Content uploaded by Barbara Sherry
Author content
All content in this area was uploaded by Barbara Sherry
Content may be subject to copyright.
CD147 facilitates HIV-1 infection by interacting with
virus-associated cyclophilin A
Tatiana Pushkarsky*, Gabriele Zybarth*
†
, Larisa Dubrovsky*, Vyacheslav Yurchenko*, Hao Tang*, Huiming Guo
‡
,
Bryan Toole
‡
, Barbara Sherry*, and Michael Bukrinsky*
§
*The Picower Institute for Medical Research, 350 Community Drive, Manhasset, NY 11030; and
‡
Tufts University Medical School, Boston, MA 02111
Edited by Anthony Cerami, The Kenneth S. Warren Laboratories, Tarrytown, NY, and approved March 23, 2001 (received for review December 8, 2000)
Cyclophilin A (CyPA) is specifically incorporated into the virions of
HIV-1 and has been shown to enhance significantly an early step of
cellular HIV-1 infection. Our preliminary studies implicated CD147
as a receptor for extracellular CyPA. Here, we demonstrate a role
for CyPA–CD147 interaction during the early steps of HIV-1 infec-
tion. Expression of human CD147 increased infection by HIV-1
under one-cycle conditions. However, susceptibility to infection by
viruses lacking CyPA (simian immunodeficiency virus or HIV-1
produced in the presence of cyclosporin A) was unaffected by
CD147. Virus-associated CyPA coimmunoprecipitated with CD147
from infected cells. Antibody to CD147 inhibited HIV-1 entry as
evidenced by the delay in translocation of the HIV-1 core proteins
from the membrane and inhibition of viral reverse transcription.
Viruses whose replication did not require CyPA (SIV or mutant
HIV-1) were resistant to the inhibitory effect of anti-CD147 anti-
body. These results suggest that HIV-1 entry depends on an
interaction between virus-associated CyPA and CD147 on a target
cell.
C
yclophilin A (CyPA) is a ubiquitously distributed intracel-
lular protein possessing peptidyl-prolyl cis-trans isomerase
activity (1). This activity enables CyPA to assist protein folding
and function as a chaperone during various cellular processes (2).
CyPA also binds with high affinity to immunosuppressive drug
cyclosporin A (CsA), and this binding is required for the
immunosuppressive effect of CsA (3). In addition to its intra-
cellular functions, CyPA can be secreted into the extracellular
environment and has been shown to induce chemotaxis of
monocytes, eosinophils, and neutrophils (4, 5). Recent studies in
our laboratory (V.Y., G.Z., M. O’Connor, W. W. Dai, T. Hao,
H.G., B.T., B.S., and M.B., unpublished data) demonstrated that
the extracellular activities of CyPA are mediated by CD147, a
type I integral membrane glycoprotein of 50–60 kDa expressed
on a wide variety of cells including hemopoietic, microglial,
endothelial, and peripheral blood cells (6–10).
CyPA has been shown to be incorporated into HIV-1 particles
during virus morphogenesis through a specific interaction with
the CA domain of the Gag precursor polyprotein (11–14) and to
play an essential role in the early steps of HIV-1 life cycle (15,
16). Indeed, viruses made CyPA deficient by growing producing
cells in the presence of CsA or by introducing specific mutations
into CA become severely attenuated in the ability to establish
productive infection (17–20). Quantitative analysis of this defect
by using one-cycle conditions demonstrated that CyPA increased
HIV-1 infection ⬇6-fold (21). Interestingly, none of the related
primate immunodeficiency viruses incorporate CyPA, and even
within the HIV-1 family, viruses from clade O do not depend on
CyPA for infection (22).
The mechanism of CyPA activity during HIV-1 infection is not
yet understood. Several reports suggested that the defect in
replication of CyPA-deficient HIV-1 occurs after virus-cell
fusion, likely at the step of uncoating (15, 16). Although the
process of HIV-1 uncoating is poorly defined, it is known to
include dissociation of CA from the viral core (23, 24). Synthesis
of these findings (CyPA binding to CA, CA dissociation from the
core, and CyPA activity in protein folding) led to a hypothesis
that CyPA, by affecting CA conformation, destabilizes the core
shell structure and thus promotes CA disassembly, leading to
uncoating of the virus. Although this view gathered some
support from in vitro studies of CA multimerization (25), most
recent reports (26, 27) dispute this model. In those papers, direct
electron microscopic and biochemical analysis of HIV-1 cores
found no differences in morphology, stability, or structure of the
cores from CyPA-containing and CyPA-deficient viruses.
Based on the finding that HIV-1 infection can be blocked by
exogenous CyPA and by anti-CyPA antibodies, we suggested that
CyPA activity might be mediated by a cellular CyPA-binding
protein (16). We speculated that CyPA might be partially
accessible on the virus surface to interact with a receptor during
virus-cell fusion. This hypothesis found support in a study by
Saphire et al. (28), who demonstrated that a small portion of viral
CyPA is accessible for protease cleavage and thus should extend
outside of the viral membrane.
Recent identification of CD147 as a cell surface receptor for
extracellular CyPA (V.Y., G.Z., M. O’Connor, W. W. Dai, T.
Hao, H.G., B.T., B.S., and M.B., unpublished data) prompted us
to hypothesize that CD147 might mediate activity of virus-
associated CyPA during HIV-1 infection. In this work, we
demonstrate that interaction between HIV-1-associated CyPA
and CD147 on target cells significantly enhances infection by
HIV-1. CD147, therefore, appears to be a cofactor that mediates
activity of virus-associated CyPA and is required for efficient
infection by HIV-1.
Materials and Methods
Cells. Chinese hamster ovary (CHO)-K1 cell line was purchased
from American Type Culture Collection and was cultured in
F-12 medium (Life Technologies, Rockville, MD) ⫹ 10%
bovine fetal serum. Primary nonadherent peripheral blood
mononuclear cells (PBMCs) and adherent monocytes were
purified from whole blood of healthy donors by Ficoll-
Hypaque centrifugation and plastic adherence and were cul-
tured as described (29). PBMC were activated by treating with
phytohemagglutinin (5
g兾ml) and IL-2 (20 units兾ml) for 3
days. Monocyte-derived macrophages were obtained by allow-
ing plastic-adherent monocytes to differentiate for 7 days
in the presence of macrophage colony-stimulating factor
(M-CSF) (2 ng兾ml).
Transfection and Infection of CHO Cells. Cells grown to 80% con-
fluency in 75-cc flasks were transfected with 10
g of CD4,
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: CyPA, cyclophilin A; CsA, cyclosporin A; A-MLV, amphotropic murine leu-
kemia virus; PBMCs, peripheral blood mononuclear cells; MA, matrix; CA, capsid; RT, reverse
transcriptase; CHO, Chinese hamster ovary; SIV, simian immunodeficiency virus; LAV,
lymphadenopathy-associated virus.
†
Present address: Biodiscovery, Central Research Division, Pfizer Inc., Groton, CT 06340.
§
To whom reprint requests should be addressed. E-mail: mbukrinsky@picower.edu.
The publication costs of this article were defrayed in part by page charge payment. This
article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.
§1734 solely to indicate this fact.
6360– 6365
兩
PNAS
兩
May 22, 2001
兩
vol. 98
兩
no. 11 www.pnas.org兾cgi兾doi兾10.1073兾pnas.111583198
CCR5, or CXCR4 expression vectors (pBABE-T4, pBABE-
CCR5, and pBABE-CXCR4, respectively) separately or in com-
bination by using Fugene (Roche Molecular Biochemicals) and
following the manufacturer-provided protocol. Efficiency of
transfection was between 5 and 20%, as revealed by flow
cytometric analysis. For infection with Env
A-MLV
-pseudotyped
viruses, cells were plated at low density (0.01–0.1 ⫻ 10
6
cells兾well
of a 24-well plate) and infected with the virus [1 ⫻ 10
5
cpm of
reverse transcriptase (RT) activity兾well].
Immunoprecipitation of CD147. Cells were lysed in a detergent
buffer (20 mM triethanolamine, pH 8.0兾300 mM NaCl兾2mM
EDTA兾20% glycerol兾1% 3-[(3-cholamidopropyl)dimethyl-
ammonio]-1-propanesulfonate兾10
g/ml leupeptin兾10
g/ml
aprotinin兾1 mM PMSF) for 30 min at 4°C with continuous
rocking and then centrifuged (15,000 g, 15 min). Goat poly-
clonal anti-CD147 antibody (R & D Systems) was added to the
supernatant and incubated for1hat4°C, followed by an
overnight incubation at 4°C with Protein G-Sepharose. This
antibody did not interfere with CyPA–CD147 interaction,
allowing coimmunoprecipitation of these two proteins. Immu-
noprecipitates were washed three times with 50 mM Tris䡠HCl,
pH 7.5, separated by SDS兾PAGE, transferred to poly(vinyli-
dene difluoride) membranes, and analyzed by Western blot-
ting by using rabbit polyclonal anti-CyPA antibody (Affinity
Bioreagents, Golden, CO) or anti-EMMPRIN mAb (Leinco
Technologies, St. Louis, MO).
Subcellular Fractionation and Western Blot Analysis. Subcellular
fractions of MT-4 cells infected with HIV-1
LAV
(lymphadenop-
athy-associated virus) were prepared by using a previously
published protocol (30) with some modifications. Cells were
pelleted and incubated on ice in a hypotonic buffer (10 mM
Hepes, pH 6.9兾10 mM KCl兾0.1 mM PMSF兾1
g/ml aprotinin)
for 15 min. Cells were disrupted by Dounce homogenization and
nuclei were pelleted at 1,500 g for 5 min and discarded. Super-
natant was removed and centrifuged at 18,000 g for 45 min.
Supernatant from this centrifugation was reserved and is re-
ferred to as the cytosolic fraction. The pellet containing cy-
toskeleton and membranes was resuspended in NTENT buffer
(150 mM NaCl兾10 mM Tris䡠HCl, pH 7.2兾1 mM EDTA兾1%
Triton X-100兾0.1 mM PMSF兾1
g/ml aprotinin) supplemented
with 1% N-octyl-

-D-glycopyranoside and was centrifuged at
18,000 g for 30 min. The supernatant from this spin was reserved
and is referred to as the membrane fraction, whereas the pellet
represents the cytoskeleton fraction. Subcellular fractions from
the equivalent of 2 ⫻ 10
6
cells兾lane were fractionated on a 12%
SDS兾PAGE and analyzed by Western blot assay by using mAbs
to matrix (MA), capsid (CA) (both from Advanced Biotechnol-
ogies, Columbia, MD), and actin (Sigma).
Results
CD147 Enhances HIV-1 Infection. Our recent work (16) suggested
that the previously described activity of virus-incorporated
CyPA during an early step of HIV-1 infection might be
mediated by a CyPA-interacting protein. Because of the
demonstrated activity of CD147 as a CyPA receptor (V.Y.,
G.Z., M. O’Connor, W. W. Dai, T. Hao, H.G., B.T., B.S., and
M.B, unpublished data), we hypothesized that CD147 might
interact with virus-associated CyPA and stimulate virus infec-
tion. For quantitative analysis of CD147 activity in HIV-1
infection, we examined one-cycle infection of CHO cells stably
expressing human CD147 (CHO.CD147). The CHO cells were
selected because flowcytometric analysis revealed that all
human cells that we tested expressed high levels of CD147 (Fig.
1A). When plated at low density to dilute the anti-retroviral
inhibitory factor produced by CHO cells (31), these cells can
be infected with recombinant luciferase-expressing HIV-1
pseudotyped with an envelope of amphotropic murine leuke-
mia virus (A-MLV). Such analysis demonstrated a 4- to 5-fold
increase of luciferase expression in CD147-transfected cells
(Fig. 1B). A similar enhancement was observed when three
stable CHO.CD147 clones were compared to clones of CHO
cells transfected with empty vector pcDNA3.1 (Fig. 1C Bot-
tom). Enhancement of HIV-1 infection correlated with CD147
expression on cells (measured by flow cytometry, Fig. 1C Top).
Importantly, the observed increase in HIV-1 infection because
of the presence of CD147 was similar to the increase provided
by CyPA (ref. 21 and results not shown), suggesting that CD147
accounts for most of CyPA activity (see below). One of these
CHO.CD147 clones (CHO.CD147.17) was used in further
experiments.
To rule out the possibility that the enhancing effect of CD147
on viral infection had something to do with the inhibitory factor
produced by CHO cells (31), two types of experiments were
performed. First, we investigated whether the enhancing effect
of CD147 correlates with the density at which cells were plated.
If CD147 counteracts the activity of the CHO-produced negative
factor, than one can expect the CD147-specific enhancing effect
to diminish when cell density decreases. However, this phenom-
enon was not the case; regardless of the plating density, the
enhancing effect of CD147 was relatively constant (Fig. 1D). In
a second series of experiments, we transiently transfected
CHO.CD147 (and control CHO.pcDNA) cells with CD4- and
CXCR4-expressing vectors and infected them with luciferase-
expressing HIV-1 construct pseudotyped with LAV envelope.
Because the negative factor produced by CHO cells is specific for
the A-MLV envelope (32), it is not expected to influence
infection by such virus. Similar to results obtained with the
A-MLV envelope, a 3- to 4-fold increase in luciferase expression
was observed in cultures of cells expressing CD147 (Fig. 1E).
No increase in HIV-1 infection was observed when the virus
was produced in the presence of CsA (Fig. 1F) and thus was
deficient in CyPA (Fig. 1F Inset). As a result of CsA treatment,
infectivity of this virus in CHO.CD147 (but not in control
CHO.pcDNA) cell cultures was significantly attenuated, consis-
tent with a previously demonstrated requirement of CyPA for
efficient infection with Env
A-MLV
-pseudotyped HIV-1 (15, 33).
In contrast, infection with Env
A-MLV
-pseudotyped simian immu-
nodeficiency virus (SIV)-based virus, which does not depend on
CyPA (34), was unaffected by the presence or absence of CD147
on the target cells (Fig. 1F).
Based on results presented in Fig. 1, we conclude that CD147
enhances HIV-1 infection. This effect is independent of the viral
envelope but requires the presence of virus-associated CyPA.
Virus-Associated CyPA Interacts with CD147. To directly demon-
strate interaction between viral CyPA and cellular CD147, we
performed coimmunoprecipitation analysis. Primary PBMC cul-
tures were infected with equal amounts (by RT activity) of
HIV-1 (produced in the presence or absence of CsA)
pseudotyped with the A-MLV envelope (to increase the effi-
ciency of infection). After a 30-min incubation at 4°C, cells were
either left at 4°C (to allow virus attachment but not fusion) or
were transferred to 37°C for 1 h (to allow attachment followed
by virus-cell fusion) and lysed, and CD147 was immunoprecipi-
tated by using an anti-CD147 polyclonal antibody. Immunopre-
cipitated material was analyzed by Western blotting with anti-
bodies to CyPA or CD147 (immunoprecipitation control). As
shown in Fig. 2, no CyPA was detected in the immunoprecipi-
tates from mock-infected cells (lane 1) or cells incubated with the
virus at 4°C (lane 2), indicating that cellular CyPA does not
interact with CD147 under such experimental conditions and
that virus attachment is not sufficient for interaction between
viral CyPA and CD147. However, CyPA was coprecipitated with
CD147 from cell lysates after the 1-h incubation of virus and cells
Pusharsky et al. PNAS
兩
May 22, 2001
兩
vol. 98
兩
no. 11
兩
6361
MEDICAL SCIENCES
at 37°C (lane 3). Under similar conditions, no CyPA coprecipi-
tated with CD147 if infection was performed with HIV-1 grown
in the presence of CsA (lane 4), indicating that CyPA detected
in this assay is derived from infecting virions. We conclude that
virus-associated CyPA interacts with CD147. This interaction
appears to be temperature-dependent, suggesting that rear-
rangement of viral and兾or cell membrane proteins is required for
stable binding of CyPA to CD147.
Anti-CD147 mAb Inhibits Infection by CyPA-Dependent HIV-1 Variants.
To investigate the mechanism of CD147 activity in more detail
in the context of primary cells, we used an anti-CD147 antibody
to block interaction between CyPA and CD147. As shown in Fig.
3A, anti-CD147 antibody, but not isotype-matched control mAb,
inhibited infection by both CCR5- (ADA) and CXCR4-
dependent (LAV) HIV-1 strains in primary PBMC cultures. The
incomplete suppression of infection by this anti-CD147 mAb
Fig. 1. CD147 stimulates HIV-1 infection. (A) Analysis of CD147 expression on different cells. CD147 expression was analyzed by flow cytometry on
uncloned CHO cells transfected with CD147 (CHO.CD147) or with an empty vector (CHO.pcDNA), on MAGI-CCR5 and HeLa cell lines, and on primary PBMC
(nonactivated or activated by a 3-day treatment with phytohemagglutinin ⫹ IL-2), monocytes, and monocyte-derived macrophages. Results for primary
cells show mean ⫾ SE for two different donors. Results are representative of two independent experiments. (B) Duplicate cultures of CHO cells transiently
transfected with CD147 (CHO.CD147) or empty vector pcDNA3.1 (CHO.pcDNA) were infected with luciferase-expressing HIV-1 pseudotyped with A-MLV
envelope. Luciferase activity was measured on day 4 postinfection. Results show mean ⫾ SE and are representative of two experiments. (C) Three
CHO.CD147 (10, 17, and 18) and CHO.pcDNA clones were infected with luciferase-expressing HIV-1 pseudotyped with A-MLV envelope. CD147 expression
was measured by flow cytometric analysis and is expressed as increase in mean CD147 fluorescence over fluorescence observed with isotype control (Upper).
Luciferase activity was measured on day 4 postinfection (Lower). Representative of three experiments. (D) CHO.pcDNA and CHO.CD147 (clone 17) cells were
plated in duplicate at indicated density in a 24-well plate and infected with Env
A-MLV
-pseudotyped HIV-1 (1.3 ⫻ 10
5
cpm兾well). Results show mean ⫾ SE
and are representative of two experiments. (E) Triplicate cultures of CHO.pcDNA and CHO.CD147 cells were transiently transfected with CD4-expressing
pBABE-T4 and CXCR4-expressing pBABE-CXCR4 and infected with Luc-HIV-1 pseudotyped with Env derived from HIV-1
LAV
. Luciferase expression is
presented as percentage of expression relative to control (CHO兾CD4兾CXCR4 cells) taken as 100%. (F) Triplicate cultures of CHO.CD147 and CHO.pcDNA
cells were infected with luciferase-expressing HIV-1 or SIV constructs pseudotyped with A-MLV envelope. CyPA-deficient virus (HIV-1–CsA) was produced
in the presence of 1
MofCsA(⫹CsA). Virus used for infection was tested by Western blotting for the amount of CyPA and CA (Inset) by using rabbit
polyclonal anti-CyPA antibody or anti-CA mAb. Representative of two experiments.
6362
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.111583198 Pusharsky et al.
(⬇50%) might be because of its poor CD147-neutralizing activ-
ity. Indeed, a much more potent inhibitory effect (⬇80%
inhibition of infection, not shown) was observed with a different
anti-CD147 antibody, E11F4 (35). Unfortunately, this antibody
was unavailable in the amounts necessary to perform all of the
experiments described in this report. Therefore, we used a less
effective commercially available antibody. Nevertheless, this
antibody allowed us to obtain valuable information about the
mechanism of CD147 activity (see below).
To prove that the effect of the anti-CD147 mAb was due to
inhibition of interaction between viral CyPA and CD147,
we took advantage of SIV constructs, which differ in the ability
to incorporate CyPA. The SIV
mac
239 construct does not
incorporate CyPA and its replication is resistant to CsA,
whereas SIV
mac
239(HIV-CA), which carries a chimeric Gag
containing a small fragment of HIV-1 CA, incorporates CyPA
and is sensitive to CsA (34). As shown in Fig. 3B, replica-
tion in human PBMC of SIV
mac
239(HIV-CA) was inhibited
Fig. 2. Viral CyPA associates with CD147. Phytohemagglutinin-activated
PBMC cultures were left untreated (lane 1) or were inoculated with equal
amounts (adjusted by RT activity) of HIV-1 produced in the absence (lanes
2 and 3) or presence (lane 4) of CsA. Aftera1hincubation at 4°C, cells were
transferred to 37°C for 40 min (lanes 1, 3, and 4) or left at 4°C (lane 2). Cells
were lysed, and CD147 was immunoprecipitated with a goat polyclonal
anti-CD147 antibody. Immunoprecipitated proteins were analyzed by
Western blotting by using rabbit polyclonal anti-CyPA antibody or anti-
CD147 mAb.
Fig. 3. Anti-CD147 mAb inhibits HIV-1 replication. Triplicate cultures of
phytohemagglutinin-activated PBMCs were infected by inoculation with rep-
lication-competent HIV-1 strains LAV or ADA (A), or SIV carrying either wild-
type (SIV) or chimeric (SIV兾HIV-CA) CA (B). Fifty
g兾ml anti-CD147 mAb
(Ancell) or isotype-matched control mAb (PharMingen) was added 30 min
before infection and was present throughout the duration of the experiment.
Virus replication was assessed by reverse RT in culture supernatants. Results
are shown as mean ⫾ SE and are representative of four (A) and two (B)
experiments.
Fig. 4. CD147 regulates an early step of HIV-1 infection. (A) The effect of
anti-CD147 mAb on HIV-1 RT. Duplicate PBMC cultures were inoculated
with HIV-1
LAV
, incubated at 37°C for 1.5 h, and then trypsinized to remove
unfused virus. Analysis of HIV-1 RT was performed at indicated times after
inoculation by using primers long terminal repeat R兾U5 specific for the
early RT products (29). Anti-CD147 mAb (50
g兾ml), anti-CD4 mAb (2
g兾ml), or isotype-matched control mAb (50
g兾ml) was added to cells 2 h
before infection. Dilutions of 8E5兾LAV cells containing one HIV-1 genome
per cell (44) were used as PCR standards. The data are representative of
three experiments. (B) The effect of anti-CD147 mAb on subcellular distri-
bution of HIV-1 proteins. MT-4 cells were inoculated at 4°C with HIV-1
LAV
in the presence of 50
g兾ml of anti-CD147 mAb (Ancell) or isotype-matched
control mAb (PharMingen). After 30 min, an aliquot (1 ⫻ 10
6
) was with-
drawn for protein analysis (0 h postinfection time point), while the inoc-
ulated cultures were transferred to 37°C and incubated for 1.5 h or 3.5 h.
Subcellular fractionation was performed as described in Materials and
Methods, and proteins in cytosolic (C), membrane (M), and cytoskeleton
(CS) fractions were revealed by Western blot and enhanced chemilumines-
cence by using mAbs to actin, CA, and MA.
Pusharsky et al. PNAS
兩
May 22, 2001
兩
vol. 98
兩
no. 11
兩
6363
MEDICAL SCIENCES
by anti-CD147 mAb, whereas replication of SIV
mac
239
was not.
Based on presented evidence, we conclude that the antibody
to CD147 inhibited HIV-1 infection by an envelope-
independent, but CyPA-dependent mechanism.
CD147 Regulates HIV-1 Entry. Using a semiquantitative PCR assay,
we analyzed the amount of HIV-1-specific RT products in
anti-CD147 mAb-treated PBMC cultures at different times after
infection. Using long terminal repeat R兾U5 primer pair (29), we
amplified a short DNA fragment that is produced both by
endogenous intravirion RT (36) and intracellularly early after
the virus entry (37) and can be used as a measurement for the
efficiency of entry兾initiation of RT. Results presented in Fig. 4A
demonstrate a reduction in the amount of this RT product in
anti-CD147 mAb-treated cells compared to control (untreated)
or isotype-treated cultures. The difference between anti-CD147-
treated and -untreated cells was greatest at early time points (1.5
and 4 h) after inoculation and decreased thereafter, suggesting
that the antibody delayed virus entry. The anti-CD4 mAb, which
inhibits HIV-1 attachment, reduced the amount of long terminal
repeat R兾U5-amplified fragment at all time points with similar
efficiency (Fig. 4A).
As a different way to look at early steps of infection, we
analyzed the intracellular distribution of HIV-1 MA and CA
proteins early after de novo infection. Previous studies (30)
demonstrated that soon after infection, MA relocates from
membrane to the cytoskeleton, whereas CA migrates into the
cytosol. This observation is consistent with earlier demonstra-
tion that virus uncoating involves dissociation of CA from the
nucleoprotein complex (23, 24). Consistent with these results, we
observed a characteristic translocation of the MA protein from
the membrane into the cytoskeleton fraction 1.5 h after infection
(Fig. 4B, isotype panels, lane CS), whereas CA was found
predominantly in the cytosol at this time point (isotype panels,
lane C). In contrast, protein rearrangement was significantly
delayed in cells treated with the anti-CD147 antibody (Fig. 4B,
␣
CD147 panels). This result indicates that in the presence of
anti-CD147 mAb rearrangement of CA and MA, presumably
associated with formation of the RT complex (30), is delayed and
supports the idea that CD147 is involved in the regulation of
HIV-1 entry.
Discussion
Taken together, results presented in this report indicate that
CD147 interacts with virus-associated CyPA and regulates an
early step in HIV-1 replication. A recent study by Saphire et
al. (28) suggested a role for CyPA in virus attachment through
binding to sulfated proteoglycans (heparans). It seems plau-
sible that CyPA interaction with CD147 is downstream of
CyPA–heparan interaction. Indeed, initial interaction of
CyPA with heparans might facilitate its subsequent binding to
CD147, similar to the situation with chemokines whose binding
to glycosaminoglycans enhances subsequent interaction with
their receptors (38, 39). Similarly, binding of CyPB to heparans
presumably presents the immunophilin for interaction with its
functional receptor (type I site) (40). Involvement of a CyPA-
interacting receptor in an early step of HIV-1 infection is
consistent with our earlier finding (16) that exogenously added
CyPA inhibits HIV-1 infection, presumably by competition for
CD147 with virus-associated CyPA.
Several mechanisms may account for the activity of CD147
during the early phase of the HIV-1 life cycle. Intracellular
signaling events initiated by CyPA–CD147 interaction (V.Y.,
G.Z., M. O’Connor, W. W. Dai, T. Hao, H.G., B.T., B.S., and
M.B, unpublished data) might induce MA phosphorylation,
which was suggested to regulate detachment of the RT complex
from the membrane (41). Signaling from CD147 might also
induce cytoskeleton rearrangements, facilitating virus entry. It is
also conceivable that CyPA binding to CD147 might affect
conformation of another CyPA-binding partner, CA, leading to
destabilization of the capsid shell. This latter mechanism is
similar to the one proposed in earlier studies (15). Finally, CyPA
binding to CD147 might promote transition from the step of
hemifusion to complete fusion (42), allowing liberation of the
RT complex into the cytoplasm. All these mechanisms are not
mutually exclusive and may well function simultaneously or
sequentially.
Interaction between virus-associated CyPA and CD147 dur-
ing HIV-1 entry suggests that different binding sites on CyPA
are engaged by CA and CD147. This notion is supported
indirectly by our finding that CyPA-CD147 binding is resistant
to CsA (V.Y., G.Z., M. O’Connor, W. W. Dai, T. Hao, H.G.,
B.T., B.S., and M.B, unpublished data), whereas CyPA–CA
interaction is sensitive to this agent (11). Our recent studies
also demonstrated an important role of the CD147 transmem-
brane domain in CyPA-induced signaling and chemotaxis
(V.Y., G.Z., M. O’Connor, W. W. Dai, T. Hao, H.G., B.T.,
B.S., and M.B, unpublished data). It should be noted that the
mechanisms of interaction between free CyPA and CD147 do
not necessarily apply to the interaction between CD147 and
HIV-1-associated CyPA. Indeed, in contrast to free CyPA
studied there, CyPA in the virus particle is bound to the capsid
protein (11). The activities and interactions of free vs. bound
CyPA may be very different. As an example, a complex
between CyPA and CsA binds and inhibits calcineurin,
whereas neither CsA nor CyPA alone are capable of binding
calcineurin (3, 43). More studies will be needed to define the
binding interactions between CD147 and CyPA during HIV-1
infection.
Expression of the human CD147 on CHO cells does not ap-
pear to be absolutely necessary for HIV-1 infection (in con-
trast to CD4, CCR5 or CXCR4), as low level infection occurs
without it. This is illustrated by only a 6- to 8-fold difference
between CyPA-positive and CyPA-deficient HIV-1 in one-
cycle infection (21). During long-term culture, this difference
is exponentially amplified with each infection cycle. The role
of CyPA–CD147 interaction might be even more pronounced
in vivo, where low amounts of transmitted infectious virus and
a low probability of successful infection make the establish-
ment of new infections critically dependent on every small
enhancement the virus can achieve. It appears that HIV-1
evolved to enhance its infection process by using not only a
cellular receptor but also its ligand, thus subverting the target
cell’s ligand–receptor interaction pathway for the virus’s own
purpose.
The use of CyPA and CD147 for infection sets HIV-1 apart
from other primate lentiviruses that do not rely on these factors.
The ability to incorporate CyPA and use CD147 might have been
acquired at the early stages of HIV evolution, during adaptation
of the virus to the human host, and might contribute to high
replication capacity and pathogenicity of HIV-1. Further de-
tailed analysis of CD147 activity during HIV-1 infection is likely
to identify the exact step regulated by CD147 and might define
novel targets for anti-HIV interventions.
We thank D. Littman for the pBABE-T4 and pBABE-CXCR4 plasmids;
J. Sodroski for the luciferase-expressing SIV construct; H. Gottlinger for
p239SpSp5⬘(HIV-CA), pEnv ADA ⌬CT, and pEnv HxB2; and B. Willi
(Novartis Pharma AG) for NIM 811. pNL-Luc and pSV-Env
A-MLV
plasmids were a gift from N. Landau (The Salk Institute). We thank our
colleagues at The Picower Institute for helpful discussions, and K.
Manogue for critical reading of the manuscript. This work was supported
in part by National Institutes of Health Grants R01 AI 38245 (to M.B.)
and R01 AI 29110 (to B.S.), and by funds from The Picower Institute for
Medical Research.
6364
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.111583198 Pusharsky et al.
1. Galat, A. (1993) Eur. J. Biochem. 216, 689–707.
2. Gothel, S. F. & Marahiel, M. A. (1999) Cell Mol. Life Sci. 55, 423–436.
3. Liu, J., Albers, M. W., Wandless, T. J., Luan, S., Alberg, D. G., Belshaw, P. J.,
Cohen, P., MacKintosh, C., Klee, C. B. & Schreiber, S. L. (1992) Biochemistry
31, 3896–3901.
4. Sherry, B., Yarlett, N., Strupp, A. & Cerami, A. (1992) Proc. Natl. Acad. Sci.
USA 89, 3511–3515.
5. Xu, Q., Leiva, M. C., Fischkoff, S. A., Handschumacher, R. E. & Lyttle, C. R.
(1992) J. Biol. Chem. 267, 11968–11971.
6. Wimazal, F., Ghannadan, M., Muller, M. R., End, A., Willheim, M., Meid-
linger, P., Schernthaner, G. H., Jordan, J. H., Hagen, W., Agis, H., et al. (1999)
Tissue Antigens 54, 499–507.
7. Inoue, H., Sawada, M., Ryo, A., Tanahashi, H., Wakatsuki, T., Hada, A., Kondoh,
N., Nakagaki, K., Takahashi, K., Suzumura, A., et al. (1999) Glia 28, 265–271.
8. Kasinrerk, W., Tokrasinwit, N. & Phunpae, P. (1999) Immunology 96, 184–192.
9. Kirsch, A. H., Diaz, L. A. J., Bonish, B., Antony, P. A. & Fox, D. A. (1997)
Tissue Antigens 50, 147–152.
10. Mutin, M., Dignat-George, F. & Sampol, J. (1997) Tissue Antigens 50, 449–458.
11. Luban, J., Bossolt, K. L., Franke, E. K., Kalpana, G. V. & Goff, S. P. (1993)
Cell 73, 1067–1078.
12. Colgan, J., Yuan, H. E., Franke, E. K. & Luban, J. (1996) J. Virol. 70,
4299–4310.
13. Thali, M., Bukovsky, A., Kondo, E., Rosenwirth, B., Walsh, C. T., Sodroski, J.
& Gottlinger, H. G. (1994) Nature (London) 372, 363–365.
14. Franke, E. K., Yuan, H. E. & Luban, J. (1994) Nature (London) 372, 359–362.
15. Braaten, D., Franke, E. K. & Luban, J. (1996) J. Virol. 70, 3551–3560.
16. Sherry, B., Zybarth, G., Alfano, M., Dubrovsky, L., Mitchell, R., Rich, D.,
Ulrich, P., Bucala, R., Cerami, A. & Bukrinsky, M. (1998) Proc. Natl. Acad. Sci.
USA 95, 1758–1763.
17. Bukovsky, A. A., Weimann, A., Accola, M. A. & Gottlinger, H. G. (1997) Proc.
Natl. Acad. Sci. USA 94, 10943–10948.
18. Yin, L., Braaten, D. & Luban, J. (1998) J. Virol. 72, 6430–6436.
19. Franke, E. K. & Luban, J. (1996) Virology 222, 279–282.
20. Braaten, D., Aberham, C., Franke, E. K., Yin, L., Phares, W. & Luban, J. (1996)
J. Virol. 70, 5170–5176.
21. Aiken, C. (1998) Virology 248, 139–147.
22. Braaten, D., Franke, E. K. & Luban, J. (1996) J. Virol. 70, 4220–4227.
23. Bukrinsky, M. I., Sharova, N., McDonald, T. L., Pushkarskaya, T., Tarpley,
W. G. & Stevenson, M. (1993) Proc. Natl. Acad. Sci. USA 90, 6125–6129.
24. Miller, M. D., Farnet, C. M. & Bushman, F. D. (1997) J. Virol. 71, 5382–5390.
25. Agresta, B. E. & Carter, C. A. (1997) J. Virol. 71, 6921–6927.
26. Grattinger, M., Hohenberg, H., Thomas, D., Wilk, T., Muller, B. & Krausslich,
H. G. (1999) Virology 257, 247–260.
27. Wiegers, K., Rutter, G., Schubert, U., Grattinger, M. & Krausslich, H. G.
(1999) Virology 257, 261–274.
28. Saphire, A. C., Bobardt, M. D. & Gallay, P. A. (1999) EMBO J. 18, 6771–6785.
29. Schmidtmayerova, H., Alfano, M., Nuovo, G. & Bukrinsky, M. (1998) J. Virol.
72, 4633–4642.
30. Bukrinskaya, A., Brichacek, B., Mann, A. & Stevenson, M. (1998) J. Exp. Med.
188, 2113–2125.
31. Miller, D. G. & Miller, A. D. (1993) J. Virol. 67, 5346–5352.
32. Tailor, C. S., Nouri, A. & Kabat, D. (2000) J. Virol. 74, 9797–9801.
33. Aiken, C. (1997) J. Virol. 71, 5871–5877.
34. Dorfman, T. & Gottlinger, H. G. (1996) J. Virol. 70, 5751–5757.
35. Ellis, S. M., Nabeshima, K. & Biswas, C. (1989) Cancer Res. 49, 3385–3391.
36. Zhang, H., Dornadula, G. & Pomerantz, R. J. (1996) J. Virol. 70, 2809–2824.
37. Zack, J. A., Arrigo, S. J., Weitsman, S. R., Go, A. S., Haislip, A. & Chen, I. S.
(1990) Cell 61, 213–222.
38. Kuschert, G. S., Coulin, F., Power, C. A., Proudfoot, A. E., Hubbard, R. E.,
Hoogewerf, A. J. & Wells, T. N. (1999) Biochemistry 38, 12959–12968.
39. Ali, S., Palmer, A. C., Banerjee, B., Fritchley, S. J. & Kirby, J. A. (2000) J. Biol.
Chem. 275, 11721–11727.
40. Carpentier, M., Allain, F., Haendler, B., Denys, A., Mariller, C., Benaissa, M.
& Spik, G. (1999) J. Biol. Chem. 274, 10990–10998.
41. Bukrinskaya, A. G., Ghorpade, A., Heinzinger, N. K., Smithgall, T. E., Lewis,
R. E. & Stevenson, M. (1996) Proc. Natl. Acad. Sci. USA 93, 367–371.
42. Blumenthal, R. & Morris, S. J. (1999) Mol. Membr. Biol. 16, 43–47.
43. Liu, J., Farmer, J. D. J., Lane, W. S., Friedman, J., Weissman, I. & Schreiber,
S. L. (1991) Cell 66, 807–815.
44. Folks, T. M., Powell, D., Lightfoote, M., Koenig, S., Fauci, A. S., Benn, S.,
Rabson, A., Daugherty, D., Gendelman, H. E. & Hoggan, M. D. (1986) J. Exp.
Med. 164, 280–290.
Pusharsky et al. PNAS
兩
May 22, 2001
兩
vol. 98
兩
no. 11
兩
6365
MEDICAL SCIENCES
