(Received for publication, November 5, 1996, and in revised form, March 3, 1997)
From the Institute for Molecular Science of Medicine, Heparan-sulfate 2-sulfotransferase (HS2ST), which
catalyzes the transfer of sulfate from 3 A variety of important biological functions have been proposed for
heparin and heparan sulfate (HS)1 such as
blood coagulation (1), cell recognition (2), cell adhesions (3, 4),
viral binding and infection (5), endocytosis (6), regulation of basic
fibroblast growth factor (7-11), and developmental regulation of
neural tissues (12). Such a broad range of functions of heparin and HS
may depend on their characteristic structures, with a high degree of
microheterogeneity resulting mainly from the different position and
extent of sulfation, which may lead to the specific interactions in a
"lock-and-key" manner with various proteins involved in the above
biological events. For example, N-deacetylation and
N-sulfation of N-acetylglucosamine residues and
2-O-sulfation of iduronic acid residues in HS are essential
steps in constructing the domain structure in HS for the binding of
basic fibroblast growth factor (10), and the binding was found to be
critical for its biological activity (7-9).
Since HS is a ubiquitous component in most, if not all, tissues (13),
it is likely that the regulation and biosynthetic mechanism of HS are
implicated in various biological functions of HS as described above,
compared with those of heparin, which is found only in connective
tissue mast cells. Heparin and HS share similar biosynthetic processes
(14), beginning with the polymerization reaction that produces
polysaccharide chain composed of alternating GlcA and GlcNAc units
(15). The resulting disaccharide repeats,
(-GlcA Cloning of genes for O-sulfotransferases and obtaining
antibodies against these low abundant proteins may be important not only to clarify the possible difference in the sulfation process between HS and heparin, but also to determine whether each sulfation in
the Golgi apparatus is subcompartmentalized and how each enzyme activity is regulated within the membrane. In this study, we cloned the
cDNA encoding HS2ST in CHO cells, which catalyzes the transfer of
sulfate from 3 Partially purified HS2ST was obtained by elution from
the second 3 Peptides
1 and 2 (see Table I) were used to design degenerate oligonucleotide
outer and inner primers (sense 1s and 1si or antisense 2a and 2ai) with
deoxyinosine substitution as shown in Fig. 1A. The first
strand of cDNA was synthesized by the reverse transcriptase
reaction using poly(A)+ RNA from CHO cells as template and
oligo(dT) as primer and was then used as template for PCR. The first
PCR was carried out using a pair of outer primers (1s and 2a), and the
second PCR was carried out using the first PCR product as template and
a pair of inner primers (1si and 2ai). Reaction products were analyzed
by agarose gel electrophoresis (see Fig. 1B). An amplified
DNA band of ~90 bp was excised, and the DNA fragment was recovered
from the gel using Jetsorb (Genomed) and subcloned into the
EcoRV site of pBluescript (Stratagene). These subclones were
characterized by sequencing. The radioactive probe used for screening
the cDNA library was prepared from the PCR product amplified in the
mixture containing primers 1s and 2a, the subcloned 90-bp DNA as
template, and [ Table I.
Amino acid sequences of peptides after in situ digestion of cultured
CHO, cell HS2ST with endoproteinase Lys-C followed by endoproteinase
Asp-N
The CHO cDNA library in
the Uni-ZAP XR vector (purchased from Stratagene) was constructed from
poly(A)+ RNA from CHO cells and oligo(dT) as primer at
EcoRI and XhoI recognition sites of the
pBluescript SK pBluescript plasmid DNA was purified
from SOLR cells carrying the pBluescript plasmid using QIAGEN plasmid
kits. The nucleotide sequence of the cloned cDNA was determined by
repeated sequencing of both strands of alkaline-denatured plasmid DNA
using deaza-GTP kits with Sequenase Version 2.0 (U. S. Biochemical
Corp.). DNA synthesis was primed by T3, T7, and internal primers
situated ~250 base pairs apart. The DNA sequences thus obtained were
compiled and analyzed using GENETYX-MAC computer programs (Software
Development Co., Ltd., Tokyo). The nucleotide and deduced amino acid
sequences were compared with other protein sequences in the nucleic
acid and protein data bases (EMBL Gene Data Bank Release 44 and NBRF Protein Data Bank Release 45).
Poly(A)+ RNA
prepared from cultured CHO cells was denatured and electrophoresed on
1.2% agarose gel containing 6% (v/v) formaldehyde. After treatment
with 50 mM NaOH and neutralization in 20 × SSC, the
RNA on the gel was transferred to a Hybond N+ nylon
membrane overnight and fixed by treatment with 50 mM NaOH. The RNA fixed on the membrane was prehybridized in a solution containing 50% formamide, 5 × SSPE, 5 × Denhardt's
solution, 0.5% SDS, and 100 µg/ml denatured salmon sperm DNA.
Hybridization was carried out in the same buffer containing
32P-labeled probe (1 × 106 cpm/ml) at
42 °C for 16 h. A 1145-bp fragment at positions 113-1257 was
prepared by digestion with SmaI and AflII of
clone K3 inserted into pBluescript (see Fig. 2A) and
radiolabeled for the probe by the random oligonucleotide-primed
labeling method using [
For
the construction of pcDNA3HS2ST, the
EcoRI-AflII fragment containing the open reading
frame of 1280 bp from positions COS-7 cells were plated onto 60-mm
culture dishes containing 3 ml of Dulbecco's modified Eagle's medium
(Life Technologies, Inc.) supplemented with 50 units/ml penicillin, 50 µg/ml streptomycin, and 10% (v/v) fetal bovine serum (Cytosystems
Pty. Ltd.), and cells were grown at 37 °C. When the cells were
60-70% confluent, COS-7 cells were transfected with pcDNA3HS2ST
or pcDNA3 alone and, in another experiment, with pFLAG-CMV-2HS2ST,
pFLAG-CMV-2R, or pFLAG-CMV-2 alone. The transfection was performed
using the DEAE-dextran method (27). After incubation in Dulbecco's
modified Eagle's medium and 10% fetal bovine serum containing
antibiotics for 67 h, the spent medium was collected, and the cell
layers (2 × 106 cells) were washed with Dulbecco's
modified Eagle's medium alone, scraped, and homogenized in 1 ml of 10 mM Tris-HCl, pH 7.2, 0.5% (w/v) Triton X-100, 0.15 M NaCl, 10 mM MgCl2, 2 mM CaCl2, and 20% (v/v) glycerol. The
homogenates were centrifuged at 10,000 × g for 30 min, and the activities of HS2ST, HS6ST, and chondroitin O-sulfotransferase in the supernatant fractions (cell
extracts) were measured as described previously (22, 23). In another experiment, FLAG (-DYKDDDDK-) fusion proteins in the cell extracts were
purified by anti-FLAG M2 (Kodak) affinity chromatography according to
the method described by the manufacturer, and the HS2ST and HS6ST
activities in the FLAG affinity-bound fractions were measured.
For the determination of HS2ST and HS6ST activities, the reaction
mixture (50 µl) contained 2.5 µmol of imidazole HCl, pH 6.8, 3.75 µg of protamine chloride, 25 nmol of completely desulfated and
N-resulfated heparin (Seikagaku Corp., Tokyo), 50 pmol of 35S-labeled 3 Purified HS2ST was immobilized on a
polyvinylidene difluoride membrane after SDS-polyacrylamide gel
electrophoresis and digested sequentially with endoproteinase Lys-C and
then endoproteinase Asp-N as described under "Experimental
Procedures." The generated peptides were separated on a reverse-phase
column by HPLC and analyzed with a protein sequenator. The amino acid
sequences of the five peptides analyzed are shown in Table
I. The residue at the amino terminus is consistent with
the one (aspartic acid) expected from the specificity of endoproteinase
Asp-N.
Degenerate
oligonucleotide primers were designed from peptides 1 and 2 as shown in
Fig. 1A (sense 1s and 1si or antisense 2a and
2ai). When primers 1s and 2a were used in a PCR with the first strand
cDNA of CHO cell poly(A)+ RNA as template, one DNA
fragment of ~90 bp was obtained (Fig. 1B, first
lane). This 90-bp product appeared to be specific because PCR
using primers 1si and 2ai, which were shifted to the 3 The cDNA insert obtained from clone K3
included the entire sequence of the clone H8 insert and appeared to
have the complete coding sequence of HS2ST. The nucleotide sequence of
the HS2ST cDNA and its predicted amino acid sequence are shown in
Fig. 2A. The amino-terminal sequence contains
four in-frame ATG codons. There is a TGA stop codon in-frame at
position -21 upstream from the first ATG codon. A single open reading
frame beginning at the first ATG codon predicts a protein of 356 amino
acid residues with a molecular mass of 41,830 Da with two potential
N-linked glycosylation sites. The calculated value of the
molecular mass is somewhat higher than the apparent molecular mass (38 kDa) previously estimated by SDS-polyacrylamide gel electrophoresis of
the purified protein after N-glycanase digestion under
nonreducing conditions (22). The possible explanations for the
discrepancy are that purified HS2ST might have been partially degraded
during the purification procedures or that the protein might contain
some intramolecular disulfide bonds. Hydropathic analysis of the
predicted amino acid sequence of HS2ST revealed one prominent
hydrophobic segment composed of 14 amino acid residues extending from
positions 14 to 27 in the amino-terminal region (Fig. 2B).
All the peptide sequences that had been obtained after digestion of the
purified protein with endoproteinase Lys-C followed by endoproteinase
Asp-N were found in the predicted protein sequence, confirming that the
cDNA clones encode the purified HS2ST protein.
Poly(A)+ RNA from cultured
CHO cells was prepared and hybridized with a radioactive probe prepared
from a SmaI-AflII fragment at positions 113-1257
of clone K3 (Fig. 2A). Two bands of 5.0 and 3.0 kilobases
were observed (Fig. 3).
Direct evidence that the isolated cDNA encodes the
HS2ST protein was obtained by expressing it in COS-7 cells. COS-7 cells were transfected with pcDNA3HS2ST, a recombinant plasmid containing the isolated cDNA in the mammalian expression vector pcDNA3.
The cell extract was prepared as described under "Experimental
Procedures," and activities of HS2ST, HS6ST and chondroitin
O-sulfotransferase in the extracts were determined. COS-7
cells expressed substantial levels of endogenous HS2ST activity as
shown in Table II. When the cells were transfected with
vector alone (pcDNA3), the HS2ST activity was unchanged compared
with nontransfected cells, whereas a 2.6-fold increase in the HS2ST
activity was observed in the cells transfected with
pcDNA3HS2ST (Table II, Experiment 1). The increase in the
activity was specific to HS2ST because neither the HS6ST nor the
chondroitin O-sulfotransferase activity increased. In
another experiment using the pFLAG-CMV-2 vector (Table II, Experiment
2), a 2.3-fold increase in the HS2ST activity was also observed in the
transfectant with pFLAG-CMV-2HS2ST, indicating that the increase in the
activity was not dependent on the vectors used. There was no increase
in the control transfectant with pFLAG-CMV-2 alone or with
pFLAG-CMV-2R. Furthermore, when the cell extract was applied to the
anti-FLAG antibody column, the HS2ST activity alone was recovered only
in the affinity fraction from the pFLAG-CMV-2HS2ST transfectant,
indicating that synthesized FLAG-HS2ST fusion protein has the HS2ST
activity. In addition, no significant change in the sulfotransferase
activity was observed in the spent medium of the transfected cells
(data not shown). The results thus demonstrate that the isolated
cDNA encodes a protein with HS2ST activity and most likely the
enzyme that catalyzes heparan sulfate 2-O-sulfation in
vivo.
Table II.
Overexpression of heparan-sulfate 2-sulfotransferase in COS-7 cells
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
-phosphoadenosine
5
-phosphosulfate to L-iduronic acid at position 2 in
heparan sulfate, was purified from cultured Chinese hamster ovary (CHO)
cells to apparent homogeneity (Kobayashi, M., Habuchi, H., Habuchi, O.,
Saito, M., and Kimata, K. (1996) J. Biol. Chem. 271, 7645-7653). The internal amino acid sequences were obtained from the
peptides after digestion of the purified protein with a combination of
endoproteinases. Mixed oligonucleotides based on the peptide sequences
were used as primers to obtain a probe fragment by reverse
transcriptase-polymerase chain reaction using CHO cell
poly(A)+ RNA as template. The clone obtained from a CHO
cDNA library by screening with the probe is 2.2 kilobases in size
and contains an open reading frame of 1068 bases encoding a new protein
composed of 356 amino acid residues. The protein predicts a type II
transmembrane topology similar to other Golgi membrane proteins.
Messages of 5.0 and 3.0 kilobases were observed in Northern analysis.
Evidence that the cDNA clone corresponds to the purified HS2ST
protein is as follows. (a) The predicted amino acid
sequence contains all five peptides obtained after endoproteinase
digestion of the purified protein; (b) the characteristics
of the predicted protein fit those of the purified protein in terms of
molecular mass, membrane localization, and N-glycosylation;
and (c) when the cDNA containing the entire coding
sequence of the enzyme in a eukaryotic expression vector was
transfected into COS-7 cells, the HS2ST activity increased 2.6-fold
over controls, and the FLAG-HS2ST fusion protein purified by affinity
chromatography showed the HS2ST activity alone.
1,4-GlcNAc
1,4-)n, subsequently receive a series of
reactions as follows: N-deacetylation of GlcNAc and further
N-sulfation, C-5 epimerization of GlcA to IdceA,
2-O-sulfation of IdceA, and 6-O- and
3-O-sulfation of N-sulfoglucosamine. However, heparin and HS are different in many respects. Heparin usually contains
more N- and O-sulfated groups and more IdceA than
HS (14). These distinctive characteristics may suggest some differences in their biosynthetic process and regulation.
N-Deacetylase/N-sulfotransferases have been
purified from (HS-producing) rat liver (16) as well as from
(heparin-producing) mouse mastocytoma (17). Molecular cloning studies
have shown that they are closely related, but are different proteins
(18-20). This appears to be also the case with
O-sulfotransferases. Wlad et al. (21) have
recently purified an ~60-kDa enzyme capable of catalyzing both the
2-O- and 6-O-sulfotransferase reactions from
mouse mastocytoma tissue, whereas we have separately purified ~44-kDa
heparan-sulfate 2-sulfotransferase (HS2ST) (22) and ~52-kDa
heparan-sulfate 6-sulfotransferase (HS6ST) (23) from the cell extract
and the spent medium of (HS-producing) Chinese hamster ovary (CHO) cell
culture, respectively, suggesting that our
O-sulfotransferases and that of Wlad et al. may
be distinct from each other. Taking into consideration these
differences, the molecular organization of the O- and
N-sulfation processes seems different between heparin and
HS.
-phosphoadenosine 5
-phosphosulfate specifically to
IdceA at position 2 in HS. The possible protein structure and properties predicted from the deduced amino acid sequence were consistent with those observed with the purified enzyme. In addition, the expression of the cDNA in COS-7 cells by transfection caused a
preferential increase in the activity of HS2ST. We discuss the biological significance of this cloning.
Purification of Heparan-sulfate 2-Sulfotransferase and Amino Acid
Sequencing of Peptides Obtained by Endoproteinase
Digestion
,5
-ADP-agarose column as described previously (22). A
portion of the purified HS2ST was subjected to SDS-polyacrylamide gel
electrophoresis (10% gel) according to the method of Laemmli (24)
after reduction and denaturation in loading buffer containing 5% (v/v)
2-mercaptoethanol. The protein separated by SDS-polyacrylamide gel
electrophoresis was then electrotransferred onto ProBlott polyvinylidene difluoride membrane (Applied Biosystems). The
transferred protein was visualized with Ponceau S according to
Aebersold et al. (25). The band migrating at ~44 kDa was
excised and subjected to reduction and S-carboxymethylation
on the polyvinylidene difluoride membrane according to Iwamatsu (26)
with some modifications. Briefly, the membrane-blotted protein was
reduced at room temperature in 8 M guanidine hydrochloride,
0.5 M Tris-HCl, pH 8.8, 5% (v/v) acetonitrile, and 0.33%
(w/v) dithiothreitol and then alkylated by the addition of iodoacetic
acid in 1 N NaOH. The S-carboxymethylated protein on the membrane was then incubated at room temperature with
0.5% (w/v) polyvinylpyrrolidone in 100 mM acetic acid and 0.33% methionine and washed with 10% (v/v) acetonitrile. The minced membrane was treated with 0.3 units of N-glycanase in 50 mM Tris-HCl, pH 7.5, at 37 °C for 15 h. Sequential
digestions with endoproteinase Lys-C and then endoproteinase Asp-N were
performed in 20 mM Tris-HCl, pH 9.0, and 10% acetonitrile
at 37 °C and in 20 mM ammonium bicarbonate, pH 7.8, 25 mM CaCl2, and 10% acetonitrile, respectively,
at 40 °C for 24 h at an enzyme/substrate ratio of 1:50
(mol/mol). Each endoproteinase digest was lyophilized and dissolved in
solvent A (1% acetonitrile in 0.06% (v/v) trifluoroacetic acid). The
generated peptides were separated on a reverse-phase column (0.3 × 150 mm) by capillary high performance liquid chromatography (HPLC).
Elution of the peptides was carried out with a gradient of 2-100%
solvent B (80% acetonitrile in 0.052% trifluoroacetic acid). The
peptide fractions were collected by monitoring the absorbance at 214 nm and blotted onto polyvinylidene difluoride membrane. Their amino acid
sequences were analyzed with a Model 476A protein sequenator (Applied
Biosystems).
-32P]dCTP (Amersham Corp.).
Peptide
Amino acid
sequence
1
DL(C)AKNRYHVLH(I)
2
DQ(V)RFVK(N)(I)
3
DXYRPGLX(R)
4
D(I)V(I)XYN(R)
5
D(L)YR
Fig. 1.
Oligonucleotide primer sequences derived from
the peptides of the purified heparan-sulfate 2-sulfotransferase and
analysis of the PCR products. A, the sequences derived from
peptides 1 and 2, which were used for the PCR experiment; B,
agarose gel electrophoresis of the products. The template was the first
strand cDNA of poly(A)+ RNA from CHO cells.
[View Larger Version of this Image (25K GIF file)]
plasmid in the Uni-ZAP XR vector. The host
strain Escherichia coli XL-1 Blue cells were infected with
phage from the library and plated at 2-4 × 104
plaque-forming units/dish. Approximately 1.4 × 106
plaques were screened by hybridization with 32P-labeled
probe as recommended by the manufacturer. The pBluescript plasmid from
the positive clones was excised from the Uni-ZAP XR vector according to
the Stratagene in vivo excision protocol using ExAssist
helper phage and E. coli SOLR.
-32P]dCTP and a Ready-To-Go DNA
labeling kit (Pharmacia Biotech Inc., Uppsala). The membrane was washed
at 65 °C in 1 × SSPE and 0.1% SDS and subsequently in
0.1 × SSPE and 0.1% SDS. The membrane was then exposed to x-ray
film.
Fig. 2.
Nucleotide sequence of the heparan-sulfate
2-sulfotransferase cDNA, the predicted amino acid sequence, and
hydropathic analysis of the protein. A, nucleotide 1 corresponds to the first nucleotide of the ATG triplet coding for the
initiator methionine. The nucleotides on the 5-side of position 1 are
indicated by negative numbers. The predicted amino acid
sequence is shown below the nucleotide sequence. The locations of the
five peptide sequences obtained after digestion of purified HS2ST with
endoproteinases Lys-C and Asp-N (Table I) are underlined
with solid lines; the sequence used as a PCR probe for
library screening is underlined with a dashed
line. The putative transmembrane hydrophobic domain is
boxed. The possible sites for N-glycosylation are
shown with black dots. B, the hydrophobicity
values were obtained according to the algorithm of Kyte and Doolittle
(43). Positive values represent increased hydrophobicity.
[View Larger Version of this Image (64K GIF file)]
23 to 1257 (AflII site)
shown in Fig. 2A was excised from the K3 clone and ligated
into the EcoRI and ApaI sites of the pcDNA3 expression vector (Invitrogen). The AflII end from the K3
clone and the ApaI end from pcDNA3 were blunted with T4
DNA polymerase before the EcoRI digestion. For the
construction of pFLAG-CMV-2HS2ST, the SacII-AflII
fragment (positions
7 to 1257) from the K3 clone was blunted and
ligated into the EcoRV site of the pFLAG-CMV-2 expression
vector (Eastman Kodak Co.). The plasmid (pFLAG-CMV-2R) that contained
the cDNA fragment in the reverse orientation was also constructed.
The plasmids were analyzed by restriction mapping to confirm the
correct orientation of the inserted cDNA.
-phosphoadenosine 5
-phosphosulfate (5 × 105 cpm), and enzyme. For the determination of
chondroitin O-sulfotransferase, the same reaction mixture
was used except that the completely desulfated and
N-resulfated heparin and 3.75 µg of protamine chloride were replaced by 25 nmol of chondroitin and 1.25 µg of protamine chloride, respectively. After incubation at 37 °C for 20 min, the
resultant 35S-labeled products were isolated by ethanol
precipitation and subsequent gel chromatography on a fast desalting
HR10/10 column (Pharmacia Biotech Inc.), and the radioactivity
incorporated was measured. To determine the HS2ST and HS6ST activities
separately, the isolated 35S-labeled products were digested
with a mixture of 10 milliunits of heparitinase I, 1 milliunit of
heparitinase II, and 10 milliunits of heparitinase III (Seikagaku
Corp.) in 40 µl of 50 mM Tris-HCl, pH 7.2, 1 mM CaCl2, and 4 µg of bovine serum albumin at
37 °C for 2 h. The digested products were injected together
with standard unsaturated disaccharides into a packed polyamine column
(YMC, Kyoto, Japan). Fractions of 0.6 ml were collected, and the
radioactivity was measured.
Amino Acid Sequence of Heparan-sulfate
2-Sulfotransferase
-ends of primers
1s and 2a by 9 and 3 bp, respectively, resulted in nearly the same size
fragment (Fig. 1B, second lane). The 90-bp fragment was sequenced and found to contain both the nucleotide sequences extended from primer 1s encoding the 3 carboxyl-terminal amino acids of peptide 1 (Leu, His, and Ile) and extended from primer
2a encoding the amino-terminal Asp of peptide 2 (Fig. 1A). The 90-bp PCR product was used as a probe to screen the CHO cDNA library. Six independent positive clones were obtained. The
nucleotide sequences for the two largest cDNA inserts of 2.2 kilobase pairs (clones K3 and H8) were determined.
Fig. 3.
Northern blot analysis. Polyadenylated
RNA (10 µg) prepared from cultured CHO cells was subjected to
Northern blot analysis as described under "Experimental
Procedures." The blot was probed with the 1145-base
SmaI-AflII fragment at positions 113-1257 from
clone K3 (see Fig. 2A). Bars show the positions of size standards.
[View Larger Version of this Image (23K GIF file)]
Plasmid
Sulfotransferase activitya
HS2ST
HS6ST
Chondroitin O-sulfotransferase
pmol/min/mg protein
Exp. 1
Cell extract
None
2.2
± 0.5
1.7 ± 0.5
6.4 ± 0.1
pcDNA3
2.2 ± 0.1
1.9 ± 0.3
6.7 ± 0.5
pcDNA3HS2ST
5.7 ± 0.7
1.6 ± 0.4
7.0
± 0.1
Exp. 2
Cell extract
pFLAG-CMV-2
3.9
± 0.4
1.6 ± 0.1
NDb
pFLAG-CMV-24
4.0 ± 0.1
1.1 ± 0.2
ND
pFLAG-CMV-2HS2ST
9.1 ± 0.1
1.4 ± 0.2
ND
FLAG affinity fractionc
pFLAG-CMV-2
0
0
ND
pFLAG-CMV-2R
0
0
ND
pFLAG-CMV-2HS2ST
1.5 ± 0.1d
0
ND
a
The activities determined in the absence of exogenous
sulfate acceptors were <0.1 pmol/min/mg of protein.
b
ND, not determined.
c
The recovery of the activity in the FLAG affinity fraction
was 29% of that in the cell extract.
d
The value is expressed as activity/mg of protein in the cell
extract.
The cloned cDNA corresponds to the HS2ST protein previously purified from cultured CHO cells (22). The evidence is as follows: (a) the predicted amino acid sequence of the protein contains all five peptides obtained from the purified enzyme; (b) the characteristics of the predicted protein are consistent with those of the purified enzyme in terms of molecular mass, membrane localization, and N-glycosylation; and (c) when the cDNA was introduced into a eukaryotic expression system, an increase in the HS2ST activity only was observed, and the isolation of the synthesized fusion protein revealed that the HS2ST activity alone was associated with the fusion protein.
The amino-terminal sequence of the HS2ST cDNA contains four
in-frame ATG codons. When the sequences surrounding the first ATG codon
are compared with the eukaryotic consensus translation sequence (28),
the purine at position 3 is not conserved, but G at position +4 is
conserved. Kozak (29) showed that G at position +4 is essential for
efficient translation in the absence of a purine at position
3. The
sequence surrounding the second, the third, and the fourth ATG codons
(Met7, Met8, and Met24 in Fig.
2A) also partially fit the consensus sequence; position
3
of these ATG codons is a purine (A, A, and G, respectively), whereas
position +4 is not G (A, C, and C, respectively). The fourth ATG codon
(Met24), however, is unlikely to be an initiation site
because of its location in the amino-terminal transmembrane domain
(Fig. 2A). It remains to be determined which ATG codons
could function as the initiation codon.
The amino acid sequence predicted from the cDNA suggests that CHO cell HS2ST belongs to a type II membrane protein. There is a short cytosolic hydrophilic domain followed by a hydrophobic domain at the amino terminus, which appears to serve as an uncleaved signal sequence anchor. The hydrophobic domain is then followed by a large hydrophilic domain that may face the lumen of the Golgi apparatus. These characteristics are similar to those of other Golgi proteins including heparin and heparan-sulfate N-deacetylase/N-sulfotransferases (18-20) and chondroitin 6-sulfotransferase (30).
Northern analysis at high stringency detected two transcripts of 3.0 and 5.0 kilobases. The presence of the two transcripts of different sizes suggests that CHO cells may have two different HS2ST messages from two different genes or differently spliced variants from the same gene. Such multiple transcripts of different sizes are also observed in other sulfotransferases (18, 20, 30-32), and as far as has been investigated, they are due to the difference in the size and sequence of the untranslated regions (31, 32). If this is the case for the HS2ST transcripts and the total coding sequences of the two transcripts are identical (1068 bases), large portions of the transcripts could be untranslated. The possible existence of largely different untranslated regions may be important in the function and distribution of the transcripts (32).
The transfection of the isolated cDNA into COS-7 cells caused a specific increase in the HS2ST activity without any significant effect on the other sulfotransferase activities examined (Table II). Furthermore, the fusion protein synthesized by the transfection showed the HS2ST activity exclusively. Recently, Bai and Esko (33) reported that a mutant CHO cell line defective in 2-O-sulfation showed a somewhat elevated level of 6-O-sulfation. On the other hand, biochemical studies on heparin biosynthesis in mastocytoma tissue suggested that the 2-O- and 6-O-sulfotransferase activities may share a common subunit or reside in a single protein (21). These observations argue that the molecular organization of the O-sulfation process may differ between heparin and HS, as is also the case with heparin and heparan-sulfate N-deacetylase/N-sulfotransferases (18-20). When the molecular cloning of the enzymes for O-sulfation in heparin and HS biosyntheses is completed, the possibility has to be clarified.
The predicted HS2ST protein showed no sequence homology to other
sulfotransferases including heparin and heparan-sulfate
N-deacetylase/N-sulfotransferases (18-20),
except that a stretch of 5 amino acid residues (DLYIL) at positions
338-342 (Fig. 2A) was found in chick chondrocyte chondroitin 6-sulfotransferase at positions 179-183 (30). In addition,
the size (356 amino acid residues) was much smaller than that of the
predicted N-deacetylase/N-sulfotransferases (882 amino acid residues). Neither the sequence often observed in
sulfotransferases, GXXGXXK (18), nor LEKCGR,
which was reported as a peptide sequence at the binding site for
3-phosphoadenosine 5
-phosphosulfate in arylsulfotransferase IV (34),
was found in HS2ST. Interestingly, however, a 274-amino acid overlap
with 53.6% identity to the HS2ST protein was found in the protein data
base for Drosophila segregation distorter protein (35),
which is involved in meiosis in Drosophila. This unexpected
result might suggest another HS2ST function.
2-O-Sulfation of IdceA in HS, which is catalyzed by a protein encoded by the cloned cDNA, is an essential step in the biosynthesis of the specific structural domain in HS for the binding of acidic fibroblast growth factor (36), basic fibroblast growth factor (11, 37, 38), and hepatocyte growth factor (39, 40). Recent studies (41, 42) have revealed that the binding of basic fibroblast growth factor to HS is also critical for the binding to their high affinity receptors and subsequent signal transduction. Therefore, the expression and activity of HS2ST should be strictly regulated in cells that are required for the activation and mitogenesis of these growth factors. The regulation mechanisms remain to be investigated.
The nucleotide sequence(s) reported in this paper has been submitted to the DNA Data Bank of Japan and the GenBankTM/EMBL Data Bank with accession number(s) D88811[GenBank].
We are grateful to Dr. Akihiro Iwamatsu (Kirin Brewery Co.) for helpful suggestions on peptide sequencing. M. K. thanks Dr. Akiharu Watanabe (Toyama Medical and Pharmaceutical University), Dr. Takashi Oguri (Aichi Medical University), Dr. Masaki Saito (Hokkaido University), and Dr. Akira Shigetomi (Kojin Hospital) for continuous support and encouragement.