Download Role of N-linked oligosaccharide chains in the processing and

Journal of General Virology (1994), 75, 1043 1052. Printedin Great Britain
1043
Role of N-linked oligosaccharide chains in the processing and antigenicity
of measles virus haemagglutinin protein
Aizhong Hu,* Roberto Cattaneo,t Stefan Schwartz and Erling Norrby
Department of Virology, School of Medicine, Karolinska Institute, c/o SMI, S-105 21 Stockholm, Sweden
The effects of N-linked oligosaccharides on the haemagglutinin (H) protein of measles virus (MV) were assessed
with respect to the processing and antigenicity of the
molecule. The functional glycosylation sites on the
H protein were determined by eliminating each of the
five potential positions, Asn-168, Asn-187, Asn-200,
Asn-215 and Asn-238, for N-linked glycosylation by
oligonucleotide-directed mutagenesis on a eDNA clone.
Expression of the mutant H proteins in BHK-21 cells by
a recombinant vaccinia virus encoding T7 polymerase
indicated that the first four sites were used in the H
glycoprotein for the addition of N-linked oligosaccharide
chains. Heterogeneity of oligosaccharide processing was
demonstrated. One of the four glycosylation sites had a
different carbohydrate structure from those of the other
three glycosylation sites and this varied glycosylation
was responsible for the appearance of two forms of the
H protein. The functional glycosylation sites were
systematically removed in various combinations from
the H protein to form a panel of mutants in which the
role of carbohydrate chains, singly or in different
combinations, could be evaluated. Investigations of
these glycosylation mutants indicated that (i) two of the
four individual carbohydrate side-chains have a large
influence on the antigenicity of the molecule; (ii)
Introduction
Kornfeld, 1985). Several approaches have been used to
study the functional role of N-linked oligosaccharide
side-chains in cells. These include the use of agents that
disrupt the addition or modification of glycoproteins,
use of glycan-processing-deficient cell lines, and
oligonucleotide-directed mutagenesis of cDNAs encoding the proteins to either remove or add sites for
carbohydrates. A variety of functions for N-linked
oligosaccharides have been suggested (Olden et al., 1982;
Rademacher et al., 1988). Oligosaccharides on glycoproteins play a role in the initiation and maintenance of
folding into the biologically active conformation, in
maintenance of protein stability and solubility, in
protection of the polypeptide backbone from proteolytic
degradation, in the targeting of glycoproteins to various
subcellular compartments and to the cell surface, in
promotion of specific adhesion of cells during development and in influencing the antigenicity and immunogenicity of glycoproteins.
The haemagglutinin (H) glycoprotein of measles virus
(MV) is an integral component of the virion envelope.
The H protein plays an essential role in the initiation of
Asparagine-linked (N-linked) glycosylation is one of the
most common post-translational modifications of proteins in the exocytic pathway of eukaryotic cells. The
addition of N-linked oligosaccharides starts in the
endoplasmic reticulum (ER) when the target sequence
Asn-X-Ser/Thr (where X is any amino acid except
possibly proline and aspartic acid) is exposed in the
lumen of the ER (Kornfeld & Kornfeld, 1985). Processing of these carbohydrate moieties begins immediately following addition of the oligosaccharide in the ER
and continues as the glycoproteins pass through the
Golgi complex, thus producing two major classes of sidechains: high mannose chains containing only the sugar
residues added in the ER and complex chains resulting
from further trimming and addition of carbohydrate
chains carried out in the Golgi complex (Kornfeld &
t Present address: Institut fiir Molekularbiologie I, Universit~it
Ziirich, H6nggenberg, CH-8093Zfirich, Switzerland.
0001-2204 © 1994SGM
individual carbohydrate side-chains have little effect on
the folding and oligomerization of the molecule, and are
not sufficient or necessary alone to facilitate the
transport of the molecule to the plasma membrane; (iii)
at least two carbohydrate side-chains are required for
the H protein to move along the exocytic pathway to the
plasma membrane and various combinations of oligosaccharide side-chains, irrespective of the carbohydrate
localizations, influence equally the processing of the
molecule.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
1044
A. Hu and others
infection by binding to a specific receptor on susceptible
cells. An acute infection is followed by lifelong immunity
in which efficient neutralizing antibodies against the H
protein play a pivotal role (Norrby & Oxman, 1990). H
glycoprotein is an oligomeric protein that is thought to
be a homodimer or oligomer (Hardwick & Hussell,
1978). From studies of the intracellular processing and
antigenic maturation of the H protein in virus-infected
cells, we have shown that the H protein, like paramyxovirus haemagglutinin-neuraminidase (Mottet et al.,
1986; Ng et al., 1989; Vidal et al., 1989; Waxham et al.,
1986), oligomerizes relatively slowly (tl/2 about 30 min)
(A. Hu, J. K6vamees & E. Norrby, unpublished results)
compared with some other viral glycoproteins, such as
vesicular stomatitis virus G and influenza virus haemagglutinin (tl/2 6 to 10 min) (Gething et al., 1986; Copeland
et al., 1986; Kreis & Lodish, 1986). N-linked glycosylation is possibly involved in the formation of
discontinuous epitopes since monoclonal antibodies
(MAbs) that react with such epitopes fail to recognize the
H protein from MV-infected cells treated with tunicamycin (TM) and neutralization escape mutants selected
by one MAb inactivate an N-linked glycosylation site on
the MV H protein (Hu et al., 1993; A. Hu, J. K6vamees
& E. Norrby, unpublished results).
In this report, we analysed the role of N-linked
glycosylation on the processing and antigenicity of
MV H glycoprotein from the Edmonton strain.
Oligonucleotide-directed mutagenesis was used to generate a panel of MV H glycosylation mutants which were
subsequently analysed by transient expression in BHK21 cells. Our data showed that four of the five potential
sites were used for N-linked carbohydrate additions in
the H glycoprotein. Based on the functional sites, a panel
of H glycosylation mutants were constructed and this
allowed us to evaluate the effect of carbohydrate chains,
individually or in various combinations, on the processing and antigenicity of MV H glycoprotein.
Methods
Cells and virus. BHK-21 cells were grown in Eagle's minimum
essential medium (MEM) supplemented with 10 % inactivated fetal calf
serum (FCS). Propagation and purification of the recombinant vaccinia
virus encoding T7 polymerase (vTF7-3) (Fuerst et al., 1986) grown in
BHK-21 cells was carried out essentially as described previously
(Mackett et al., 1985).
Oligonucleotide-directed mutagenesis. Plasmid pert1 containing the
full-length eDNA encoding the MV H protein has been described
previously (Eschle, 1988). Five 25-mer oligonucleotides with the desired
mismatched nucleotide in the middle of sequences synthesized to be
complementary to H cDNA sequences encoding the five N-linked
glycosylation consensus sites (Asn-X Ser/Thr) were utilized to
substitute codons for serine residues in place of those encoding
asparagiue residues in the five consensus sequences present in MV H
protein (Table 1).
Single-stranded DNA was prepared according to a standard protocol
(Sambrook et al., 1989). Oligonucleotide-directed site-specific mutagenesis was performed using a kit according to the instructions of the
supplier (Bio-Rad). To obtain mutant cDNA lacking a particular single
natural consensus site, we performed oligonucleotide-directed mutagenesis using a single oligonucleofide designed to produce the required
change. The combinations of mutation sites were generated by
performing successive site-directed mutagenesis procedures on previously mutagenized templates. Confirmation of successful mutagenesis
was obtained by dideoxynucleotide chain termination sequencing of
alkali-denatured plasmid DNA according to the protocol of the
supplier (Pharmacia). The four sites used for N-linked glycosylation
(see Results) are designated gl, g2, g3 and g4, corresponding to Asn
found at H protein residues 168, 187, 200 and 215, respectively
(Alkhatib & Briedis, 1986). A total of 13 N-linked glycosylation
mutants were generated, each designated Hg(n), where n is a set of
numbers defining the N-linked glycosylation sites that are used: e.g.,
a mutant lacking the gl site is Hg234 and the non-glycosylated mutant
is Hg0 (see Fig. 1).
Transfection. BHK-21 cells were passaged the day before transfection
and grown to about 70 % confluence in MEM supplemented with 10 %
FCS in 25 cm 2 tissue culture plates, The medium was removed and the
cells were washed three times with serum-free MEM. The cells were
then infected with the recombinant vaccinia virus vTF7-3 at a
multiplicity of infection of about 15 p.f.u./cell and incubated at 37 °C
for 45 rain. The virus inoculum was then removed and replaced with
1 ml of fresh Opti-MEM (Gibco BRL). The cells were transfected with
4 lag of plasmid DNA using 15 lag lipofectin as instructed by the
supplier (Gibco BRL).
Metabolic labelling and pulse-chase analysis. At 5 h post-transfection,
the medium was replaced with MEM containing 1/20 of the normal
content of methionine and the ceils were incubated at 37 °C for 30 min.
The cells were then labelled with 100 laCi/ml of [35S]methionine for
30 min and the medium was removed at the end of pulse radiolabelling.
The monolayer was washed three times with cold PBS, overlaid with
MEM containing 0.5% FCS and incubated for 2 h. The labelled cells
were harvested in 1 ml radioimmunoprecipitation assay (RIPA) buffer
(2% Triton X-100, 0-15M-NaC1, 0-6M-KCI, 0-5M-MgCla, 5mMEDTA, I mM-PMSF, 1% aprotinin, 0.01 M-Tri~HCI pH 7.8). Nuclei
and cell debris were removed by centrifugation at 15000g in a
microfuge and the supernatants were kept for immunoprecipitation.
RIPA. This technique has been described in detail previously
(Sheshberadaran et al., 1983). Lysates (100 lal) from the transfected
Table 1. Nucleotide substitutions and corresponding
amino acid alterations in mutants with changes in the
potential N-linked glycosylation consensus sequences
Glycosylation Amino acid
site
residue
g1
168
g2
187
g3
200
g4
215
g --
238
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
Wild-type
sequence
Mutant
sequence
As~Se~Thr
AAC TCA-ACT
Asn-Cys-Ser
AAC-TGC-TCA
Asn-Met-Ser
A A C ~ T G TCG
Asn Val Ser
AAT-GTG-TCA
Asn-Leu-Ser
AAT-CTG-AGC
Ser-Ser-Thr
AGC-TCA-ACT
Ser-Cy~Ser
AGC-TGC-TCA
Ser-Me~Ser
AGC-ATG-TCG
Ser-Val-Ser
A G T - G T G TCA
Ser-Leu-Ser
AGT-CTG-AGC
Measles virus haemagglutinin N-linked glycans
cells were mixed with 2 gl MAb and the final volume was adjusted to
500 ~tl using RIPA buffer. The mixture of antigen and antibody was
adsorbed onto Protein A bound to Sepharose CL-4B (Pharmacia),
washed three times with RIPA washing buffer (RIPA buffer without
protease inhibitors) and once with 0.15M-NaC1, 0.01 M-Tri~HC1
pH 8-0. The purified immune complexes were dried, mixed with protein
sample buffer [3 % SDS, 3 % 2-mercaptoethanol (2-ME), 10 % glycerol
and 0'1% EDTA], boiled for 3mix, and finally fractionated by
SDS-PAGE. For the analysis of non-reduced proteins, 2-ME was
omitted from the sample buffer. The gels were analysed by fluorography. Quantification of signals on gels was done by laser scanning
densitometry.
Drug treatment. Endoglycosidase H (Endo H) digestion was
performed as described previously (Hu et al., 1993). For tunicamycin
(TM) treatment, the drug (final concentration of 2 gg/ml) was added at
4 h post-transfection and remained present throughout the transfection.
Indirect immunofluorescence. This technique has been described
previously (Norrby et at., 1982). BHK-21 cells were grown to 70%
confluence on microscope coverslips before transfection. Five hours
post-transfection, the cells were washed with PBS and fixed with 3 %
paraformaldehyde in PBS for 30 mix. After washing three times with
PBS, the cells were permeabilized with 0.1% Triton X-100 in PBS for
30 rain. Following three further washes with PBS, cells were treated
with primary antibody at room temperature for 30 min. Cells were then
rinsed with PBS and incubated with goat anti-mouse immunoglobulin
(IgG) conjugated with fluorescein isothiocyanate (FITC; Cappel
Laboratories) for 30 rain. Cells were washed again with PBS before
mounting of the coverslips on glass slides for photography.
The H glycosylation mutants were expressed using
recombinant vaccinia virus encoding the T7 RNA
polymerase. The expression of the mutant proteins was
analysed by metabolic labelling, RIPA and SDS-PAGE.
Transmembrane
anchor
I
gl
g2
g3
g4
Y
Y Y Y
168
187 200 215
NH 2
Y Y Y
Hg134
Y
g124
Y
Y
Hgl23--
Y
Y Y
Y
Y
Hgl2
Hg34
Determination of functionaI N-linked glycosylation sites
in the H glycoprotein and construction of a panel of
glycosylation mutants
Hgl4
rg2
Hg3
Hg4
Y Y
Y
Y
t
Y
Y
Hg24
Hgl
238
COOH
Hg234
Results
The predicted amino acid sequence indicates that the H
protein from the Edmonston strain contains five potential
N-linked glycosylation sites (Alkhatib & Briedis, 1986).
Partial Endo H digestion shows that four of these are
used (Cattaneo & Rose, 1993). However, it remains to be
determined which sites are actually used. To address this
question, we constructed a panel of H site-directed
mutants in which the consensus sequences for N-linked
glycosylation were altered singly and these mutants were
expressed transiently in BHK-21 cells. Oligonucleotidedirected mutagenesis was employed to introduce a onenucleotide alteration in the codon for asparagine,
resulting in a single amino acid substitution at each
potential glycosylation site. The addition of N-linked
carbohydrates was blocked by altering the Asn-X-Ser/
Thr at asparagine residues 168, 187, 200, 215 and 238 to
Ser-X-Ser/Thr. The asparagine-encoding codons at
each site were replaced with serine-encoding codons,
since an asparagine to serine alteration is conservative
due to the similarity in bulk and polar nature of the sidechains. Five single-site mutants were generated. The
derivation of the nomenclature for each mutant is
described in Methods (see also Fig. 1).
1045
Y
Y
%%_
%__
%%-
5"
Y
/ ~ -
x(,
/ / ~
Fig. 1. Schematic diagram of MV H gene products. The H wild-type
(H) protein (617 amino acids) is shown on the top of the figure.
Symbols: Y, functional N-linked glycosylation sites; v, unused
glycosylation site. The designation of the carbohydrate chains (gl, g2,
g3 and g4) are indicated above each site. H glycosylation mutants are
represented with the designated name shown at the left of each
diagram. The residue nmnbers indicate the Asn residue in the H amino
acid sequence. The sites are numbered sequentially from the N to C
terminus of the protein. Figures are not drawn to scale.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
1046
A. Hu and others
(a)
-TM
(a)
+TM
eq
~
H
Hg234
Hg134
Hg124
Hgl2
Hg34
Hg24
Hgl4
,,~
,,b
~b
,3
Hg123
eq
(b)
(b)
-TM
+TM
TM
¢'q
+TM
~
c'q
(c)
(c)
TM
%
%
Hgl
Hg2
Hg3
Hg4
Hg0
+TM
% ~
%
%
%
%
~
%
Fig. 2. Expression of H protein glycosylation mutants. The transfected
cells were metabolically labelled, at 5 h post-transfection, with
[35S]methionine for 3 h in the absence ( - T M ) or presence (+TM) of
TM (2 gg/ml). Equal amounts of cell lysates containing labelled
proteins were immunoprecipitated with MAb 1-29 and analysed by
SDS-PAGE. (a) Single-site mutants. (b) Double-site mutants. (c)
Triple- and quadruple-site mutants. Only the relevant section of the
autoradiogram is shown.
Wild-type H expressed two forms of the glycoprotein of
214,. 77K and 79K which represented heterogeneity of
glycosylation modifications at different sites in BHK-21
cells (see below). The glycosylation-site mutant polypeptides Hg234, Hg134, Hg124 and Hg123 (asparagine
to serine alteration at positions 168, 187, 200 and 215,
respectively) showed an increased electrophoretic mobility compared with wild-type H protein (Fig. 2a,
- T M ) . Synthesis of these mutant polypeptides in the
presence of TM (Fig. 2a, + T M ) indicated that the
differences in mobility resulted from differences in
carbohydrate addition, since all the mutants showed the
same mobility as unglycosylated wild-type H protein.
This provides evidence that sites gl, g2, g3 and g4 are
used for N-linked glycosylation. Deletion of the carbohydrate attachment site at position 238 did not cause an
alteration in mobility on SDS-PAGE compared with
wild-type H protein (Fig. 2 a, H g - ) , suggesting that this
site is not glycosylated. The mutant H g - will not be
Fig. 3. Immunoreactivity of glycosylation mutant proteins with a panel
of MAbs. Equal amount of cell lysates containing labelled proteins
were immunoprecipitated with different MAbs as illustrated and the
polypeptides were analysed by SDS-PAGE. (a) Single-site mutants. (b)
Double-site mutants. (c) Triple- and quadruple-site mutants. Only the
relevant section of the autoradiogram is shown.
described further in this report, but it was examined in
further assays and found to be indistinguishable in every
way from wild-type H, which suggests that at least at this
site the substitution from asparagine to serine does not
have a deleterious effect on the integrity of the H protein.
With the knowledge that four sites on the H
glycoprotein are used for N-linked glycosylation, further
mutants in which two, three or all four sites were
mutated were generated as described in Methods and are
schematically illustrated in Fig. 1. Expression of the
double-, triple- and quadruple-site glycosylation mutant
polypeptides showed alterations in electrophoretic mobility compared with wild-type H protein (Fig. 2 b and c,
- T M ) , which are in agreement with the deletion of two,
three or all four carbohydrate chains. When the mutant
polypeptides were synthesized in the presence of TM
(Fig. 2 b and c, + TM), they all showed the same mobility
as non-glycosylated wild-type H protein, indicating that
the mobility differences seen in the absence of TM are
caused by differences in glycosylation.
The triple-site glycosylation mutant proteins showed a
single protein band, indicating that the individual
oligosaccharide structure is relatively homologous. The
mutant Hg4 displayed a slightly lower mobility than the
mutant Hgl, Hg2 or Hg3 (Fig. 2c, - T M ) . When the
mutant polypeptides were produced in the presence of
TM (Fig. 2 c, + TM), the same electrophoretic mobility
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
Measles virus haemagglutinin N-linked gIycans
Table 2. MAb reactivity, Endo H sensitivity and cell
surface expression of wild-type and mutant proteins
MAb reactivity*
1-44
1-41
Endo H Cell surface
sensitivitytexpression:~
Protein
16-DE6
H
Hg234
Hg134
Hg124
Hg123
Hgl2
Hg34
Hg24
Hgl4
Hgl
Hg2
Hg3
Hg4
HgO
++++
+++
+++
+++
+++
++
+++
+++
++
+
+
++++
++++
++++
++++
++++
+++
++++
++++
+++
++
++
++
++
++++
++++
++++
+++
++++
++
++++
+++
++
++
++
R
R
R
R
R
R
R
R
R
S
S
S
S
-
+
-
NA
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
* Where there were two protein bands, the combination of the two
signals was calculated. The data are presented as the amount of protein
immunoprecipitatedby MAbs relativeto that precipitated by MAb 1-29:
+ + + + , 70 to 100%; + + + , 40 to 70%; + + , 20 to 40%, +,
5 to 20 %; - , < 5 %. Data are from three independent experiments.
tAbbreviations: S, carbohydrate chains were sensitive to Endo H
digestion; R, some carbohydrates were found to be resistant to Endo
H digestion; NA, not applicable. Data are from three independent
experiments.
Cell surface expression was analysed by indirect immunofluorescence with MAb 1-29. Data are from three independent experiments.
was observed, suggesting that the oligosaccharide at
glycosylation site g4 had a different structure from those
at the rest of the glycosylation sites. This heterogeneity in
glycosylation processing was further reflected in the
mutants in which glycosylation site g4 was involved. All
the mutants containing this site (Hgl4, Hg24, Hg34,
Hg234, Hg134 and Hg124) showed some effects on the H
protein migration characteristics. The combination of g2
and g4 showed the most pronounced synergistic effect.
This effect was further observed in the mutant Hg124
which showed different migration properties compared
with the other single-site glycosylation mutants (Fig. 2 a,
- T M ) . The mutant Hg123 lacking the glycosylation site
g4 showed a relatively homogeneous protein band (Fig.
2 a, - T M ) , which again supported the assumption that
heterogeneity in glycosylation was mainly dependent on
that of site g4.
Immunoreactivity of the glycosylation mutant proteins
with a panel of MAbs
The contribution of the individual carbohydrate chains
to the folding and antigenicity of MV H protein was
determined using a panel of H protein-specific M A b s
which recognize continuous ( M A b 1-29) or discontinuous
(MAbs 16-DE6, 1-41 and 1-44) epitopes (A. Hu, J. K6vamees & E. Norrby, unpublished results). The transfected
1047
cells were pulse-labelled for 30 min and incubated in the
chase medium for 2 h for the processing and maturation
of the protein. Cells were lysed, the labelled glycosylation
mutant proteins were immunoprecipitated with individual MAbs and the polypeptides were analysed by
S D S - P A G E . The autoradiograms were quantitatively
interpreted by densitometry. To measure the immunological reactivity of the glycosylation mutant proteins,
we compared the amount of the H protein immunoprecipitated by the M A b s reacting with discontinuous
epitopes with that precipitated by the M A b recognizing
a continuous epitope. The results are shown in Fig. 3 and
summarized in Table 2. All mutants, including the
unglycosylated mutant Hg0, were found to react with
M A b 1-29 as well as M A b 1-44, but the reactivity with
the latter was weaker. Removal of any single glycosylation site at positions gl, g2, g3 or g4 had no effect on
the ability of the H protein to react with the MAbs
specific for conformational epitopes (Fig. 3 a), indicating
that no single carbohydrate chain appears to be critical
for the formation of such epitopes. When the double-site
mutants were examined, mutants Hg34 and Hg24 were
found to retain nearly full M A b reactivity, whereas
mutants H g l 2 and H g l 4 showed a markedly reduced
reactivity (Fig. 3b). Further analysis of the triple-site
mutants which each contain a single functional glycosylation site showed that the presence of the gl or g2
carbohydrate chains alone did not cause retention of the
M A b reactivity (Fig. 3 c). However, the mutants Hg3 and
Hg4, which respectively have the g3 and g4 carbohydrate
chains alone, showed approximately 2 0 % of M A b
reactivity. These results suggest that these carbohydrate
chains have a large influence on the conformational
epitopes of the H protein.
Folding and oligomerization of the glycosylation mutant
proteins
The folding state of the glycosylation mutant proteins
was investigated using non-reducing gels. BHK-21 cells
were transfected with plasmids containing H glycosylation mutant cDNAs, labelled with [35S]methionine
for 30 rain and incubated in the chase medium for 2 h.
Cells were lysed and the H proteins were immunoprecipitared with M A b 1-29 which recognizes a continuous
epitope on the H protein, and the immunoprecipitates
were separated by S D S - P A G E under reducing (Fig. 2
and Fig. 3) and non-reducing conditions (Fig. 4). Wildtype and mutant H proteins migrated differently in the
presence and absence of 2-ME. In the absence of 2-ME,
little m o n o m e r form of the H protein was observed and
most of the molecules displayed the migration pattern of
dimers or of larger aggregates. The single- and doublesite mutant H proteins were all folding-competent,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
1048
A. Hu and others
(a)
.,
~
=
,
=
=
~
=
,
=
=
%
,*
%
-*
~
V~ %
%
"
=
Aggregate ]
-- 2 0 0 K
Dimer -
-- 92-5K
Monomer -- 69K
(b)
Aggregate
I
- 200K
Dimer -
Fig. 4. Analysis of H glycosylation mutants under non-reducing conditions. (a) Equal amounts of cell lysates were immunoprecipitated
with MAb 1-29. The samples were analysed by SDS-PAGE under non-reducing conditions. Positions of the monomer, dimer, and
aggregate of the glycosylation mutants are indicated on the left. The positions of Mr markers are indicated on the right. (b) Same gel
as (a), but only the top section of the film which was deliberately overexposed is shown. Positions of the dimer and aggregate of the
glycosylation mutants are indicated on the left.
ee5
Monomer- :
Hg0 - i~,
e¢~
¢',1
¢',1
¢',1
~
~l"
'~"
-80K
- 69K
Fig. 5. Processing of wild-type and glycosylation mutant H proteins.
Cell lysates were prepared and were immunoprecipitated with MAb 1-29.
The immune complexes were digested with Endo H overnight and
the polypeptides were analysed by SDS-PAGE. The positions of
unglycosylated H protein (Hg0) and the monomer form of the H
protein are indicated on the left. The positions of M r markers are
shown on the right.
although the quantity o f folded proteins was reduced
c o m p a r e d with the wild-type protein. W h e n three
glycosylation sites were inactivated, properly folded
protein could be seen only after overexposure o f the film
(Fig. 4b), indicating that these m u t a n t H proteins were
unstable a n d / o r susceptible to proteolytic degradation.
N o dimer f o r m o f the m u t a n t protein could be observed
when the m u t a n t Hg0 protein was analysed under nonreducing conditions and this m u t a n t protein was aggregated on the top o f the gel, suggesting that the m u t a n t
protein was totally defective in folding and m o s t o f the
molecules existed in the f o r m o f aggregates. A protein
b a n d o f Mr approximately 200K was observed in all
lanes; this p r o b a b l y represented a complex between the
H protein and a cellular protein. Aggregates at the top o f
the gel were observed a m o n g all mutants. The aggregate
materials were greatly reduced in a m o u n t u n d e r reducing
conditions (data not shown), which is consistent with the
assumption that they are disulphide-linked. There was a
large difference in the ratio o f the a m o u n t o f the dimers
to that o f the aggregates between the m u t a n t s having two
or m o r e than two glycosylation sites and the mutants
with a single glycosylation site or without any such site.
Intracellular transport of the glycosylation mutant
proteins
Pulse-chase analysis has been used to examine the
m a t u r a t i o n o f the H protein in MV-infected cells (A. Hu,
J. K 6 v a m e e s & E. N o r r b y , unpublished results). To
extend these studies and determine what effect removal
o f a specific N-linked glycosylation site had on the
m a t u r a t i o n o f the glycoprotein, the processing o f the
m u t a n t proteins was analysed. There is evidence that
p r o p e r folding and oligomerization are prerequisites for
transport o f integral m e m b r a n e proteins f r o m the E R
(Copeland et al., 1986, 1988; Gething et al., 1986; Kreis
& Lodish, 1986). To determine whether the glycosylation
m u t a n t proteins acquire resistance to E n d o H digestion,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
Measles virus haemagglutinin N-linked glycans
a marker of the modification of carbohydrate chains
from simple to complex form in the medial Golgi
complex, transfected cells were labelled with
[35S]methionine for 30 min, followed by incubation in
chase medium for 2 h to allow processing and maturation
of the mutant molecules. H proteins were immunoprecipitated with MAb 1-29, digested with Endo H and
analysed by SDS-PAGE. The results are shown in Fig.
5 and summarized in Table 2. The Endo H-resistant form
of the H protein was occasionally observed in the triplesite glycosylation mutants only after overexposure of the
film and hence these mutant proteins were considered to
be the molecules retained in the ER, although the
occurrence of a small portion of the molecules in the
Golgi could not be excluded. The results of most of the
mutant H proteins showed a good correlation between
the acquisition of reactivity to the MAbs detecting
conformational epitopes, representing folding and oligomerization, and the conversion of some H carbohydrate
chains to the Endo H-resistant forms (Table 2), indicating intracellular transport from the ER to the medial
Golgi complex.
Internal
1049
Surface
Intracellular localization of the glycosylation mutant H
proteins
The subcellular localization of the glycosylation mutant
proteins was examined by indirect immunofluorescence.
Immunostaining was performed on fixed intact or fixed
and permeabilized transfected BHK-21 cells using MAb
1-29 and FITC-conjugated goat anti-mouse IgG. An
example representing wild-type, single-, double-, tripleand quadruple-site mutants is shown in Fig. 6. Cells
expressing wild-type H protein exhibited internal staining
throughout the cytoplasmic reticulum as well as in the
juxtanuclear region (Fig. 6a), and cell surface staining
was readily detectable (Fig. 6b). In contrast, the
unglycosylated H protein showed no surface staining
(Fig. 6j), and the internal staining pattern showed that
this protein was limited to reticular perinuclear structures
(Fig. 6 i), indicating the presence of unfolded molecules
in the ER. The results of the indirect immunofluorescence
analysis for the entire panel of mutants are presented in
Table 2. We found that the H glycosylation mutants
(Hg234, Hg134, Hg124, Hg123, Hgl2, Hg34, Hg24 and
Hg 14) that had molecules recognizable by MAbs reacting
with conformational epitopes and that acquired some
Endo H-resistant carbohydrate chains could be detected
on the cell surface. The mutants (Hg3 and Hg4) that had
molecules recognizable by the MAbs reacting with
conformational epitopes and that did not acquire Endo
H-resistant forms of carbohydrate chains did not express
a detectable level of H protein at the cell surface. This
should not be surprising since conformational antigenic
Fig. 6. Indirect immunofluorescenceof wild-type and mutant H
proteins. TransfectedBHK-2I cells on coverslips were fixed in 3%
paraformaldehydein PBS for cell surfacefluorescence(b, d,f, h and j)
or fixedthen permeabilizedin 0.1% Triton X-100 (a, c, e, g and 0. H
proteins were stained with MAb 1-29, followedby FITC-conjugated
goat anti-mouseIgG. (a, b) wild-type;(c, d) Hg234;(e,f) Hgl2; (g, h)
Hgl ; (i,j) Hg0.
epitope formation of the MV H protein occurs in the ER
(A. Hu, J. K6vamees & E. Norrby, unpublished results)
and the locally folded conformation-competent protein
does not necessarily indicate a transport-competent
molecule.
Discussion
Accumulated data have shown that not all the N-linked
glycosylation target sequences Asn-X-Ser/Thr are actually used (Kornfeld & Kornfeld, 1985). The amino acid
sequence of the MV H glycoprotein contains five
potential sites for the addition of N-linked oligosaccharide chains. Deletion and analysis of the individual
sites indicate that four sites are used (asparagine 168,
187, 200 and 215) and that asparagine residue 238 is not
utilized. The reason for this remains unclear. It is
generally considered that a polypeptide influences its
own glycosylation by controlling accessibility of oligosaccharyltransferase. Inspection of the amino acid
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
1050
A. H u and others
sequence of MV H protein reveals that this potential Nlinked glycan acceptor site is flanked by three positively
charged amino acids, residues Lys-236, Lys-242 and Arg243 (Alkhatib & Briedis, 1986). The ability of the
tripeptide signal to achieve a favourable conformation
by interacting with neighbouring amino acids may have
a significant effect on the realization, rate and extent of
glycosylation (Kornfeld & Kornfeld, 1985). This positively charged fragment may convey an unfavourable
element to the conformation of the target sequence and
hence the acceptor site at asparagine 238 may not be
accessible to the glycosylation enzyme.
Carbohydrate modifications at individual glycosylation sites were relatively homogeneous. However,
heterogeneity of oligosaccharide structure was observed
at the glycosylation site g4, which was responsible for the
appearance of two populations of the H glycoprotein in
BHK-21 cells. The heterogeneity may be a result of
altered carbohydrate processing, since pulse-chase analysis showed that the 79K form of the H protein could
be seen in the processed samples, but not in the pulselabelled samples (data not shown). Heterogeneity of
glycosylation is also observed when the same plasmid
(pert1) is used for transfection in HeLa cells (Cattaneo
& Rose, 1993). The heterogeneity in glycosylation is
therefore unlikely to be host cell-dependent, but is an
intrinsic property dictated by the primary amino acid
sequence of the H protein. Two types of H protein
caused by altered glycosylation processing have also
been observed in virus-infected ceils (Graves, 1981;
Ogura et al., 1991). Thus, heterogeneity in H protein
glycosylation probably also exists in virus lyric infections.
This type of observation has been made previously, e.g.
in studies of herpesvirus glycoproteins (Sodora et al.,
1989; Tikoo et al., 1993) and of the cellular glycoprotein
Thy-1 (Rademacher et al., 1988). However, the significance of glycosylation heterogeneity remains unclear.
A number of studies have indicated that N-linked
glycans can influence proper folding, disulphide bond
formation and oligomerization of glycoproteins (Gallagher et al., 1992; Grigera et al., 1991; Guan et al.,
1988; Machamer & Rose, 1988; Machamer et al., 1985;
Matzuk & Boime, 1988; Ng et al., 1990; Sodora et al.,
1989; Taylor & Wall, 1988; Tikoo et al., 1993; Vidal
et al., 1989; Wright et al., 1989). Some studies have
suggested that N-linked oligosaccharides affect the
conformational integrity of certain proteins, whereas
some proteins appear relatively unaffected (Hannink &
Donoghue, 1986; Williams & Lamb, 1986). The mutant
Hg0 lacking any glycosylation sites was found to produce
molecules that formed disulphide-linked aggregates that
did not fold properly, did not oligomerize, accumulated
in the ER, and were not transported to the cell surface,
as was found previously for inappropriately disulphide-
linked aggregates of vesicular stomatitis virus G protein,
influenza virus haemagglutinin and simian virus 5
haemagglutinin-neuraminidase (Doms et al., 1988;
Machamer & Rose, 1988; Hurtley et al., 1989; Ng et al.,
1990). Even MV H proteins with a single side-chain
could not efficiently fold or dimerize, and were not
efficiently transported to the cell surface.
The neutralization escape mutants selected by MAb
1-44 show a single amino acid alteration, Ser-189 to Pro
(Sheshberadaran & Norrby, 1986; Hu et al., 1993). This
change inactivates the glycosylation of site g2. In this
study, the mutant Hg134 lacking g2 with alteration of
Asn-187 to Ser was found to react with MAb 1-44 to the
same extent as the wild-type. The observed phenotype of
the neutralization escape mutants may therefore result
from the change in the amino acid near the glycosylation
site rather than the absence of the N-linked oligosaccharide. The glycosylation g2 might thus not be
directly involved in this epitope, although a possible
indirect influence could not be excluded.
The presence of glycosylation gl, g2 alone, or a total
lack of glycosylation yielded a significantly reduced
immunoreactivity with the MAbs recognizing discontinuous epitopes, indicating that the g 1 or g2 individually
have little effect on the conformation of the discontinuous epitopes. In contrast, the presence of glycosylation g3 or g4 alone allowed moderate immunoreactivity of the H protein to the MAbs specific for the
conformational epitopes, suggesting that glycosylation
g3 or g4 alone can contribute to the discontinuous
epitopes. Interestingly, these two sites are in close
proximity to glycosylation site g3 of the corresponding
attachment protein of simian virus 5 which has been
shown to have an important role in folding, assembly
and intracellular transport (Ng et al., 1990).
Mutants containing two oligosaccharide chains were
folded and transported to the cell surface and maintained
moderate immunoreactivity with MAbs specific for the
conformational epitopes, albeit less efficiently than the
wild-type protein. The H protein from one case of
subacute sclerosing panencephalitis (SSPE) MV displayed the same glycosylation phenotype as the mutant
Hgl4. The mutant Hgl4 protein was transport-competent whereas the protein of the SSPE case A is poorly
transported to the cell surface (Cattaneo & Rose, 1993),
indicating some other mutations besides those causing
the inactivation of the N-linked glycosylation are also
involved in the process. Alternatively, it may be a result
of different expression efficiency in different cell lines.
Mutants containing three oligosaccharide chains were
found to have biological activity indistinguishable from
the wild-type H protein, suggesting that individual
carbohydrate side-chains are not necessary for the
expression of the structure and function of the H
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
M e a s l e s virus haemagglutinin N - l i n k e d glycans
molecules. Therefore, the ability of H glycosylation
mutant proteins to be transported through the exocytic
pathway is not dependent on the presence of any
particular carbohydrate chain but on the collective
carbohydrate chains that substantially aid in the folding
of the H protein to acquire the native structure.
Our sincere thanks go to Drs Yihai Cao, Anita Bergstrrm, Lars
Melin and Ralf Pettersson for helping with establishment of the
vaccinia virus expression system and Dr Susan Cox for careful reading
of the manuscript. We also thank Dr Jan K6vamees for helpful
discussion. This work was supported by Swedish Medical Research
Council (B93-16X-00116-29A).
References
ALKHATm,G. & BRIEDIS,D. J. (1986). The predicted primary structure
of the measles virus hemagglutinin. Virology 150, 479-490.
CATTANEO, R. & ROSE, J.K. (1993). Cell fusion by the envelope
glycoprotein of persistent measles virus which caused lethal human
brain disease. Journal of Virology 67, 1493 1502.
COPELAND, C. S., DOMS, R.W., BOLZAU, E. M., WEBSTER, R.G. &
HELENIUS, A. (1986). Assembly of influenza hemagglutinin trimers
and its role in intracellular transport. Journal of Cell Biology 103,
1179 1191.
COPELAND, C.S., ZIMMER,K.P., WAGNER,K.R., HEALEY, G.A.,
MELLMAN, I. & HELENIUS,A. (1988). Folding, trimerization, and
transport are sequential events in the biogenesis of influenza virus
hemagglutinin. Cell 53, 197-209.
DOMS, R. W., RUUSALA,A., MACHAMER,C. M., HELENIUS,A. & ROSE,
J. K. (1988). Differential effects of mutations in three domains on
folding, quaternary structure, and intracellular transport of vesicular
stomatitis virus G protein. Journal of Cell Biology 107, 8%99.
ESCHLE,D. (1988). Cloning and reconstruction of the entire measles virus
genome. Diploma thesis, University of Zurich.
FUERST, T.R., NINES, E.G., STUDIER, F.W. & Moss, B. (1986).
Eukaryotic transient-expression system based on recombinant
vaccinia virus that synthesizes bacteriophage T7 RNA polymerase.
Proceedings of the National Academy of Sciences, U.S.A. 83,
812~8126.
GALLAGHER, P.J., HENNEBERRY,J.M., SAMBROOK, J. & GETHING,
M. J. (1992). Glycosylation requirements for intracellular transport
and function of the hemagglutinin of influenza virus. Journal of
Virology 66, 7136-7145.
GETHING,M. J., McLAMMON,K. & SAMBROOK,J. (1986). Expression of
wild type and mutant forms of influenza hemagglutinin: the role of
folding in intracellular transport. Cell 46, 939-950.
GRAVZS, M.C. (1981). Measles virus polypeptides in infected cells
studied by immune precipitation and one-dimensional peptide
mapping. Journal of Virology 38, 224-230.
GRIGERA, P. R., MATHIEU,M. E. & WAGNER, R. R. (1991). Effect of
glycosylation on the couformational epitopes of the glycoprotein of
vesicular stomatitis virus (New Jersey serotype). Virology 180, 1 9.
GUAN, J. L., CAO, H. & ROSE, J. K. (1988). Cell-surface expression of
a membrane-anchored form of the human chorionic gonadotropin
subunit. Journal of Biological Chemistry 263, 5306-5313.
HANNrNK, M. & DONO6Ht~, D. J. (1986). Cell surface expression of
membrane-anchored V-sis gene products: glycosylation is not
required for cell surface transport. Journal of Cell Biology 103,
2311-2322.
HARDWICK, J. M. & HUSSELL,R. H. (1978). Glycoproteins of measles
virus under reducing and nonreducing conditions. Journal of
Virology 25, 687-692.
Hu, A., S~SHBERADARAN,H., NORRBY, E. & K6VAMEES,J. (1993).
Molecular characterization of epitopes on the measles virus
hemagglutinin protein. Virology 192, 351-354.
HURTLEY, S. M., BOLE, D.G., HOO\~ER-LtTTY, H., HELENIUS,A. &
COPELAND,C. S, (1989). Interactions of misfolded influenza virus
1051
hemagglutinin with binding protein (Bip). Journal of Cell Biology
168, 2117-2126.
KORNFELO,R. & KORNEELD,S. (1985). Assembly of asparagine-linked
oligosaccharides. Annual Review of Biochemistry 54, 631-664.
KREIS, T. E. & LODISH, H. F. (1986). Oligomerization is essential for
transport of vesicular stomatitis viral glycoprotein to the cell surface.
Cell 46, 929-937.
MACHAtC~R,C. E., FLORKIEWICZ,R. Z. & ROSE, J. K. (1985). A single
N-linked oligosaccharide at either of the two normal sites is sufficient
for transport of vesicular stomatitis virus G protein to the cell
surface. Molecular and Cellular Biology 5, 3074-3083.
MACHAS~R, C. E. & ROSE, J. K. (1988). Vesicular stomatitis virus G
proteins with altered glycosylation sites display temperature-sensitive
intracellular transport and are subject to aberrant intermolecular
disulfide bonding. Journal of Biological Chemistry 263, 5955 5960.
MACKETT,M., SMITH,G. L. & Moss, B. (1985). The construction and
characterization of vaccinia virus recombinants expressing foreign
genes. In DNA Cloning, vol. 2, pp. 191-211. Edited by D. Rickwood
& B. D. Haines. Washington, D.C. : IRL Press.
MATZUK, M. M. & BOIME,I. (1988). The role of the asparagine-linked
oligosaccharides of the ~ subunit on the secretion and assembly of
human chorionic gonadotropin. Journal of Cell Biology 106,
1049 1059.
MOTTET,G., PORTNER,A. & ROU×, L. (1986). Drastic immunoreactivity
changes between the immature and mature forms of the Sendai virus
HN and F0 glycoprotein. Journal of Virology 59, 132-141.
NG, D. T. W., RANDALL,R. E. & LAMB, R.A. (1989). Intracellular
maturation and transport of the SV5 type II glycoprotein hemagglutinin-neuraminidase: specific and transient association with
GRP78-Bip in the endoplasmic reticulum and extensive internalization from the cell surface. Journal of Cell Biology 109, 3273-3289.
NG, D. T. W., HIEBERT,S. W. & LAMB,R. A. (1990). Different roles of
individual N-linked oligosaccharide chains in folding, assembly, and
transport of the simian virus 5 hemagglutinin-neuraminidase.
Molecular and Cellular Biology 10, 1989-2001.
NORRBY, E. & OXMAN,M. (1990). Measles virus. In Virology, 2nd edn,
pp. 1013-1044. Edited by B. N. Fields & D. M. Knipe. New York:
Raven Press.
NORRBY, E., CHEN, S.N., TOGASHI, T., SHESHBERADARAN,H. &
JoIa~SON, K. P. (1982). Five measles virus antigens demonstrated by
use of mouse hybridoma antibodies in productively infected tissue
cells. Archives of Virology 71, 1-11.
OGURA, H., SATO, H., KAMIYA,S. d~ NAKAMURA,S. (1991). Glycosylation of measles virus haemagglutinin protein in infected cells.
Journal of General Virology 72, 2679-2684.
OLOEN, K., PARENT, J.B. & WHJTE, S.L. (1982). Carbohydrate
moieties of glycoproteins. A re-evaluation of their function.
Biochimica et biophysica acta 650, 209-232.
RADEMACHER,Z. W., PAREKH, R. B. & DWEK, R.A. (1988). Glycobiology. Annual Review of Biochemistry 57, 785-838.
SAMBROOK, J., FRITSCH, E.F. & MANIATIS, T. (1989). Molecular
Cloning: A Laboratory Manual, 2rid edn. New York: Cold Spring
Harbor Laboratory.
SrI~SHBERADARAN,H., CVXN, S. H. & NORRBY, E. (1983). Monoclonal
antibodies against five structural components of measles virus. I.
Characterization of antigenic determinants on nine strains of measles
virus. Virology 128, 341-353.
SI~SHBER~a~ARAN, H. & NORRBY, E. (1986). Characterization of
epitopes on the measles virus hemagglutinin. Virology 152, 58-65.
SODORA, D. L., COrrEN, G. H. & EImqBERG, R. J. (1989). Influence of
asparagine-linked oligosaccharide on antigenicity, processing, and
cell surface expression of herpes simplex virus type I glycoprotein D.
Journal of Virology 63, 5184-5193.
TAYLOR,A. K. & WALL, R. (1988). Selective removal of alpha heavychain glycosylation sites causes immunoglobulin A degradation and
reduced secretion. Molecular and Cellular Biology 8, 4197-4203.
TIKOO, S. K., PARKER, M. D., VAN DEN HURK, J. V., KOWALSKI,J.,
ZAMB, T. J. & BABtUK, L. A- (1993). Role of N-linked glycaus in
antigenicity, processing, and cell surface expression of bovine
herpesvirus 1 glycoprotein glV. Journal of Virology 67, 726-733.
VtDAL, S., MOTTET,G., KOLAKOFSKY,D. & RotJx, L. (1989). Addition
of high-mannose sugars must precede disulfide bond formation for
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32
1052
A. Hu and others
proper folding of Sendai virus glycoproteins. Journal of Virology 63,
892-900.
WAXHAM, M. N., MERZ, D. C. & WOLINSKY, J. S. (1986). Intracellular
maturation of mumps virus hemagglutinin-neuraminidase glycoprotein: conformational changes detected with monoclonal antibodies. Journal of Virology 59, 392-400.
WILLIAM, M.A. & LAMB, R.A. (1986). Determination of the
orientation of an integral membrane protein and sites of glycosylation by oligonucleotide-directed mutagenesis: influenza B virus
NB glycoprotein lacks a cleavable signal sequence and has an
extracellular NH2-terminal region. Molecular and Cellular Biology 6,
4317-4328.
WRIaIJT, K. E., SALVATO, M. S. & B~:CH~IER, M. J. (1989). Neutralization epitopes of lymphocytic choriomeningitis virus are conformational and require both glycosylation and disulfide bonds for
expression. Virology 171, 417-426.
(Received 29 October 1993; Accepted 12 December 1993)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 02:25:32