Journal of General Virology (2004), 85, 1153–1165
DOI 10.1099/vir.0.19769-0
The small hydrophobic (SH) protein accumulates
within lipid-raft structures of the Golgi complex
during respiratory syncytial virus infection
Helen W. McL. Rixon,1 Gaie Brown,1 James Aitken,2 Terence McDonald,1
Susan Graham1 and Richard J. Sugrue1
Correspondence
Richard J. Sugrue
[email protected]
Received 30 October 2003
Accepted 19 January 2004
1
MRC Virology Unit, Institute of Virology, Church Street, Glasgow G11 5JR, UK
2
Division of Virology, University of Glasgow, Institute of Virology, Church Street, Glasgow
G11 5JR, UK
The cellular distribution of the small hydrophobic (SH) protein in respiratory syncytial virus
(RSV)-infected cells was examined. Although the SH protein was distributed throughout the
cytoplasm, it appeared to accumulate in the Golgi complex within membrane structures that
were enriched in the raft lipid, GM1. The ability of the SH protein to interact with lipid-raft
membranes was further confirmed by examining its detergent-solubility properties in Triton
X-100 at 4 6C. This analysis showed that a large proportion of the SH protein exhibited
detergent-solubility characteristics that were consistent with an association with lipid-raft
membranes. Analysis of virus-infected cells by immuno-transmission electron microscopy revealed
SH protein clusters on the cell surface, but only very low levels of the protein appeared to be
associated with mature virus filaments and inclusion bodies. These data suggest that during
virus infection, the compartments in the secretory pathway, such as the endoplasmic reticulum
(ER) and Golgi complex, are major sites of accumulation of the SH protein. Furthermore, although
a significant amount of this protein interacts with lipid-raft membranes within the Golgi complex,
its presence within mature virus filaments is minimal.
INTRODUCTION
Respiratory syncytial virus (RSV) is the major cause of
lower respiratory tract disease in young children. Although
RSV causes severe bronchiolitis in infants, milder respiratory tract infections are generally observed in most adults.
However, in certain high-risk groups within the adult
population (e.g. the elderly and immunocompromised),
disease progression can lead to more severe complications
and currently no effective vaccine exists to protect such
individuals within these high-risk groups from RSV infection.
The formation of the RSV envelope occurs within lipid-raft
platforms on the surface of virus-infected cells (Brown et al.,
2002b; McCurdy & Graham, 2003; Jeffree et al., 2003) and
the mature virus can be visualized as long filamentous
structures on the cell surface (Parry et al., 1979; Roberts
et al., 1995; Brown et al., 2002a). The virus encodes three
membrane-bound glycoproteins, namely the fusion (F),
attachment (G) and small hydrophobic (SH) proteins. The
F protein mediates fusion of the virus and cell membranes,
and the G protein is involved in virus attachment. The
biological properties of the F and G glycoproteins and the
role that they play during virus replication are relatively
well understood, but the functional significance of the SH
protein during virus replication remains unclear.
0001-9769 G 2004 SGM
Printed in Great Britain
The amino acid sequence of the SH protein is highly
conserved among all RSV A subtypes (Chen et al., 2000) and
the SH protein of the RSV A2 strain is expressed as several
different forms in virus-infected cells (Olmsted & Collins,
1989; Collins et al., 1990; Anderson et al., 1992). In the
human RSV A2 subtype, the full-length translated SH
protein contains 65 amino acids and appears in a variety
of forms depending upon its glycosylation status. These are
a 7?5 kDa non-glycosylated form (SH0), a 13–15 kDa Nlinked glycosylated form (SHg) and a polylactosaminoglycanmodified form of the protein (SHp) that ranges in size from
21–30 kDa. Evidence suggests that SH0 is first modified
by N-linked glycosylation to SHg, which is subsequently
transformed into SHp by the addition of polylactosaminoglycan. In addition, a fourth form of the SH protein can
be detected in which initiation of translation occurs at an
alternative methionine, giving rise to a 4?6 kDa truncated
form of the non-glycosylated protein (SHt).
The role played by the SH protein during RSV replication
is at present unclear. Studies which have employed reverse
genetics to produce viruses in which the SH gene has been
deleted, appear to suggest that it is dispensable for virus
growth but may play a role in evading the host’s immune
system (Bukreyev et al., 1997). In contrast, several publications have indicated that this protein may play an ancillary
1153
H. W. McL. Rixon and others
role in virus-mediated cell fusion (Heminway et al., 1994;
Perez et al., 1997; Techaarpornkul et al., 2001) and in this
regard, the ability of the SH protein to interact with the
F protein in virus-infected cells has been demonstrated
(Feldman et al., 2001). A greater understanding of its
biological properties should aid our understanding of the
role played by the SH protein during virus replication. This
report examines the cellular distribution of the SH protein
in virus-infected cells. The results provide evidence for an
association of the SH protein with lipid-raft membrane
structures, although only low levels of the protein are
associated with the envelope of mature virus filaments.
Furthermore, the Golgi complex appears to be a major site
of SH protein accumulation within the cell.
METHODS
Cells and viruses. The RSV A2 strain was used throughout this
study. Vero C1008 cells were purchased from the European Cell and
Culture Collection and maintained in Dulbecco’s Modified Eagle’s
Medium (DMEM) with 10 % foetal calf serum (FCS) and antibiotics.
Antibodies. The SH protein monoclonal antibody (MAbSH) was
prepared as follows. A branched peptide (SH52–64) corresponding
to the C-terminal 13 amino acids (NKTFELPRARVNT) of the RSV
A2 strain (Collins & Wertz, 1985) was synthesized on a tetravalent
(Lys)2-Lys-b-Ala-Wang resin (Novabiochem) using standard 9fluorenylmethoxycarbonyl solid-phase chemistry and an Advanced
ChemTech 348 (automated peptide synthesizer). The resulting
peptide, SH52–64, was coupled to a mouse T-cell epitope from sperm
whale myoglobin (NKALELFRKDIAAKYKE) using standard procedures. This peptide was used to immunize BALB/c mice and
monoclonal antibodies were prepared using standard protocols. The
tissue culture medium was harvested from MAbSH-expressing
hybridoma cells, concentrated and filtered through a 0?2 mm filter
prior to use.
The F protein monoclonal antibody (MAb169) was prepared from
recombinant F protein expressed in E. coli (H. Rixon, S. Graham and
R. Sugrue, unpublished data). MAb19 and anti-M were gifts from
Geraldine Taylor (IAH, Compton, UK) and Paul Yeo (MRC Virology
Unit, Glasgow, UK), respectively. The Golgi-specific marker, GM130,
was provided by Martin Lowe (School of Biological Sciences,
University of Manchester, UK), the Calnexin antibody was purchased
from Stressgen and cholera toxin B subunit conjugated to FITC
(CTX-B–FITC) was purchased from Sigma.
Radioimmunoprecipitation (RIP). The radiolabelling and RIP
procedures were performed as previously described (Rixon et al.,
2002). Briefly, mock- and RSV-infected Vero cells were incubated at
33 uC for 18 h, after which the DMEM plus 2 % FCS was replaced
with DMEM minus either methionine or glucose and the cells were
incubated for a further 18 h in the presence of either 100 mCi
(3?70 MBq) [35S]methionine ml21 or 150 mCi (5?55 MBq) [3H]glucosamine ml21, respectively. Immunoprecipitated proteins were separated by using 15 % SDS-PAGE and the [35S]methionine-labelled
proteins were detected by using a Bio-Rad personal Fx phosphorimager while the [3H]glucosamine-labelled proteins were detected by
fluorography.
Immunofluorescence. Cells were seeded on 13 mm glass coverslips and incubated overnight at 37 uC prior to infection with RSV.
Mock- or RSV-infected cells were fixed with 3 % paraformaldehyde
for 30 min at 4 uC. The fixative was removed and the cells were
washed once with PBS+1 mM glycine and then four times with
1154
PBS. The cells were incubated at 25 uC for 1 h with primary
antibody after which they were washed and incubated for a further
1 h either with anti-mouse or anti-rabbit IgG (whole molecule)
conjugated either to FITC, TRITC or cy5 (1/100 dilution). The
stained cells were mounted on slides using Citifluor and visualized
in a Zeiss Axioplan 2 confocal microscope. The images were processed using LSM 510 v2.01 software.
Flotation gradient analysis of detergent-insoluble lipid-raft
membranes. Cell sheets were washed twice with PBSA (20 mM
sodium phosphate, 150 mM NaCl pH 7?2) and then twice with TM
buffer (10 mM Tris/HCl pH 7?4; 1 mM MgCl2). The cells (16108)
were scraped into TM buffer supplemented with an EDTA-free
protease-inhibitor mixture (Roche Molecular Biochemicals) and
Dounce homogenized (80 strokes). Unbroken cells and nuclei were
removed by centrifugation at 1000 g for 5 min. The 1000 g supernatant was centrifuged at 45 000 g for 15 min to pellet the cell
membranes, which were then washed in TM buffer. The membranes
were resuspended at 4 uC in PBSA containing 1 % Triton X-100 by
using a Dounce homogenizer, after which they were incubated for
a further 1?5 h at 4 uC. The homogenate was added to an equal
volume of 80 % (w/v) sucrose in 10 mM Tris/HCl, pH 7?4; 150 mM
NaCl; 1 mM EDTA (TNE). The solubilized membranes (in 40 %
sucrose) were placed at the bottom of a centrifuge tube and overlaid
successively with 7 ml 35 % sucrose and 2 ml 5 % sucrose (in TNE).
After centrifugation at 210 000 g for 18 h, the gradient was fractionated and the individual fractions were analysed by Western blotting
for the presence of specific proteins using appropriate antibodies.
Western blotting. Protein samples were separated using SDSPAGE and were transferred by Western blotting onto a PVDF
membrane. After transfer, membranes were washed with PBSA and
blocked for 18 h at 4 uC in PBSA containing 1 % Marvel and 0?05 %
Tween 20. They were then washed twice in PBSA, prior to incubation with the appropriate primary antibody for 1 h, washed four
times in PBSA containing 0?05 % Tween 20 and probed either with
goat anti-mouse or anti-rabbit IgG (whole molecule) peroxidase
conjugate (Sigma). The protein bands were visualized by using the
ECL protein detection system (Amersham). Apparent molecular
masses were estimated using Rainbow protein markers (Amersham)
in the molecular mass range 14?3–220 kDa.
electron microscopy (I-TEM). Cell
monolayers were pelleted in BEEM capsules and fixed for 5 h at 4 uC
with 0?5 % glutaraldehyde in PBS after which the fixed pellet was
extensively washed with PBS and dehydrated by transfer through a
gradient of 30, 50, 70, 90 and 100 % ethanol. The cell pellet was subsequently infiltrated with Unicryl (TAAB Laboratories) and the resin
polymerized by UV irradiation at 215 uC. Ultrathin sections were
placed on nickel-coated grids and incubated either with MAbSH
(1/50 dilution), or anti-M (1/100 dilution) for 5 h at 25 uC. The
grids were washed with PBS and incubated for 2 h either with antirabbit or anti-mouse IgG (whole molecule) 5 or 10 nm colloidal
gold conjugate (Sigma) or anti-rabbit IgG (whole molecule) 20 nm
colloidal gold conjugate (British BioCell) as appropriate. They were
then washed in PBS and fixed in osmium tetroxide vapour for 2 h
prior to staining with uranyl acetate (saturated in 50 : 50 ethanol/
water), washing in distilled water and counter-staining with lead citrate.
Immuno-transmission
RESULTS AND DISCUSSION
The specificity of MAbSH in RSV-infected Vero
C1008 cells
The data presented in this report were obtained with the
Vero C1008 cell line and the SH protein was detected in
Journal of General Virology 85
Accumulation of SH protein during RSV infection
RSV-infected cells by using the monoclonal antibody
MAbSH (see Methods). This antibody was raised against
a peptide whose sequence is present within the C terminus
of the SH protein and which represents a part of the protein that is orientated on the extracellular side of the cell
membrane (Collins & Mottet, 1993).
The specificity of MAbSH for the SH protein was confirmed
by Western blotting analysis (Fig. 1A). Probing RSVinfected cell lysates with MAbSH revealed the presence
of a major protein band of approximately 10 kDa and a
minor band of approximately 5 kDa. The protein profiles
and sizes of the products detected are similar to the
SH protein species that have been previously described
(Olmsted & Collins, 1989; Collins et al., 1990; Anderson
et al., 1992). The size of the 10 kDa species is consistent
with it being the non-glycosylated form of the SH protein
(SH0) and the 5 kDa species is similar to the expected size
of the truncated form of the SH protein (SHt). No protein
bands corresponding to SHg or SHp were detected by
Western blotting, confirming previous findings that these
glycosylated forms represent minor amounts of the total
expressed SH protein (Olmsted & Collins, 1989; Anderson
et al., 1992). As expected, no protein bands were detected
in mock-infected cells by Western blotting, thus demonstrating the specificity of MAbSH.
Although no significant SHg or SHp was detected by
Western blotting, low levels of these proteins could be
detected by immunoprecipitation of the SH protein from
lysates prepared from RSV-infected cells. In RSV-infected
cells radiolabelled with [35S]methionine (Fig. 1B), only SH0
Fig. 1. Specificity of MAbSH and MAb169
in RSV-infected cells. (A) Western blot analysis of mock- (M) and RSV-infected (I) Vero
cells at 36 h p.i. using MAbSH. (B and C).
MAbSH is able to detect the low levels of
the glycosylated SH protein forms by radioimmunoprecipitation (RIP). Mock- and RSVinfected Vero cells were labelled with
[35S]methionine (B) or [3H]glucosamine (C)
and the SH immunoprecipitated with
MAbSH as described in Methods. It should
be noted that the detection of the
[3H]glucosamine-labelled SH protein species
(i.e. SHg and SHp) by autoradiography
used much longer exposure times (typically
2 months) compared to those required for
detection of the [35S]methionine-labelled SH
protein species (i.e. SH0) (typically 2 days
when using fluorography). The molecular
mass markers are indicated as are the various SH protein species. Major non-specific
host-cell proteins detected in the immunoprecipitation reactions are highlighted ( ).
(D). Western blot analysis of mock- and
virus-infected cells at 36 h p.i. using
MAb169.
*
http://vir.sgmjournals.org
1155
H. W. McL. Rixon and others
1156
Journal of General Virology 85
Accumulation of SH protein during RSV infection
could be detected clearly, whereas in RSV-infected cells
radiolabelled with 3[H]glucosamine (Fig. 1C), low levels
of a 15 kDa labelled protein band together with diffusely
labelled protein species between 20 and 30 kDa could be
detected. Similar protein species have been previously
detected (Olmsted & Collins, 1989; Collins et al., 1990;
Anderson et al., 1992) and these presumably represent SHg
and SHp, respectively. This therefore shows that while
SH0 is the major form of the SH protein detected in RSVinfected cells, MAbSH is also able to recognize the glycosylated forms of the SH protein. In this report, the SH0 form
will be referred to as the SH protein unless otherwise stated.
In this study, two monoclonal antibodies were used to detect
the RSV F protein, namely MAb19, whose use has been previously described (Taylor et al., 1992; Brown et al., 2002a, b),
and MAb169 (see Methods). Analysis of virus-infected cell
lysates by Western blotting with MAb169 showed a single
protein band whose size is consistent with that of the
F1 subunit of the F protein (Fig. 1D). No protein bands
were detected in lysates prepared from mock-infected cells,
thus demonstrating the specificity of this reagent.
The SH protein interacts with lipid-raft
membrane structures
Although several publications have shown that the SH protein is an integral membrane protein (Olmsted & Collins,
1989; Collins & Mottet, 1993), there is little information
about how the protein is distributed within the virusinfected cell. Mock- and virus-infected cells were labelled
with MAbSH and analysed by fluorescence microscopy
(Fig. 2A). No significant staining was observed in mockinfected cells whereas in virus-infected cells, although
MAbSH exhibited a diffuse staining pattern across the
cell, the staining appeared to concentrate in the perinuclear
region. This observation, together with the known association of the SH protein with cellular membranes (Olmsted
& Collins, 1989), suggested that the SH protein may
accumulate at specific locations in the secretory pathway.
Therefore, infected cells were examined by fluorescence
microscopy after double-labelling with MAbSH and cellular
markers specific for the endoplasmic reticulum (ER) and
Golgi complex. Cells stained with GM130 exhibited a
perinuclear pattern of labelling that was characteristic for
Golgi staining (Fig. 2B). Although the SH protein partly
co-localized with ER markers such as Calnexin (Fig. 2C),
the co-localization between the SH protein and the Golgi
marker GM130, was much stronger. The strong level of
co-localization between GM130 and the SH protein
(Fig. 2B, yellow staining pattern) suggested that despite
being detected in the cytoplasm, a large proportion of the
total expressed SH protein was located in the early compartments of the secretory pathway and in particular the
Golgi compartment.
The F protein, like the SH protein, is transported through
the secretory pathway of the cell. Therefore, the doublelabelling pattern of the F protein and GM130 (Fig. 2D,
plate A) was compared with that of the SH protein and
GM130 shown in Fig. 2(B). In contrast to the strong colocalization observed between the SH protein and GM130,
the F protein showed very little co-localization with GM130
(Fig. 2D, plate A), which is consistent with its rapid transit
through the secretory pathway (Collins & Mottet, 1991). In
addition, no significant co-localization of the M protein, a
major structural protein of the mature virus, with the Golgi
was detected, confirming previous observations (Henderson
et al., 2002). This, therefore, suggests that co-localization of
the SH protein with the Golgi compartment is not just a
consequence of its transport through the Golgi complex,
but reflects more an accumulation of the SH protein at this
site in the cell.
Recent evidence has suggested that RSV assembly occurs
within lipid-raft structures on the cell surface (Brown et al.,
2002a, b; McCurdy & Graham, 2003; Jeffree et al., 2003) and
several reports have demonstrated that specific proteins
are sorted into lipid-raft domains in the Golgi network
(reviewed by van Meer & Simons, 1988; Brown & London,
1998; Ikonen, 2001). It was therefore interesting to determine if the SH protein was able to interact with lipid rafts,
since very little of the protein was detected in structures
which are associated with mature virus (see below). In this
experiment, the cholera toxin B subunit (CTX-B) was used
to detect the presence of the raft lipid, GM1, within the
cell. CTX-B binds to GM1 at the cell surface and is then
transported via endocytosis to the Golgi apparatus (Lencer
et al., 1992, 1995, 1999; Le & Nabi, 2003). Exposure of
cells to CTX-B at 4 uC allows the detection of cell surface
GM1 since at this temperature CTX-B is not internalised.
At higher temperatures (e.g. between 25 uC and 37 uC), the
GM1-bound CTX-B is internalized and transported to
the Golgi complex. In this way, CTX-B-conjugated to FITC
(CTX-B–FITC) can be used to locate GM1-enriched microdomains at both the cell surface and within the Golgi
compartment depending upon the temperature at which
the cells are exposed to CTX-B. Virus-infected cells were
exposed to CTX-B–FITC under normal culturing conditions (i.e. 33 uC), which allowed internalization and
uptake of CTX-B–FITC, after which the cells were fixed
and labelled with MAbSH and GM130 and visualized using
a three-colour analysis that has been previously described
(Henderson et al., 2002). In this way, the distribution of
Fig. 2. Distribution of the SH protein in virus-infected Vero cells examined by fluorescence microscopy. (A) Mock- and virusinfected cells were labelled with MAbSH. (B and C) RSV-infected cells were labelled with MAbSH (green) and (B) Golgi
marker, GM130 (red) or (C) ER marker, Calnexin (red). (D) Distribution of GM130 (red) in relation to the F (plate A) and the M
(plate B) proteins (green) in virus-infected cells showing the merged images. In all cases, co-localization of antigens is shown
by the yellow staining pattern in the merged images.
http://vir.sgmjournals.org
1157
H. W. McL. Rixon and others
GM1, the Golgi compartment and the SH protein was
compared in the same cell. This revealed that GM1 (green),
GM130 (blue) and the SH protein (red) co-localized
strongly in the perinuclear region of the cell (Fig. 3A,
indicated by white staining pattern in the merged image),
providing evidence that the SH protein is located within
regions of the Golgi compartment that contain the raft
lipid GM1. The presence of the SH protein within GM1rich regions of the cell can be seen clearly by comparing
the SH protein and GM1 distribution within the cell,
revealing a significant level of co-localization between
CTX-B–FITC and MAbSH (Fig. 3B, indicated by yellow
staining in the merged image). However, no significant
co-localization of the SH protein with GM1 was detected at
the surface of virus-infected cells following exposure of the
cells to CTX-B–FITC at 4 uC (data not shown).
The ability of the SH protein to associate with lipid-raft
structures was investigated by examining its detergent
solubility. Although proteins associated with lipid-raft
membranes can be solubilized with detergents such as
Nonidet P 40 (NP40) and octyl b-glucoside (O-b-G), they
are resistant to solubilization when treated with other
detergents such as Triton X-100 or Brij 58 at 4 uC (Brown &
Fig. 3. SH protein is located in GM1-enriched microdomains within the Golgi complex. Virus-infected cells were incubated at
33 6C with CTX-B–FITC and processed for fluorescence microscopy using procedures previously described (Brown et al.,
2002b). The treated cells were then labelled with MAbSH and anti-GM130. (A) Distribution of the SH protein (red), GM1
(green) and the Golgi marker GM130 (blue) using a three-colour analysis routine. In the merged image, co-localization of all
three colours is indicated by the white staining pattern. (B) Cellular distribution of the SH protein (red) and GM1 (green) is
shown and co-localization is indicated by the yellow staining pattern in the merged image.
1158
Journal of General Virology 85
Accumulation of SH protein during RSV infection
London, 1998). In this way, the detergent solubility of the
SH protein was compared with those of three other proteins
which are known to reside in lipid rafts, namely caveolin-1
(cav-1), flotillin-1 (flot-1) and the RSV F glycoprotein.
Infected cell monolayers were extracted with the respective
detergent and separated by centrifugation into a detergentsoluble (S) and an insoluble cell pellet (P) fraction. Each
fraction was examined for the presence of the F protein
(indicated by the F1 subunit), cav-1 and the SH protein by
Western blotting (Fig. 4A). In lysates prepared with Triton
X-100 and Brij 58, both the F protein and cav-1 were found
to be in the insoluble pellet fraction, whereas in lysates
prepared with NP40/O-b-G, these proteins were located
almost entirely in the soluble fraction. Similarly, the SH
protein was soluble in NP40/O-b-G, but although this
protein remained largely associated with the pellet fraction
following Triton X-100 extraction, a significant proportion of the SH protein was soluble in Triton X-100. This
suggests that, in contrast to the F protein, flot-1 and cav-1,
a population of the SH protein exists which is distributed
in non-raft membranes. However, these results indicate
that significant levels of the SH protein appear to show
detergent-solubility characteristics which are consistent
with it being able to associate with raft membranes.
The detergent-solubility characteristics of the SH protein
were further examined by flotation gradient analysis.
Total membranes, prepared from virus-infected cells, were
extracted with Triton X-100 at 4 uC and placed at the
bottom of a step gradient containing 35 % and 5 % sucrose
Fig. 4. Examination of the detergent-solubility characteristics of the SH protein. (A) Virus-infected cells were extracted either
with 1 % Triton X-100, 1 % Brij 58 or 1 % NP40/60 mM O-b-G for 20 min at 4 6C. The lysates were then separated by
centrifugation (13 000 g, 30 min at 4 6C) into soluble (S) and pellet (P) fractions. Comparable amounts of each fraction were
analysed by Western blotting to detect the presence of cav-1, F and SH proteins. (B) Flotation gradient analysis. Total cell
membranes were prepared from virus-infected cells and extracted with 1 % Triton X-100 (a and b) or 1 % NP40/60 mM O-bG (c). The gradients were harvested and specific proteins were located in gradient fractions by Western blot analysis using
relevant antibodies. The profile of the SH protein can be seen and the presence of the F1 subunit indicates the location of the
F protein in the gradient. (C) The specific raft markers, cav-1 and flot-1 were predominantly located in the peak raft fraction
(fraction 3). 1 is the top fraction, 10 is the bottom fraction and 11 is the pellet (P) fraction.
http://vir.sgmjournals.org
1159
H. W. McL. Rixon and others
(see Methods). After centrifugation, the gradient was
fractionated and examined by Western blotting for the
presence of the SH and F proteins. Although some of each
protein was detected in the pellet fraction (fraction 11),
presumably as a result of protein aggregation, a significant
proportion of each protein was found in the fractions at, or
close to, the 5–35 % sucrose interface [Fig. 4B, (a) and (b)].
This corresponds to the region of the gradient at which
lipid-raft membranes migrate due to their higher buoyant
density (fractions 3–5). The endogenous raft-associated
cellular proteins, cav-1 and flot-1, were also detected at the
5–35 % sucrose interface (Fig. 4C). In contrast, analysis of
flotation gradients in which cell membranes from virusinfected cells had been extracted with NP40/O-b-G showed
that the SH protein was restricted entirely to the lower
fractions (fractions 10 and 11) [Fig. 4B, (c)], which is consistent with its solubilization in this detergent. Although
the F and SH proteins migrated into the region of the
flotation gradient at which raft membranes were present, a
proportion of the SH protein could also be detected in the
Triton-X-100-soluble fraction (fractions 9 and 10). This
supports the above suggestion that a proportion of the
SH protein resides in non-raft membrane systems. However,
the fluorescent microscopy data and detergent-solubility
characteristics of the SH protein suggest that a significant
proportion accumulates within lipid-raft membranes of
the Golgi complex.
Low levels of the SH protein are incorporated
into the virus envelope
It is currently unclear to what extent the SH protein interacts with mature RSV filaments. The cellular distribution
of the SH protein in infected cells was therefore examined
using two complementary techniques, fluorescence microscopy and immuno-transmission electron microscopy
(I-TEM).
Previously, fluorescent staining of RSV filaments had been
observed following labelling of infected cells with antibodies
against the F and G glycoproteins (Roberts et al., 1995;
Brown et al., 2002a; Jeffree et al., 2003) and the M protein
(Henderson et al., 2002). However, in the fluorescence
microscopic analysis described above (Fig. 2), no obvious
staining of the virus filaments with MAbSH was observed.
The absence of detectable SH protein in virus filaments
studied by confocal microscopy was confirmed by comparing the surface-labelling pattern of the SH and F proteins
(Fig. 5A, plates A and B). Non-permeabilized virus-infected
cells were labelled either with MAbSH or MAb19 and the
labelling pattern visualized by confocal microscopy. Lowlevel surface-labelling for the SH protein was observed,
which appeared to be punctate in appearance rather than
filamentous. This contrasted with the surface-labelling of
cells stained with MAb19 which revealed the presence of
clearly defined virus filaments.
The labelling pattern of the M protein, a major virus
structural protein, was compared with that of the SH
1160
protein in permeabilized virus-infected cells using confocal
microscopy. In this analysis, a Z-stack gallery of images
from the same cells at different focal planes was obtained,
thus allowing a comparison of the internal and surfacelabelling patterns (Fig. 5B, plates A–C). As expected, anti-M
efficiently labelled both the virus filaments and the inclusion
bodies (IB). The latter structures contain a large amount of
the N and P proteins in addition to the M protein (Garcia
et al., 1993; Garcia-Barreno et al., 1996; Ghildyal et al.,
2002). In contrast, no significant labelling of the filaments
was observed with MAbSH but a low level of SH protein
staining could be detected within IBs, observed as a weak
co-localization between anti-M and MAbSH (Fig. 5B, visualized as speckled yellow staining within the IB structures).
The fluorescence microscopy data suggested that the SH
protein was not present within mature virus filaments.
However, a major limitation of light microscopy is its lower
resolution compared with that of electron microscopy
and thus it seemed possible that fluorescence microscopy
may not detect low levels of virus-associated SH protein.
Previous studies which employed an indirect biochemical
analysis suggested that low levels of the SH0 and SHp forms
may be present within purified RSV particles (Collins et al.,
1990; Anderson et al., 1992). However, a major problem
in interpreting the results with this type of approach is
that RSV-infectivity remains largely cell-associated (Roberts
et al., 1995) and purification of infectious RSV is a difficult
process. It is possible that low levels of contaminating virus
proteins may also co-purify with the RSV particles, which
is particularly relevant since the failure to detect the SH
protein in virus filaments by fluorescence microscopy
suggested that if it was present in mature virus filaments,
it was only present at low levels. To overcome these two
potential problems, immuno-transmission electron microscopy (I-TEM) was employed to analyse RSV-infected cells
in situ. In this study, infected cells were processed for I-TEM
as previously described (Brown et al., 2002a) and the SH
protein was detected using MAbSH. In the I-TEM images,
the mature virus filaments were visible as were the virus
inclusion bodies (Fig. 6A and B). The thin sections prepared
from virus-infected cells were labelled with MAbSH and
the presence of bound antigen was detected using 5 nm
colloidal gold-labelled second antibody. Although it was
possible to visualize the SH protein in virus filaments, only
between one and four gold particles per filament were
detected, despite extensive examination of the samples. The
significance of this low level of gold signal within the virus
filaments is unclear. However, it is unlikely to be due to
non-specific binding of the gold conjugate since gold label
was not observed in sections in which either MAbSH was
omitted or MAbSH was replaced with a non-specific mouse
IgG (Fig. 6C and D).
The SH protein within IBs appeared to be more readily
detected than in virus filaments. This is presumably a
reflection of the larger surface area of the IBs within the
thin sections compared with that of the virus filaments,
Journal of General Virology 85
Accumulation of SH protein during RSV infection
A
B A
B
C
Fig. 5. SH protein is not detected in mature virus filaments by fluorescence microscopy. (A) Non-permeabilized virus-infected
cells were labelled with MAbSH (plate A) and MAb19 (plate B) and the labelling pattern observed by fluorescence
microscopy. (B) Permeabilized virus-infected cells were labelled with anti-M (red) and MAbSH (green) and examined by
fluorescence microscopy. The presence of the virus filaments (VF) and inclusion bodies (IB) can be seen. The images are
taken from a Z-stack series recorded through three individual cells. Three individual images from the gallery at different focal
planes are shown which represent (plate A) internal-labelling, (plate B) surface-labelling and (plate C) the virus filaments
above the cell surface. A low level of co-localization between the M and SH proteins can be seen in the inclusion bodies
(plate A, yellow staining pattern, white arrow).
which allows easier visualization of the SH protein. In
addition, the I-TEM data revealed regions on the infectedcell surface, in the absence of virus filaments, that exhibited
a relatively higher level of SH protein labelling (Fig. 6B). In
many instances, this was visualized as small clusters of gold
particles, suggesting that the SH protein is concentrated at
specific regions of the plasma membrane. This is consistent
with the data which show surface expression of the SH
protein (Fig. 5A, plate A) and with previous reports in
which expression of the SH protein on the surface of virusinfected cells has been observed (Olmsted & Collins, 1989).
I-TEM was used to examine and compare the relative
distribution of the SH and M proteins. Although such a
http://vir.sgmjournals.org
comparison may not provide an accurate quantitative
measure of SH and M protein levels, it was reasoned that the
levels of each protein in the virus-induced structures, i.e.
virus filaments and IBs, could be compared qualitatively.
The relative distribution of the M and SH proteins within
virus filaments was examined in sections labelled with
both anti-M and MAbSH, and the presence of the bound
antibody detected using 5 and 10 nm colloidal goldconjugated second antibodies, respectively (Fig. 7A and B).
An abundance of labelling with anti-M was identified by
the relatively large amount of 5 nm gold probe (highlighted
by white arrow) associated with mature virus filaments
and newly budded virus. In contrast, very little 10 nm
gold probe (highlighted by black arrow) appeared to be
1161
H. W. McL. Rixon and others
Fig. 6. SH protein is detected in mature virus filaments (VF) and inclusion bodies (IB) by I-TEM, but at low levels. (A and B).
Virus-infected cells were prepared for I-TEM (see Methods), labelled with MAbSH, and the presence of bound antibody
detected using 5 nm colloidal gold (indicated by black arrows). (B) The presence of SH protein clusters can be seen at the
cell surface in the absence of visible virus structures (highlighted by ). (C and D) Representative images from sections,
showing both VF and IB, incubated in the presence of secondary antibody conjugated to 5 nm colloidal gold only. Bar,
200 nm.
*
associated either with the mature virus filaments or with
newly budding virus.
A similar analysis to that above was performed to examine
the relative amounts of each protein in the virus IBs. In this
case, the presence of the SH protein was detected using a
gold probe of larger size (20 nm) (Fig. 7C). It is clear that
while there is an abundance of the 5 nm gold probe (white
arrow, representing the M protein), there are many fewer
20 nm gold particles (black arrow, representing the SH
protein), indicating that the SH protein is present in IBs
but in low amounts. The presence of the SH protein, an
1162
integral membrane protein, suggests that host-cell derived
membranes may be associated with IBs. It should be noted
that the results of the labelling experiment were consistent
irrespective of the size of the gold probe used for detection
since this analysis has been performed extensively using
a variety of gold conjugate sizes (between 5 and 20 nm) to
detect each of the two proteins.
Although direct evidence has been provided that a low level
of the SH protein is present within mature RSV filaments,
our results show that raft membranes within the Golgi
complex are a major site of SH protein accumulation
Journal of General Virology 85
Accumulation of SH protein during RSV infection
Fig. 7. Relative distribution of the M and
SH proteins within mature virus filaments.
Virus-infected cells were labelled with anti-M
and MAbSH and the relative distribution of
the M and SH and proteins compared in
virus filaments (VF) (A) and newly budding
virus (VB) (B). The distribution of the M and
SH proteins was detected using 5 and
10 nm colloidal gold, respectively. (C) Relative distribution of the M and SH proteins in
an inclusion body (IB). The presence of the
M and SH proteins was detected using 5
and 20 nm colloidal gold, respectively. Inset,
high magnification image of the IB showing
the different sizes of gold probe. The 5 nm
gold particles are highlighted by white
arrows and the 10 and 20 nm gold particles
by black arrows. Bar, 200 nm.
during virus infection. Although the function of the SH
protein has not been defined, it is possible that it may
modify some property of the Golgi complex to aid virus
protein transport through the secretory pathway. In this
respect, it is interesting to note that several small membrane
proteins encoded by other viruses are able to alter the
protein transport machinery of the host-cell, a list that
includes the M2 and BM2 proteins of influenza virus. The
M2 protein of influenza A virus modifies the acidity of the
Golgi compartment (Sugrue et al., 1990; Ciampor et al.,
1992), a process that has been shown, in polarized cells, to
decrease the rate of transport of glycoproteins to the apical
cell surface (Sakaguchi et al., 1996; Henkel & Weisz, 1998;
Henkel et al., 1998, 2000). More recent work has demonstrated that the BM2 protein of influenza B virus is able to
elicit similar effects to those observed for the M2 protein
(Mould et al., 2003). However, at present, the effect exerted
by the SH protein on the physiology of the host cell, and in
http://vir.sgmjournals.org
particular the secretory pathway (e.g. the Golgi complex),
remains to be established as does its functional significance
within the virus.
ACKNOWLEDGEMENTS
We thank Duncan McGeoch for critical review of the manuscript. We
are grateful to Geraldine Taylor at the Institute of Animal Health,
Compton, UK, for providing MAb19, Paul Yeo for providing anti-M
and Martin Lowe from the University of Manchester for providing
anti-GM130.
REFERENCES
Anderson, K., King, A. M., Lerch, R. A. & Wertz, G. W. (1992).
Polylactosaminoglycan modification of the respiratory syncytial
virus small hydrophobic (SH) protein: a conserved feature among
human and bovine respiratory syncytial viruses. Virology 191,
417–430.
1163
H. W. McL. Rixon and others
Brown, D. A. & London, E. (1998). Functions of lipid rafts in
biological membranes. Annu Rev Cell Dev Biol 14, 111–136.
Brown, G., Aitken, J., Rixon, H. W. McL. & Sugrue, R. J. (2002a).
Caveolin-1 is incorporated into mature respiratory syncytial virus
particles during virus assembly on the surface of virus-infected cells.
J Gen Virol 83, 611–621.
Brown, G., Rixon, H. W. McL. & Sugrue, R. J. (2002b). Respiratory
syncytial virus assembly occurs in GM1-rich regions of the host-cell
membrane and alters the cellular distribution of tyrosine phosphorylated caveolin-1. J Gen Virol 83, 1841–1850.
Bukreyev, A., Whitehead, S. S., Murphy, B. R. & Collins, P. L. (1997).
Recombinant respiratory syncytial virus from which the entire SH
gene has been deleted grows efficiently in cell culture and exhibits
site-specific attenuation in the respiratory tract of the mouse. J Virol
71, 8973–8982.
Chen, M. D., Vazquez, M., Buonocore, L. & Kahn, J. S. (2000).
Conservation of the respiratory syncytial virus SH gene. J Infect Dis
182, 1228–1233.
Ciampor, F., Bayley, P. M., Nermut, M. V., Hirst, E. M. A., Sugrue,
R. J. & Hay, A. J. (1992). Evidence that the amantadine-induced
M2-mediated conversion of influenza A virus hemagglutinin to the
low pH conformation occurs in an acidic trans Golgi compartment.
Virology 188, 14–24.
Collins, P. L. & Wertz, G. W. (1985). The 1A protein gene of human
respiratory syncytial virus: nucleotide sequence of the mRNA and a
related polycistronic transcript. Virology 141, 283–291.
Collins, P. L. & Mottet, G. (1991). Post-translational processing and
oligomerization of the fusion glycoprotein of human respiratory
syncytial virus. J Gen Virol 72, 3095–3101.
Collins, P. L. & Mottet, G. (1993). Membrane orientation and
oligomerization of the small hydrophobic protein of human
respiratory syncytial virus. J Gen Virol 74, 1445–1450.
Collins, P. L., Olmsted, R. A. & Johnson, P. R. (1990). The small
hydrophobic protein of human respiratory syncytial virus: comparison
between antigenic subgroups A and B. J Gen Virol 71, 1571–1576.
Henkel, J. R., Apodaca, G., Altschuler, Y., Hardy, S. & Weisz, O. A.
(1998). Selective perturbation of apical membrane traffic by expres-
sion of influenza M2, an acid-activated ion channel, in polarized
Madin–Darby canine kidney cells. Mol Biol Cell 9, 2477–2490.
Henkel, J. R., Gibson, G. A., Poland, P. A., Ellis, M. A., Hughey, R. P.
& Weisz, O. A. (2000). Influenza M2 proton channel activity
selectively inhibits trans-Golgi network release of apical membrane
and secreted proteins in polarized Madin–Darby canine cells. J Cell
Biol 148, 495–504.
Ikonen, E. (2001). Roles of lipid rafts in membrane transport. Curr
Opin Cell Biol 13, 470–477.
Jeffree, C. E., Rixon, H. W. McL., Brown, G., Aitken, J. & Sugrue, R. J.
(2003). Distribution of the attachment (G) glycoprotein and GM1
within the envelope of mature respiratory syncytial virus filaments
revealed using field emission scanning electron microscopy. Virology
306, 254–267.
Le, P. U. & Nabi, I. R. (2003). Distinct caveolae-mediated endocytic
pathways target the Golgi apparatus and the endoplasmic reticulum.
J Cell Sci 116, 1059–1071.
Lencer, W. I., Delp, C., Neutra, M. R. & Madara, J. L. (1992).
Mechanism of cholera toxin action on a polarized human
intestinal epithelial cell line: role of vesicular traffic. J Cell Biol
117, 1197–1209.
Lencer, W. I., Constable, C., Moe, S., Jobling, M. G., Webb, H. M.,
Ruston, S., Madara, J. L., Hirst, T. R. & Holmes, R. K. (1995).
Targeting of cholera toxin and Escherichia coli heat labile toxin in
polarized epithelia: role of COOH-terminal KDEL. J Cell Biol 131,
951–962.
Lencer, W. I., Hirst, T. R. & Holmes, R. K. (1999). Membrane traffic
and the cellular uptake of cholera toxin. Biochim Biophys Acta 1450,
177–190.
McCurdy, L. H. & Graham, B. S. (2003). Role of plasma membrane
lipid microdomains in respiratory syncytial virus filament formation.
J Virol 77, 1747–1756.
Feldman, S. A., Crim, R. L., Audet, S. A. & Beeler, J. A. (2001).
Mould, J. A., Paterson, R. G., Takeda, M., Ohigashi, Y.,
Venkataraman, P., Lamb, R. A. & Pinto, L. H. (2003). Influenza B
Human respiratory syncytial virus surface glycoproteins F, G and SH
form an oligomeric complex. Arch Virol 146, 2369–2383.
virus BM2 protein has ion channel activity that conducts protons
across membranes. Dev Cell 5, 175–184.
Garcia, J., Garcia-Barreno, B., Vivo, A. & Melero, J. A. (1993).
Olmsted, R. A. & Collins, P. L. (1989). The 1A protein of respiratory
syncytial virus is an integral membrane protein present as multiple,
structurally distinct species. J Virol 63, 2019–2029.
Cytoplasmic inclusions of respiratory syncytial virus-infected cells:
formation of inclusion bodies in transfected cells that coexpress the
nucleoprotein, the phosphoprotein, and the 22K protein. Virology
195, 243–247.
Garcia-Barreno, B., Delgado, T. & Melero, J. A. (1996). Identification
of protein regions involved in the interaction of human respiratory
syncytial virus phosphoprotein and nucleoprotein: significance for
nucleocapsid assembly and formation of cytoplasmic inclusions.
J Virol 70, 801–808.
Ghildyal, R., Mills, J., Murray, M., Vardaxis, N. & Meanger, J. (2002).
Respiratory syncytial virus matrix protein associates with nucleocapsids in infected cells. J Gen Virol 83, 753–757.
Heminway, B. R., Yu, Y., Tanaka, Y., Perrine, K. G., Gustafson, E.,
Bernstein, J. M. & Galinski, M. S. (1994). Analysis of respiratory
syncytial virus F, G, and SH proteins in cell fusion. Virology 200,
801–805.
Henderson, G., Murray, J. & Yeo, R. P. (2002). Sorting of the
respiratory syncytial virus matrix protein into detergent-resistant
structures is dependent on cell-surface expression of the glycoproteins. Virology 300, 244–254.
Henkel, J. R. & Weisz, O. A. (1998). Influenza virus M2 protein slows
traffic along the secretory pathway. pH perturbation of acidified
compartments affects early Golgi transport steps. J Biol Chem 273,
6518–6524.
1164
Parry, J. E., Shirodaria, P. V. & Pringle, C. R. (1979). Pneumoviruses:
the cell surface of lytically and persistently infected cells. J Gen Virol
44, 479–491.
Perez, M., Garcia-Barreno, B., Melero, J. A., Carrasco, L. &
Guinea, R. (1997). Membrane permeability changes induced in
Escherichia coli by the SH protein of human respiratory syncytial
virus. Virology 235, 342–351.
Rixon, H. W., Brown, C., Brown, G. & Sugrue, R. J. (2002). Multiple
glycosylated forms of the respiratory syncytial virus fusion protein
are expressed in virus-infected cells. J Gen Virol 83, 61–66.
Roberts, S. R., Compans, R. W. & Wertz, G. W. (1995). Respiratory
syncytial virus matures at the apical surfaces of polarized epithelial
cells. J Virol 69, 2667–2673.
Sakaguchi, T., Leser, G. P. & Lamb, R. A. (1996). The ion channel
activity of the influenza virus M2 protein affects transport through
the Golgi apparatus. J Cell Biol 133, 733–747.
Sugrue, R. J., Bahadur, G., Zambon, M. C., Hall-Smith, M., Douglas,
A. R. & Hay, A. J. (1990). Specific structural alteration of the
influenza haemagglutinin by amantadine. EMBO J 9, 3469–3476.
Taylor, G., Stott, E. J., Furze, J., Ford, J. & Sopp, P. (1992). Protective
epitopes on the fusion protein of respiratory syncytial virus
Journal of General Virology 85
Accumulation of SH protein during RSV infection
recognized by murine and bovine monoclonal antibodies. J Gen Virol
73, 2217–2223.
lacking the small hydrophobic and/or attachment glycoprotein gene.
J Virol 75, 6825–6834.
Techaarpornkul, S., Barretto, N. & Peeples, M. E. (2001). Functional
van Meer, G. & Simons, K. (1988). Lipid polarity and sorting in
analysis of recombinant respiratory syncytial virus deletion mutants
epithelial cells. J Cell Biochem 36, 51–58.
http://vir.sgmjournals.org
1165