus was collected during April to September
at the shore near Grenå, Jutland, Denmark, at a depth of 2-5 meters.
Freshly collected fronds were rinsed with cold water and stored on ice
during transport to the laboratory (<24 h). The seaweed was either
air-dried immediately or stored frozen at 18 °C. The dried
material was ground to fine powder in a Waring commercial blendor
(Waring, New Hartford, CT). Approximately 100 g of C. crispus powder was mixed with 500 ml of 20 mM
Tris-HCl, pH 7.0, and extracted for 6-8 days at 4 °C. The extract
was collected by filtration through several layers of gauze, and the
seaweed residue was then subjected to 2-3 further extractions. The
filtrate was clarified by centrifugation, filtered through Whatman
chromatography paper (chr 1), diluted with water to a conductivity of
7-8 mS/cm, and adjusted to pH 7.5 with NaOH. The purification
procedure described below was carried out at 4 °C, except the final
chromatofocusing step which was performed at room temperature. (i) The
extract was immediately applied to a 250-ml DEAE-Sepharose Fast Flow
column (Pharmacia) equilibrated with 20 mM Tris-HCl, pH
7.5. Adsorbed proteins were eluted with a linear gradient of 0-500
mM NaCl in equilibration buffer, and collected fractions
were assayed for hexose oxidase activity. (ii) Hexose
oxidase-containing fractions from 3-6 DEAE-Sepharose chromatography
runs were pooled and concentrated to ~50 ml by ultrafiltration using
a membrane cell of nominal molecular weight 30,000 limit (Millipore).
The enzyme preparation was further concentrated to 10-20 ml in
Centriprep-30 concentrators (Amicon), clarified by centrifugation, and
filtered through a filter unit of 0.22-µm pore size. (iii) Aliquots
of 3.0-4.0 ml of the concentrated preparation of hexose oxidase were
loaded onto a 350-ml column of Sephacryl S-200 HR (Pharmacia, 2.6 × 70 cm) equilibrated with 20 mM Tris-HCl, 500 mM NaCl, pH 7.5, at a flow rate of 0.5 ml/min. Fractions of
2.5 ml were collected and assayed for hexose oxidase activity. (iv) The
pool from gel filtration was applied to a ~2.0-ml column of
phenyl-Sepharose 6 Fast Flow High Substitution (Pharmacia) equilibrated
in the buffer used for gel filtration. Hexose oxidase activity was
recovered in the run-through fraction. (v) The enzyme preparation from
the phenyl-Sepharose step was desalted on Sephadex G-25 columns (PD-10, Pharmacia) equilibrated and eluted with 25 mM
piperazine-HCl, pH 5.5, and loaded onto a 1-ml Mono P column
(Pharmacia) equilibrated in the same buffer. A gradient from pH 5.0 to
3.5 was generated by elution with 11 ml of 10% Polybuffer 74 (Pharmacia) adjusted to pH 3.5 with HCl. Collected fractions were
assayed for hexose oxidase activity.
Nondenaturing gradient gels containing 8-25% polyacrylamide were run and silver stained on a PhastSystem (Pharmacia). Staining for hexose oxidase activity was carried out as described for glucose oxidase (6).
SDS-PAGEMinigels containing 12% acrylamide/bisacrylamide (37.5:1) were run in a Mini-Protean II apparatus (Bio-Rad) according to Laemmli (7) and stained with Coomassie Brilliant Blue R-250.
SDS-PAGE followed by Staining for CarbohydratePurified hexose oxidase was subjected to SDS-PAGE, blotted to nitrocellulose, and stained for carbohydrate with the DIG Glycan Detection Kit (Boehringer Mannheim).
Isoelectric FocusingPurified hexose oxidase was run on IsoGel-agarose plates, pH 3-10 (FMC Bioproducts). A mixture of pI markers (FMC Bioproducts) was run in parallel with the hexose oxidase sample. The gels were stained with Coomassie Brilliant Blue R-250. The enzyme was also analyzed by isoelectric focusing on pre-cast polyacrylamide gels, pH 3.5-9.5 (Pharmacia, Ampholine PAG plates). These gels were stained for enzyme activity as described above.
Protein AssayProtein concentration was determined
according to Lowry (8) with the DC protein assay (Bio-Rad). Samples
from chromatofocusing were analyzed by the method of Bradford (9). The
standard was bovine -globulin.
The molecular mass of native hexose oxidase was determined by gel filtration on a column packed with Sephacryl S-200, as described above. The standard proteins used for calibration were ovalbumin (Mr 43,000), albumin (67,000), catalase (158,000), and aldolase (252,000).
Determination of KmThe initial reaction rate of hexose oxidase at varying concentrations of glucose and galactose at pH 6.3 and 25 °C was measured and plotted in a Hanes plot. The apparent Km values were calculated after linear regression.
Metal AnalysisThe metal content of purified hexose oxidase was determined by high resolution inductively coupled plasma mass spectrometry.
Preparation and Purification of Cyanogen Bromide Fragments of Hexose OxidasePurified hexose oxidase was digested with cyanogen bromide in 70% (v/v) formic acid. The resulting peptide fragments were separated by high resolution SDS-PAGE (10), transferred to PVDF membrane (Problott, Applied Biosystems) in a Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad), and stained with Coomassie Brilliant Blue R-250.
Preparation and Purification of Proteolytic Fragments of Hexose OxidasePurified hexose oxidase was concentrated by ultrafiltration, run on 12% SDS-polyacrylamide gels as described above, transferred to PVDF membrane (Immobilon P, Millipore), and stained with Coomassie Brilliant Blue R-250. Digestion of hexose oxidase polypeptides bound to PVDF with Endo Lys-C and extraction of the resultant proteolytic peptides was performed as described by Fernandez et al. (11, 12). The peptide fragments obtained by digestion of Mr 40,000 and 29,000 polypeptides of hexose oxidase were purified by reversed phase chromatography on a SMART system (Pharmacia) equipped with a µRPC C2/C18 SC2.1/10 narrow-bore column (Pharmacia) using a gradient of 0-80% acetonitrile in 0.1% trifluoroacetic acid.
Amino Acid Analysis and Protein SequencingAmino acid analysis was carried out by ion exchange chromatography and post-column derivatization with o-phtaldialdehyde (13). Protein sequencing was carried out by Edman degradation on an automated protein sequencer from Applied Biosystems (model 477A or Procise).
Isolation of RNA from C. crispusTotal RNA from C. crispus was isolated essentially as described by Chomczynski and Sacchi (14). The RNA was further purified by precipitation in ~3.5 M sodium acetate, pH 5.5. The precipitated RNA was pelleted by centrifugation, washed with 70% ethanol, and resuspended in water. Polyadenylated RNA was isolated from total RNA using magnetic beads containing oligo(dT) essentially as described by the manufacturer (Dynal, Norway).
cDNA SynthesisDegenerate oligonucleotide primers were
designed on the basis of the peptide sequences shown in Table II. The
primers were synthesized as mixtures of oligonucleotides including the
universal base deoxyinosine. Deoxyinosine was incorporated at 4-fold
redundancy positions in the 5-part of the primers, resulting in a
reduced complexity of the primer mixture. The 3
-end of the hexose gene was synthesized and amplified with the 3
-Amplifinder kit (Clontech) using the oligo(dT) primer of the kit and a sense primer (HOX4+: 5
-GARGGIAAYGAYGGIGARCTNTT-3
)2 designed on
the basis of the HOX-4 peptide sequence. The PCR reaction mixture was
fractionated on an agarose gel, and relevant fragments were purified
and sequenced. At least three independently derived fragments were
sequenced to exclude errors caused by misincorporations in PCR.
|
The 5-end of the hexose oxidase gene was isolated using the
5
-Amplifinder kit from Clontech. The primer hox3-
(5
-AACCAIARRTTIGANGCDATNAC-3
), specific for the HOX-3 peptide
sequence, was used in first strand cDNA synthesis. After ligation
of the anchor primer of the kit, PCR was performed with the
gene-specific primer and the anchor primer. Analysis of the PCR
reaction mixture was carried out as described above.
Total RNA from C. crispus
was fractionated on a denaturing formaldehyde-agarose gel and blotted
onto a Hybond C filter (Amersham) as described (15). A 400-base pair
cDNA fragment was synthesized by PCR using polyadenylated RNA as
template and the primers hox2+ (5-YTIGTIGARWSIGGNTAYGA-3
) and hox3
,
specific for the sequences of peptides HOX-2 and HOX-3, respectively.
This fragment was labeled with 32P (15) and used to probe
the Northern blot. After hybridization and washing, the filter was
subjected to autoradiography.
Primers hox5 (5
-ATGGCTACTCTTCCCCAGAAAG-3
) and hox3
(5
-ACCAAGTTTATAAAAAGCAACCATCAC-3
), specific for the 5
- and 3
-ends of the hexose oxidase gene, respectively, were used to synthesize and
amplify the entire gene encoding hexose oxidase. The gene was isolated
on a 1.9-kilobase pair fragment and inserted into pT7Blue. The
recombinant plasmid was cut with the restriction enzyme NdeI
and treated with Klenow polymerase. After heat-inactivation of the
Klenow polymerase the plasmid was restricted further with the enzyme
EcoRI. The excised hexose oxidase gene was inserted into the
P. pastoris expression vector pPIC3 on a
blunt-end/EcoRI fragment (plasmid pUPO153) and transformed
into P. pastoris strain KM71 using electroporation as
described in the procedures from Research Corporation Technologies.
P. pastoris cells containing the integrated plasmid pUPO153 were grown at 30 °C in shaker flasks in MD (1.34 g/liter of yeast nitrogen base (Difco), 0.4 mg/liter of biotin, 0.1% arginine, and 20 g/liter of glucose). At a density of OD600 = 15-20, the cells were harvested and shifted to induction medium, MM (1.34 g/liter of yeast nitrogen base, 0.4 mg/liter of biotin, 0.1% arginine, and 5% methanol). Three or four days after induction the cells were harvested by centrifugation, resuspended in 50 mM Tris-HCl, pH 7.5, and disrupted in a French Press (SLM Instruments, Inc., Rochester, NY) at an internal pressure of 20,000 p.s.i. The cell extract was cleared by centrifugation and the hexose oxidase containing supernatant was used for enzyme purification.
Purification of Recombinant Hexose Oxidase(i) Clarified extract of P. pastoris was loaded onto a 5-ml column of Q-Sepharose High Performance (HiTrap-Q, Pharmacia) equilibrated in 20 mM Tris-HCl, pH 7.5, and eluted with a linear gradient of 0-750 mM NaCl in equilibration buffer. Collected fractions were assayed for hexose oxidase activity as described above, and for alcohol oxidase activity in an assay which was identical to the hexose oxidase assay except that 0.5% methanol was used as substrate. (ii) The pool from Q-Sepharose chromatography was concentrated in Centriprep-30 concentrators before gel filtration on a Sephacryl S-200 HR column as described above. Collected fractions were assayed for hexose oxidase and alcohol oxidase activity. (iii) Pooled fractions were desalted by gel filtration on Sephadex G-25 columns (PD-10, Pharmacia) and applied to a 1-ml Mono Q column (Pharmacia) equilibrated in 20 mM Tris-HCl, pH 7.5, and eluted with a linear gradient of 0-500 mM NaCl in equilibration buffer. Collected fractions were assayed for hexose oxidase activity. (iv) The pool from the Mono Q column was purified by chromatofocusing on a Mono P column as described above.
Hexose oxidase was purified from fronds of the red alga C. crispus. Initial attempts to reproduce the previously published purification procedure (2) were unsuccessful, and therefore an alternative method was developed. This protocol comprised (i) anion exchange chromatography on DEAE-Sepharose, (ii) gel filtration on Sephacryl S-200, (iii) hydrophobic interaction chromatography on phenyl-Sepharose, and (iv) chromatofocusing on a Mono P column (Table I). The strongly colored and fluorescent extract contained large amounts of the red protein phycoerythrin (2), which, however, was almost completely removed during the gel filtration step. Residual amounts of this and other colored substances were adsorbed and thus eliminated by passing the enzyme preparation over a column of phenyl-Sepharose before the final Mono P step. The recovery of hexose oxidase activity was 13%, the enzyme was enriched 889-fold, and the specific activity of the final preparation of hexose oxidase was 16 × 103 units/mg. Protease inhibitors, like phenylmethylsulfonyl fluoride, reduced the assay response of hexose oxidase, and were therefore not included during extraction and purification of the enzyme.
|
Purified hexose oxidase migrated as a
single band in native, non-dissociating PAGE (Fig.
1A). The identity of the band was confirmed
by enzyme staining of the gel using glucose or galactose as substrate
in the dye-forming enzyme reaction (results not shown). In isoelectric
focusing two bands of pI 4.3 and 4.5, respectively, were observed (Fig.
2, lane 1). Both of these enzyme variants were enzymatically active, as shown by enzyme staining of the gel (Fig.
2, lane 2). The relative molecular mass of the enzyme, as
determined by gel filtration, was approximately 110,000. In SDS-PAGE
purified hexose oxidase showed two bands at Mr
29,000 and 40,000, respectively, and a faint band at
Mr 62,000 (Fig. 1B). The migration of
these bands was not affected by reduction of the sample before running
the gel (data not shown). None of the bands seen in SDS-PAGE were
glycosylated, as seen by staining for carbohydrate (data not shown).
Purified hexose oxidase was strongly inhibited by sodium
diethyldithiocarbamate. At an inhibitor concentration of 0.1 mM the residual activity was ~5%. An excess amount of
copper completely neutralized the inhibitory effect. Metal analysis of
the enzyme showed a copper content of 0.11%, corresponding to 2 copper
atoms per enzyme molecule of Mr approximately 110,000. The apparent Km values of hexose oxidase
were 2.7 mM for D-glucose and 3.8 mM for D-galactose. To determine the N-terminal
amino acid sequence of the hexose oxidase polypeptides, a sample of the
enzyme was subjected to SDS-PAGE and blotting to PVDF membrane. The
N-terminal amino acid sequence of the 29,000 polypeptide was TSYMH-.
The N-terminal amino acid of the 40,000 polypeptide was blocked, and
none of the attempts to deblock the N terminus (16) were successful.
Similarly, no N-terminal sequence was obtained for the 62,000 polypeptide. The 29,000 and 40,000 polypeptide bands from the PVDF blot
were also used for amino acid analysis (data not shown).
Amino Acid Sequencing of Peptide Fragments
Purified hexose oxidase was digested with cyanogen bromide. After separation of the peptide fragments by high resolution SDS-PAGE and blotting to PVDF membrane, a strong band of Mr ~9,000 was selected for amino acid sequencing (Table II) and amino acid analysis (data not shown). To obtain multiple amino acid sequences which were known to be derived from the same polypeptide chain, a sample of purified hexose oxidase was run on SDS-polyacrylamide gel and blotted to PVDF membrane. The polypeptide bands of Mr 40,000 and 29,000 were digested with Endo Lys-C, and the proteolytic peptides were purified by reversed phase high performance liquid chromatography. Five peptide peaks from the 40,000 polypeptide and two from the 29,000 polypeptide were selected for amino acid sequencing (Table II).
Isolation and Sequence Analysis of cDNA Encoding Hexose Oxidase from C. crispusPolyadenylated RNA was isolated from C. crispus and used as template in reverse transcription. The 3-end
of the hexose oxidase cDNA was isolated and analyzed as described
under "Experimental Procedures." The DNA sequence of the 3
-end
fragment contained an open reading frame with a coding capacity of 356 amino acids followed by two stop codons in tandem. The deduced amino
acid sequence contained, from its N-terminal end, the HOX-4, HOX-5, and
HOX-3 peptide sequences derived from the 40,000 polypeptide, the
N-terminal amino acid sequence of the 29,000 polypeptide, the HOX-7 and
HOX-8 peptide sequences from the 29,000 polypeptide, and the HOX-1
sequence of the 9,000 CNBr fragment (Table II). The 5
-end of the
hexose oxidase cDNA was isolated similarly, and DNA sequence
analysis of the 5
-end fragment showed a translation start codon
followed by an open reading frame with a coding capacity of 234 amino
acids. All the peptide sequences from the 40,000 polypeptide (HOX-6,
HOX-2, HOX-4, HOX-5, and HOX-3) were found in this open reading frame.
The overlap between the open reading frames encoded by the 5
-end and
3
-end fragments was 44 amino acid codons. The combined sequences
formed a contiguous 1,881-base pair cDNA sequence encoding a
polypeptide of 546 amino acids with a predicted
Mr of 61,898 (Fig. 3). All the
peptide sequences determined by amino acid sequencing were found in the
cDNA sequence, and their position in the cDNA sequence showed
that the Mr 40,000 and 29,000 polypeptides are
derived from the N- and C-terminal regions of hexose oxidase,
respectively.
Northern Blot Analysis
Total RNA from C. crispus was fractionated on a denaturing agarose gel and used for Northern blotting. The size of the hexose oxidase-specific transcript was approximately 1.9 kilobases.
Expression of the C. crispus Hexose Oxidase Gene in P. pastorisThe open reading frame from nucleotide numbers 1-1,638 (Fig. 3) was inserted into the P. pastoris expression vector pPIC3 between the strong methanol inducible promotor and the transcriptional termination signal of the alcohol oxidase gene, aox1, of P. pastoris. The hexose oxidase expression vector was inserted into the chromosome of P. pastoris strain KM71 which is defective in aox1, the primary alcohol oxidase gene. During growth of the recombinant P. pastoris cells, the aox1 promotor expressing the hexose oxidase gene was induced by shifting the cells to a medium containing methanol. The secondary aox2 gene of P. pastoris was also induced by methanol. The activity of hexose oxidase and alcohol oxidase increased during the first 3 days after induction and reached a maximum on days 3-4.
Purification of Recombinant Hexose Oxidase Expressed in P. pastorisThis purification procedure was similar to the one used for purification from C. crispus. The protocol included (i) anion exchange chromatography on Q-Sepharose, (ii) gel filtration on Sephacryl S-200, (iii) anion exchange chromatography on a Mono Q column, and (iv) chromatofocusing on a Mono P column. The separation of hexose oxidase from alcohol oxidase was an important aspect in the development of a purification procedure. The two oxidases co-eluted from the Q-Sepharose column, but were readily separated by gel filtration, since alcohol oxidase is a very large protein of Mr ~600,000 (17).
Characterization of Recombinant Hexose Oxidase from P. pastorisThe specific activity of purified, recombinant hexose oxidase from P. pastoris was identical to that of native enzyme from C. crispus. Similarly, the elution volumes of the two enzyme variants during gel filtration were identical. In SDS-PAGE purified, recombinant hexose oxidase migrated as two bands with molecular masses corresponding to the Mr 29,000 and 40,000 polypeptides of the native enzyme from C. crispus (Fig. 1C). The faint band at Mr 62,000 observed after purification from C. crispus was not seen on Coomassie-stained gels, but it was clearly visible after blotting to PVDF membrane and Coomassie staining of the blot (results not shown). To provide evidence that the amino acid sequence of hexose oxidase from C. crispus was in fact expressed in P. pastoris and responsible for the observed, recombinant enzyme activity, the Mr 40,000 and 29,000 polypeptides of recombinant hexose oxidase seen in SDS-PAGE were digested with Endo Lys-C as described above. After purification of the proteolytic peptides, one peptide from the 40,000 polypeptide and three peptides from the 29,000 polypeptide were subjected to amino acid sequencing. The peptide sequence derived from the 40,000 polypeptide corresponded to Asp8-Asn17 in the deduced amino acid sequence of the hexose oxidase gene. Similarly, the three peptide sequences derived from the 29,000 polypeptide corresponded to amino acid sequences in the C-terminal region of the deduced amino acid sequence (Leu434-Glu443, Tyr388-Thr397, and Trp452-Tyr461, respectively).
Substrate SpecificityThe substrate specificity of recombinant hexose oxidase was almost identical to that of the native enzyme (Table III). However, although the relative rate among disaccharides decreased for both enzyme forms in the order maltose, cellobiose, and lactose, the recombinant enzyme seemed less selective in its ability to oxidize these disaccharides. The results for the native enzyme were almost identical to the data reported earlier (2).
|
Recombinant hexose oxidase from P. pastoris was compared with the native enzyme with respect to inhibition by this compound. The two variants of the enzyme were inhibited to the same extent: inhibitor concentrations of 0.1 and 0.01 mM caused an inhibition of ~95 and ~40%, respectively. This result was in accordance with the previously published data for native hexose oxidase (2).
The purification scheme presented in this study resulted in a preparation of hexose oxidase with a specific activity 3-4-fold higher than those obtained in previous studies (2, 18). The high degree of purity may explain the fact that no carbohydrate was found in purified hexose oxidase, in contrast to the earlier report of a sugar content of 70% (2). The Mr of approximately 110,000 found by gel filtration in this study is similar to the Mr of 130,000 reported earlier (2). Since the open reading frame of the isolated cDNA corresponds to a polypeptide of Mr ~62,000, the enzyme might be of a homodimeric structure, as previously suggested (18). The presence of two copper atoms per enzyme molecule, as determined in this work, supports the assumption that hexose oxidase is a homodimer, since each of the identical subunits would then contain one atom of copper. This structure would resemble that of copper amine oxidases, in which each of the two identical subunits carries one Cu(II) ion (19, 20). The apparent Km values of the purified C. crispus enzyme were 2.7 mM for D-glucose and 3.8 mM for D-galactose. Corresponding values of 4 and 8 mM, respectively, have been reported earlier (2). In comparison, the functionally related enzyme glucose oxidase has a Km for glucose of 33 mM (21). These values indicate that hexose oxidase has a 10-fold higher substrate affinity than glucose oxidase.
Although purified hexose oxidase migrated as a single band in native
PAGE, three bands of Mr 29,000, 40,000, and
62,000 were observed in SDS-PAGE. These molecular masses suggested that
the two smaller polypeptides might be cleavage products derived
from the 62,000 polypeptide. This assumption was supported by the
cDNA sequence and the N-terminal amino acid sequence of the 29,000 polypeptide, which indicate that a 62,000 precursor polypeptide is
proteolytically cleaved into two smaller polypeptides of predicted Mr 37,000 and 25,000. The N-terminal amino acid
sequence of the 29,000 polypeptide showed that cleavage occurs at
the C-terminal side of a positively charged sequence in which five out
of nine residues are either lysine or arginine (-RPRKRHTSKTSYMH-).
This structure is similar to the well known proteolytic activation of
zymogens. However, further experiments should be carried out to
determine whether hexose oxidase is in fact activated by cleavage in
this position.
The amino acid compositions of the 40,000 and 29,000 polypeptides and the 9,000 CNBr fragment were in good agreement with the compositions predicted from the corresponding regions of the cDNA sequence (Met1-Lys337, Thr338-Lys546, and Tyr461-Lys546, respectively). Compared with an average protein (22), the enzyme is rich in aspartic acid, tryptophan, and tyrosine, and low in arginine, alanine, and serine. The fact that polypeptides with predicted Mr of 37,000 and 25,000 migrated in SDS-PAGE with apparent Mr of 40,000 and 29,000, respectively, might reflect their acidic amino acid composition, which may reduce binding of SDS and, thus, electrophoretic mobility in SDS-PAGE. The prevalence of acidic amino acids was confirmed by the low pI observed in isoelectric focusing (Fig. 2). The presence of two catalytically active variants of pI 4.3 and 4.5, respectively, might be associated with cleavage of the Mr 62,000 putative precursor polypeptide, but the polypeptide composition of the two pI variants is presently unknown.
The open reading frame of the isolated hexose oxidase cDNA
comprised 1,638 nucleotides encoding 546 amino acid residues. The nucleotide sequence around the Met start codon at position +1 (TTCACGGCT) is very similar to other C. crispus initiation sequences and to a plant consensus sequence
(TAAACAGCT) (23, 24). Another ATG codon is located at
nucleotide 34. The putative translation product from this open
reading frame, however, is only 25 amino acid residues, and the
sequence around the Met codon at
34 (CCGTGTTCA) shows
no similarity to the plant consensus sequence or to the known C. crispus initiation sequences. In addition, it has been shown that
such upstream, out-of-frame ATG codons will only reduce initiation at
the normal site if the context around the out-of-frame codon is close
to the optimal consensus sequence (25). Therefore, we assume that the
ATG triplet at nucleotide +1 is used as translation start codon.
Northern blot analysis showed that the length of the mRNA encoding
hexose oxidase is 1.9 kilobases. This size is in accordance with the
cDNA of 1,881 base pairs, which was the longest cDNA isolated
(Fig. 3). Therefore, we conclude that the DNA sequence shown in Fig. 3
represents the full-length cDNA sequence of the C. crispus hexose oxidase gene.
The hexose oxidase amino acid sequence deduced from the nucleotide
sequence (Fig. 3) was compared with the sequences present in the
GenBank and European Molecular Biology Laboratory (EMBL) data bases.
None of the data base sequences, however, were clearly similar to
hexose oxidase. The highest scoring protein was the berberine bridge
enzyme ((S)-reticuline:oxygen oxidoreductase (methylene-bridge-forming), EC 1.5.3.9) from the plant
Eschscholtzia californica (27), which showed a low, but
hardly significant, sequence similarity to hexose oxidase. One of the
best matching regions in the two sequences was a tetrapeptide motif,
-SGGH-, assumed to be involved in covalent binding of flavin in
berberine bridge enzyme and in 6-hydroxy-D-nicotine oxidase
(EC 1.5.3.6) (28). However, UV/VIS scanning spectroscopy of purified
hexose oxidase showed no peaks in the 380 and 460 nm regions
characteristic of FAD (data not shown). Similarly, no FAD was found by
chemical and spectral analysis in previous studies (2). Since hexose oxidase contains copper atoms, a search for consensus sequences of
copper-binding domains was carried out. None of the known
copper-binding sites (20, 29-31) were observed in the hexose oxidase
amino acid sequence (Fig. 3). However, a -HYH- motif is present at
amino acid 297-299, but it remains to be demonstrated whether these histidine residues are involved in copper binding in C. crispus hexose oxidase. The deduced amino acid sequence of hexose
oxidase contains two potential N-glycosylation sites
(Asn95 and Asn358). No glycosylation, however,
was detected by SDS-PAGE and staining for carbohydrate. This result was
supported by amino acid sequencing of peptide HOX-2, in which an
unmodified asparagine was released at position 95 (Table II). The
predicted amino acid sequence (Fig. 3) suggests that hexose oxidase
might be an extracellular enzyme, since a putative C. crispus transit peptide cleavage site (24, 26) is located at amino
acids 29-33 (-PSMK-). However, the facts (i) that recombinant
hexose oxidase was exclusively found in the cytoplasmic fraction of
P. pastoris, (ii) that the amino acid sequence of the
putative transit peptide (residue numbers 1-31) does not resemble
other known transit peptide sequences, and (iii) that the peptide
sequence HOX-6 was found in the active enzyme, all support the
assumption that the enzyme does not contain a functional transit
peptide.
Induction of recombinant P. pastoris cells activated synthesis of C. crispus hexose oxidase. In addition, large amounts of alcohol oxidase were produced, due to activation of the aox2 gene of P. pastoris. The expression level of hexose oxidase was rather low compared with that of alcohol oxidase. The low level of expression may reflect differences in codon usage between highly expressed P. pastoris genes and the hexose oxidase gene of C. crispus. In the nucleotide sequence of the hexose oxidase gene (Fig. 3) the frequencies of CpG doublets and CpNpG triplets do not deviate significantly from their expected values, resulting in an almost equal distribution of codons for the degenerate codon families. This is in contrast to the high degree of codon bias in highly expressed P. pastoris genes. For example, a comparison of the codon usage in four highly expressed P. pastoris genes (aox1 (32), aox2 (32), das (dihydroxy-acetone synthase, Ref. 33), and gap (glyceraldehyde-3-phosphate dehydrogenase),3 and in the hexose oxidase gene shows that the codons CCG (proline), GCG (alanine), and CGC (arginine) do not occur in the P. pastoris genes. Their frequencies in the hexose oxidase gene, however, are 0.31, 0.28, and 0.38, respectively. A codon usage similar to the one observed in the hexose oxidase gene was previously found in the GapA and GapC genes of C. crispus (24). Thus, our results support the hypothesis that CpG and CpNpG methylation is absent in red algae (24).
The chromatographic properties of recombinant hexose oxidase observed during purification were similar to those of the native enzyme, and almost identical purification procedures could be used for the two enzyme variants. Similarly, (i) the specific activity, (ii) the molecular mass determined by gel filtration, and (iii) the sensitivity to inhibition by sodium diethyldithiocarbamate of the recombinant enzyme was identical to that of the native enzyme. In SDS-PAGE, recombinant hexose oxidase showed two bands corresponding to the Mr 29,000 and 40,000 bands of the native enzyme (Fig. 1C). This pattern suggests that the recombinant protein undergoes cleavage, similar to the native enzyme. Furthermore, the identity of the recombinant enzyme was confirmed by amino acid sequencing of peptide fragments derived from the recombinant hexose oxidase. All the peptide sequences corresponded to sequences in the amino acid sequence deduced from the cDNA sequence. The substrate specificity of the recombinant enzyme was almost identical to that of the native hexose oxidase. The relative rate among disaccharides decreased for both enzyme variants in the order maltose, cellobiose, and lactose, although the recombinant enzyme seemed less selective in its ability to oxidize these disaccharides (Table III).
Based on our comparison of the two enzyme variants, we conclude that hexose oxidase from C. crispus can be expressed in P. pastoris, and that the structural and catalytic properties of the recombinant enzyme are similar, if not identical, to those of the native enzyme. The availability of recombinant hexose oxidase will greatly facilitate further structural characterization of the enzyme and also permit further investigations of the analytical and industrial applicability of the enzyme.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U89770[GenBank] (Fig. 3).
We thank Ulla Poulsen and Britt Grønvold Olsen for skillful technical assistance; Arne L. Jensen, Department of Protein Chemistry, University of Copenhagen, Bente Isbye and Anne Blicher, Department of Biochemistry and Nutrition, Technical University of Denmark, for performing the amino acid analysis and protein sequencing; Dr. Stefan Stürup, Environmental Science and Technology Department, Risø National Laboratory, Denmark, for performing the determination of copper content; and our colleagues Drs. Lars Bjarne Nielsen and José Arnau for valuable discussions and critically reading the manuscript.