Full text of DOI:10.1093/glycob/11.11.997

Glycobiology vol. 11 no. 11 pp. 997–1008, 2001
The polysialyltransferase ST8Sia II/STX: posttranslational processing and role of
autopolysialylation in the polysialylation of neural cell adhesion molecule
Brett E. Close, Jennifer M. Wilkinson, Tracy J. Bohrer,
Clement P. Goodwin, Lucy J. Broom, and Karen J. Colley1
and developmentally regulated event (Kiss and Rougon, 1997;
Muhlenhoff et al., 1998). Though this unusual carbohydrate
structure is found in neuroinvasive bacteria (Robbins et al.,
1974; Vimr et al., 1995) and in fish and sea urchin eggs (Inoue
and Iwasaki, 1978; Kitazume et al., 1994), the expression of
polysialic acid in mammalian cells is of particular interest
because of its established role as an antiadhesive agent during
development and in metastasis. Though many forms of sialic
acid have been identified, the predominant form of polysialic
acid found in mammals is a linear homopolymer of 5-N-acetyl-
neuraminic acid (Neu5Ac) linked by α2,8-glycosidic bonds
(Troy, 1992). Mammalian polysialic acid is unique in that it is
found on a small group of proteins including the α-subunit of
the rat brain voltage-sensitive sodium channel (Zuber et al.,
1992) and the polysialyltransferase enzymes themselves
(Close and Colley, 1998). However, the most abundant carrier
of polysialic acid in mammals is the neural cell adhesion
molecule (NCAM) (Edelman and Crossin, 1991; Troy, 1995).
During embryogenesis, NCAM is present in a highly poly-
sialylated form, with sialic acid comprising ∼30% of its molecular
mass (Rothbard et al., 1982; Goridis and Brunet, 1992). High
levels of polysialylated NCAM are observed in developing
tissues, such as heart and muscle (Finne et al., 1987; Watanabe
et al., 1992; Dubois et al., 1994), kidney (Roth et al., 1987,
1988b; Troy, 1995), and brain (Finne et al., 1983; Berardi et al.,
1995). However, the majority of adult tissues express very low
levels or no polysialic acid (Edelman, 1985; Rutishauser et al.,
1988; Troy, 1995). Many lines of evidence indicate that the
presence of highly polysialylated NCAM at the cell surface
disrupts the homophilic binding properties of NCAM, as well
as general cell adhesion (Sadoul et al., 1983; Rutishauser et al.,
1985; Rutishauser, 1996). This in turn facilitates cell migration
during development (Rutishauser et al., 1988; Acheson et al.,
1991; Rutishauser and Landmesser, 1996) and neurite
outgrowth (Kiss and Rougon, 1997; Franceschini et al., 2001).
Polysialylated NCAM is also reexpressed on some metastatic
cancers, such as neuroblastoma (Livingston et al., 1988;
Moolenaar et al., 1990; Seidenfaden and Hildebrandt, 2001),
small cell lung carcinoma (Kibbelaar et al., 1989; Nilsson,
1992; Miyahara et al., 2001), and the highly metastatic kidney
tumor, Wilms’ tumor (Roth et al., 1988a; Gluer et al., 1998).
Akin to its role during development, this reexpression of poly-
sialic acid is thought to enhance the metastatic potential of
tumor cells by decreasing their adhesion and facilitating their
migration (Michalides et al., 1994; Scheidegger et al., 1994;
Fukuda, 1996).
Department of Biochemistry and Molecular Biology, University of Illinois
College of Medicine, 1819 West Polk Street M/C 536, Chicago, IL 60612,
USA
Received on June 25, 2001; revised on August 15, 2001; accepted on
August 17, 2001
The presence of α2,8-linked polysialic acid on the neural
cell adhesion molecule (NCAM) is known to modulate cell
interactions during development and oncogenesis. Two
enzymes, the α2,8-polysialyltransferases ST8Sia IV/PST
and ST8Sia II/STX are responsible for the polysialylation
of NCAM. We previously reported that both ST8Sia IV/PST
and ST8Sia II/STX enzymes are themselves modified by
α2,8-linked polysialic acid chains, a process called auto-
polysialylation. In the case of ST8Sia IV/PST, autopolysialyl-
ation is not required for enzymatic activity. However,
whether the autopolysialylation of ST8Sia II/STX is
required for its ability to polysialylate NCAM is unknown.
To understand how autopolysialylation impacts ST8Sia II/STX
enzymatic activity, we employed a mutagenesis approach.
We found that ST8Sia II/STX is modified by six Asn-linked
oligosaccharides and that polysialic acid is distributed
among the oligosaccharides modifying Asn 89, 219, and
234. Coexpression of a nonautopolysialylated ST8Sia II/STX
mutant with NCAM demonstrated that autopolysialylation
is not required for ST8Sia II/STX polysialyltransferase
activity. In addition, catalytically active, nonautopoly-
sialylated ST8Sia II/STX does not polysialylate any endo-
genous COS-1 cell proteins, highlighting the protein
specificity of polysialylation. Furthermore, immunoblot
analysis of NCAM polysialylation by autopolysialylated
and nonautopolysialylated ST8Sia II/STX suggests that the
NCAM is polysialylated to a higher degree by autopoly-
sialylated ST8Sia II/STX. Therefore, we conclude that
autopolysialylation of ST8Sia II/STX, like that of ST8Sia
IV/PST, is not required for, but does enhance, NCAM
polysialylation.
Key words: autopolysialylation/glycosylation/NCAM/
polysialyltransferase/ST8Sia II
Introduction
The addition of sialic acid polymers (polysialic acid) to the
termini of glycoprotein oligosaccharides is a posttranslational
Biosynthesis of polysialic acid in mammalian cells is carried
out by two closely related enzymes, ST8Sia IV/PST
(Nakayama et al., 1995; Yoshida et al., 1995; Eckhardt et al.,
1995; Phillips et al., 1997) and ST8Sia II/STX (Livingston and
1To whom correspondence should be addressed
© 2001 Oxford University Press
997

B.E. Close et al.
Paulson, 1993; Kojima et al., 1995; Scheidegger et al., 1995).
Whereas ST8Sia IV/PST and ST8Sia II/STX both catalyze the
same polysialylation reaction, their expression patterns and
polysialylation site preferences on NCAM differ. ST8Sia IV/PST
and ST8Sia II/STX mRNA are both highly expressed in fetal
brain (Livingston and Paulson, 1993; Nakayama et al., 1995;
Yoshida et al., 1995; Kurosawa et al., 1997). However, though
ST8Sia II/STX mRNA is absent in the adult brain (Livingston
and Paulson, 1993), ST8Sia IV/PST mRNA expression
persists in a number of somatic tissues, as well as in areas of
the adult brain that require continuing neurogenesis and neural
plasticity (Nakayama et al., 1995; Yoshida et al., 1995; Phillips
et al., 1997). The distinct temporal and tissue-specific expression
patterns of the polysialyltransferases suggest independent
regulation at the transcriptional level (Seidenfaden et al.,
2000), and imply that ST8Sia II/STX is responsible for the
embryonic polysialylation of NCAM (Angata et al., 1997),
whereas ST8Sia IV/PST is the dominant polysialyltransferase
in the adult brain (Hildebrandt et al., 1998).
In addition to the differences in their expression patterns,
ST8Sia IV/PST and ST8Sia II/STX also exhibit distinct preferences
with regard to the NCAM oligosaccharides they polysialylate.
Members of the immunoglobulin superfamily of proteins, the
NCAM 140 and 180 isoforms are composed of five Ig-like
domains, two fibronectin type III repeats, and a trans-
membrane spanning segment followed by a cytoplasmic tail
(Cunningham et al., 1987; Goridis and Brunet, 1992). NCAM
is modified by six Asn-linked oligosaccharides, three of which
reside within the fifth Ig-like domain (numbered 4 through 6)
and serve as the acceptors for polysialic acid addition (Nelson
et al., 1995; Franceschini et al., 2001). Work by Angata et al.
(1998) demonstrated that ST8Sia IV/PST preferentially adds
polysialic acid to the oligosaccharide on Asn 6, with a small
amount added to the oligosaccharide on Asn 5. In contrast,
ST8Sia II/STX evenly distributes polysialic acid between the
oligosaccharides on Asn 5 and Asn 6 (Angata et al., 1998).
Furthermore, ST8Sia IV/PST synthesizes a larger amount of
polysialic acid on NCAM than ST8Sia II/STX (Angata et al.,
1998). It is interesting to note that we have also observed that
ST8Sia IV/PST is autopolysialylated to a greater degree than
ST8Sia II/STX (Close and Colley, 1998).
We previously reported that the ST8Sia IV/PST and ST8Sia
II/STX polysialyltransferases are modified by α2,8-linked
polysialic acid chains when expressed in COS-1 cells (Close
and Colley, 1998), a process termed autopolysialylation
(Muhlenhoff et al., 1996). These autopolysialylated enzymes
localize to the Golgi, the cell surface, and are found soluble in
the extracellular space. In addition, following their expression
in COS-1 cells, we found that these polysialyltransferase
proteins are the only polysialylated proteins expressed,
suggesting that polysialylation is a protein-specific modification
(Close and Colley, 1998; Close et al., 2000). Subsequent anal-
ysis revealed that ST8Sia IV/PST is modified by five Asn-
linked oligosaccharides, and that the oligosaccharide modi-
fying Asn 74 possesses the majority of the polysialic acid
(Close et al., 2000). Furthermore, when nonautopolysialylated
mutants of ST8Sia IV/PST were coexpressed with full-length
and soluble NCAM, we found that ST8Sia IV/PST autopolysi-
alylation is not required for NCAM polysialylation, but that it
did seem to increase the amount of polysialic acid added to
NCAM (Close et al., 2000). Both ST8Sia IV/PST and ST8Sia
II/STX share 57% amino acid identity and catalyze the same
polysialylation reaction on NCAM. This high degree of simi-
larity suggests that the glycosylation pattern and activity
requirements of ST8Sia II/STX may parallel that of ST8Sia
IV/PST. Here we report on the carbohydrate modifications of
ST8Sia II/STX and identify the requirements for ST8Sia II/
STX autopolysialylation, as well as the requirements for poly-
sialylation of NCAM by ST8Sia II/STX.
Results
ST8Sia II/STX is modified by six Asn-linked oligosaccharides
In order to determine whether autopolysialylation of ST8Sia II/STX
is a prerequisite for its enzymatic activity, we first had to determine
which N-glycosylation consensus sites of ST8Sia II/STX are
utilized. Human ST8Sia II/STX (Scheidegger et al., 1995) has
six Asn residues positioned in potential Asn-X-Ser/Thr glyco-
sylation sites (Figure 1A), and a seventh Asn residue (Asn 206)
found in a disallowed Asn-Pro-Ser/Thr sequence (Kornfeld
and Kornfeld, 1985). Asn residues in each consensus glyco-
sylation site were mutated to Ser, as described in Materials and
methods. COS-1 cells expressing wild-type and mutant ST8Sia
II/STX proteins were metabolically labeled with 35S-Express
protein labeling mix for only 1 h to detect N-glycosylated but
nonautopolysialylated species (Close and Colley, 1998). V5
epitope-tagged enzymes were immunoprecipitated from cell
lysates using the anti-V5 epitope tag antibody (V5 Ab) and
separated on sodium dodecyl sulfate (SDS)–polyacrylamide
gels.
Empirically, a single nonautopolysialylated Asn-linked
oligosaccharide contributes approximately 4 kDa to the molecular
mass of a protein. Thus, mutagenesis of the Asn residue in a
glycosylated consensus sequence will result in a protein that
migrates ∼4 kDa smaller than the wild-type protein. Analysis of
wild-type glycosylated but nonautopolysialylated ST8Sia II/STX
by SDS–polyacrylamide gel electrophoresis (PAGE) revealed
that the largest form of the enzyme migrated with an apparent
molecular mass of 67 kDa. This corresponds to the expected
molecular mass of human ST8Sia II/STX modified by the V5
epitope tag and six Asn-linked oligosaccharides (Figure 1B,
WT ST8Sia II/STX). Interestingly, six additional lower molecular
mass forms decreasing in size by ∼4 kDa increments were also
observed in the lane containing the immunoprecipitated wild-
type protein. Based on their molecular masses (43–63 kDa), it
is likely that these bands represent differentially glycosylated
forms of the enzyme with zero to six Asn-linked oligosaccharides
added. Some or all of these lower-molecular-mass forms were
also observed in lanes containing the mutant ST8Sia II/STX
proteins described below, and these observations indicate some
inefficiency in ST8Sia II/STX glycosylation in the COS-1 cell
system. Individual mutations of Asn 60, 72, 89, 134, 219, and
234 to Ser resulted in proteins that migrated with reduced
molecular mass (63 kDa) in comparison to wild-type ST8Sia
II/STX (Figure 1B, compare the largest forms of WT ST8Sia
II/STX to those of N60S, N72S, N89S, N134S, N219S, and
N234S). This demonstrated each of these Asn residues is
modified by an Asn-linked oligosaccharide in the wild-type
ST8Sia II/STX protein. In contrast, no change in molecular
mass occurred when Asn 206 was converted to Ser, indicating
that this residue is not glycosylated (data not shown). These
998

ST8Sia II/STX autopolysialylation and enzymatic activity requirements
nonautopolysialylated form and a polydisperse form with an
apparent molecular mass ranging from ∼68 kDa to ∼160 kDa
(Figure 2A, WT). Previous work established that the 60-kDa
form represents a population of ST8Sia II/STX secreted from
the cell modified only by high mannose oligosaccharides
(Figure 2A, see arrow), whereas the larger form of ST8Sia II/STX
represents the autopolysialylated enzyme (Close and Colley,
1998). Mutation of Asn 0, 1, or 3 to Ser resulted in proteins that
appeared to be autopolysialylated to the same extent as the
wild-type ST8Sia II/STX (Figure 2A, mutants 0, 1, and 3). In
Fig. 1. ST8Sia II/STX is modified by five Asn-linked oligosaccharides.
(A) Schematic representation of the ST8Sia II/STX and ST8Sia IV/PST
enzymes. The numeric position of each Asn residue in the consensus
N-glycosylation sites is given. Also shown are the transmembrane domain
(TMD) and the sialyl motifs (L, S, and VS). The question mark (?) denotes a
possible site of autopolysialylation of ST8Sia IV/PST. (B) Wild-type (WT) and
mutant V5-tagged ST8Sia II/STX proteins were expressed in COS-1 cells,
metabolically labeled for 1 h with 35S-Express protein labeling mix and
immunoprecipitated from cell lysates with V5 Ab. The samples were separated
on 7.5% SDS–polyacrylamide gels, and radiolabeled protein bands visualized
by fluorography. The molecular mass marker shown is 50.7 kDa, ovalbumin.
data demonstrate that human ST8Sia II/STX is modified by six
Asn-linked oligosaccharides. For the purpose of directly
comparing analogous glycosylation sites of ST8Sia II/STX and
ST8Sia IV/PST (see Close et al., 2000), we will refer to Asn
residues 60, 72, 89, 134, 219, and 234 as Asn 0, 1, 2, 3, 4, and
5, respectively (Figure 1B, bottom).
Fig. 2. Elimination of glycosylation site 2 (Asn 89), site 4 (Asn 219), or site 5
(Asn 234) decreases the autopolysialylation of ST8Sia II/STX. (A) Wild-type
(WT) and single Ser mutant V5-tagged ST8Sia II/STX proteins were transiently
expressed in COS-1 cells, metabolically labeled with 35S-Express protein
labeling mix for 1 h and immunoprecipitated from 6 h chase medium with the
V5 Ab. The immunoprecipitated samples were electrophoresed on 7.5%
SDS–polyacrylamide gels, and radiolabeled protein bands visualized by
fluorography. The arrow indicates the position of the form of ST8Sia II/STX
possessing only high-mannose oligosaccharides. (B) A parallel experiment was
conducted without metabolic labeling of the polysialyltransferases. Wild-type
(WT) and mutant V5-tagged ST8Sia II/STX proteins were immunoprecipitated
from COS-1 cell lysates and medium 18 h posttransfection, separated on 7.5%
SDS–polyacrylamide gels, and electrophoretically transferred to nitrocellulose
membranes. The membranes were subjected to immunoblot analysis using
OL.28 Ab. Molecular mass markers are as follows: 203 kDa, myosin; 123 kDa,
β-galactosidase; 83 kDa, BSA; 50.7 kDa, ovalbumin.
Glycosylation of Asn 89 (site 2), Asn 219 (site 4), and Asn 234
(site 5) is necessary for ST8Sia II/STX autopolysialylation
To establish whether the ST8Sia II/STX glycosylation mutants
were autopolysialylated like wild-type enzyme, we expressed
these mutant proteins in COS-1 cells that lack endogenous
polysialyltransferases and their known glycoprotein substrates
(Close and Colley, 1998). The expressing cells were meta-
bolically labeled with 35S-Express protein labeling mix for 1 h and
chased with unlabeled medium for 6 h; the mutant ST8Sia II/STX
proteins were immunoprecipitated from the cell medium and
analyzed by SDS–PAGE and fluorography. Wild-type ST8Sia
II/STX appeared as two forms: a 60-kDa glycosylated but
999

B.E. Close et al.
contrast, mutation of Asn 2, Asn 4, or Asn 5 to Ser resulted in a
substantial reduction in the high-molecular-mass autopolysialylated
form of the enzyme, suggesting that the oligosaccharides attached
to these Asn residues were modified by polysialic acid (Figure 2A,
mutants 2, 4, and 5). To confirm that the polydisperse, high-
molecular-mass forms of wild-type and mutant ST8Sia II/STX
proteins are modified by α2,8-linked polysialic acid and that
autopolysialylation of the mutant 2, 4, and 5 proteins is
reduced, we performed a parallel immunoblot with the OL.28
antibody (Figure 2B). Proteins were expressed in COS-1 cells,
immunoprecipitated from unlabeled cell lysates and medium,
and analyzed by immunoblotting with the OL.28 anti–poly-
sialic acid antibody (OL.28 Ab) that is specific for α2,8-linked
polyNeu5Ac of 5 units or longer (Close et al., 2000; Sato et al.,
2000). We found that wild-type ST8Sia II/STX and the mutant
proteins lacking the oligosaccharides on sites 0, 1, and 3
migrated with similar polysialylated molecular masses and
exhibited equal intensities upon immunoblotting with OL.28
Ab (data not shown), suggesting that none of the oligosaccharides
attached to these sites are polysialylated. Interestingly,
although the autopolysialylation of the cell-associated mutant
2, 4, and 5 proteins was only marginally reduced compared to
that of the wild-type ST8Sia II/STX, the autopolysialylation of
the secreted mutant 2, 4, and 5 proteins was dramatically
reduced relative to that of wild-type ST8Sia II/STX (Figure 2B,
mutants 2, 4, and 5, Cell Lysates and Media).
From these data, we could not be sure whether the decrease
in autopolysialylation of these three mutant proteins (mutants
2, 4, and 5) reflected the absence of polysialic acid attachment
sites or problems with their intracellular trafficking. We
considered two possibilities. First, were these proteins folding
in such a way that they were slow to leave the endoplasmic
reticulum (ER) or slow to leave the Golgi, resulting in their
reduced appearance in the cell medium? If this was the case,
could this be due to the absence of an oligosaccharide during
the folding process or the presence of the foreign Ser residue?
Second, was the polysialic acid divided evenly between the
oligosaccharides attached to these three sites? If it were, would
we observe a more substantial decrease in autopolysialylation
if we mutated two or three glycosylation sites at one time?
To address these issues we replaced the Asn 2, 4, and 5 residues
with Gln rather than Ser to test the possibility that the nature of
the amino acid at this site was having an impact on folding, and
that this was leading to differences in trafficking and/or
enzymatic activity. To confirm that the oligosaccharides on
Asn 2, 4, and 5 possess all the polysialic acid–modifying
ST8Sia II/STX, we also made double glycosylation site
mutants (mutants 2.4, 2.5, and 4.5) and triple glycosylation site
mutants (mutant 2.4.5) using both Ser and Gln to replace Asn.
We first analyzed these single, double, and triple glycosylation
site mutants by immunofluorescence microscopy (Figures 3
[Ser mutants] and 4 [Gln mutants]). We used both the V5 Ab to
localize the proteins within the cell, and the OL.28 Ab to
evaluate the level of autopolysialylation of the mutant proteins, as
well as where these autopolysialylated proteins were localized.
Previous work by Close et al. (2000) demonstrated that no
endogenous COS-1 cell proteins are substrates for ST8Sia IV/PST,
and we expected the same situation for ST8Sia II/STX.
Staining with the V5 Ab revealed that, like the wild-type
ST8Sia II/STX protein, all the single, double, and triple
mutants (either Ser or Gln replacements) were localized in the
Fig. 3. Glycosylation sites 2, 4, and 5 are necessary for ST8Sia II/STX
autopolysialylation: Localization and autopolysialylation of wild-type ST8Sia
II/STX and Ser mutants. COS-1 cells transiently expressing wild-type ST8Sia
II/STX (WT) or single, double, or triple Asn to Ser mutant ST8Sia II/STX
proteins were analyzed by indirect immunofluorescence microscopy using
either V5 Ab (internal staining) or OL.28 Ab (internal and surface staining), and
the appropriate FITC-conjugated secondary antibodies. Prior to staining, cells
were fixed and permeabilized with methanol to visualize internal structures or
fixed with 3% paraformaldehyde to visualize only cell surface.
Immunofluorescence was visualized using a Nikon Axiophot fluorescence
microscope and a 60× oil immersion Plan Apochromat objective.
Magnification, 750×.
Golgi. As more glycosylation sites were mutated, however, we
did observe more ER staining in addition to distinct Golgi
staining. This suggested that the presence of oligosaccharides
at specific positions in the protein might aid the folding
process. Notably, no differences were seen when we compared
the localization of the Ser versus the Gln mutants.
Comparison of the autopolysialylated forms of these two sets
of mutant proteins revealed only subtle differences in the level
of autopolysialylation and localization. As was observed on the
immunoblots in Figure 2B, the level of autopolysialylation of
the intracellular single glycosylation mutants (Ser or Gln
replacements) did not appear substantially decreased relative
to wild-type ST8Sia II/STX (Figures 3 and 4, compare ST8Sia
II/STX, Mut 2, Mut 4, and Mut 5, OL.28 Ab internal staining).
Notably, the presence of these autopolysialylated mutant
ST8Sia II/STX proteins was decreased on the cell surface
relative to wild-type ST8Sia II/STX (Figures 3 and 4, compare
ST8Sia II/STX, Mut 2, Mut 4, and Mut 5, OL.28 Ab surface
staining). This again suggested that these mutants are slow to
leave the Golgi, after which we would expect a smaller amount to
move to the cell surface and the majority to enter a post-Golgi
compartment where cleavage would occur leading to secretion
1000

ST8Sia II/STX autopolysialylation and enzymatic activity requirements
Fig. 5.Autopolysialylation of Asn to Ser ST8Sia II/STX mutant proteins. Wild-type
ST8Sia II/STX (WT), double, or triple Asn to Ser mutant ST8Sia II/STX proteins
were transiently expressed in COS-1 cells without metabolic labeling and
immunoprecipitated from COS-1 cell lysates and medium 18 h posttransfection.
Immunoprecipitated proteins were separated on 7.5% SDS–polyacrylamide
gels, electrophoretically transferred to nitrocellulose membranes, and subjected
to immunoblot analysis using OL.28 Ab. Molecular mass markers are as
follows: 203 kDa, myosin; 123 kDa, β-galactosidase; 83 kDa, BSA.
Ser, which still possesses an oligosaccharide on Asn 89 (site 2)
exhibits the most autopolysialylation of the three double
glycosylation site mutants in both the immunoblot and
immunofluorescence experiments (Figure 3 and Figure 5, Mut
4.5). This suggested that the oligosaccharide attached to Asn
89 (site 2) may, like the analogous oligosaccharide attached to
ST8Sia IV/PST, possess the majority of the polysialic acid.
Fig. 4. Glycosylation sites 2, 4, and 5 are necessary for ST8Sia II/STX
autopolysialylation: Localization and autopolysialylation of wild-type ST8Sia
II/STX and Gln mutants. COS-1 cells transiently expressing wild-type ST8Sia
II/STX (WT) or single, double, or triple Asn to Gln mutant ST8Sia II/STX
proteins were analyzed by indirect immunofluorescence microscopy using
either V5 Ab (internal staining) or OL.28 Ab (internal and surface staining), and
the appropriate FITC-conjugated secondary antibodies. Prior to staining, cells
were fixed and permeabilized with methanol to visualize internal structures or
fixed with 3% paraformaldehyde to visualize only cell surface.
Retention of at least two Asn-linked oligosaccharides is
sufficient for ST8Sia II/STX autopolysialylation in COS-1 cells
To evaluate the relative levels of polysialylation of the
oligosaccharides attached to the oligosaccharides modifying
Asn 89 (site 2), Asn 219 (site 4), and Asn 234 (site 5), we
generated a series of mutants retaining one, two, or all three of
these Asn-linked glycosylation sites (the “alone” mutants).
These mutant proteins were analyzed by immunofluorescence
microscopy to determine their intracellular localization and
level of autopolysialylation, as done with the glycosylation site
mutants in Figures 3 and 4. All the mutants retaining one, two,
or three Asn-linked oligosaccharides were transported to the
Golgi apparatus (Figure 6, V5 Ab internal). However, immuno-
fluorescence microscopy experiments using the OL.28 Ab
revealed that none of mutants retaining only one of the three
Asn-linked oligosaccharides was detectably autopolysialylated
(Figure 6, 2 Alone, 4 Alone, and 5 alone, OL.28 Ab internal
and surface). In contrast, mutants retaining two of the three
Asn-linked oligosaccharides, those modifying Asn 2 and 4
(2.4 Alone), Asn 2 and 5 (2.5 Alone), or Asn 4 and 5 (4.5 Alone),
were found in their autopolysialylated forms in the Golgi and
at the cell surface (Figure 6, OL.28 Ab internal and surface).
Likewise, the ST8Sia II/STX protein retaining Asn residues 2, 4,
and 5 (2.4.5 Alone) was autopolysialylated to a level comparable
to wild-type ST8Sia II/STX (Figure 6, OL.28 Ab internal and
surface, compare WT to 2.4.5 Alone). This experiment
suggested that there were no substantial differences in the
autopolysialylation of the three ST8Sia II/STX Asn-linked
oligosaccharides in these “alone” mutant proteins. In addition,
the results did support the previous conclusion that the
Immunofluorescence was visualized using a Nikon Axiophot fluorescence
microscope and a 60× oil immersion Plan Apochromat objective.
Magnification, 750×.
(see Ma et al., 1997). Interestingly, we observed more cell
surface autopolysialylated protein when Asn 2 or Asn 4 was
replaced with Ser than when these amino acids were replaced
by Gln, suggesting that the Gln mutants may be either more
stably retained in the Golgi or have lower catalytic activity
than the analogous Ser mutants (compare Mut 2 and Mut 4
OL.28 Ab surface staining in Figure 3 to the analogous panels
in Figure 4).
Evaluation of the double and triple glycosylation site
mutants verified that the polysialic acid on ST8Sia II/STX is
modifying the oligosaccharides on Asn 89 (site 2), Asn 219
(site 4), and Asn 234 (site 5). For both the Ser and Gln double
glycosylation mutants, we observed a significant decrease in
both intracellular and cell surface polysialic acid (Figures 3
and 4, Mut 2.4, Mut 2.5, and Mut 4.5, OL.28 Ab internal and
surface staining). Replacing all three glycosylation sites with
either Ser or Gln, led to proteins that were primarily Golgi
localized but that exhibited no autopolysialylation (Figures 3
and 4, Mutant 2.4.5). These results were also supported by
OL.28 immunoblot experiments shown in Figure 5 for the
double and triple Ser glycosylation mutants. Notably, Mut 4.5
1001

B.E. Close et al.
5 Alone), we evaluated their ability to polysialylate NCAM
when coexpressed with this substrate in COS-1 cells. First, to
determine whether the 2 Alone, 4 Alone, and 5 Alone mutant
ST8Sia II/STX proteins were active, we coexpressed these
proteins with full-length NCAM and evaluated NCAM poly-
sialylation by immunofluorescence microscopy with the OL.28
Ab. We found that all three of these mutant proteins were
capable of NCAM polysialylation, even though they were not
autopolysialylated at a detectable level (Figure 7A). This indi-
cated that the inability of these mutant ST8Sia II/STX proteins
to autopolysialylate did not reflect a loss of catalytic activity or
mislocalization to an early Golgi compartment.
Next we evaluated the ability of both the double and triple
glycosylation site mutants to polysialylate NCAM in a similar
coexpression assay. Because the double glycosylation site
mutants were autopolysialylated to some degree, we could not
use immunofluorescence microscopy to evaluate NCAM poly-
sialylation. Instead, we coexpressed soluble NCAM-Fc with
wild-type ST8Sia II/STX and the double and triple mutant
ST8Sia II/STX proteins (Ser and Gln replacements) in COS-1
cells. Expressing cells were radiolabeled for 1 h, chased with
unlabeled medium for 3 h, and soluble NCAM-Fc was precipi-
tated from the medium with protein A-Sepharose as described
in Materials and methods. The NCAM-Fc monomer migrates
on SDS–polyacrylamide gels with an apparent molecular mass
of 180 kDa, and polysialylated NCAM-Fc appears as a poly-
disperse band with a molecular mass ranging from 180 kDa to
above 200 kDa (Figure 7B). We confirmed that the poly-
disperse appearance of NCAM-Fc reflected polysialylation by
immunoblotting NCAM-Fc with the OL.28 Ab (Figure 7C).
As expected, coexpression of wild-type ST8Sia II/STX with
NCAM-Fc resulted in the polysialylation of the NCAM-Fc
(Figures 7B and 7C, WT). Likewise, coexpression of the Ser
and Gln double glycosylation mutants with NCAM-Fc also
resulted in the polysialylation of NCAM-Fc (Figures 7B and
7C, Ser and Gln mutants 2.4, 2.5, and 4.5). Strikingly, we
observed that the Mut 2.4 Gln and Mut 4.5 Gln proteins add
less polysialic acid to NCAM than the comparable Ser double
mutants (Figure 7C, compare especially the Gln and Ser
mutants 2.4 and 4.5). Most significant was the finding Mut
2.4.5 Ser, which was shown to be nonautopolysialylated
(Figures 3, 4, and 5) was still able to polysialylate NCAM-Fc
in our coexpression assay, whereas the Mut 2.4.5 Gln appeared
to be completely inactive (Figures 7B and 7C, Ser and Gln
mutants 2.4.5).
Whether inability of the Mut 2.4.5 Gln to polysialylate
NCAM reflects the protein’s inactivity or its inability to
compartmentalize correctly is not clear. More important, the
ability of the nonautopolysialylated Mut 2.4.5 Ser protein to
polysialylate NCAM nicely parallels our previous report that
nonautopolysialylated ST8Sia IV/PST is capable of NCAM
polysialylation (Close et al., 2000) by demonstrating that
autopolysialylation is not a requirement for ST8Sia II/STX
enzymatic activity as well. In sum, our data in this work establish
that autopolysialylation of ST8Sia II/STX, like that of ST8Sia
IV/PST, is not a prerequisite for its ability to polysialylate
NCAM. In addition, the mutant ST8Sia II/STX proteins
analyzed in these experiments showed us that the level of
NCAM polysialylation appears to depend on not only their
autopolysialylation (as seen previously with the ST8Sia IV/PST
glycosylation/autopolysialylation mutants [Close et al., 2000])
Fig. 6. Retention of two or more of the identified glycosylation sites (Asn 89,
and/or Asn 219, and/or Asn 234) is sufficient for ST8Sia II/STX
autopolysialylation. A series of mutants was generated that retain only one, two,
or all three of the Asn 89, Asn 219, and/or Asn 234 glycosylation sites. The
localization and autopolysialylation of these mutants was analyzed by indirect
immunofluorescence microscopy using V5 Ab (internal staining), OL.28 Ab
(internal and surface staining), and the appropriate FITC-conjugated secondary
antibodies. Immunofluorescence was visualized using a Nikon Axiophot
fluorescence microscope and a 60× oil immersion Plan Apochromat objective.
Magnification, 750×.
oligosaccharides on Asn 89, Asn 219, and Asn 234 possess the
polysialic acid–modifying ST8Sia II/STX.
Why was the presence of at least two of the three glyco-
sylation sites required for ST8Sia II/STX autopolysialylation?
One obvious possibility is that the mutant proteins modified by
only one glycosylation site were folding into forms that were
able to exit the ER and be transported to the Golgi but lacked
enzymatic activity. In this type of in vivo assay, a lack of
enzymatic activity could either reflect the protein folding into
an inactive state or the protein folding into a form that
localized too early in the Golgi to be functional, such as in an
early Golgi compartment that does not contain the CMP-NeuAc
sugar nucleotide donor. These same issues were concerns for
the triple glycosylation mutants, Mut 2.4.5 Ser and Mut 2.4.5 Gln,
that were also nonautopolysialylated when expressed in COS-1
cells (Figures 3, 4, and 5).
ST8Sia II/STX autopolysialylation is not a requirement for
polysialylation of NCAM
To evaluate the polysialyltransferase activity of both the
ST8Sia II/STX triple glycosylation site mutants (Mut 2.4.5 Ser
and Mut 2.4.5 Gln), as well as the mutant proteins possessing
only one Asn-linked oligosaccharide (2 Alone, 4 Alone, and
1002

ST8Sia II/STX autopolysialylation and enzymatic activity requirements
but also other aspects of ST8Sia II/STX folding that can be
altered by specific amino acid replacements (as observed with
the Gln double mutants 2.5 and 4.5).
Discussion
Previously, we reported that both ST8Sia IV/PST and ST8Sia
II/STX are autopolysialylated on complex type Asn-linked
oligosaccharides when expressed in mammalian cells (Close
and Colley, 1998). This was notable because polysialylation
appears to be an extremely protein-specific event, with poly-
sialic acid previously found only on the oligosaccharides
attached to NCAM and the α subunit of the voltage sensitive
sodium channel (Zuber et al., 1992). The presence of polysialic
acid on ST8Sia IV/PST and ST8Sia II/STX led us to question
what role this modification plays in their enzymatic activity.
We found that abrogation of ST8Sia IV/PST autopolysialylation
by elimination of critical glycosylation sites or by a nonasparagine
point mutation did not abolish its ability to polysialylate
NCAM (Close et al., 2000; Close and Colley, unpublished
data). However, autopolysialylated ST8Sia IV/PST did seem to
add more polysialic acid to NCAM, suggesting that polysialic
acid may somehow stabilize the interaction of ST8Sia IV/PST
with NCAM.
In this work we have demonstrated that ST8Sia II/STX has
polysialic acid modifying the oligosaccharides attached to Asn
89, Asn 219, and Asn 234. These ST8Sia II/STX glycosylation
sites are analogous to glycosylation sites 2, 4, and 5 in ST8Sia
IV/PST (Close et al., 2000). Elimination of combinations of
two of these glycosylation sites (Mut 2.4, Mut 2.5, and Mut
4.5) led to significant decreases in the autopolysialylation of
the mutant ST8Sia II/STX proteins, and elimination of all three
sites completely abolished ST8Sia II/STX autopolysialylation
(see Figures 3, 4, and 5). The observation that the Ser triple
Mut 2.4.5 protein was not autopolysialylated yet still was able
to polysialylate NCAM demonstrated that this mutant protein
was still catalytically active and that autopolysialylation was
not required for NCAM polysialylation (Figure 7).
ST8Sia IV/PST and ST8Sia II/STX share 57% amino acid
identity and catalyze the same polysialylation of NCAM oligo-
saccharides, with some slight differences in site preference and
length of polysialic acid chains added (Angata et al., 1998;
Kitazume-Kawaguchi et al., 2001). Therefore, it was unexpected
that the glycosylation and autopolysialylation of ST8Sia II/STX
would be significantly different from that of ST8Sia IV/PST.
Initially we were surprised that three sites in ST8Sia II/STX
(rather than two sites as originally found for ST8Sia IV/PST)
were autopolysialylated and that only one of these sites (site 2,
Asn 74 in ST8Sia IV/PST and Asn 89 in ST8Sia II/STX) was
the same for the two enzymes. In our original analysis of
ST8Sia IV/PST polysialylation, we showed that the bulk of the
polysialic acid (∼80%) was found on the oligosaccharide
attached to Asn 74 (site 2) (Close et al., 2000). However, as
clearly shown and discussed in that study, our data indicated
that there is a residual amount of polysialic acid present on
ST8Sia IV/PST after mutation of Asn 74. Experiments
described in that work and from results of later unpublished
work suggested that there was also polysialic acid modifying
the oligosaccharide attached to Asn 119 (site 3). More recent
experiments show that a myc-tagged version of the ST8Sia IV/PST
Fig. 7. Autopolysialylation of ST8Sia II/STX is not a prerequisite for
polysialylation of NCAM. (A) The localization, autopolysialylation, and
NCAM polysialylation activity of V5-tagged ST8Sia II/STX Ser mutant
proteins that retain only one Asn-linked glycosylation consensus sequence
(2 Alone, 4 Alone, and 5 Alone) were analyzed by indirect immunofluorescence
microscopy using V5 Ab (internal staining), OL.28 Ab (internal staining), and
the appropriate FITC-conjugated secondary antibodies. Immunofluorescence
was visualized using a Nikon Axiophot fluorescence microscope and a 60× oil
immersion Plan Apochromat objective. Magnification, 750×. (B) COS-1 cells
transiently coexpressing soluble NCAM-Fc and wild-type ST8Sia II/STX
(WT), double, or triple Asn to Ser and Asn to Gln mutant ST8Sia II/STX
enzymes were metabolically labeled with 35S-Express protein labeling mix for
1 h and chased with unlabeled medium for 3 h. NCAM-Fc protein was
precipitated from the chase medium with protein A-Sepharose beads and
separated on 5% SDS–polyacrylamide gels. The radiolabeled proteins were
visualized by fluorography. (C) To confirm the presence of polysialic acid on
NCAM-Fc, a parallel coexpression experiment was performed without
metabolic labeling. NCAM-Fc was precipitated from the cell medium 18 h
posttransfection and immunoblotted with OL.28 Ab. Molecular mass markers
are as follows: 203 kDa, myosin; and 123 kDa, β-galactosidase. The (–) symbol
denotes the control transfection of NCAM-Fc.
1003

B.E. Close et al.
Mut 2.3 protein, in contrast to the previously analyzed V5-tagged
protein, exhibits weak autopolysialylation. This suggests that a
third site, possibly Asn 219 (site 5), possesses a small amount
of polysialic acid and that autopolysialylation is very sensitive
to subtle folding differences, even those generated by the
addition of a short peptide epitope to the enzyme’s carboxy-
terminus (Close and Colley, unpublished data). Considered
together, these data suggest that the autopolysialylation of the
two polysialyltransferases may be more similar than we
originally suspected.
ST8Sia II/STX resulted in a catalytically inactive polysialyl-
transferase (see triple Gln mutant 2.4.5, Figure 7A, B, and C).
Thus, it is likely that the inability of Muhlenhoff et al. (2001)
to detect any NCAM polysialylation activity of their nonauto-
polysialylated ST8Sia II/STX mutants is the result of replace-
ment of Asn residues with Gln residues and the concomitant
subtle perturbation of protein structure that leads to an inactive
enzyme, rather than being suggestive of a functional link
between autopolysialylation and catalytic activity.
From the work presented herein, it is clear that the folding of
ST8Sia II/STX is much more sensitive to replacement of
amino acids and the elimination of glycosylation sites than that
of ST8Sia IV/PST. Even the alteration of one Asn residue and
the resultant elimination of one glycosylation site significantly
slowed the cleavage and secretion of the ST8Sia II/STX
protein and decreased the amount observed at the cell surface
(Figures 2–5). These effects were even more severe when two
or more glycosylation sites were eliminated (Figures 3–5). In
addition, ST8Sia II/STX catalytic activity was clearly compro-
mised when Gln residues were used to replace Asn 89, 219,
and 234. This was surprising because Gln is structurally more
similar to Asn than is Ser. The double Asn to Gln mutants (Mut
2.4 Gln and Mut 4.5 Gln) exhibited a decreased ability to add
polysialic acid to NCAM relative to identical Ser double
mutants, and NCAM proteins polysialylated by these mutant
ST8Sia II/STX proteins migrated with a reduced molecular
mass compared to NCAM proteins polysialylated by wild-type
ST8Sia II/STX (Figure 7C). This result suggested that the
presence of a Gln residue at amino acid 219 was particularly
damaging to the ability of ST8Sia II/STX to fully polysialylate
NCAM. Similarly, while a triple Asn to Ser replacement
mutant (Mut 2.4.5 Ser) was able to polysialylate NCAM,
although to a lesser extent than wild-type ST8Sia II/STX, an
identical triple Asn to Gln replacement mutant (Mut 2.4.5 Gln)
was completely inactive in NCAM polysialylation.
The “alone” mutant series was originally generated to
confirm the polysialylation of the oligosaccharides attached to
Asn 89, Asn 219, and Asn 234 (sites 2, 4, and 5, respectively),
and to determine which Asn-linked oligosaccharides were
sufficient for ST8Sia II/STX autopolysialylation. We found
that mutant proteins retaining the single glycosylation sites
Asn 89, Asn 219, or Asn 234 exhibited no detectable auto-
polysialylation (Figure 6). This was unexpected because previ-
ously we found a mutant ST8Sia IV/PST protein retaining a
single glycosylation site at Asn 74 (2 Alone ST8Sia IV/PST)
was polysialylated to nearly wild-type levels (Close et al.,
2000). Localization studies demonstrated the presence of these
ST8Sia II/STX “alone” mutant proteins in the Golgi (Figure 6),
and coexpression with full-length NCAM demonstrated that
they were able to polysialylate NCAM despite their lack of
detectable autopolysialylation (Figure 7A). These results
suggest that the process of ST8Sia II/STX autopolysialylation
may be cooperative, requiring more than one oligosaccharide
to interact with the enzyme, or extremely sensitive to alterations in
enzyme structure generated by the lack of oligosaccharides
during the folding process.
Work with other sialyltransferases, such as ST8Sia IV/PST
and the STtyr isoform of the α2, 6-sialyltransferase (ST6Gal
I), has shown us that either the absence of an oligosaccharide
or the presence of a foreign amino acid (Ser or Gln) at a
consensus N-glycosylation site can have profound effects on
protein trafficking and/or catalytic activity. ST8Sia IV/PST,
ST8Sia II/STX, and the STtyr isoform of ST6Gal I are all
transiently localized in the Golgi. By analogy with the STtyr
isoform of the ST6Gal I, we believe that the majority of
ST8Sia IV/PST and ST8Sia II/STX ultimately migrate to a
post-Golgi compartment where it undergoes cleavage in the
stem region and is then secreted. A smaller portion bypasses
the cleavage event and is transported to the cell surface (see
Ma et al., 1997; Close and Colley, 1998). We found that
replacement of Asn 158 (one of two glycosylation sites) of the
ST6Gal I STtyr isoform with a Gln residue generated a
misfolded protein that was retained in the ER and rapidly
degraded (Chen and Colley, 2000). However, replacement of
the same Asn residue with Ser generated an active protein that
was stably localized in the Golgi and never cleaved and
secreted into the cell medium (Chen and Colley, 2000). In stark
contrast, replacing Asn 119 (glycosylation site 3) of ST8Sia
IV/PST with a Ser rendered it completely inactive without
slowing its ER to Golgi trafficking, whereas replacing the
same Asn residue with Gln generated a completely active
polysialyltransferase (Close et al., 2000; Bharaty et al., unpub-
lished data).
Recently, Gerardy-Schahn and colleagues (Muhlenhoff et al.,
2001) reported that autopolysialylation of murine ST8Sia II/STX
was dependent on the presence of specific N-glycans attached
to Asn 89 and Asn 219. This correlates well with the data
presented here that demonstrates the oligosaccharides
modifying Asn 89, Asn 219, and Asn 234 are necessary for
human ST8Sia II/STX autopolysialylation (see Figures 2, 3,
and 5). Muhlenhoff et al. (2001) also reported that hamster
ST8Sia IV/PST is modified by five Asn-linked oligosaccha-
rides and that the bulk of the of polysialic acid is present on the
oligosaccharide attached to Asn 74. This nicely parallels our
previously published report that established that human
ST8Sia IV/PST is modified by five Asn-linked oligosaccha-
rides and that the oligosaccharide modifying Asn 74 is the
major acceptor of polysialic acid (Close et al., 2000). Based on
the inability of their Gln replacement mutant, ST8SiaII∆3,5,6
(analogous to sites 2, 4, and 5 in our numbering system), to
polysialylate NCAM, Muhlenhoff et al. (2001) again
suggested that elimination of ST8Sia II/STX autopolysialyl-
ation leads to an abrogation of polysialyltransferase activity. In
contrast, we found that a nonautopolysialylated mutant of
ST8Sia II/STX, triple Ser mutant 2.4.5, is still active to poly-
sialylate NCAM (see Figure 7A, B, and C). Moreover, we
found that substitution of Gln residues for Asn residues in
In sum, we have definitively shown that autopolysialylation of
both polysialyltransferases, ST8Sia IV/PST and ST8Sia II/STX,
is not required for their ability to polysialylate NCAM.
Gerardy-Schahn and colleagues (Muhlenhoff et al., 1996;
Windfuhr et al., 2000) have suggested a functional link between
1004

ST8Sia II/STX autopolysialylation and enzymatic activity requirements
autopolysialylation and NCAM polysialylation. However, our
previous (Close et al., 2000) and present conclusions have been
further substantiated by a recent report by Fukuda and
colleagues (Angata et al., 2001), in which they show that
exogenous ST8Sia IV/PST expressed in insect cells is not
autopolysialylated, yet is an active polysialyltransferase.
Presently, the mechanism of enzyme autopolysialylation is not
well understood. It is clear that polysialic acid chains are not
preassembled on the oligosaccharides of the polysialyltrans-
ferase and then transferred to the oligosaccharides of NCAM
(Muhlenhoff et al., 1996). However, it is not known if auto-
polysialylation is a cis event, with one molecule of enzyme
polysialylating its own oligosaccharides, or a trans event, with
one molecule of enzyme polysialylating a neighbor’s oligo-
saccharides. In addition, why the presence of polysialic acid on
ST8Sia IV/PST and ST8Sia II/STX enhances the amount of
polysialic acid added to NCAM, and whether this reflects
increased stability of the enzyme–substrate interaction is not
clear. More work is clearly required to answer these questions
and elucidate the mechanisms of autopolysialylation and
NCAM polysialylation.
Site-directed mutagenesis of ST8Sia II/STX cDNA
Consensus N-glycosylation site Asn residues in full-length
V5-tagged ST8Sia II/STX cDNA (Close and Colley, 1998)
were mutated to Ser or Gln using the QuikChangesite-directed
mutagenesis system according to the manufacturer’s protocol
(Stratagene). The following primers were used for mutation of
the Asn codons to the Ser codons, with the mutating codon in
bold:
• Mutant 0, GCTGAAGTTGTAATATCCGGCTCCTCATCAC
and GTGATGAGGAGCCGGATATTACAACTTCAGC
• Mutant 1, GTTGTTGACAGAAGTTCCGAAAGCATCAAGC
and GCTTGATGCTTTCGGAACTTCTGTCAACAAC
• Mutant 2, CCAAATGGAGACATTCCCAGACGCTCTCTC
and GAGAGAGCGTCTGGGAATGTCTCCATTTGG
• Mutant 3, GACAGCACCATGTCCGTGTCCCAGAACC
and GTTCTGGGACACGGACATGGTGCTGTCTC
• Mutant 4, GAGGACTTGGTCTCCGCCACGTGGCG and
CGCCACGTGGCGGAGACCAAGTCCTC
• Mutant 5, GCACAGCCTCTCCGGCAGCATCCTG and
GGATGCTGCCGGAGAGGCTGTGCAG.
The following primers were used for mutation of the Asn
codons to the Gln codons, with the mutating codon in bold:
• Mutant 2, CCAAATGGAGACATCAACAGACGCTCTCTC
and GAGAGAGCGTCTGTTGATGTCTCCATTTGG
• Mutant 4, GAGGACTTGGTCCAAGCCACGTGGCG and
CGCCACGTGGCTTGGACCAAGTCCTC
• Mutant 5, GCACAGCCTCCAAGGCAGCATCCTG and
GGATGCTGCCTTGGAGGCTGTGCAG.
Each mutation was confirmed using the Sequenase version 2.0
DNA Sequencing kit (United States Biochemical) and the
appropriate sequencing primer.
Materials and methods
Tissue culture media and reagents, including Dulbecco’s
modified Eagle’s medium (DMEM), Opti-MEM I, Lipofectin,
and oligonucleotides were purchased from Life Technologies,
a division of Invitrogen Corporation (Carlsbad, CA). Fetal
bovine serum (FBS) was purchased from Atlanta Biologicals
(Norcross, GA). Nitrocellulose membranes were purchased
from Schleicher and Schuell (Keene, NH). SuperSignal West
Pico chemiluminescence reagent was obtained from Pierce
Chemical (Rockford, IL). Protein molecular mass standards
(myosin, 203 kDa; β-galactosidase, 123 kDa; bovine serum
albumin [BSA], 83 kDa; ovalbumin, 50.7 kDa) were purchased
from Bio-Rad Laboratories (Hercules, CA). The cDNA for
human ST8Sia II/STX was a kind gift from Dr. John Lowe
(University of Michigan, Ann Arbor). Murine NCAM-Fc
cDNA was a generous gift from Dr. Genevieve Rougon
(CNRS, Marseilles, France). Full-length NCAM 140 cDNA
(human) and the OL.28 Ab hybridomas (mouse IgM) were
gifts from Nancy Kedersha (Brigham and Women’s Hospital,
Boston, MA). The OL.28 Ab is specific for α2,8-linked
polyNeu5Ac of 5 units or longer (Close et al., 2000; Sato et al.,
2000). The QuikChange site-directed DNA mutagenesis kit
and Pfu DNA polymerase were purchased from Stratagene
(La Jolla, CA). V5 Ab (mouse IgG) was purchased from
Invitrogen. Sequenase version 2.0 DNA sequencing kit was
purchased from United States Biochemical (Cleveland, OH).
DNA purification kits were obtained from Qiagen (Valencia,
CA). Protein-A Sepharose and α35S-dATP for DNA
sequencing were obtained from Amersham Pharmacia Biotech
(Piscataway, NJ). 35S-Express protein labeling mix was
purchased from Perkin Elmer Life Sciences (Boston, MA).
Fluorescein isothiocyanate (FITC)–conjugated and horse-
radish peroxidase (HRP)–conjugated goat anti-mouse anti-
bodies were purchased from Jackson Laboratories (West
Grove, PA). Other chemicals and reagents were obtained from
Sigma Chemical (St. Louis, MO) and Fisher Scientific
(Hanover Park, IL).
Transient transfection of COS-1 cells with mutant ST8Sia II/STX
cDNAs
COS-1 cells maintained in DMEM, 10% FBS were plated on
100-mm tissue culture plates or 12-mm glass coverslips and
grown in a 37°C, 5% CO2 incubator until 50–70% confluent.
Lipofectin transfections were performed according to the
protocols provided by Life Technologies, as described
previously (Close et al., 2000).
Immunofluorescence localization of mutant ST8Sia II/STX
proteins
COS-1 cells were plated on glass coverslips, transiently
transfected with wild-type or mutant V5-tagged ST8Sia II/STX
cDNA, and processed for immunofluorescence microscopy as
described previously (Colley et al., 1992; Close et al., 2000).
V5 Ab was diluted 1:100; the OL.28 Ab, FITC-conjugated
secondary antibodies, goat anti-mouse IgG, and goat anti-mouse
IgM, were diluted 1:200 in 5% normal goat serum/phosphate
buffered saline (PBS) blocking buffer prior to use.
Metabolic labeling of cells and immunoprecipitation of
ST8Sia II/STX proteins
Following transient transfection with wild-type or mutant V5-
tagged ST8Sia II/STX cDNA and expression of these proteins
for 18 h, COS-1 cells plated in 100-mm dishes were labeled for
1 h with 100 µCi/ml 35S-Express protein labeling mix (Perkin
Elmer Life Sciences) and chased for either 0 h (cell lysates
only) or 6 h as described previously (Close et al., 2000). The
1005

B.E. Close et al.
mutant and wild-type V5-tagged ST8Sia II/STX proteins were
then immunoprecipitated from the cell lysates and 6 h chase
media using 2 µg of V5 Ab and protein A-Sepharose
(Amersham Pharmacia Biotech) as previously described
(Colley et al., 1989; Close et al., 2000). To avoid breakdown of
the polysialic acid, the boiling step was omitted and the immuno-
precipitation beads were resuspended in 50 µl of Laemmli
sample buffer containing 5% β-mercaptoethanol (BME) and
directly loaded into the gel wells. Immunoprecipitated proteins
were separated on 7.5% separating/3% stacking SDS–poly-
acrylamide gels (Laemmli, 1970). Radiolabeled proteins were
visualized by fluorography using 10% 2,5-diphenyloxalzole in
dimethyl sulfoxide (Bonner and Lasky, 1974), and gels were
exposed to Kodak BioMax MR film at –80°C.
dimers. Gels were processed for fluorography as described
above. For the second set of plates, unlabeled NCAM-Fc was
precipitated from the medium 18 h posttransfection as
described above. The precipitated NCAM-Fc proteins were
subjected to immunoblotting with OL.28 Ab as described
above.
Acknowledgments
We would like to thank Dr. John Lowe for his generous gift of
the human ST8Sia II/STX cDNA. We would also like to thank
Dr. Nancy Kedersha for her kind gift of NCAM 140 cDNA and
Dr. Genevieve Rougon for her generous gift of the NCAM-Fc
cDNA. Sialyltransferase nomenclature according to Tsuji et al.
(1996). This work was supported by National Institutes of
Health Research Grant GM48134 (to K.J.C.). K.J.C. is an
established investigator of the American Heart Association.
Immunoprecipitation and immunoblot of ST8Sia II/STX
proteins
COS-1 cells were transiently transfected with wild-type or
mutant V5-tagged ST8Sia II/STX cDNA as previously
described (Close et al., 2000). Eighteen hours posttransfection,
cell medium was collected. The V5-tagged ST8Sia II/STX
mutant and wild-type proteins were immunoprecipitated from
the cell lysates and media and electrophoresed as described above.
Following electrophoresis, proteins were subjected to immuno-
blotting as previously described (Close et al., 2000). OL.28 Ab
(IgM) was used at a 1:500 dilution, and the HRP-conjugated
secondary antibody, goat anti-mouse IgM, was used at a
1:8000 dilution.
Abbreviations
BME, β-mercaptoethanol; BSA, bovine serum albumin;
DMEM, Dulbecco’s modified Eagle’s medium; ER, endo-
plasmic reticulum; FBS, fetal bovine serum; FITC, fluorescein
isothiocyanate; HRP, horseradish peroxidase; NCAM, neural
cell adhesion molecule; OL.28 Ab, anti–polysialic acid antibody;
PAGE, polyacrylamide gel electrophoresis; PBS, phosphate
buffered saline; SDS, sodium dodecyl sulfate; V5 Ab, anti–V5
epitope tag antibody.
Indirect immunofluorescence microscopy of COS-1 cells
coexpressing full-length NCAM and mutant ST8Sia II/STX
proteins
References
COS-1 cells maintained in DMEM, 10% FBS were plated on
12-mm glass coverslips and grown in a 37°C, 5% CO2
incubator until 50–70% confluent and essentially transfected
as previously described (Close et al., 2000). For each cover-
slip, the ratio of V5-tagged mutant ST8Sia II/STX to NCAM
140 plasmid DNA transfected was 1:1 (0.5 µg:0.5 µg). Cells
were fixed and processed for immunofluorescence microscopy
as previously described (Close et al., 2000). Primary anti-
bodies (V5 Ab and OL.28 Ab) and secondary antibodies
(FITC-conjugated goat anti-mouse IgG and goat anti-mouse
IgM) were diluted 1:200 in blocking buffer (5% normal goat
serum in PBS) prior to use.
Acheson, A., Sunshine, J. L., and Rutishauser, U. (1991) NCAM polysialic
acid can regulate both cell–cell and cell-substrate interactions. J. Cell Biol.,
114, 143–153.
Angata, K., Nakayama, J., Fredette, B., Chong, K., Ranscht, B., and Fukuda, M.
(1997) Human STX polysialyltransferase forms the embryonic form of the
neural cell adhesion molecule. J. Biol. Chem., 272, 7182–7190.
Angata, K., Yen, T.-Y., El-Battari, A., Macher, B.A., and Fukuda, M. (2001)
Unique disulfide bond structures found in ST8Sia IV polysialyltransferase
are required for its activity. J. Biol. Chem., 276, 15369–15377.
Angata, K., Suzuki, M., and Fukuda, M. (1998) Differential and cooperative
polysialylation of the neural cell adhesion molecules by two polysialyl-
transferases, PST and STX. J. Biol. Chem., 273, 28524–28532.
Berardi, M., Hindelang, C., Laurent-Huck, F.M., Langley, K., Rougon, G.,
Felix, J.M., and Stoekel, M.E. (1995) Expression of neural cell adhesion
molecules, NCAMs, and their polysialylated forms, PSA-NCAMs, in the
developing rat pituitary gland. Cell Tissue Res., 280, 463–472.
Bonner, W.M. and Lasky, R.A. (1974) A film detection method for tritium-labelled
proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem., 46, 83–88.
Chen, C. and Colley, K.J. (2000) Minimal structural and glycosylation requirements
for ST6Gal I activity and trafficking. Glycobiology, 10, 531–538.
Close, B.E. and Colley, K.J. (1998) In vivo autopolysialylation and localization of
the polysialyltransferases PST and STX. J. Biol. Chem., 273, 34586–34593.
Close, B.E., Tao, K., and Colley, K.J. (2000) Polysialyltransferase-1 auto-
polysialylation is not requisite for polysialylation of neural cell adhesion
molecule. J. Biol. Chem., 275, 4484–4491.
Colley, K.J., Lee, E. U., Adler, B., Browne, J.K., and Paulson, J.C. (1989)
Conversion of a Golgi apparatus sialyltransferase to a secretory protein by
replacement of the NH2-terminal signal anchor with a signal peptide.
J. Biol. Chem., 264, 17619–17622.
Colley, K.J., Lee, E.U., and Paulson, J.C. (1992) The signal anchor and stem
regions of the β-galactoside α2, 6-sialyltransferase may each act to localize
the enzyme to the Golgi apparatus. J. Biol. Chem., 267, 7784–7793.
Cunningham, B.A., Hemperly, J.J., Murray, B.A., Prediger, E.A., Brackenbury, R.,
and Edelman, G.M. (1987) Neural cell adhesion molecule: Structure,
Precipitation and immunoblot analysis of soluble NCAM-Fc
from COS-1 cells coexpressing mutant ST8Sia II/STX proteins
In duplicate, COS-1 cells plated on 100-mm tissue culture
plates were transiently cotransfected with wild-type and
mutant V5-tagged ST8Sia II/STX and NCAM-Fc plasmid
DNA at a ratio of 4:1, respectively (20 µg:5 µg). For the first
set of plates, cells were radiolabeled for 1 h with 100 µCi/ml
35S-Express protein labeling mix (Perkin Elmer Life Sciences)
and chased for 3 h with 4 ml unlabeled DMEM, 10% FBS as
previously described (Close et al., 2000). NCAM-Fc was
precipitated from the medium using protein-A Sepharose, and
the precipitated proteins were separated on 5% separating/3%
stacking SDS–polyacrylamide gels after the addition of
Laemmli sample buffer containing 5% BME and incubation of
the protein A-Sepharose beads at 65°C for 5 min. This heating
step was added to eliminate the formation of NCAM-Fc
1006

ST8Sia II/STX autopolysialylation and enzymatic activity requirements
immunoglobulin-like domains, cell surface modulation, and alternative
RNA splicing. Science, 236, 799–806.
Dubois, C., Figarella-Branger, D., Pastoret, C., Rampini, C., Karpati, G., and
Rougon, G. (1994) Expression of NCAM and its polysialylated forms
during mdx mouse muscle regeneration and in vitro myogenesis.
Neuromusc. Disord., 4, 171–182.
Martersteck, C.M., Kedersha, N.L., Drapp, D.A., Tsui, T.G., and Colley, K.J.
(1996) Unique α2, 8-polysialylated glycoproteins in breast cancer and
leukemia cells. Glycobiology, 6, 289–301.
Michalides, R., Kwa, B., Springall, D., van Zandwijk, N., Koopman, J.,
Hilkens, J., and Mooi, W. (1994) NCAM and lung cancer. Intl. J. Cancer
Suppl., 8, 34–37.
Miyahara, R., Tanaka, F., Nakagawa, T., Matsuoka, K., Isii, K., and Wada, H.
(2001) Expression of neural cell adhesion molecules (polysialylated form
of neural cell adhesion molecule and L1-adhesion molecule) on resected
small cell lung cancer specimens: In relation to proliferation state. J. Surg.
Oncol., 77, 49–54.
Eckhardt, M., Muhlenhoff, M., Bethe, A., Koopman, J., Frosch, M., and
Gerardy-Schahn, R. (1995) Molecular characterization of eukaryotic poly-
sialyltransferase-1. Nature, 373, 715–718.
Edelman, G.M. (1985) Cell adhesion and the molecular process of morphogenesis.
Annu. Rev. Biochem., 54, 135–169.
Moolenaar, C.E., Muller, E.J., Schol, D.J., Figdor, C.G., Bock, E., Bitter-Suermann, D.,
and Michalides, R.J. (1990) Expression of neural cell adhesion molecule-related
sialoglycoprotein in small cell lung cancer and neuroblastoma cell lines
H69 and CHP-212. Cancer Res., 50, 1102–1106.
Muhlenhoff, M., Eckhardt, M., Bethe, A., Frosch, M., and Gerardy-Schahn, R.
(1996) Autocatalytic polysialylation of polysialyltransferase-1. EMBO J.,
15, 6943–6950.
Muhlenhoff, M., Eckhardt, M., and Gerardy-Schahn, R. (1998) Polysialic acid:
three-dimensional structure, biosynthesis, and function. Curr. Opin. Struct.
Biol., 8, 558–564.
Muhlenhoff, M., Manegold, A., Windfuhr, M., Gotza, B., and Gerardy-
Schahn, R. (2001) The impact of N-glycosylation on the functions of poly-
sialyltransferases. J. Biol. Chem., 276, 34066–34073.
Nakayama, J., Fukuda, M.N., Fredette, B., Ranscht, B., and Fukuda, M. (1995)
Expression cloning of a human polysialyltransferase that forms the
polysialylated neural cell adhesion molecule present in embryonic brain.
Proc. Natl Acad. Sci. USA, 92, 7031–7035.
Nelson, R.W., Bates, P.A., and Rutishauser, U. (1995) Protein determinants
for specific polysialylation of the neural cell adhesion molecule. J. Biol.
Chem., 270, 17171–17179
Nilsson, O. (1992) Carbohydrate antigens in human lung carcinomas. Apmis
Suppl., 27, 149–161.
Phillips, G.R., Krushel, L.A., and Crossin, K.L. (1997) Developmental
expression of two rat sialyltransferases that modify the neural cell adhesion
molecule, N-CAM. Dev. Brain Res., 102, 143–155.
Edelman, G.M. and Crossin, K.L. (1991) Cell adhesion molecules: implications
for a molecular histology. Annu. Rev. Biochem., 60, 155–190.
Finne, J., Finne, U., Deagostini-Bazin, H., and Goridis, C. (1983) Occurrence
of α2, 8-linked polysialosyl units in a neural cell adhesion molecule.
Biochem. Biophys. Res. Commun., 112, 482–487.
Finne, J., Bitter-Suermann, D., Goridis, C., and Finne, U. (1987) An IgG
monoclonal antibody to group B meningococci cross-reacts with develop-
mentally regulated polysialic acid units of glycoproteins in neural and
extraneural tissues. J. Immunol., 138, 4402–4407.
Franceschini, I., Angata, K., Ong, E., Hong, A., Doherty, P., and Fukuda, M.
(2001) Polysialyltransferase ST8Sia II (STX) polysialylates all of the
major isoforms of NCAM and facilitates neurite outgrowth. Glycobiology,
11, 231–239.
Fukuda, M. (1996) Possible roles of tumor-associated carbohydrate antigens.
Cancer Res., 56, 2237–2244.
Gluer, S., Schelp, C., Gerardy-Schahn, R., and von Schweinitz, D. (1998)
Polysialylated neural cell adhesion molecule as a marker for differential
diagnosis in pediatric tumors. J. Pediatr. Surg., 33, 1516–1520.
Goridis, C. and Brunet, J.-F. (1992) NCAM: structural diversity, function, and
regulation of expression. Sem. Cell Biol., 3, 189–197.
Hildebrandt, H., Becker, C., Murau, M., Gerardy-Schahn, R., and Rahman, H.
(1998) Heterogeneous expression of the polysialyltransferases ST8Sia II
and ST8Sia IV during postnatal rat brain development. J. Neurochem., 71,
2339–2348.
Inoue, S. and Iwasaki, M. (1978) Isolation of a novel glycoprotein from the
eggs of rainbow trout: Occurrence of disialosyl groups on all carbohydrate
chains. Biochem. Biophys. Res. Commun., 83, 1018–1023.
Kibbelaar, R.E., Mooleanaar, C.E.C., Michalides, R.J.A.M., Bitter-Suermann, D.,
Addis, B.J., and Mooi, W. J. (1989) Expression of the embryonal neural
cell adhesion molecule N-CAM in lung carcinoma. Diagnostic usefulness
of monoclonal antibody 735 for the distinction between small cell lung
cancer and non-small cell lung cancer. J. Pathol., 159, 23–38.
Kiss, J.Z. and Rougon, G. (1997) Cell biology of polysialic acid. Curr. Opin.
Neurobiol., 7, 640–646.
Robbins, J.B., McCracken, G.H. Jr., Gotschlich, E.C., Orskov, F., Orskov, I.,
and Hanson, L.A. (1974) Escherichia coli K1 capsular polysaccharide
associated with neonatal meningitis. New Engl. J. Med., 290, 1216–1220.
Roth, J., Taatjies, D.J., Bitter-Suermann, D., and Finne, J. (1987) Polysialic
acid units are spatially and temporally expressed in developing postnatal
rat kidney. Proc. Natl Acad. Sci. USA, 84, 1969–1973.
Roth, J., Zuber, C., Wagner, P., Taatjies, D.J., Weiserber, C., Heitz, P.U.,
Goridis, C., and Bitter-Suermann, D. (1988a) Reexpression of poly(sialic
acid) units of the neural cell adhesion molecule in Wilms’ tumor.
Proc. Natl Acad. Sci. USA, 85, 2999–3003.
Kitazume, S., Kitajima, K., Inoue, S., Troy, F.A., Cho, J.-W., Lennarz, W.J.,
and Inoue, Y.(1994) Identification of polysialic acid-containing glyco-
protein in the jelly coat of sea urchin eggs: occurrence of a novel type of
polysialic acid structure. J. Biol. Chem., 269, 22712–22718.
Kitazume-Kawaguchi, S., Kabata, S., and Arita, M. (2001) Differential biosynthesis
of polysialic or disialic acid structure by ST8Sia II and ST8Sia IV. J. Biol.
Chem., 276, 15696–15703.
Kojima, N., Yoshida, Y., Kurosawa, N., Lee, Y.-C., and Tsuji, S. (1995)
Enzymatic activity of a developmentally regulated member of the sialyl-
transferase family (STX): evidence for α2, 8-sialyltransferase activity
toward Asn-linked oligosaccharides. FEBS Lett., 360, 1–4.
Kornfeld, S. and Kornfeld, R. (1985) Assembly of asparagine-linked oligo-
saccharides. Annu. Rev. Biochem., 54, 631–664.
Kurosawa, N., Yoshida, Y., Kojima, N., and Tsuji, S. (1997) Polysialic acid
synthase (ST8Sia II/STX) mRNA expression in the developing mouse
central nervous system. J. Neurochem., 69, 494–503.
Laemmli, U.K. (1970) Cleavage of structural proteins during assembly of the
head of bacteriophage T4. Nature, 227, 680–685.
Livingston, B.D. and Paulson, J.C. (1993) Polymerase chain reaction of a
developmentally regulated member of the sialyltransferase gene family.
J. Biol. Chem., 268, 11504–11507.
Livingston, B.D., Jacobs, J L., Glick, M.C., and Troy, F.A. (1988) Extended
polysialic acid chains (n > than 55) in glycoproteins from human
neuroblastoma cells. J. Biol. Chem., 263, 9443–9448.
Ma, J., Qian, R., Rausa, F.M., and Colley, K.J. (1997) Two naturally occurring
α2, 6-sialyltransferase forms with a single amino acid change in the
catalytic domain differ in their catalytic activity and proteolytic processing.
J. Biol. Chem., 272, 672–679.
Roth, J., Zuber, C., Wagner, P., Blaha, I., Bitter-Duermann, D., and Heitz, P.U.
(1988b) Presence of the long chain form of polysialic acid of the neural cell
adhesion molecule in Wilms’ tumor. Identification of a cell adhesion
molecule as an oncodevelopmental antigen and implications for tumor
histogenesis. Am. J. Pathol., 133, 227–240.
Rothbard, J.B., Brackenbury, R., Cunningham, B.A., and Edelman, G.M. (1982)
Differences in the carbohydrate structures of neural cell-adhesion molecules
from adult and embryonic brain. J. Biol. Chem., 257, 11064–11069.
Rutishauser, U. (1996) Polysialic acid and the regulation of cell interactions.
Curr. Opin. Cell Biol., 8, 679–684.
Rutishauser, U. and Landmesser, L. (1996) Polysialic acid in the vertebrate
nervous system: A promoter of plasticity in cell–cell interactions. Trends
Neurosci., 19, 422–427.
Rutishauser, U., Watanabe, M., Silver, J., Troy, F.A., and Vimr, E.R. (1985)
Specific alteration of NCAM-mediated cell adhesion by an endoneuraminidase.
J. Cell Biol., 101, 1842–1849.
Rutishauser, U., Acheson, A., Hall, A.K., Mann, D.M., and Sunshine, J. (1988)
The neural cell adhesion molecule (NCAM) as a regulator of cell–cell
interactions. Science, 240, 53–557.
Sadoul, R., Hirn, M., Deagostini-Bazin, H., Rougon, G., and Goridis, C.
(1983) Adult and embryonic mouse neural cell adhesion molecules have
different binding properties. Nature, 304, 347–349.
Sato, C., Fukuoka, H., Ohta, K., Matsuda, T., Koshino, R., Kobayashi, K.,
Troy, F.A., and Kitajima, K. (2000) Frequent occurrence of pre-existing
α2, 8-linked disialic and oligosialic acids with chain lengths up to 7 sia
residues in mammalian brain glycoproteins: Prevalence revealed by highly
sensitive chemical methods and anti-di-, oligo-, and poly-sia antibodies
specific for defined chain lengths. J. Biol. Chem., 275, 15422–15431.
1007

B.E. Close et al.
Scheidegger, E.P., Lackie, P.M., Papay, J., and Roth, J. (1994) In vitro and
in vivo growth of clonal sublines of human small cell lung carcinoma is
modulated by polysialic acid of the neural cell adhesion molecule. Lab.
Invest., 70, 95–106.
Scheidegger, E.P., Sternberg, L.R., Roth, J., and Lowe, J.B. (1995) A human
STX confers polysialic acid expression in mammalian cells. J. Biol. Chem.,
270, 22685–22688.
Seidenfaden, R. and Hildebrandt, H. (2001) Retinoic acid-induced changes in
polysialyltrnasferase mRNA expression and NCAM polysialylation in
human neuroblastoma cells. J. Neurobiol., 46, 11–28.
Seidenfaden, R., Gerardy-Schahn, R., and Hildebrandt, H. (2000) Control of
NCAM polysialylation by the differential expression of polysialyltrans-
ferases ST8Sia II and ST8Sia IV. Eur. J. Cell Biol., 79, 680–688.
Troy, F.A. (1992) Polysialylation: from bacteria to brains. Glycobiology, 2, 5–23.
Troy, F.A. (1995) Sialobiology and the polysialic acid glycotope: occurrence,
structure, synthesis, and glycopathology. In Rosenberg, A. (ed), Biology of
the sialic acids. New York: Plenum Press, pp. 95–144.
Tsuji et al. (1996) Glycobiology, 6, v–xiv.
Vimr, E., Steenbergen, S., and Cieslewicz, M. (1995) Biosynthesis of the polysialic
acid capsule in Escherichia coli K1. J. Ind. Microbiol., 15, 352–360.
Watanabe, M., Timm, M., and Fallah-Najmabadi, H. (1992) Cardiac expression of
polysialylated NCAM in the chicken embryo: correlation with the
ventricular conduction system. Dev. Dyn., 194, 128–141.
Windfuhr, M., Manegold, A., Muhlenhoff, M., Eckhardt, M., and Gerardy-
Schahn, R. (2000) Molecular defects that cause loss of polysialic acid in
the complementation group 2A10. J. Biol. Chem., 275, 32861–32870.
Yoshida, Y., Kojima, N., and Tsuji, S. (1995) Molecular cloning and charac-
terization of a third type of N-glycan α2, 8-sialyltransferases from mouse
lung. J. Biochem., 118, 658–664.
Zuber, C., Lackie, P.M., Catterall, W.A., and Roth, J. (1992) Polysialic acid is
associated with sodium channels and the neural cell adhesion molecule N-CAM
in adult rat brain. J. Biol. Chem., 267, 9965–9971.
1008