the paramyxoviruses sv5 and mumps virus recruit host...
TRANSCRIPT
1
The Paramyxoviruses SV5 and Mumps Virus Recruit 2
Host Cell CD46 to Evade Complement-Mediated Neutralization 3
4
5
6
7
John B. Johnson1, Ken Grant
2 and Griffith D. Parks
1* 8
9
10
Department of Microbiology and Immunology1
and Pathology2, 11
Wake Forest University School of Medicine, Winston-Salem, NC 27157-1064 12
13
14
15
Running title: Paramyxovirus incorporate host cell CD46 16
17
*Corresponding Author. Mailing address: Department of Microbiology and Immunology, Wake Forest 18
University School of Medicine, Medical Center Blvd., Winston-Salem, NC 27157-1064, Tel: (336) 716-19
9083, Fax: (336) 716-9928, Electronic mail address: [email protected] 20
21
22
23
Copyright © 2009, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.J. Virol. doi:10.1128/JVI.00713-09 JVI Accepts, published online ahead of print on 20 May 2009
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
2
ABSTRACT 1
The complement system is a critical component of the innate immune response that all animal viruses 2
must face during natural infections. Our previous results have shown that treatment of the paramyxovirus 3
Simian Virus 5 (SV5) with human serum results in deposition of complement C3-derived polypeptides on 4
virion particles. Here we show that the virion-associated C3 component included the inactive form iC3b, 5
suggesting that SV5 may have mechanisms to evade the host complement system. Electron microscopy, 6
gradient centrifugation and western blot analysis indicated that purified SV5 virions derived from human 7
A549 cells contained CD46, a plasma membrane-expressed regulator of complement that acts as a 8
cofactor for cleavage and inactivation of C3b into iC3b. In vitro cleavage assays with purified 9
complement components showed that SV5 virions had C3b cofactor activity, resulting in specific Factor 10
I-mediated cleavage of C3b into inactive iC3b. SV5 particles generated in CHO cells which do not 11
express CD46 did not have cofactor activity. Conversely, virions derived from a CHO cell line that was 12
engineered to overexpress human CD46 contained elevated levels of virion-associated CD46 and 13
displayed enhanced C3b cofactor activity. By comparison with C3b, purified SV5 virions had very low 14
cofactor activity against C4b, consistent with the known preference of CD46 for C3b versus C4b. Similar 15
results were found for the closely related Mumps virus (MuV), except that MuV particles derived from 16
CHO-CD46 cells had higher C4b cofactor activity compared to SV5 virions. In neutralization assays with 17
human serum, SV5 and MuV containing CD46 showed slower kinetics and more resistance to 18
neutralization compared to SV5 and MuV that lacked CD46. Our results support a model whereby the 19
rubulaviruses SV5 and MuV incorporate cell surface complement inhibitors into progeny virions as a 20
mechanism to limit complement-mediated neutralization. 21
22
23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
3
INTRODUCTION 1
The complement system constitutes a complex group of both soluble and cell-associated 2
proteins that together form an integral part of the innate host defense against pathogens (reviewed in 7, 3
9, 11, 31). Complement can serve to link innate and adaptive immunity through a large number of 4
activities, including recognition of viruses, direct neutralization of infectivity, recruitment and 5
stimulation of leukocytes, opsonization by immune cells, and activation of T and B cell responses (9, 6
11, 27). Complement activation and the ability of viruses to counteract complement can play important 7
roles in viral pathogenesis, as well as the design of more effective vaccines and therapeutic vectors (6, 8
9, 17, 36, 43). The overall goal of the work described here was to determine the mechanism by which 9
the paramyxoviruses simian virus 5 (SV5) and mumps virus (MuV) limit activation of complement 10
pathways. 11
The complement cascade can be initiated through three main pathways: the classical pathway, 12
lectin pathway or alternative pathway (11, 40). These three pathways converge on a central component 13
C3, which is activated by cleavage into C3a and C3b. C3a serves as an anaphylatoxin to promote 14
inflammation. C3b can bind covalently to viral components to aid in opsonization and phagocytosis. In 15
addition, C3b can associate with other factors such as Factor B to form the C3 convertase (e.g. 16
C3bBb), and this functions to amplify the initially deposited C3b signal by further cleavage of C3 17
molecules in a feedback loop. Likewise, C4 can be activated by cleavage into the anaphylatoxin C4a 18
and the C4b fragment which links the classical and lectin pathways with the alternative pathway. The 19
association of C3b with further downstream components such as C6 through C9 can lead to formation 20
of the membrane attack complex (MAC) which is capable of lysing virus particles or infected cells 21
(reviewed in 7, 11, 28). 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
4
The complement system needs to be highly regulated to prevent inappropriate activation and 1
potential damage to normal cells and healthy tissues (e.g., 3). Self regulation of complement pathways 2
involves the highly concerted actions of a family of soluble and cell-associated proteins called 3
regulators of complement activation (RCA). RCA proteins can limit inappropriate complement 4
activation by two major mechanisms: 1) by accelerating the disassociation of C3 or C5 convertases or 5
2) by acting as a cofactor to promote proteolytic cleavage of C3b or C4b by the complement protease 6
Factor I. Examples of RCA proteins include Factor H, CD46, Complement Receptor 1 (CR1 or CD35), 7
and C4 binding protein (14, 19, 24, 45). 8
CD46 or Membrane Cofactor Protein (MCP) is an integral membrane RCA protein that is 9
expressed on a wide range of tissues and cell types (32). CD46 is an N- and O-linked glycosylatated 10
protein expressed at the plasma membrane as multiple isoforms that are derived from alternative 11
splicing (32, 33, 39, 41). CD46 selectively binds to both C3b and C4b on cell surfaces, where it acts as 12
a cofactor to promote efficient cleavage by complement protease Factor I (44; reviewed in 5, 32). For 13
C3b, CD46 and Factor I combine to mediate inactivation to iC3b, and this is a major mechanism for 14
limiting amplification of low basal levels of C3b that arise from spontaneous alternative pathway 15
activation. CD46 also serves as a cofactor for Factor I-mediated cleavage of C4b into C4c and C4d, but 16
this cofactor activity is less efficient than that seen for C3b cleavage (35, 45). 17
Viruses have evolved a number of mechanisms to inhibit or to delay the neutralizing effects of 18
complement (9, 16). Large DNA viruses have coding capacities that allow them to encode a variety of 19
mimics of host cell RCA proteins, and these viral homologs often function to inactivate complement 20
components by supplying cofactor activity or by accelerating the decay of convertases (reviewed in 2, 21
7, 31, 29). For example, herpes saimiri expresses a complement control protein that inhibits the C3 22
convertase (20). As an alternative mechanism to counteract complement, a number of enveloped DNA 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
5
viruses and retroviruses have been shown to recruit cell-associate RCA proteins into budding particles 1
(15, 47, 48). Examples of this include vaccinia virus (VACV) and human immunodeficiency virus type 2
1 (HIV 1) which incorporate CD55, CD59 and CD46 into progeny virions (16, 42, 48). 3
In contrast to retroviruses and large DNA viruses, mechanisms that are employed to limit or 4
evade host cell complement pathways have not been described for the paramyxovirus family of 5
negative strand RNA viruses (30). It has been known for many years that complement is an important 6
factor in paramyxovirus neutralization (18, 23, 34, 49, 50). For example, the closely related 7
paramyxoviruses Simian Virus 5 (SV5) and Mumps virus (MuV) preferentially activate the 8
complement alternative pathway in vitro, and this activation can contribute to the efficiency of 9
neutralization by human serum (23, 26). These findings raised the question of whether negative strand 10
RNA viruses have mechanisms to limit activation and/or amplification of the complement cascade. 11
We have previously shown that treatment of SV5 and MuV particles with normal human serum 12
led to deposition of C3-derived components on virions, but the virion-associated C3 molecules had 13
properties of the inactive form iC3b and not the intact C3b (26). In this study, we have tested the 14
hypothesis that SV5 and MuV incorporate host cell RCA proteins into budding virions as a mechanism 15
to limit complement activity through inactivation of C3b. CD46 was found associated with purified 16
SV5 and MuV virions that were derived from cells expressing CD46 and these particles displayed C3b 17
cofactor activity in vitro. Consistent with this inactivation of C3b, CD46-containing virus was more 18
resistant to in vitro neutralization by human serum compared to virus derived from CD46-deficient 19
cells. Our results support a model whereby these closely related paramyxoviruses incorporate at least 20
one cell surface RCA protein into progeny virions as a mechanism to evade complement-mediated 21
neutralization. 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
6
MATERIALS AND METHODS 1
Cells and Viruses. A549 and CV1 monolayer cultures were grown in Dulbecco modified Eagle 2
medium (DMEM) supplemented with 10% heat inactivated fetal bovine serum (FBS), 100 µg/ml 3
streptomycin and 200 mM L-Glutamine. Chinese hamster ovary (CHO) cells that overexpress the 4
CYT2 isoform of human CD46 (CHO-CD46) and the control drug-resistant CHO cells were kindly 5
provided by Dr. Denise Gerlier (21) and were maintained in DMEM supplemented with 4.5 g/l 6
glucose, 10 mM HEPES pH 7.2, 10 ug/ml Gentamycin, L-glutamine, 1% non essential amino acid and 7
6% FBS. Recombinant WT SV5 (W3A strain) or the Enders strain of MuV (ATCC, VR-1379) were 8
grown in A549, CHO-CD46 or control CHO cells in the presence of heat inactivated serum. Virus was 9
purified by sucrose gradient centrifugation and titrated as previously described (26). 10
Ultracentrifugation and Western blotting. Sucrose gradient purified particles alone or 11
particles that had been incubated with NHS at a ratio of 1:1 (v/v) were layered on top of 15-60 % 12
sucrose gradients and subjected to ultracentrifugation as described earlier (26; 37). Fractions collected 13
from the bottom of the tube (250 ul) were analyzed by SDS PAGE followed by Western blotting with 14
an antibody against the SV5 P protein (26) or with a polyclonal goat antibody against human C3 (MP 15
Biomedics, Cappell, CA) at 1:1000 dilution. For CD46, gradient fractions were concentrated by TCA 16
precipitation before analysis by Western blotting with rabbit anti human CD46 polyclonal antibody 17
(Santa Cruz Biotechnology, CA) at a dilution of 1:500. SV5 infected A549 cell lysates were used as an 18
electrophoretic marker for position of CD46. The blots were treated with Super Signal West Pico 19
Chemiluminescent substrate (Thermo Scientific) and the signal was detected by exposing the blots to 20
film. 21
Complement reagents. Normal human serum (NHS) was collected, processed as described 22
previously (26) and stored at -80oC in small aliquots. Results were consistent among serum from 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
7
multiple donors. Purified complement proteins C3b, C4b, Factor H and Factor I were from 1
Complement Technologies (Tyler, TX) and were used in standard assays as described (8). Soluble CR1 2
was a kind gift from Dr. Henry Marsh (Celldex Therapeutics; Needham, MA). 3
Electron Microscopy. Sucrose gradient purified SV5 or MuV particles (10 ul, ~105 pfu/ml) 4
that had been generated either in A549, CHO or CHO-CD46 cells were analyzed for the presence of 5
CD46 by adsorbing on carbon-coated 200 mesh gold grids (Electron Microscopy Sciences, PA) and 6
incubated at room temperature in a humidified chamber for 5 min. The grids were blocked with PBS 7
containing 1% BSA and the adsorbed virus were probed with a mouse anti-human CD46 monoclonal 8
antibody (R&D Systems, MN) at a dilution of 1 ug/10 ul and detected with 12 nm colloidal gold goat 9
anti-mouse antibody (Jackson Immunoresearch Laboratories, PA). The labeled particles were subjected 10
to negative staining with 2% phosphotungstic acid (pH 6.6) and analyzed with a Philips TEM400 11
transmission electron microscope as described previously (26). 12
Factor I cofactor activity assay. In vitro cofactor activity was assayed as described previously 13
(8). SV5 or MuV particles generated either in A549, CHO or CHO-CD46 cells were purified by 14
sucrose gradient centrifugation and the protein concentrations determined by BCA. Seven ug of viral 15
particles was incubated with 3 ug of either C3b or C4b along with 100 ng of Factor I in a total volume 16
of 20 ul. Incubation was in PBS (pH 7.4) at 37OC for times as indicated in the Figure legends. 17
Reactions were stopped by adding 5 ul of SDS-PAGE sample buffer containing mercaptoethanol and 18
boiling. C3b reaction products were analyzed on 9 % SDS-PAGE gels, while C4b reaction products 19
were analyzed on 10% gels. The gels were stained with Gelcode Blue Stain reagent (Thermo 20
Scientific, IL) to visualize proteins. In some cases, reaction products were analyzed by Western 21
blotting with polyclonal goat anti-human C3 at a dilution of 1:1000. Factor H which is known to have 22
cofactor activity for cleavage of C3b into iC3b served as a control for the C3b assay. Soluble CR1 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
8
which is known to act as a cofactor for promoting cleavage of C4b into C4c and C4d was used for the 1
C4b assay. 2
Virus neutralization assay. Time- or concentration-dependent neutralization assays were 3
carried out as described previously (26). 100 PFU of SV5 or MuV grown in either CHO or CHO-CD46 4
cells were treated at 37OC with varying concentrations of NHS (1 hr) or dilutions of NHS for varying 5
times. After incubation, viral titers were determined by plaque assays as described previously (26). 6
Reported results were the average of six reactions, with the significance of data points calculated using 7
the student’s t-test. 8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
9
RESULTS 1
iC3b cleavage products and cellular CD46 are associated with purified SV5 virions. C3b is 2
composed of an alpha’ chain and a beta chain. In the presence of Factor I and a cofactor such as CD46, 3
the alpha’ chain is cleaved into 67 and 43 kDa fragments but the C3b beta chain remains intact (1). To 4
determine the form of C3 that was associated with serum-treated SV5, purified virions produced in 5
A549 cells were left untreated or were treated for 1 h with NHS as a source of complement. Samples 6
were centrifuged on 15-60% sucrose gradients, and fractions were collected and analyzed for the 7
position of virions by western blotting with antiserum specific for the SV5 P protein. As shown in Fig. 8
1A, untreated SV5 virions sedimented with a peak in fractions 10-12, while NHS-treated virions 9
sedimented further down the gradient to fractions 7-9 as described previously (26). The more rapid 10
sedimentation of NHS-treated SV5 particles was dependent on the presence of C3 (data not shown). 11
When fractions from the NHS-treated SV5 sample were probed with antibodies to C3, two main C3 12
fragments the size of iC3b proteins were detected that cosedimented to the same peak position as SV5 13
virions (bottom panel, Fig. 1A). When compared to purified C3b and iC3b marker standards (Fig. 1B), 14
the C3 proteins that cosedimented with SV5 particles did not show an intact C3b alpha’ chain, but 15
instead showed a polypeptide profile matching that of the beta chain of C3b and the 43 kDa fragment 16
which is produced when C3b is cleaved to iC3b (1). These data demonstrate that serum-treated SV5 17
particles are associated with iC3b, the inactivated form of C3b. 18
The above finding of iC3b cleavage products cosedimenting with NHS-treated virion particles 19
raised the hypothesis that an RCA protein was associated with SV5, with CD46 being the most likely 20
candidate. In the presence of NHS, CD46 would promote Factor I-mediated cleavage to produce iC3b. 21
In support of this hypothesis, Fig 1C shows that the same gradient fractions from NHS-treated virus 22
that contained cleaved C3b also contained cellular CD46. Using immunogold electron microscopy, 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
10
purified SV5 particles showed strong labeling for CD46 (Fig. 1D). Together, these data indicate that 1
the C3 protein associated with serum-treated SV5 virions is the inactivated iC3b form, and that the 2
cellular cofactor CD46 is associated with purified SV5 particles. 3
4
Purified SV5 virions display cofactor activity that promotes Factor I-mediated cleavage of C3b 5
into iC3b. To directly test the hypothesis that SV5 virions had C3b cofactor activity, SV5 was grown 6
in A549 cells and purified in the absence of NHS. Factor I-mediated cleavage of C3b was reconstituted 7
in vitro using purified SV5 virions and purified commercially available components as described 8
previously (8). The appearance of iC3b cleavage products was monitored by SDS-PAGE and 9
coomassie blue staining. As shown in the positive control samples in Fig. 2A, the C3b alpha’ chain 10
was cleaved into 67 kDa and 43 kDa fragments when incubated with purified Factor I and Factor H 11
(compare lanes 1-3). When SV5 virions were tested for cofactor activity by themselves, C3b cleavage 12
was not detected (lane 4). However, incubation of SV5 particles with Factor I and C3b resulted in a 13
time-dependent disappearance of the C3b alpha’ chain and corresponding appearance of both the 43 14
and 67 kDa fragments (compare lane 4 with lanes 9 and 10). 15
The above results were supported by western blot analysis of reaction products (Fig. 2B), 16
which demonstrated the loss of the alpha’ fragment and appearance of the 43 and 67 kDa proteins in 17
the case of the positive control sample which included C3b, Factor I and Factor H (lane 3). C3b alpha’ 18
cleavage was also seen in the case of SV5 virions incubated with C3b and Factor I (lane 5). In the case 19
of the SV5 samples, C3b cleavage produced both a 43 and 46 kDa fragment as reported elsewhere (1). 20
These data provide direct evidence that purified SV5 particles contain a functional cofactor activity 21
that promotes C3b cleavage and are consistent with the association of CD46 with SV5 virions (Fig. 1C 22
and D). 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
11
1
Increasing the level of virion-associated CD46 leads to increased C3b cofactor activity. To further 2
test the hypothesis that virion-associated CD46 contributes to C3b cofactor activity, SV5 was grown 3
and purified from CHO cells which are deficient in CD46 or in a CHO cell line that was engineered to 4
overexpress the CYT2 isoform of human CD46 (CHO-CD46; 21). Virions from these two cell lines 5
had similar gradient sedimentation profiles (data not shown). As shown in Fig. 3A, CHO cells express 6
very low to undetectable levels of CD46, and no detectable CD46 was found in purified virions 7
derived from CHO cells. By contrast, CHO-CD46 cells express high levels of CD46 (Fig. 3A) and 8
SV5 virus purified from CHO-CD46 cells had abundant CD46. Using immunogold electron 9
microscopy, purified virions from CHO-CD46 cells showed high labeling for CD46 around the outside 10
edges of particles (Fig. 3B), whereas virus from control CHO cells gave only background labeling. 11
CD46 in virions derived from CHO, CHO-CD46 and A549 cells was compared by western 12
blotting. As shown in Fig. 3C, abundant CD46 was detected in virions from CHO-CD46 cells but not 13
from CHO cells. In the case of virions from A549 cells, the level of CD46 was greatly reduced 14
compared to virions from CHO-CD46 cells. In addition, there was a clear difference in the isoform 15
which was associated with SV5 particles from A549 cells compared to CHO-CD46 cells, consistent 16
with the CHO-CD46 cell line stably expressing the CYT2 form from a plasmid (Gerlier et al., 1994). 17
To determine if virus grown in CHO-CD46 cells was associated with inactivated C3b, purified 18
virus was incubated with NHS as a source of complement for 1 h and then sedimented on sucrose 19
gradients as described for Fig. 1 above. Peak fractions containing virus were pooled and analyzed for 20
C3b fragments. By comparison with marker lanes in Fig. 3D, the C3 protein fragments that co-21
sedimented with SV5 virions (fractions 7-9) were found to be the iC3b fragments (43 and 67 kDa 22
proteins) and not the intact C3b proteins. 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
12
The above results predict that SV5 derived from the CHO-CD46 cells would have increased 1
C3b cofactor activity compared to virus from the CD46-deficient CHO cell line. To test this, Factor I-2
mediated cleavage of C3b was reconstituted in vitro using purified commercially available components 3
and reactions were monitored by SDS-PAGE and coomassie blue staining. As shown in Fig. 4A, 4
purified SV5 particles derived from CHO-CD46 cells by themselves did not promote C3b cleavage 5
(lane 4). Incubation of SV5 particles with Factor I and C3b resulted in a rapid time-dependent 6
disappearance of the C3b alpha’ polypeptide and corresponding appearance of both the 43 and 67 kDa 7
fragments (lanes 5-10). The kinetics and extent of cofactor activity was much greater than that seen 8
with SV5 derived from A549 cells which express a lower level of CD46 than that seen with the CHO-9
CD46 cells (compare rate of cleavage between Fig. 2A and Fig 4A). These results were supported by 10
western blot analysis using antibody specific for C3 polypeptides (Fig. 4B). By contrast to the results 11
with CHO-CD46 cells, SV5 derived from the control CD46-deficient CHO cell line did not show 12
appreciable C3b cofactor activity (Fig. 4C). Together, these data indicate that growing SV5 in cell 13
lines that differ in their level of CD46 expression results in virus particles with different levels of 14
CD46 and these particles differ in their ability to promote cleavage of C3b into iC3b. 15
16
Changes in the level of SV5-associated CD46 alters the extent and kinetics of neutralization in 17
vitro. The above data on C3b cofactor activity leads to the prediction that SV5 containing CD46 would 18
be more resistant to neutralization by human serum than virus lacking CD46. To test this, one hundred 19
PFU of SV5 grown in either CHO-CD46 or in control CHO cells were incubated for 1 h with PBS or 20
with dilutions of NHS. Remaining infectivity was determined by plaque assay. As shown in Fig. 5A, 21
virus derived from both cell lines was effectively neutralized at serum dilutions of 1:10 and 1:20, since 22
few if any plaques could be detected from these samples. However, at dilutions of 1:40 and 1:60 virus 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
13
grown in CHO-CD46 cells (black bars) was not neutralized as efficiently as SV5 from CHO cells 1
(hatched bars). The number of plaques was significantly higher at these two dilutions (p<0.0001) for 2
SV5 grown in CHO-CD46 cells compared to control CHO cells. 3
SV5 containing CD46 was also neutralized at a slower rate compared to virus deficient in 4
CD46. This is evident in the timecourse of neutralization shown in Fig. 5B, where the number of PFU 5
of CHO-CD46 grown virus did not decrease until 30 min of incubation with NHS. By comparison, 6
~40% of the virus from CHO control cells was neutralized by 5 min with NHS, and only ~40 and 7
~20% of infectivity remained by 15 and 30 min of incubation with NHS. Differences in neutralization 8
of the two virus preparations was statistically significant at all timepoints after addition of NHS 9
(p<0.0001). These data support the proposal that virion-associated CD46 functions to decrease the 10
efficiency of SV5 neutralization by NHS. 11
12
SV5 virions have low C4b cofactor activity. CD46 can function as a cofactor for Factor I-mediated 13
cleavage of the C4b alpha’ chain into a 25 kDa fragment and C4d. To determine if purified SV5 14
particles have C4b cofactor activity, C4b cleavage assays were reconstituted in vitro from purified 15
commercially available components and assayed by SDS-PAGE and coomassie blue staining. As 16
shown in the positive control lane 3 of Fig. 6, the C4b alpha’ chain was efficiently cleaved into C4d 17
and the 25 kDa fragments when incubated with purified Factor I and soluble CR1. Addition of Factor I 18
and C4b to purified SV5 derived from CHO-CD46 cells resulted in a very low level of C4b cleavage, 19
which is most clearly evident in Fig. 6 by the appearance of low levels of C4d (arrow, lane 8). SV5 20
from CHO control cells did not have detectable C4b cofactor activity (Fig. 6B). These data indicated 21
that SV5 particles have only very low levels of C4b cofactor activity, even when associated with large 22
amounts of CD46 due to growth in CHO-CD46 cells. 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
14
1
MuV particles containing CD46 inactivate C3b and C4b and are resistant to complement 2
mediated neutralization. To determine if the above results with SV5 extend to other related 3
paramyxoviruses, MuV was grown in control CHO cells or CHO-CD46 cells. As shown in the Western 4
blot in Fig. 7A, purified MuV was associated with CD46 when grown in CHO-CD46 cells but not in 5
control CHO cells. Similarly, immunogold electron microscopy of MuV particles showed CD46 6
labeling of virions from CHO-CD46 cells, but not from CHO cells (Fig. 7B). As show in Fig. 7B, a 7
large fraction of MuV particles displayed a concentration of immunogold labeling on one side, but the 8
significance of this is not known (see discussion). 9
To determine if MuV particles had C3b or C4b cofactor activity, Factor I-mediated cleavage 10
reactions were reconstituted in vitro using purified commercially available components and reactions 11
were monitored by SDS-PAGE and coomassie blue staining. As shown in Fig. 7C, purified MuV from 12
CHO-CD46 cells had efficient C3b cofactor activity (lane 5), which is most evident by the 13
disappearance of the C3b-alpha’ chain and corresponding appearance of the 43 kDa fragment. MuV 14
from control CHO cells lacked cofactor activity (data not shown). In contrast to SV5 particles, MuV 15
particles containing CD46 had high C4b cofactor activity. This is most evident in the timecourse shown 16
in Fig. 7D, where there is a substantial loss of the C4b-alpha’ chain and appearance of the 25 kDa 17
fragment as early as 2 h after incubation. This contrasts with the low C4b activity associated with SV5 18
virions (Fig. 6A), even after 24 h of incubation. 19
Similar to the results with SV5 (Fig. 5 above), MuV derived from CHO-CD46 cells was more 20
resistant to in vitro neutralization by human serum than virus derived from the control CHO cells. This 21
is evident in the results from a timecourse of in vitro neutralization by serum from two donors shown in 22
Fig. 8. 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
15
DISCUSSION 1
Complement is an important mediator of innate response to paramyxovirus infections. A 2
number of paramyxoviruses have been shown to be neutralized by complement through either the 3
classical or alternative pathways (18, 23, 34, 49, 50). However to our knowledge, no previous study 4
has addressed mechanisms by which paramyxoviruses can counteract or limit the effects of 5
complement. This study was initiated by our previous finding that C3 protein fragments are associated 6
with SV5 and MuV virions after exposure to NHS in vitro (26). However, the virion-associated C3 7
fragments were the inactive iC3b species which results from cleavage of C3b by Factor I in 8
combination with a cofactor such as CD46 or Factor H. This finding raised the hypothesis tested here 9
that SV5 and MuV virions contain cofactor activity that promotes inactivation of complement 10
pathways. As described below, our results support a model whereby the closely related 11
paramyxoviruses SV5 and MuV recruit the cellular RCA protein CD46 from the plasma membrane 12
during the budding process. Together with Factor I derived from serum, these CD46-enriched virions 13
mediated cleavage of C3b and C4b into their inactive forms. 14
Our results are consistent with the proposal that SV5 and MuV incorporate CD46 into progeny 15
virions, as evidenced by co-sedimentation on sucrose gradients, immunogold electron microscopy and 16
differential levels of CD46 in virions grown in cells that differ in CD46 expression. Preliminary data 17
(not shown) indicate that our findings of SV5- and MuV-associated CD46 also extend to a third 18
rubulavirus Human Parainfluenza Virus type 2 (HPIV2). The signals that direct some membrane 19
proteins into budding paramyxovirus particles or exclude other cell surface proteins are not completely 20
understood. For CD46, different cell types express distinct isoforms through differential splicing (39, 21
41), and this results in CD46 molecules that differ in two protein domains: a serine- and threonine- rich 22
ectodomain segment which is likely to be heavily glycosylated and length of the cytoplasmic C-23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
16
terminal tail (32). In our studies, we have used either A549 cells which express one predominant 1
isoform (see Fig. 1A and 3C) or a stable CHO cell line engineered to overexpress the CYT2 isoform 2
from a transfected plasmid. Since the length and sequence of the CD46 cytoplasmic domain could 3
influence incorporation into budding SV5 or MuV particles, it is not known whether budding of SV5 4
or MuV particles preferentially incorporates a particular CD46 isoform. It is notable that in our 5
electron microscopy studies that most SV5 and MuV particles derived from CHO-CD46 cells 6
contained an enrichment of CD46 on one face of spherical particles (see Figs. 3B and 7B). In the case 7
of filamentous particles, CD46 staining was predominantly on one tip of the particle (not shown). 8
Further studies are required to determine if this reflects a selective enrichment of CD46 at sites that 9
initiate budding. 10
In our in vitro assays using purified complement components, both SV5 and MuV particles had 11
cofactor activity needed for cleavage of C3b to the inactive iC3b. Cofactor activity was also seen 12
against C4b, but it was less efficient than that seen for C3b cleavage as described previously (35, 45). 13
Interestingly, C4b cofactor activity associated with MuV particles was much more efficient than that 14
seen with SV5 particles. One explanation for this difference could be that CD46 in MuV is more 15
concentrated or in a different conformation which allows a higher preference for inactivation of both 16
C3b and C4b. Alternatively, MuV may have mechanisms to recruit additional complement inhibitors, 17
and this could promote more effective cofactor activity against C4b. 18
It is not clear whether SV5 and MuV recruit other cellular RCA proteins or inhibitors of 19
complement in addition to CD46. A recent study by Shaw et al (46) showed that influenza virus can 20
incorporate the inhibitory protein CD59 which acts downstream at the assembly of the MAC, but 21
CD46 was not found in purified virions. This raises the interesting possibility that enveloped RNA 22
viruses may have selective incorporation of host inhibitors of complement depending on the particular 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
17
pathway activated by an individual virus type or sites of budding at the plasma membrane. Consistent 1
with this hypothesis, influenza virus can activate the lectin or classical pathways (22, 25) and does not 2
incorporate CD46 (46), which acts preferentially on the alternative pathway (35, 45). Conversely, SV5 3
activates the alternative pathway, incorporates at least CD46, and has low activity against the classical 4
and lectin pathway factor C4b. It has recently been shown that the flavivirus West Nile Virus evades 5
complement activity by recruiting the soluble Factor H from serum (13). This finding raises the 6
possibility that SV5 and MuV may also recruit soluble serum-associated RCA proteins in addition to 7
membrane-bound factors such as CD46. 8
It is important to note that the effect of SV5- and MuV-associated cofactor activity on 9
neutralization was not absolute, but instead resulted in a delay in the kinetics of neutralization and a 10
decrease in the efficiency of inactivation (Fig. 5). We interpret this to show that the potency of 11
complement activities eventually overcome any viral inhibitory mechanisms, and the virus-associated 12
RCA proteins only act to delay but not completely block neutralization. A similar proposal has been 13
made for HIV, and it is thought that delaying neutralization confers an advantage to virus growth (4). 14
All of our data have been obtained using human serum, a conventional approach that takes advantage 15
of the ease of obtaining serum and the powerful reagents such as serum depleted of specific factors. 16
Parainfluenza viruses such as SV5 are typically restricted to the respiratory tract (30) and are generally 17
shielded from the high concentrations of complement in serum. It is known that complement levels and 18
pathways in the respiratory tract differ significantly from serum (10). Thus, it is possible that virus-19
associated RCA proteins will be more potent factors in the inhibition of virus neutralization when 20
examined in the context of immunity specific for the respiratory tract. 21
CD46 is expressed on the surface of nearly all human cells, but the level of expression can vary 22
between different cell types or tissues (32). SV5 does not induce a global shutdown of cellular 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
18
transcription or translation (38), raising the possibility that the level of CD46 incorporation into 1
budding SV5 particles is determined by the constitutive level in a particular cell type. CD46 and other 2
cell RCA proteins can function to inactivate complement at the cell surface. Since SV5 establishes a 3
largely noncytopathic infection of human epithelial cells (12), constitutive expression of RCA proteins 4
could contribute to establishing noncytopathic persistent infections. 5
In summary, our results demonstrate that the paramyxoviruses SV5 and MuV activate the 6
alternative complement pathway (26), but also incorporate the cellular RCA protein CD46 to decrease 7
the efficiency of virus neutralization. This is consistent with the general principle that viruses cannot 8
avoid activation of complement, but instead limit the effects of this activation by employing 9
mechanisms that target a downstream inhibitory step in the complement cascade (9, 16). 10
11
12
13
14
15
ACKNOWLEDGEMENTS 16
We thank members of the Parks lab and Dr Doug Lyles for comments on the manuscript. We are 17
grateful to Ellen Young for excellent technical help, Dr. Denis Gerlier (University of Lyon) for the 18
CHO cells lines and Dr. Henry Marsh (Celldex Therapeutics) for the kind gift of purified sCR1. This 19
work was supported by NIH grant AI081022. 20
21
22
23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
19
REFERENCES 1
2
1. Adams, E. M., M. C. Brown, M. Nunge, M. Krych and J. P. Atkinson. 1991. Contribution of the 3
repeating domains of membrane cofactor protein (CD46) of the complement system to ligand 4
binding and cofactor activity. J. Immunol. 147:3005-3011. 5
2. Ahmad, M., K. Pyaram, J. Mullick and A. Sahu. 2007. Viral complement regulators: the expert 6
mimicking swindlers. Indian J. Biochem. Biophys. 44:331-343. 7
3. Atkinson, J. P., M. K. Liszewski, A. Richards, D. Kavanagh, and E. A. Moulton. 2005. 8
Hemolytic uremic syndrome: an example of insufficient complement regulation on self tissue. 9
Ann NY Acad Sci. 1056:144-152 10
4. Banki, Z., H. Stoiber, M. P. Dierich. 2005. HIV and human complement: inefficient virolysis and 11
effective adherence. Immunology letters 97:209-214. 12
5. Barilla-LaBarca, M. L., M.K. Liszewski, J. D. Lambris, D. Hourcade and J. P. Atkinson. 2002. 13
Role of membrane cofactor protein (CD46) in regulation of C4b and C3b deposited on cells. J. 14
Immunol. 168:6298-6304. 15
6. Bergmann-Leitner, E. S., W. W. Leitner, and G. C. Tsokos. 2006. Complement 3d: from 16
molecular adjuvant to target of immune escape mechanisms. Clin Immunol. 121:177-185. 17
7. Bernet, J., J. Mullick, A. K. Singh and A. Sahu. 2003. Viral mimicry of the complement system. 18
J. Biosci. 28(3):249-64. 19
8. Bernet, J., J. Mullick, Y. Panse, P. B. Parab, and A. Sahu. 2004. Kinetic analysis of the 20
interactions between vaccinia virus complement control protein and human complement 21
proteins C3b and C4b. J. Virol. 78:9446-9457 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
20
9. Blue, C.E., O.B. Spiller and D.J. Blackbourn. 2004. The relevance of complement to virus 1
biology. Virology 319:176-184. 2
10. Bolger, M. S., D. S. Ross, H. Jiang, M. M. Frank, A. J. Ghio, D. A. Schwartz, and J. R. 3
Wright. 2007. Complement levels and activity in the normal and LPS-injured lung. Am J. 4
Physiol. Lung Cell Mol Physiol. 292:L748-L759. 5
11. Carroll M. C., 2004. The complement system in regulation of adaptive immunity. Nat. Immunol. 6
5:981-6. 7
12. Choppin, P. W. 1964. Multiplication of a myxovirus (SV5) with minimal cytopathic effects and 8
without interference. Virology 23:224-233 9
13. Chung, K. M, M. K. Liszewski, G. Nybakken, A. E. Davis, R. R. Townsend, D. H. Fremont, J. 10
P. Atkinson, and M. S. Diamond. 2006. West Nile virus nonstructural protein NS1 inhibits 11
complement activation by binding the regulatory factor H. Proc. Natl. Acad. Sci. 103:19111. 12
14. Cooper, N. R. 1975. Isolation and analysis of the mechanism of action of an inactivator of C4b in 13
normal human serum. J. Exp. Med. 141:890-903. 14
15. Cooper, N. R. 1998 Complement and viruses; in The human complement system in health and 15
disease (eds) J.E. Volanakis and M.M. Frank (New York:Marcel Dekker) pp 393-407. 16
16. Cummings, K. L., S. N. Waggoner, R. Tacke, and Y. S. Hahn. 2007. Role of complement in 17
immune regulation and its exploitation by virus. Viral Immunology. 20:505 18
17. Delgado, M. F. and F. P. Polack. 2004. Involvement of antibody, complement and cellular 19
immunity in the pathogenesis of enhanced respiratory syncytial virus disease. Expert Rev. 20
Vaccines. 3:693-700. 21
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
21
18. Devaux, P., D. Christiansen, S. Plumet and D. Gerlier. 2004. Cell surface activation of the 1
alternative complement pathway by the fusion protein of measles virus. J. Gen. Virol. 85:1665-2
1673. 3
19. Fearon, D. T. 1979. Regulation of the amplification C3 convertase of human complement by an 4
inhibitory protein isolated from human erythrocyte membrane. Proc. Natl. Acad. USA. 5
76:5867-5871. 6
20. Fodor, W. L., S. A. Rollins, S. Bianco-Carson, R. P. Rother, E. R. Guilmette, W. V. Burton, 7
J.-C. Albrecht, B. Fleckenstein, and S. P. Squinto. 1995. The complement control protein 8
homolog of herpesvirus saimiri regulates serum complement by inhibiting C3 convertase 9
activity. J. Virol. 69:3889-3892. 10
21. Gerlier, D., B. Loveland, G. Varior-Krishnan, B. Thorley, I. F. C. McKenzie and C. 11
Rabourdin-Combe. 1994. Measles virus receptor properties are shared by several CD46 12
isoforms differing in extracellular regions and cytoplasmic tails. J. Gen. Virol., 75:2163-71. 13
22. Hartshorn, K. L., Sastry, K., White, M. R., Anders, E.M., Super, M., Ezekowitz, R. A. and A. 14
I. Tauber. 1993. Human mannose-binding protein functions as an opsonin for influenza A 15
viruses. J Clin Invest. 91:1414-20. 16
23. Hirsch, R. L., J. S. Wolinsky and J. A. Winkelstein. 1986. Activation of the alternative 17
complement pathway by mumps infected cells: Relationship to viral neuraminidase activity. 18
Arch. of Virol. 87:181-190. 19
24. Iida, K. and V. Nussenzweig. 1981. Complement receptor is an inhibitor of the complement 20
cascade. J. Exp. Med. 153:1138-1150. 21
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
22
25. Jayasekara, J. P., Moseman E.A., and M. C. Carroll. 2007. Natural antibody and complement 1
mediate neutralization of influenza virus in the absence of prior immunity. J. Virol. 81:3487-2
3494. 3
26. Johnson, J. B., G. A. Capraro and G. D. Parks. 2008. Differential mechanisms of complement-4
mediated neutralization of the closely related paramyxoviruses simian virus 5 and mumps 5
virus. Virology 376:112-23. 6
27. Kemper, C. and J. P. Atkinson. 2007. T-cell regulation: with complements from innate 7
immunity. Nat. Rev. Immunol. 7:9-18. 8
28. Kerr, M. A. 1980. The human complement system: assembly of the classical pathway C3 9
convertase. Biochem. J. 189:173-181. 10
29. Kotwal, G. J. and B. Moss. 1988. Vaccinia virus encodes a secretory polypeptide structurally 11
related to complement control proteins. Nature. 335:176-8. 12
30. Lamb, R. A., and G. D. Parks. 2007. “Paramyxoviridae: The viruses and their replication” in 13
Fields Virology. Fifth Edition, Ed. by David M. Knipe and Peter M. Howley; Published by 14
Wolters Kluwer and Lippincott Williams and Wilkins, 1449-1496. 15
31. Lambris, J. D., D. Ricklin and B. V. Geisbrecht. 2008. Complement evasion of human 16
pathogens. Nat. Rev. Microbiol. 6:132-142. 17
32. Liszewski, M. K., T. W. Post and J. P. Atkinson. 1991. Membrane cofactor protein (MCP or 18
CD46): Newest member of the regulators of complement activation gene cluster. Annu. Rev. 19
Immunol. 9:431-55. 20
33. McNearney, T., L. Ballard, T. Seya and J. P. Atkinson. 1989. Membrane cofactor protein of 21
complement is present on human fibroblast, epithelial and endothelial cells. J. Clin. Invest. 22
84:538-545. 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
23
34. McSharry, J. J., R. J. Pickering and L. A. Caliguiri. 1981. Activation of the alternative 1
complement pathway by enveloped viruses containing limited amounts of sialic acid. Virology 2
114:507-515. 3
35. Masaki, T., M. Matsumoto, I. Nakanishi, R. Yasuda and T. Seya. 1992. Factor I-dependent 4
inactivation of human complement C4b of the classical pathway by C3b/C4b receptor (CR1, 5
CD35) and membrane cofactor protein (MCP, CD46). J. Biochem. 111:573-578. 6
36. Morrison, T. E., R. J. Fraser, P. N. Smith, S. Mahalingam and M. T. Heise. 2007. Complement 7
contributes to inflammatory tissue destruction in a mouse model of Ross River virus-induced 8
disease. J. Virol. 81:5132-5143. 9
37. Paterson, R. G., G. P. Leser, M. A. Shaughnessy, and R. A. Lamb. 1995. Paramyxovirus SV5 V 10
protein binds two atoms of zinc and is a structural component of virions. Virology 208:121-11
131. 12
38. Peluso, R. W., R. A. Lamb, and P. W. Choppin. 1977. Polypeptide synthesis in simian virus 5-13
infected cells. Virology 23:177-187. 14
39. Post, T. W., M. K. Liszewski, E. M. Adams, I. Tedja, E. A. Miller. 1991 Atkinson. Membrane 15
cofactor protein of the complement system: alternative splicing of serine/threonine/proline-rich 16
exons and cytoplasmic tails produces multiple isoforms and correlate with protein phenotype. J. 17
Exp. Med. 174:93-102. 18
40. Roozendaal, R. and M. C. Carroll. 2006. Emerging patterns in complement-mediated pathogen 19
recognition. Cell. 125:29-32. 20
41. Russell, S. R., R. L. Sparrow, I. F. C McKenzie and D. F. J. Purcell. 1992. Tissue-specific and 21
allelic expression of the complement regulator CD46 is controlled by alternative splicing. Eur. 22
J. Immunol. 22:1513-1518. 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
24
42. Saifuddin, M., T. Hedayati, J. P. Atkinson, M. H. Holguin, C.J. Parker and G. T. Spear. 1997. 1
Human immunodeficiency virus type 1 incorporates both glycosyl phosphatidyinositor-2
anchored CD55 and CD59 and integral membrane CD46 at levels that protect from 3
complement mediated destruction. J. Gen. Virol. 78:1907-1911. 4
43. Schauber-Plewa, C., A. Simmons, M. J. Tuerk, C. D. Pacheco, and G. Veres. 2005. 5
Complement regulatory proteins are incorporated into lentiviral vectors and protect particles 6
against complement inactivation. Gene Therapy 12:238-245. 7
44. Seya, T., T. Hara, M. Matsumoto, Y. Sugita and H. Akedo. 1990. Complement-mediated tumor 8
cell damage induced by antibodies against membrane cofactor protein (MCP, CD46). J. Exp. 9
Med. 172:1673. 10
45. Seya, T., J. R. Turner and J. P. Atkinson. 1986. Purification and characterization of a membrane 11
protein (gp45-70) that is a cofactor for cleavage of C3b and C4b. 163:837-855. 12
46. Shaw, M. L., K. L. Stone, C. M. Colangelo, E. E. Gulcicek, and P. Palese. 2008. Cellular 13
proteins in influenza virus particles. PLoS Pathogens 4:e1000085. 14
47. Stoiber, H., C. Pinter, A. G. Siccardi, A. Clivio and M. P. Dierich. 1996. Efficient destruction 15
of human immunodeficiency virus in human serum by inhibiting the protective action of 16
complement factor H and decay accelerating factor (DAF, CD55). J. Exp. Med. 183:307-10. 17
48. Vanderplasschen, A., E. Mathew, M. Hollinshead, R. B. Sim and G.L. Smith. 1998. 18
Extracellular enveloped vaccinia virus is resistant to complement because of incorporation of 19
host complement control proteins into its envelope. Proc. Natl. Acad. Sci. USA.95:7544-7549. 20
49. Vasantha, S., K. L. Coelingh, B. R. Murphy, R. R. Dourmashkin, C. H. Hammer, M. M. 21
Frank and L. Fries. 1988. Interaction of a nonneutralizing IgM antibody and complement in 22
parainfluenza virus neutralization. Virology 167:433-41. 23
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
25
50. Welsh, R. M. Jr., 1977. Host cell modification of Lymphocytic choriomeningitis virus and 1
Newcastle disease virus altering viral inactivation by human complement. J. Immunol. 2
118:348-354. 3
4
5
6
FIGURE LEGENDS 7
Figure 1. CD46 and iC3b are associated with purified SV5 particles. A) Gradient sedimentation. 8
Purified SV5 was incubated alone or with NHS for 1 h and then analyzed by centrifugation through 9
15-60% sucrose gradients. Fractions were collected and analyzed for viral proteins by western blotting 10
with antiserum specific for the SV5 P protein (α-P; top and middle panels) or for C3-related proteins 11
(α-C3; bottom panel). B and C) Virus-associated iC3b and CD46. Fractions 7-9 from the SV5 + NHS 12
gradient shown in panel A were pooled and analyzed by western blotting for C3 proteins (panel B) or 13
CD46 (panel C). Markers are purified C3b and iC3b (panel B) or a lysate from SV5 infected A549 14
cells (panel C). The positions of alpha’ and beta chains of C3b, the 67 kDa and 43 kDa fragments of 15
iC3b and CD46 are indicated. D) EM analysis. Sucrose gradient purified virus was treated with anti-16
CD46 antibody followed by 12 nm colloidal gold goat anti-mouse antibody (bottom panels) or with 17
secondary antibody alone (top panels). Samples were analyzed by EM at a magnification of 55,000 X 18
(bar represents 0.02 um). 19
Figure 2. Purified SV5 virions have C3b cofactor activity. A) Timecourse of C3b-alpha’ cleavage. 20
Purified SV5 was incubated for 1 to 24 h with purified C3b and Factor I as indicated and as detailed in 21
Materials and Methods. The cleavage of the alpha’ chain of C3b into the 67 and 43 kDa products was 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
26
monitored by SDS-PAGE and coomassie blue staining. Lanes 1-3 are controls showing C3b-alpha’ 1
cleavage with Factor I and Factor H. Lane 4 is from a control reaction of C3b plus SV5 particles 2
incubated without Factor I. The positions of SV5 L and M proteins are denoted. B) Western blot 3
analysis of C3b cofactor reactions. Purified C3b (lane 1) was incubated with purified Factor I (lane 2) 4
or Factor I plus Factor H (lane 3). After 24 h incubation, samples were analyzed by western blotting 5
with antibody specific for C3b. Lanes 4 (SV5 plus C3b) and 5 (SV5 plus C3b and Factor I) are samples 6
analyzed in parallel. The position of alpha’ and beta chains of C3b, as well as the 67 and 43/46 kDa 7
cleavage products are indicated. 8
Figure 3. Increased incorporation of CD46 into SV5 particles derived from cells overexpressing 9
CD46. A) Western blotting. Lysates from CHO or CHO-CD46 cells and SV5 virus derived from these 10
two cell lines were analyzed by western blotting for CD46. Levels of actin and P protein were also 11
analyzed as load controls for cell and virus samples, respectively. B) EM analysis. Sucrose gradient 12
purified virus grown in control CHO or CHO-CD46 cells was treated with anti-CD46 antibody 13
followed by 12 nm colloidal gold goat anti-mouse antibody (right panels) or with secondary antibody 14
alone (left panels). Samples were analyzed by EM at a magnification of 55,000X (bar represents 0.02 15
um). C) CD46 levels in SV5 virions. Purified SV5 derived from CHO, CHO-CD46 or A549 cells was 16
analyzed by Western blotting for CD46 or viral P protein. D) iC3b fragments associated with virus 17
from CHO-CD46 cells. SV5 grown in CHO-CD46 cells was incubated with NHS and then purified by 18
centrifugation through 15-60% sucrose gradients. Peak fractions containing virus were analyzed by 19
western blotting for C3b cleavage products. Marker lanes are purified standards of C3b and iC3b. The 20
position of alpha’ and beta chains of C3b, as well as the 67 and 43 kDa cleavage products are 21
indicated. 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
27
Figure 4. SV5 virions derived from CHO-CD46 cells have enhanced C3b cofactor activity. A) 1
Purified SV5 derived from CHO-CD46 cells was incubated from 1-24 h with purified C3b and Factor I 2
as indicated in the lower box and detailed in Fig. 2B. The cleavage of C3b-alpha’ into the 67 and 43 3
kDa products was monitored by SDS-PAGE and coomassie blue staining. Lanes 1-4 are control 4
samples as described in the legend to Fig. 2. B and C) Western blot (panel B) or coosmassie staining 5
(pane C) of C3b cofactor reactions. Purified SV5 derived from CHO-CD46 cells (panel B) or control 6
CHO cells (panel C) were incubated with purified C3b (lane 4) or with purified C3 plus Factor I (lane 7
5). After 24 h incubation, samples were analyzed by western blotting with antibody specific for C3b. 8
Lanes 1-3 are controls of purified C3b alone (lane 1) or C3b plus purified Factor I (lane 2) or plus 9
Factor I and Factor H (lane 3). 10
Figure 5. Reduced efficiency and rate of in vitro neutralization of SV5 derived from CHO-CD46 11
cells. A) Effect of serum dilution. One hundred PFU of SV5 grown in CHO-CD46 cells (striped bars) 12
or control CHO cells (black bars) was incubated for 1 h with PBS (left side) or with the indicated 13
dilutions of NHS. Remaining infectious titers were determined by plaque assays. B) Timecourse of 14
neutralization. One hundred PFU of SV5 grown in CHO-CD46 cells (striped bars) or control CHO 15
cells (black bars) was incubated for the indicated times with a 1/40 dilution of NHS. Remaining 16
infectious titers were determined by plaque assays. For both panel A and B, results are the average of 17
six reactions, with error bars representing standard deviations. *; p value < 0.001; #, p value < 0.01 18
comparing corresponding values from CHO versus CHO-CD46. 19
Figure 6. SV5 virions have low C4b cofactor activity. Purified SV5 derived from CHO-CD46 cells 20
(panel A) or from control CHO cells (panel B) was incubated with the indicated combinations of C4b 21
and Factor I. As a positive control, purified C4b was incubated with Factor I and sCR1. The cleavage 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
28
of C4b alpha’ chain into C4d and 25 kDa fragments was monitored by SDS-PAGE and coomassie blue 1
staining. Lanes 5-8 of panel A represent incubations of 8, 12, 16, or 24 h, respectively. In panel A, 2
Lanes 1-4 are samples from a 24 h incubation, while lanes 5-8 were incubated for 8, 12, 16 and 24 h, 3
respectively. Samples shown in panel B are from 24 h incubation. The positions of C4b components 4
and the C4d cleavage products are indicated. 5
Figure 7. MuV particles containing CD46 inactivate complement. A) Western blotting. Purified 6
MuV derived from control CHO cells or CHO-CD46 cells were analyzed by western blotting for 7
CD46 or for viral P protein. B) EM analysis. Sucrose gradient purified MuV was treated with anti-8
CD46 antibody followed by 12 nm colloidal gold goat anti-mouse antibody (bottom panels). Samples 9
were analyzed by EM at a magnification of 55,000 X (bar represents 0.02 um). C) C3b cofactor 10
activity. Purified MuV grown in CHO-CD46 cells was assayed for C3b cofactor activity as described 11
in the legend to Fig. 2. The cleavage of the alpha’ chain of C3b into the 67 and 43 kDa products was 12
monitored by SDS-PAGE and coomassie blue staining. Lanes 1-3 are controls showing C3b-alpha’ 13
cleavage with Factor I and Factor H. D) C4b cofactor activity. Purified MuV derived from CHO-CD46 14
cells was assayed for C4b cofactor activity by incubating for the indicated times at 37oC as described 15
in the legend to Fig. 4. 16
Figure 8. Reduced efficiency and rate of in vitro neutralization of MuV containing CD46. One 17
hundred PFU of MuV grown in CHO-CD46 cells (gray bars) or control CHO cells (black bars) was 18
incubated for the indicated times with a 1/60 (donor #1) or 1:20 (donor #2) dilution of NHS. 19
Remaining infectivity was determined by plaque assays. Results are the average of six reactions, with 20
error bars representing standard deviations. *; p value < 0.001; #, p value < 0.01 comparing MuV from 21
CHO versus CHO-CD46 cells. 22
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
A.
SV5 + NHS
α-C3
141210864 topbottom
SV5 + NHS
α-P
SV5 alone
α-P
P
P
C3b-β
iC3b (43 kDa)
B. C.
Cell Mark
er
Fracti
ons 7-9
CD46
D.
iC3b 43kDa
C3b
C3b-βiC3b 67kDa
C3b α’
iC3b
Fracti
ons 7-9
0.02 µµµµm
20 Only
0.02 µµµµm
α-CD46
α-CD46
0.02 µµµµm
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
L
C3b α’
C3b β67 kDa
43 kDa
M
1 2 4 6 12 24
SV5 + I + C3b
1 2 63 4 5 87 9 10
C3b α’
C3b
Factor I
Factor H
SV5
+ + + + + + + + + +
+ + + ++++
+ ++
+ + + +++
Factor H
A.
B.
C3b β
C3b α’
67 kDa
43 kDa
46 kDa
1 2 3 4 5
Time in hr
C3b
Factor I
Factor H
SV5
+ + + + +++
+
+
+
+
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
SV5 fromCHO-CD46
SV5 from
Control CHO
0.02 µµµµm 0.02 µµµµm
2o Only 1o plus 2o
B.
A.
Actin
CD46
CHOCHO-C
D46
P protein
CD46
CHOCHO-C
D46
Cells Virions C.
C3b
iC3b 43kDa
C3b-β
iC3b 67kDa
C3b α’
Virus
iC3b
0.02 µµµµm0.02 µµµµm
D.
P protein
CD46
CHOCHO-C
D46
A549
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
1 2 63 4 5 87 9 10
C3b
Factor I
Factor H
SV5
+ + + + + + + + + +
+ + + ++++
+ ++
+ + + +++
A.
Factor H
C3b α’
C3b β67 kDa
43 kDa
C3b α’
1 2 4 6 12 24
SV5 + I + C3b
M
Time in hr
Factor H
C3b β 67 kDa
43 kDa
M
C3b α’
1 2 3 4 5
B. C.
C3b β
C3b α’
67 kDa
43 kDa
46 kDa
1 2 3 4 5
SV5 from CHO-CD46SV5 from Control CHO
C3b
Factor I
Factor H
SV5
+ + + + +++
+
+
+
+
C3b
Factor I
Factor H
SV5
+ + + + +++
+
+
+
+
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
20
40
60
80
100
120
Virus+
PBS
1:10 1:20 1:40 1:60 1:80 1:160
Serum Dilution (Virus + NHS)
SV5 CHO-CD46
SV5 CHO
A. B.
20
40
60
80
100
120
5 15 30 45 60
Time in Min (Virus + NHS)
Nu
mb
er
of
pla
qu
es
Virus+
PBS
*
*
*
*#
SV5 CHO-CD46
SV5 CHO
Nu
mb
er
of
pla
qu
es
*
*
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
+ + + + ++ +
++
+
C4b
SV5
sCR1
Factor I ++ + + + + + + +
+ +++
++
+ + ++
C4b
SV5
sCR1
Factor I + +
C4d
sCR1
α’
25 kDa
β
γ
M
1 2 3 4 5
A. B.
1 2 63 4 5 87
C4d
sCR1
α’
25 kDa
β
γ
M
SV5 from CHO-CD46 SV5 from Control CHO
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
C3b
Factor I
Factor H
MuV CD46
+ + + + + + + + + +
+ + + ++++
+ ++
+ + + +++
1 2 63 4 5 87 9 10
C4b
Factor I
sCR1
MuV CD46
+ + + + +
+ +
+ ++
+
1 2 63 4 5 87 9 10
P protein
CD46
CHOCHO-C
D46
MuV Virions
A. MuV from Control CHO MuV from CHO-CD46 MuV from CHO-CD46B.
C. D.
0.02 µµµµm 0.02 µµµµm 0.02 µµµµm
Factor H
C3b α’
C3b β 67 kDa
43 kDa
C3b α’
1 2 4 6 12 24
MuV CD46 + I + C3b
M
t in hr0 1 2 4 6 12 24
MuV CD46 + I + C4b
t in hr
C4d
sCR1
α’
27 kDa
β
γ
0
+
+
++
+
++
+
++
+
++
+
+
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from
20
40
60
80
100
120
5 15 30 45 60
Time in Min (Virus + NHS)
Nu
mb
er
of
Pla
qu
es
Virus+
PBS
*
MuV CHO-CD46
MuV CHO
*#
20
40
60
80
100
120
Nu
mb
er
of
Pla
qu
es
5 15 30 45 60
Time in Min (Virus + NHS)
Virus+PBS
*
*
#
MuV CHO-CD46
MuV CHO
Donor #1 Donor #2
on May 8, 2018 by guest
http://jvi.asm.org/
Dow
nloaded from