JOURNAL OF VIROLOGY, Nov. 2008, p. 11228–11238
0022-538X/08/$08.00⫹0 doi:10.1128/JVI.00981-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 82, No. 22
Retroviruses Human Immunodeficiency Virus and Murine Leukemia
Virus Are Enriched in Phosphoinositides䌤†
Robin Chan,1 Pradeep D. Uchil,3 Jing Jin,3 Guanghou Shui,1 David E. Ott,4‡
Walther Mothes,3‡ and Markus R. Wenk1,2‡*
Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore1;
Department of Biological Sciences, National University of Singapore, Singapore, Singapore2; Department of Microbial Pathogenesis,
Yale University School of Medicine, New Haven, Connecticut3; and SAIC-Frederick, Inc.,
National Cancer Institute at Frederick, Frederick, Maryland4
Received 12 May 2008/Accepted 28 August 2008
Retroviruses acquire a lipid envelope during budding from the membrane of their hosts. Therefore, the
composition of this envelope can provide important information about the budding process and its location.
Here, we present mass spectrometry analysis of the lipid content of human immunodeficiency virus type 1
(HIV-1) and murine leukemia virus (MLV). The results of this comprehensive survey found that the overall
lipid content of these viruses mostly matched that of the plasma membrane, which was considerably different
from the total lipid content of the cells. However, several lipids are enriched in comparison to the composition
of the plasma membrane: (i) cholesterol, ceramide, and GM3; and (ii) phosphoinositides, phosphorylated
derivatives of phosphatidylinositol. Interestingly, microvesicles, which are similar in size to viruses and are
also released from the cell periphery, lack phosphoinositides, suggesting a different budding mechanism/
location for these particles than for retroviruses. One phosphoinositide, phosphatidylinositol 4,5-bisphosphate
[PI(4,5)P2], has been implicated in membrane binding by HIV Gag. Consistent with this observation, we found
that PI(4,5)P2 was enriched in HIV-1 and that depleting this molecule in cells reduced HIV-1 budding. Analysis
of mutant virions mapped the enrichment of PI(4,5)P2 to the matrix domain of HIV Gag. Overall, these results
suggest that HIV-1 and other retroviruses bud from cholesterol-rich regions of the plasma membrane and
exploit matrix/PI(4,5)P2 interactions for particle release from cells.
Retroviruses rely on their host for many essential parts of
the viral replication cycle. Biochemical and antibody-based
analyses of the replication cycle and proteins found in the
virions have revealed many details of the molecular interactions between human immunodeficiency virus (HIV) and its
host (20). In contrast, the role of lipids has been less well
studied. With the increasing recognition that lipids play an
important role in cellular signaling, it is no coincidence that
lipid factors are slowly gaining prominence in our understanding of retroviral replication.
Retroviruses, including HIV and murine leukemia virus
(MLV), acquire their lipid coats by budding through host
plasma membranes. Two important issues arise when considering the roles of lipids in retrovirus assembly and budding.
First, the idea that HIV and other retroviruses bud from lipid
rafts has gained widespread acceptance (39, 45). Lipid rafts are
liquid ordered domains that exist within the liquid disordered
phase of the bulk cell membrane. These dynamic lipid-protein
assemblies are characterized by high levels of cholesterol,
sphingolipids, saturated glycerophospholipids, and raft proteins. Because the half-lives for lipid rafts are extremely short
* Corresponding author. Mailing address: Centre for Life Sciences
(CeLS), Yong Loo Lin School of Medicine, National University of
Singapore, 28 Medical Drive, Level 04-21, Singapore 117607, Singapore. Phone: 65 6516 3624. Fax: 65 6777 3271. E-mail: bchmrw@nus
.edu.sg.
† Supplemental material for this article may be found at http://jvi
.asm.org/.
‡ D.E.O., W.M., and M.R.W. are co-senior authors of the paper.
䌤
Published ahead of print on 17 September 2008.
(50), the assignment of HIV to lipid rafts is commonly established through the colocalization of HIV proteins with putative
raft proteins and the preponderance of raft lipids, including
cholesterol, sphingomyelin (SM), dihydrosphingomyelin (dhSM),
ceramide (Cer), and glucosylceramide (Glu-Cer) (8, 36).
The second issue is the role of phosphatidylinositol 4,5bisphosphate [PI(4,5)P2] in retrovirus assembly. Although
PI(4,5)P2 comprises only a small fraction of the total phospholipids in a typical animal cell, it plays multiple roles in regulating many cell signaling pathways. One such function is in the
targeting of proteins with polybasic amino acid clusters to the
plasma membrane (23, 60). Gag proteins, whose matrix domain (MA) contains a polybasic cluster in the globular domain,
rely on a similar targeting mechanism for its accumulation at
the plasma membrane prior to budding (10, 35). Indeed, the
depletion of PI(4,5)P2 inhibits virus assembly and leads to an
accumulation of Gag at late endosomes and multivesicular
bodies (38). Further, it has been argued that PI(4,5)P2 acts as
both a trigger for myristate exposure and membrane anchor
(49).
The composition of the retroviral lipid envelope provides
important information about the assembly and budding process, especially regarding the nature of the budding site. Early
work in this field examined the lipid composition of Rous
sarcoma virus (42–44) and HIV (1) but at low resolution using
thin-layer chromatography. A more recent study by Brugger et
al. used electrospray ionization mass spectrometry (ESI-MS)
to compare the lipid composition of HIV with that of total cell
membrane and suggested that HIV buds from a unique or
specialized type of lipid raft (8). To better determine what
11228
VOL. 82, 2008
RETROVIRUSES HIV AND MLV ARE ENRICHED IN PHOSPHOINOSITIDES
lipids are enriched or depleted in virions during budding, we
analyzed the total lipid composition from highly purified samples of HIV and MLV by MS and compared it to that of the
membrane from which these viruses bud, the plasma membrane. In addition, we have expanded the scope of the previous
analyses to include more lipid classes and quantify the relative
levels of glycerophosphatidylinositol (PI), glycerophosphatidylinositol monophosphate (PIP), glycerophosphatidylinositol
bisphosphate (PIP2), ether glycerophosphatidylcholine (ePC),
and glycosphingolipid GM3. We find that retroviral envelopes
resemble the lipid composition of plasma membrane except
that they are highly enriched in specific raft lipids and phosphoinositides PIP and PIP2. In addition, we demonstrate that
the source of PIP2 enrichment in the HIV envelope maps to
the globular head of the MA domain of HIV Gag which contains the polybasic cluster, suggesting that the electrostatic
interaction between MA and PIP2 is a conserved function in
proper retroviral assembly and budding.
MATERIALS AND METHODS
Reagents. All cell culture medium and supplements were purchased from
GIBCO, Invitrogen (Carlsbad, CA). Lipid standards were purchased from
Avanti Polar Lipid Inc. (Alabaster, AL) and Echelon Biosciences Inc.(Salt Lake
City, UT). All other reagents, including high-performance liquid chromatography (HPLC)-grade methanol, chloroform, and piperidine, were purchased from
Sigma Aldrich (St. Louis, MO).
Cell lines. Uninfected H9 cells and the chronically HIV type 1 HIV-1MNinfected cell line, clone 4 (41), were cultured in RPMI 1640 medium with 10%
(vol/vol) fetal bovine serum, 2 mM L-glutamine, penicillin G at 100 U/ml, and
streptomycin sulfate at 100 ␮g/ml (complete medium). Uninfected and chronically MLV-infected rat embryo fibroblast (REF) cell lines were cultured in
Dulbecco modified Eagle medium with 10% fetal bovine serum, 2 mM L-glutamine, and penicillin G at 100 U/ml, and streptomycin sulfate at 100 ␮g/ml
(complete medium).
Isolation and culture of macrophages. Elutriated monocytes from HIV-negative donor leukopacs were obtained from the NIH Transfusion Branch and
cultured at 2 ⫻ 106 to 3 ⫻ 106 cells per well on ultralow-attachment six-well
Costar plates (catalog no. 3471; Corning, Acton, MA) in RPMI 1640 medium
with the supplements described above (complete medium) for 7 days to generate
monocyte-derived macrophages (MDMs). MDMs were infected with the CCR5tropic NLAD8 (18) overnight followed by two washes with phosphate-buffered
saline to remove nonadhered virus. The infected cells were then cultured, and
supernatants were periodically removed for virus isolation.
Virus stock preparation. Concentrated MLV was produced in roller culture
bottles from chronically infected REFs. The culture supernatants were passed
through a 0.45-␮m filter and ultracentrifuged at 25,000 rpm at 4°C for 90 min
through a 15% sucrose layer to obtain purified virus. For HIV produced from
clone 4 and MDMs, the culture supernatants were passed through a 0.45-␮m
filter and sedimented through 15% sucrose cushion, and the resulting HIV pellet
was further purified away from contaminating microvesicles using anti-CD45
depletion (55). All virus and microvesicle (prepared from uninfected cultures)
stocks were stored at ⫺80°C until use.
Preparation of VLP. Constructs expressing wild-type and ⌬MA HIV-Gag in
the absence of Env were made based on pNL4-3/KFS (a gift from Eric Freed,
NCI, Frederick, MD). Pol was deleted, and a hemagglutinin tag was added to the
C terminus of Gag to study the release of virus-like particles (VLP) in the
absence of protease. To make a ⌬MA HIV-Gag expression vector, the globular
head (amino acids 8 to 126) was deleted from the modified NL4-3/KFS clone.
HEK293 cells were transfected with wild-type and ⌬MA HIV expression vector.
At 48 h posttransfection, the supernatants were collected and passed through a
0.45-␮m filter. The clarified supernatants were ultracentrifuged at 25,000 rpm at
4°C for 3 h through a 15% sucrose layer to obtain purified virus.
Plasma membrane extraction from cells. Plasma membrane fractions were
purified using cationic colloidal silica beads (29, 53). Briefly, at least 106 cells
were typically used for each extraction experiment. Cells were bound sequentially
to cationic colloidal silica beads (a gift from Donna Beer Stolz, McGowan
Institute for Regenerative Medicine, University of Pittsburgh) at 1% (wt/vol) and
then to polyacrylic acid (average molecular weight of 100,000; Aldrich, St. Louis,
11229
MO). The bound cells were then incubated in a hypotonic lysis buffer for 30 min
before being lysed using a cell homogenizer followed by centrifugation at 900 ⫻
g to remove released internal membranes. The plasma membranes in the resulting pellet were purified away from the nuclei by sedimentation through a 70%
Histodenz (Aldrich) cushion using a TW41 rotor at 20,000 rpm for 30 min. The
purified plasma membrane pellet was then washed thoroughly with the lysis
buffer to remove any residual Histodenz.
Lipid preparation. Total lipid samples were prepared using a modified version
of the method of Bligh and Dyer (4). All buffers and reagents were prechilled in
an ice bath. Virus or cells were washed and resuspended in 50 ␮l phosphatebuffered saline. A portion (0.6 ml) of a chloroform-methanol (1:2) mixture was
added to the sample, and the mixture was vortexed vigorously three times for 1
min each time with a 5-min interval between the vortexing steps. Next, 0.3 ml
chloroform and 0.2 ml of 1 M KCl was added to the tube, and the mixture was
again vortexed, three times for 30 s each time with 1-min intervals between the
vortexing steps. The mixture was then centrifuged for 2 min at 9,000 rpm to
separate the phases. The lower organic layer was transferred to a clean microcentrifuge tube and dried under a stream of N2 gas or via a speed vacuum.
Phosphoinositide-enriched lipid samples were prepared by replacing the 1 M
KCl solution with 1 M HCl (58). Sphingolipid samples were prepared via the
alkaline hydrolysis method described by Merrill (32). Briefly, 0.75 ml of chloroform-methanol (1:2) was added to the sample, and the mixture was sonicated
until the sphingolipid samples appeared evenly dispersed and then incubated
overnight at 48°C. Next the samples were cooled and mixed with 75 ␮l of 1 M
KOH in methanol and incubated for 2 h at 37°C with shaking. The sample was
then neutralized with 6 ␮l of glacial acetic acid. Last, 0.5 ml of chloroform and
0.3 ml of water were added to the tube, and the mixture was then centrifuged for
2 min at 9,000 rpm to separate the phases. The lower organic layer was transferred to a clean microcentrifuge tube and dried under a stream of N2 gas or via
a speed vacuum.
Analysis of lipids using ESI-MS. Qualitative lipid profiling via ESI-MS was
carried out with a Waters Micromass Q-TOF micromass spectrometer with an
upfront Waters CapLC inlet (Waters Corp., Milford, MA) as described previously (52). The capillary voltage and sample cone voltage were maintained at 3.0
kV and 50 V, respectively. The source temperature was 80°C, and the desolvation
temperature was set at 250°C. Mass spectra were acquired in the negative ion
mode with an acquisition time of 3 min, using chloroform-methanol (1:1, vol/vol)
at a flow rate of 15 ␮l/min as the mobile phase. Typically, samples were dissolved
in the mobile phase to give an appropriate concentration, and 2 ␮l of sample was
injected for analysis.
For quantitative analysis, we used a triple quadrupole instrument ABI 4000QT
(Applied Biosystems, Foster City, CA) in the multiple reaction monitoring mode.
This method is optimized to detect only specified ions of interest, thus increasing
sensitivity and selectivity of detection (14, 32). In our experiments, the internal
standards used included 1,2-dimyristoyl-glycero-phosphoserine, 1,2-dimyristoylglycero-3-phosphoethanolamine, 1,2-dimyristoyl-glycero-3-phosphocholine, lauroyl sphingomyelin, N-heptadecanoyl-D-erythro-sphingosine (C17 ceramide), and
D-glucosyl-ß1-1⬘-N-octanoyl-D-erythro-sphingosine (C8 glucosyl ceramide) (Avanti
Polar Lipids), which allowed the measurement of glycerophosphatidylserine (PS),
glycerophosphatidylethanolamine (PE) and plasmalogen-PE (pl-PE), glycerophosphatidylcholine (PC) and ePC, SM, and Cer and Glu-Cer, respectively (see
Fig. S5A in the supplemental material). PI, PIP, and PIP2 levels were referenced
to 1,2-dioctanoyl-glycero-3-phosphoinositol, 1,2-dioctanoyl-glycero-3-(phosphoinositol-4-phosphate), and 1,2-dioctanoyl-glycero-3-(phosphoinositol-4,5-bisphosphate)
(Echelon Biosciences Inc.), respectively (see Fig. S5B in the supplemental material). Since a suitable standard was not available for GM3, we normalized GM3
levels to SM levels. The total lipid and sphingolipid extracts were dissolved in
chloroform-methanol (1:1, vol/vol), and typically, 10 to 15 ␮l of the sample was
injected via an autosampler. For the phosphoinositide samples, the lipids were
dissolved in chloroform-methanol (1:1), spiked with 1/10 volume of 300 ppm
piperidine solution, and directly infused into the mass spectrometer (58). The
m/z transitions used were published previously (7, 32, 58), and we optimized the
declustering potential and collision energy using the Quantitative Optimization
function available for the Analyst 1.4.1 software. The instrument was calibrated
using polypropyleneglycol standards provided by the manufacturer (Applied
Biosystems), and mass tolerance was adjusted to ⬃100 ppm or 0.1 Da. The signal
intensity obtained for each lipid species was converted to their level (moles) in
each fraction by normalization to the appropriate internal standard. The final
lipid molar percentage (the amount of each lipid was measured in moles and
then converted to a percentage of the total) of each sample was obtained by cross
normalizing the lipid level of each fraction measured (25). The standard deviations shown represent the differences in at least three replicate experiments, i.e.,
n ⱖ 3.
11230
CHAN ET AL.
J. VIROL.
For the measurement of cholesterol, we employed the same ABI 4000QT
instrument connected to an HPLC system using a sensitive HPLC–ESI-MS
method (G. Shui et al., unpublished data). Briefly, lipids were separated on an
Agilent Zorbax Eclipse XDB-C18 column (inner diameter, 4.6 mm; length, 150
mm) (Agilent Technologies, Santa Clara, CA) at 30°C using chloroform–methanol–0.1 M ammonium acetate (100:100:4 [vol/vol/vol]) as a mobile phase at a
flow rate of 0.4 ml. min⫺1 and an injection volume of 30 ␮l. MS was recorded at
both positive and negative ESI modes in enhanced MS scan mode with a Turbo
spray source voltage of ⫹5,000 V and ⫺4,500 V, respectively, and a source
temperature of 250°C. A total run time of 30 min was utilized to elute both polar
lipids and neutral lipids. The cholesterol-to-PC molar ratio (cholesterol/PC ratio) was used to compare the cholesterol levels in the different samples (8). The
cholesterol levels were calculated by normalizing the 369 m/z peak intensity to
the total PC peak intensity, both of which were measured at positive mode.
Virus budding and release assays. To assess HIV release, 2.5 ⫻ 105 HEK293
cells in 24 wells were transfected with 150 ng of replication-competent HIV
pNL4-3 construct (NIH AIDS Reagent Program) together with plasmids expressing 5-phosphatase IV or catalytically inactive 5-phosphatase ⌬1 (gifts from
Eric Freed, NCI, Frederick with permission from P. Majerus, Washington University School of Medicine, St. Louis, MO) or with an empty vector control. For
MLV release assays, 200 ng of plasmid MLV Env-GFP encoding full-length
Friend 57 MLV genome with a green fluorescent protein (GFP) insertion into
the envelope protein (28, 51) was cotransfected together with plasmids encoding
5-phosphatase IV or catalytically inactive 5-phosphatase ⌬1 or with a vector
control as described above. At 48 h posttransfection, the released virus infectivity
was measured by titering serial dilutions of the culture supernatants onto target
cells (TZM-bl [12] and DFJ8 [2] cell lines for HIV and MLV, respectively).
Luciferase activity was measured 36 to 48 h postinfection in TZM-bl cell lysates
to determine HIV infectivity. For MLV, GFP-positive DFJ8 cells were enumerated after additional 36 to 48 h using a fluorescence-activated cell sorter (Becton
Dickinson) for MLV infectivity. The experiment was carried out with three
replicates per trial on two separate days. The data were normalized to virus
infectivity released from samples transfected with empty vector and presented as
inhibition of infectious virus released. In parallel, triplicate samples of cells and
culture supernatants from the experiments described above were pooled and
processed for Western blot analysis and probed with antibodies to HIV capsid
(NIH AIDS Reagent Program) or MLV capsid (Camden, NJ).
RESULTS
Lipid profiles of HIV and other retroviruses. In order to
analyze retroviral lipids, we purified HIV and MLV from culture supernatants of chronically infected cells by ultracentrifugation through 15% sucrose cushions. This resulted in highly
purified MLV particles lacking detectable microvesicles (see
Fig. S1 in the supplemental material). HIV preparations isolated from culture supernatants of H9 cells (T-cell line) and
MDMs still contained nonviral particles and were further purified using anti-CD45 immunodepletion to remove these microvesicles (55) (see Fig. S1 in the supplemental material).
Due to the different chemistry of different lipid classes, we
prepared our samples in three different fractions using a variation of the chloroform-methanol extraction method (4). Total
lipid (PS, PC, ePC, PE, pl-PE, PI, and SM) fractions were
prepared by the standard method. Phosphoinositides (PI, PIP,
and PIP2) were extracted by partitioning into the organic phase
at low pH (58). Sphingolipids (SM, dhSM, Cer, and Glu-Cer)
were made accessible following alkaline hydrolysis of the total
lipid extract (32). Finally, we employed ESI-MS and multiple
reaction monitoring to profile and quantify a total of ⬃250
individual lipid species in the complex mixtures. To allow for
the quantitation of the relative amounts of lipids, each sample
was spiked with a relevant set of internal standards (see Materials and Methods). By referencing to these internal standards, we were able to calculate the molar percentages of the
individual lipid classes in each fraction and then through cross
normalization, the total detectable lipid classes (Fig. 1A). The
lipid data obtained in our work were in agreement with data
generated in previous studies (1, 8).
To reveal the potential enrichment or exclusion of lipids in
the virion, we compared viral lipids to total cellular lipids of
their producer cells (Fig. 1B). The HIV-H9 lipid composition
and the comparative ratio of viral lipids to total cellular lipids
confirmed the enrichment of PS, pl-PE, cholesterol, SM,
dhSM, and Glu-Cer previously reported by Brugger et al. (8),
while PC and ceramide (Cer) were reduced (Fig. 1A and B) (1,
7, 8). Interestingly, PI was consistently low, while phosphoinositides were highly elevated in HIV envelopes. Qualitatively, parallel experiments with virions produced from H9
cells infected with a mutant found that the lipid composition of
HIV was not changed in the absence of the viral envelope
glycoprotein (Env) (Fig. 2C) (see Fig. S2A, B, and C in the
supplemental material), suggesting that even though the Env
protein is thought to associate with lipid rafts (46), Env does
not assist the specific incorporation of any lipid. Instead, our
data suggest that Gag alone dictates the major lipid components incorporated into HIV, consistent with the fact that Gag
alone is responsible for lipid raft association (3) and is sufficient for particle assembly and release (33).
HIV derived from H9 cells or MDMs were also virtually
identical, with the exception that dhSM is elevated in HIV
from MDM cells (Fig. 1A). While the differences are subtle,
they could be a consequence of the precise site of assembly in
both cell types. HIV from H9 cells buds at the plasma membrane, while HIV originates at deeply invaginated membrane
structures that appear to be derived from the plasma membrane in MDMs (11, 27, 57). Their overall similarity in virion
lipid composition is in agreement with the emerging evidence
that both assembly sites are continuous with the plasma membrane (11, 27, 47, 57). Microvesicles released from uninfected
MDMs mostly share the lipid composition of HIV released
from these cells but interestingly had very low levels of phosphoinositides, while HIV from the same cells contained these
species (Fig. 1A). Lipid profiles of MLV from REFs showed
remarkable similarity to that of HIV even with the differences
in virus species and cell type (Fig. 1A and B and 2B).
The lipid composition of retroviruses resembles that of
plasma membranes. The lipid composition of the plasma
membrane is distinct from that of other cellular membranes,
exhibiting higher levels of cholesterol and sphingolipids (56).
Because HIV and MLV both use the plasma membrane for
budding, we analyzed the lipids present in the plasma membrane rather than total cellular lipids to better assess enrichment. We enriched the plasma membrane via the use of cationic silica beads that adhere electrostatically to the plasma
membrane (29, 53). To monitor the efficacy of these cationic
beads for plasma membrane preparation, we followed enrichment of raft (flotillin and caveolin) and nonraft markers (transferrin receptor [TrF]) (Fig. 2A). Actin (cytoplasmic protein)
and Rab5 (endosomal protein) served as indicators for plasma
membrane purity (Fig. 2A). Membranes are relatively plastic
and can be altered by outside agents. Therefore, before proceeding, we investigated whether the concentration of cationic
beads had any artificial effects on the membrane preparations
that could lead to artifacts. At a low bead concentration of 1%,
all three markers were present to similar extents in adherent
VOL. 82, 2008
RETROVIRUSES HIV AND MLV ARE ENRICHED IN PHOSPHOINOSITIDES
11231
FIG. 1. Lipid composition of retroviruses in comparison to the lipid composition of total membranes of producer cells. (A) Lipid composition
of different retrovirus envelopes and microvesicles (MV) produced from various cell types as described in the text. Values are expressed as molar
percentages of a given lipid to the total lipid measured (except cholesterol and GM3, which were normalized to the total PC and SM signal
intensities, respectively). PA, glycerophosphates; n/a, not available. (B) Ratio of retroviral lipids to total membrane lipids of host cells. Lipids that
are significantly enriched (⬎1.5-fold) or reduced (⬍1.5-fold) in viral envelopes are highlighted in red and green, respectively. Each experiment was
performed at least three times (n ⱖ 3). Values that were significantly different (P ⬍ 0.05) from total membranes of producer cells are indicated
by an asterisk.
REFs (Fig. 2A). TrF levels in the plasma membrane fractions
appear to be constant with increasing bead concentration and
less abundant than the total membrane levels. The fact that
TrF recycles between plasma membrane and endosomes in a
cell (22) explains why REF plasma membrane preparations
will contain less transferrin than observed in the REF total
membrane preparations. Surprisingly, the raft markers flotillin
and caveolin appeared to be enriched over the nonraft marker
TrF with increasing amounts of beads used, suggesting that
these beads have a propensity for inducing lipid raft fractions
at high concentrations (Fig. 2A). Therefore, we decided to
apply the 1% cationic bead solution, which does not induce this
artifact, to prepare plasma membranes from all the other adherent primary cell lines, such as MDMs. We also found that
the condition works well for the suspension cell line H9, which
showed equal levels of transferrin and flotillin (Fig. 2A).
Unbiased lipid profiling of purified viral particles and
plasma membrane fractions revealed that the MS patterns of
11232
CHAN ET AL.
J. VIROL.
FIG. 2. The lipid composition of retroviruses resembles that of plasma membranes. (A) Plasma membrane (PM) fractions were isolated using
cationic beads. Western blots with sera against raft (flotillin and caveolin) and nonraft markers (transferrin receptor [TrR]) were used to assess
the quality of the plasma membranes, while actin (cytoplasmic protein) and Rab5 (endosomal protein) served as indicators for the purity of the
plasma membranes. Conditions equivalent to 1% bead solution were used for the determination of lipid contents. TM, total membranes; ⫺, no
beads. (B to G) Time of flight ESI-MS (negative mode) are shown for lipid mixtures from MLV (B), REF plasma membrane (D), and REF
VOL. 82, 2008
RETROVIRUSES HIV AND MLV ARE ENRICHED IN PHOSPHOINOSITIDES
MLV and HIV-H9 lipids were highly similar to their plasma
membrane lipids, but not total cellular lipids (Fig. 2B to G).
Both virus spectra were dominated by PS ions, namely, PS 36:1,
PS 34:1, and PS 38:3 or PS 38:2 in the negative polarity ESI
mode (Fig. 2B and C). This can be attributed to the abundance
of PS in the lipid extract and the high ionization efficiency of
PS as well as the exclusion of PI in viral envelopes. In addition
to PS, other minor ions in the mass spectra corresponded to
numerous species of PE, pl-PE, and GM3 which are not labeled in the figures. REF and H9 plasma membrane spectra
were also similar to the viruses in having high PS levels but
differed in having higher levels of PI (Fig. 2D and E). The
results of the detailed analysis of fatty acid side chains is presented in Fig. 3 for HIV-H9 and in Fig. S3 and S4 in the
supplemental material for MLV and HIV-MDM, respectively,
indicating that retroviral envelopes have a tendency to harbor
the shorter and more saturated fatty acyl chains of the cell.
This distribution of retroviral lipids is highly similar to the
plasma membrane while being distinctly different from the
total membrane of their respective host cell (Fig. 4A and Fig.
3) (see Fig. S3 and S4 in the supplemental material).
The resulting comparison of viral lipids to plasma membrane
lipids illustrates the striking similarity of retroviruses with
plasma membrane lipids (Fig. 1B). In contrast, when total cell
lipids are used for comparison, PS, pl-PE, SM, and dhSM, are
significantly enriched, while PI, PC, PE, and Cer are reduced in
viral envelopes (Fig. 4B). Thus, the plasma membrane is the
appropriate cellular reference for comparison. Cholesterol,
Cer, and GM3 remain enriched in both retroviral envelopes
and microvesicles over their host plasma membrane (Fig. 4B)
(see Fig. S4 in the supplemental material). More significantly,
PIP and PIP2 are enriched in the retrovirus envelopes compared to the plasma membranes (Fig. 2H) and microvesicles
(see Fig. S4 in the supplemental material). These data demonstrate that while retroviral and microvesicle lipids resemble
plasma membrane lipids, the presence of phosphoinositides
distinguishes retroviruses from microvesicles.
The incorporation of PIP2 into HIV is reduced in HIV lacking the MA domain. We next investigated the source of PIP2
enrichment in retroviral envelopes. The MA domain of retrovirus Gag protein contains a polybasic patch which is argued to
interact electrostatically with negatively charged PI(4,5)P2 at
the membrane surface during virus assembly (9, 35, 38). To test
the role of the polybasic patch in PIP2 incorporation, we produced and purified VLP from HEK293 cells expressing either
wild-type HIV Gag or a mutant HIV Gag lacking the polybasic
globular head of MA domain but still containing the N-terminal myristylation signal (⌬MA HIV-Gag) and compared their
phosphoinositide profiles (Table 1). Strikingly, PIP2 was reduced twofold in ⌬MA HIV-Gag VLP, near the level of microvesicles, suggesting that the MA domain of HIV-Gag is
responsible for the enrichment of PIP2 in the retroviral envelope. Consistent with a role of PIP2 in retrovirus release, HIV
11233
and MLV release were sensitive to the lowering of PIP2 levels
at the plasma membrane. The enzymatically active 5-phosphatase interfered with both HIV and MLV release from
HEK293 cells, with a stronger effect seen in MLV (Fig. 5A and
B). While the inactive form did not affect virus release significantly, we observed that at high transfection levels, the inactive form of the enzyme also interfered with MLV release,
suggesting pleiotropic side effects under these conditions (Fig.
5B). Together, these results suggest that the dependence of
HIV budding on PIP2 is likely due to the regulatory role of MA
in HIV release.
DISCUSSION
With this work, we present an extensive analysis of the lipidome of HIV produced from both T cells and macrophages
along with the corresponding lipid contents of the plasma
membrane from which the virus buds. This study greatly expands the coverage of lipid species previously presented by
Aloia et al. (1) and Brugger et al. (8), including the bioactive
lipids PIP, PIP2, and GM3. Additionally, we analyzed the oncoretrovirus MLV to provide a more global assessment of
retroviral envelopes. We report that HIV and MLV have similar lipid composition despite being produced from different
cell types. Importantly, when retroviral lipids were compared
to plasma membrane lipids rather than to total cellular lipids,
the lipidome of retroviruses largely resembled the composition
of the plasma membrane. Thus, our data do not support the
idea that HIV buds from a raft with unusual lipid composition
as proposed by Brugger et al. (8). The differences in our data
are clearly due to the source used for lipid comparison. Brugger et al. (8) used total lipids, which includes lipids from organelles and other membrane structures that are unlikely to
support retroviral budding. In contrast, our comparison between viral envelopes and plasma membranes is more virologically relevant, given the current data that support HIV budding from the plasma membrane (11, 17, 27, 57). It is
important to note that we fully recapitulate the results of
Brugger et al. (8) when we analyze total cell membranes.
Because the condition of the cellular reference used for
comparison with the viral envelope is crucially important, we
carefully considered and tested the preparation procedure for
our plasma membrane preparations. One major concern is the
contamination of the plasma membrane preparations with lipids from cellular organelles which have different lipid compositions (56). Hence, the use of plasma membranes for comparative lipid analysis purpose depends critically on the purity of
the plasma membrane preparations. While the method we
used, cationic silica bead capture of the plasma membrane, can
artificially induce lipid rafts, we carefully crafted and monitored our conditions to eliminate this possibility.
Despite the similarity between the plasma membrane and
the viral envelope, retroviral lipids were still distinct from
total membranes (TM) (F). The lipids extracted from CD45-depleted HIV (C), H9 plasma membrane (E), and H9 total membranes (G) were
analyzed similarly. The representative spectra shown are normalized to the highest peak within the m/z range. Prominent ions which were
characterized by tandem MS are labeled. (H) Enrichment of phosphoinositides in retroviral envelopes. Precursor ion scanning for m/z 241
(dehydrated inositol fragment) was used to detect PI and phosphoinositides in MLV and plasma membrane.
11234
CHAN ET AL.
J. VIROL.
FIG. 3. Quantification of individual species of glycerophospholipids and sphingolipids of HIV and H9 host cells. Abundance is represented as
the molar percentages (y axis) of a given lipid (x axis) to total lipid measured, except for GM3, which was normalized to the total SM levels. Lipids
were extracted from purified virus (black bars), total cell membrane, and plasma membrane fractions and quantified via MS using multiple reaction
monitoring. The percentages were calculated with relevant internal standards. GM3 quantification is represented in relative levels due to the lack
of suitable internal standards. Sphinogolipids are presented as sphinogoid base residue/fatty acyl residue.
VOL. 82, 2008
RETROVIRUSES HIV AND MLV ARE ENRICHED IN PHOSPHOINOSITIDES
11235
TABLE 1. Phosphoinositide composition of wild-type HIV Gag
VLP, ⌬MA HIV Gag VLP, and microvesicles
Phosphoinositide composition (%)a in:
Phosphoinositide
WT HIV
Gag VLPb
⌬MA HIV
Gag VLP
MVc
PI
PIP
PIP2
31.95 ⫾ 5.67
28.29 ⫾ 5.23
39.76 ⫾ 5.41
63.15 ⫾ 3.07
18.40 ⫾ 2.97
18.45 ⫾ 6.02
67.32 ⫾ 1.39
17.58 ⫾ 2.44
15.10 ⫾ 3.60
a
Values are expressed as molar percentages, and data presented are the
averages ⫾ standard deviations for three independent experiments.
b
WT, wild type.
c
MV, microvesicles.
viral assembly. Cer molecules dramatically change the biophysical properties of rafts. In vivo, the accumulation of Cer at the
plasma membrane occurs as a result of activation and surface
translocation of acid sphingomyelinase (21). Not only is Cer a
strong promoter of lipid raft formation, Cer-rich rafts appear
to spontaneously coalesce to form larger macrodomains or
platforms through fusion (5). Cer further exerts its effects by
selectively displacing cholesterol in the rafts and interfering
FIG. 4. Lipid composition of retroviruses in comparison to the lipid
composition of the plasma membranes of producer cells. (A) Values
are expressed as molar percentages (except for cholesterol and GM3).
(B) Ratio of retroviral and microvesicle (MV) lipid composition to
plasma membrane lipid composition. Lipids that are significantly enriched (⬎1.5-fold) or reduced (⬍1.5-fold) in viral envelopes are highlighted in red and green, respectively. Each experiment was performed
at least three times (n ⱖ 3). Values that were significantly different
(P ⬍ 0.05) from plasma membrane fractions of producer cells are
indicated by an asterisk.
plasma membrane lipids in a number of ways. First, retrovirus
envelopes are highly enriched in raft lipids, such as cholesterol
and GM3 and in most cases Cer (with the exception of HIV
produced from H9 cells). The enrichment of cholesterol and
GM3 comes without surprise, as it has been well established
that HIV buds selectively from cholesterol-enriched (40) and
glycolipid-enriched (36) membrane rafts that potentially originate during initial Gag-membrane interactions (6). Moreover,
HIV has been shown to specifically control the enrichment of
both these lipids through its accessory Nef protein (34, 62).
However, the enrichment of Cer rather than SM or dhSM (as
previously reported by Brugger et al. [8]) signifies a fundamental difference in the type of lipid raft that forms during retro-
FIG. 5. Effects of PI(4,5)P2 depletion on HIV (A) and MLV (B) release from HEK293 cells. The effect of transient expression of 5PaseIV
and 5Pase⌬1 was determined by normalizing virus infectivity released
from samples transfected with empty vector and presented as inhibition of infectious virus released. The ratio data were obtained by
normalizing the 5-phosphatase IV (5PaseIV) to 5Pase⌬1 values. The
Western blots show the levels of HIV (A) and MLV (B) Gag expression in cells and released virus detected in culture supernatants.
11236
CHAN ET AL.
with the association of cholesterol binding/interacting proteins
with these platforms (31, 61). Thus, Cer-enriched membrane
platforms appear to function as a tool that reorganizes receptor and signaling molecules in and at the cell membrane to
facilitate and amplify signaling processes.
It was recently demonstrated that Cer-enriched exosomes
bud into multivesicular bodies, triggered by localized accumulation of Cer through sphingomyelinase action (54). This is
consistent with the enrichment of Cer shown in our own data
for both microvesicles and retroviruses. However, it appears
that the level of Cer has to be carefully regulated in order to
ensure continued reinfection. Interestingly, increasing Cer levels in cells by pharmacological or enzymatic means inhibits
HIV infectivity (16), supposedly by inducing CD4 clustering
and preventing coreceptor engagement and HIV fusion (15). It
is intriguing to consider how the supposed formation of Cerenriched macrodomains would fit into a dynamic model for the
formation of microvesicle and retrovirus particles.
An association with tetraspanins represents another parallel
phenomenon seen between retroviruses, microvesicles, and/or
exosomes. For exosomes, the enrichment of tetraspanins, including CD37, CD63, CD81, CD82, and CD86, has been well
established (13, 59). Likewise, it was shown that HIV Gag and
Env colocalize with distinct tetraspanin-enriched microdomains containing CD9, CD53, CD63, CD81, and CD82 during
particle assembly at the plasma membrane (11, 26, 37). At the
same time, components of the cellular budding machinery,
including TSG101 and VSP28, are also recruited to tetraspanin-enriched microdomains to facilitate viral budding
(33, 37). In this context, it is possible that the precise regulation
of plasma membrane Cer and cholesterol levels helps control
the dimensions of membrane macrodomain structures required to accommodate all these membrane-associated protein
structures during the retrovirus assembly and budding process.
The second distinction drawn from our lipid comparison is
that retroviruses lack PI, while phosphoinositides PIP and PIP2
are highly enriched compared to plasma membrane levels. We
note that it is likely that phosphatase activity may have reduced
the native levels of PIP2 in the plasma membrane during isolation and resulted in a corresponding increase in PI levels.
Nevertheless, the key distinguishing feature that differs between retroviruses and microvesicles is the level of phosphoinositides, particularly PIP2. Microvesicles likely reflect the
PIP2 levels in the native plasma membrane, thereby supporting
an enrichment of this lipid in the virions. Strikingly, we determine that the MA domain of HIV-Gag is responsible for the
enrichment of PIP2. Taken together, these data strongly suggest that efficient retrovirus assembly and release depend on
the electrostatic interaction between both the polybasic MA
domain and divalent negatively charged PI(4,5)P2 molecules at
the plasma membrane.
Unfortunately, it is not feasible to discriminate stereoisomers of PIs by MS, but the majority of PIP2 at the plasma
membrane in resting mammalian cells appears to be present as
PI(4,5)P2 (48). Our findings are consistent with the notion that
PI(4,5)P2 acts in targeting of the HIV-1 Gag protein via an
interaction with the MA domain of Gag (38) to sites at the
plasma membrane (17, 27). Recent structural analysis of
HIV-1 Gag and its interaction with PI(4,5)P2 strongly suggest
that this interaction results in the exposure of the saturated
J. VIROL.
myristic acid into the budding membrane and the equivalent
flipping out of the polyunsaturated 2⬘-fatty acid of PI(4,5)P2,
resulting in the formation of an extended lipid conformation
(49). At least in the case of HIV, this interaction is specific for
PI(4,5)P2 and not other phosphoinositides (49). While all retroviral Gag types possess an electropositive surface patch on its
MA component (35), it remains to be determined which precise phosphoinositide species is involved in targeting Gag for
other retroviruses and cell types, in particular PI(3,4,5)P3 (23).
In addition to phosphoinositides, monovalent negative lipids
like PS and glycerophosphate may also contribute to the localization of Gag molecules through electrostatic interaction (60).
While it is not clear how PI(4,5)P2 is mechanistically enriched in the virus envelope, current theories of how PI(4,5)P2
is spatially regulated may provide some clues. It is currently
believed that there are two separate pool of PI(4,5)P2 in the
plasma membrane. About two-thirds is believed to be electrostatically sequestered by protein buffers with clusters of basic
residues, such as myristylated alanine-rich C kinase substrate,
to be released only in response to specific stimuli, such as an
increase in local calcium ion concentration (19). The remainder of PI(4,5)P2 is unbound and free to diffuse in the plasma
membrane milieu. Thus, it is unlikely that PI(4,5)P2 is able to
form concentrated spots locally on their own through random
diffusion. A more likely scenario lies in the initial electrostatic
interaction of the Gag molecule with PI(4,5)P2 followed by the
lateral sequestering of PI(4,5)P2 due to Gag-Gag multimerization.
Besides initiating Gag assembly, further roles for PI(4,5)P2
may well be a distinct possibility. Intriguingly, PI(4,5)P2 is
intimately involved in the inward and outward bending of the
plasma membrane in other biological systems. During endocytosis, BAR domain proteins bind to PI(4,5)P2-rich membranes
to form inward invaginations (24, 63). Conversely, during the
formation of filopodia, MIM and IRSp53, proteins which contain BAR-like domains, can lead to the formation of outward
bending of PI(4,5)P2-rich membranes (30). Considering the
strong enrichment of PIP2 lipids in the viral envelope, we hypothesize that binding of retroviral Gag proteins to PI(4,5)P2 may
contribute to the induction of membrane curvature during
virus assembly and budding.
In conclusion, our experimental approach in comparing retroviral lipids to the plasma membrane represents a more relevant picture of the enrichment of lipids in the viral envelope
over a comparison to total cell membrane (8). Taking a
broader view, this approach should be equally useful in the
study of other medically important enveloped viruses. The list
of enriched lipids presented herein should make a strong case
for continued investigation into their contribution toward retroviral assembly and budding. However, beyond analyzing the
obvious enrichment of these lipids, it must be noted that other
lipids that occur at high levels but are less enriched, such as PS,
pl-PE, SM, and even PC, are nonetheless important in the
formation of the native retroviral envelope and deserve further
investigation as well.
ACKNOWLEDGMENTS
We thank Maik Lehmann for initial virus preparations; Donna Beer
Stolz for silica beads; Kunio Nagashima and Marc Pypaert for electron
microscopy; and Pietro de Camilli for valuable suggestions.
VOL. 82, 2008
RETROVIRUSES HIV AND MLV ARE ENRICHED IN PHOSPHOINOSITIDES
M.R.W. was supported in part by grants from the Singapore National Research Foundation under CRP award no. 2007-04, the National University of Singapore via the Office of Life Science (R-183000-607-712), the Academic Research Fund (R-183-000-160-112), the
Biomedical Research Council of Singapore (R-183-000-134-305), and
the Novartis Institute for Tropical Diseases (R-183-000-166-592).
W.M. was supported by a grant from the National Institutes of Health
(NIH) (R21 AI065284), and P.D.U. was supported by an Anna Fuller
Fellowship in Cancer Research. This project was funded in part with
federal funds from the National Cancer Institute, National Institutes of
Health, under contract no. ND1-CO-12400.
REFERENCES
1. Aloia, R. C., H. Tian, and F. C. Jensen. 1993. Lipid composition and fluidity
of the human immunodeficiency virus envelope and host cell plasma membranes. Proc. Natl. Acad. Sci. USA 90:5181–5185.
2. Barsov, E. V., W. S. Payne, and S. H. Hughes. 2001. Adaptation of chimeric
retroviruses in vitro and in vivo: isolation of avian retroviral vectors with
extended host range. J. Virol. 75:4973–4983.
3. Bhattacharya, J., A. Repik, and P. R. Clapham. 2006. Gag regulates association of human immunodeficiency virus type 1 envelope with detergentresistant membranes. J. Virol. 80:5292–5300.
4. Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extraction
and purification. Can. J. Biochem. Physiol. 37:911–917.
5. Bollinger, C. R., V. Teichgraber, and E. Gulbins. 2005. Ceramide-enriched
membrane domains. Biochim. Biophys. Acta 1746:284–294.
6. Briggs, J. A., T. Wilk, and S. D. Fuller. 2003. Do lipid rafts mediate virus
assembly and pseudotyping? J. Gen. Virol. 84:757–768.
7. Brugger, B., G. Erben, R. Sandhoff, F. T. Wieland, and W. D. Lehmann.
1997. Quantitative analysis of biological membrane lipids at the low picomole level by nano-electrospray ionization tandem mass spectrometry. Proc.
Natl. Acad. Sci. USA 94:2339–2344.
8. Brugger, B., B. Glass, P. Haberkant, I. Leibrecht, F. T. Wieland, and H. G.
Krausslich. 2006. The HIV lipidome: a raft with an unusual composition.
Proc. Natl. Acad. Sci. USA 103:2641–2646.
9. Chukkapalli, V., I. B. Hogue, V. Boyko, W. S. Hu, and A. Ono. 2008. Interaction between the human immunodeficiency virus type 1 Gag matrix domain and phosphatidylinositol-(4,5)-bisphosphate is essential for efficient
Gag membrane binding. J. Virol. 82:2405–2417.
10. Dalton, A. K., D. Ako-Adjei, P. S. Murray, D. Murray, and V. M. Vogt. 2007.
Electrostatic interactions drive membrane association of the human immunodeficiency virus type 1 Gag MA domain. J. Virol. 81:6434–6445.
11. Deneka, M., A. Pelchen-Matthews, R. Byland, E. Ruiz-Mateos, and M.
Marsh. 2007. In macrophages, HIV-1 assembles into an intracellular plasma
membrane domain containing the tetraspanins CD81, CD9, and CD53.
J. Cell Biol. 177:329–341.
12. Derdeyn, C. A., J. M. Decker, J. N. Sfakianos, X. Wu, W. A. O’Brien, L.
Ratner, J. C. Kappes, G. M. Shaw, and E. Hunter. 2000. Sensitivity of human
immunodeficiency virus type 1 to the fusion inhibitor T-20 is modulated by
coreceptor specificity defined by the V3 loop of gp120. J. Virol. 74:8358–
8367.
13. Escola, J. M., M. J. Kleijmeer, W. Stoorvogel, J. M. Griffith, O. Yoshie, and
H. J. Geuze. 1998. Selective enrichment of tetraspan proteins on the internal
vesicles of multivesicular endosomes and on exosomes secreted by human
B-lymphocytes. J. Biol. Chem. 273:20121–20127.
14. Fei, W., G. Shui, B. Gaeta, X. Du, L. Kuerschner, P. Li, A. J. Brown, M. R.
Wenk, R. G. Parton, and H. Yang. 2008. Fld1p, a functional homologue of
human seipin, regulates the size of lipid droplets in yeast. J. Cell Biol.
180:473–482.
15. Finnegan, C. M., S. S. Rawat, E. H. Cho, D. L. Guiffre, S. Lockett, A. H.
Merrill, Jr., and R. Blumenthal. 2007. Sphingomyelinase restricts the lateral
diffusion of CD4 and inhibits human immunodeficiency virus fusion. J. Virol.
81:5294–5304.
16. Finnegan, C. M., S. S. Rawat, A. Puri, J. M. Wang, F. W. Ruscetti, and R.
Blumenthal. 2004. Ceramide, a target for antiretroviral therapy. Proc. Natl.
Acad. Sci. USA 101:15452–15457.
17. Finzi, A., A. Orthwein, J. Mercier, and E. A. Cohen. 2007. Productive human
immunodeficiency virus type 1 assembly takes place at the plasma membrane. J. Virol. 81:7476–7490.
18. Freed, E. O., G. Englund, and M. A. Martin. 1995. Role of the basic domain
of human immunodeficiency virus type 1 matrix in macrophage infection.
J. Virol. 69:3949–3954.
19. Gambhir, A., G. Hangyas-Mihalyne, I. Zaitseva, D. S. Cafiso, J. Wang, D.
Murray, S. N. Pentyala, S. O. Smith, and S. McLaughlin. 2004. Electrostatic
sequestration of PIP2 on phospholipid membranes by basic/aromatic regions
of proteins. Biophys. J. 86:2188–2207.
20. Goff, S. P. 2007. Host factors exploited by retroviruses. Nat. Rev. Microbiol.
5:253–263.
21. Grassme, H., A. Jekle, A. Riehle, H. Schwarz, J. Berger, K. Sandhoff, R.
Kolesnick, and E. Gulbins. 2001. CD95 signaling via ceramide-rich membrane rafts. J. Biol. Chem. 276:20589–20596.
11237
22. Green, S. A., K. P. Zimmer, G. Griffiths, and I. Mellman. 1987. Kinetics of
intracellular transport and sorting of lysosomal membrane and plasma membrane proteins. J. Cell Biol. 105:1227–1240.
23. Heo, W. D., T. Inoue, W. S. Park, M. L. Kim, B. O. Park, T. J. Wandless, and
T. Meyer. 2006. PI(3,4,5)P3 and PI(4,5)P2 lipids target proteins with polybasic clusters to the plasma membrane. Science 314:1458–1461.
24. Itoh, T., and P. De Camilli. 2006. BAR, F-BAR (EFC) and ENTH/ANTH
domains in the regulation of membrane-cytosol interfaces and membrane
curvature. Biochim. Biophys. Acta 1761:897–912.
25. Ivanova, P. T., B. A. Cerda, D. M. Horn, J. S. Cohen, F. W. McLafferty, and
H. A. Brown. 2001. Electrospray ionization mass spectrometry analysis of
changes in phospholipids in RBL-2H3 mastocytoma cells during degranulation. Proc. Natl. Acad. Sci. USA 98:7152–7157.
26. Jolly, C., and Q. J. Sattentau. 2007. Human immunodeficiency virus type 1
assembly, budding, and cell-cell spread in T cells take place in tetraspaninenriched plasma membrane domains. J. Virol. 81:7873–7884.
27. Jouvenet, N., S. J. Neil, C. Bess, M. C. Johnson, C. A. Virgen, S. M. Simon,
and P. D. Bieniasz. 2006. Plasma membrane is the site of productive HIV-1
particle assembly. PLoS Biol. 4:e435.
28. Lehmann, M. J., N. M. Sherer, C. B. Marks, M. Pypaert, and W. Mothes.
2005. Actin- and myosin-driven movement of viruses along filopodia precedes their entry into cells. J. Cell Biol. 170:317–325.
29. Mason, P. W., and B. S. Jacobson. 1985. Isolation of the dorsal, ventral and
intracellular domains of HeLa cell plasma membranes following adhesion to
a gelatin substrate. Biochim. Biophys. Acta 821:264–276.
30. Mattila, P. K., A. Pykalainen, J. Saarikangas, V. O. Paavilainen, H. Vihinen,
E. Jokitalo, and P. Lappalainen. 2007. Missing-in-metastasis and IRSp53
deform PI(4,5)P2-rich membranes by an inverse BAR domain-like mechanism. J. Cell Biol. 176:953–964.
31. Megha, and E. London. 2004. Ceramide selectively displaces cholesterol
from ordered lipid domains (rafts): implications for lipid raft structure and
function. J. Biol. Chem. 279:9997–10004.
32. Merrill, A. H., Jr., M. C. Sullards, J. C. Allegood, S. Kelly, and E. Wang.
2005. Sphingolipidomics: high-throughput, structure-specific, and quantitative analysis of sphingolipids by liquid chromatography tandem mass spectrometry. Methods 36:207–224.
33. Morita, E., and W. I. Sundquist. 2004. Retrovirus budding. Annu. Rev. Cell
Dev. Biol. 20:395–425.
34. Mujawar, Z., H. Rose, M. P. Morrow, T. Pushkarsky, L. Dubrovsky, N.
Mukhamedova, Y. Fu, A. Dart, J. M. Orenstein, Y. V. Bobryshev, M. Bukrinsky, and D. Sviridov. 2006. Human immunodeficiency virus impairs reverse
cholesterol transport from macrophages. PLoS Biol. 4:e365.
35. Murray, P. S., Z. Li, J. Wang, C. L. Tang, B. Honig, and D. Murray. 2005.
Retroviral matrix domains share electrostatic homology: models for membrane binding function throughout the viral life cycle. Structure 13:1521–
1531.
36. Nguyen, D. H., and J. E. Hildreth. 2000. Evidence for budding of human
immunodeficiency virus type 1 selectively from glycolipid-enriched membrane lipid rafts. J. Virol. 74:3264–3272.
37. Nydegger, S., S. Khurana, D. N. Krementsov, M. Foti, and M. Thali. 2006.
Mapping of tetraspanin-enriched microdomains that can function as gateways for HIV-1. J. Cell Biol. 173:795–807.
38. Ono, A., S. D. Ablan, S. J. Lockett, K. Nagashima, and E. O. Freed. 2004.
Phosphatidylinositol (4,5) bisphosphate regulates HIV-1 Gag targeting to
the plasma membrane. Proc. Natl. Acad. Sci. USA 101:14889–14894.
39. Ono, A., and E. O. Freed. 2005. Role of lipid rafts in virus replication. Adv.
Virus Res. 64:311–358.
40. Ono, A., A. A. Waheed, and E. O. Freed. 2007. Depletion of cellular cholesterol inhibits membrane binding and higher-order multimerization of human
immunodeficiency virus type 1 Gag. Virology 360:27–35.
41. Ott, D. E., S. M. Nigida, Jr., L. E. Henderson, and L. O. Arthur. 1995. The
majority of cells are superinfected in a cloned cell line that produces high
levels of human immunodeficiency virus type 1 strain MN. J. Virol. 69:2443–
2450.
42. Pessin, J. E., and M. Glaser. 1980. Budding of Rous sarcoma virus and
vesicular stomatitis virus from localized lipid regions in the plasma membrane of chicken embryo fibroblasts. J. Biol. Chem. 255:9044–9050.
43. Quigley, J. P., D. B. Rifkin, and E. Reich. 1971. Phospholipid composition of
Rous sarcoma virus, host cell membranes and other enveloped RNA viruses.
Virology 46:106–116.
44. Quigley, J. P., D. B. Rifkin, and E. Reich. 1972. Lipid studies of Rous
sarcoma virus and host cell membranes. Virology 50:550–557.
45. Rajendran, L., and K. Simons. 2005. Lipid rafts and membrane dynamics.
J. Cell Sci. 118:1099–1102.
46. Rousso, I., M. B. Mixon, B. K. Chen, and P. S. Kim. 2000. Palmitoylation of
the HIV-1 envelope glycoprotein is critical for viral infectivity. Proc. Natl.
Acad. Sci. USA 97:13523–13525.
47. Rudner, L., S. Nydegger, L. V. Coren, K. Nagashima, M. Thali, and D. E.
Ott. 2005. Dynamic fluorescent imaging of human immunodeficiency virus
type 1 Gag in live cells by biarsenical labeling. J. Virol. 79:4055–4065.
48. Rusten, T. E., and H. Stenmark. 2006. Analyzing phosphoinositides and their
interacting proteins. Nat. Methods 3:251–258.
11238
CHAN ET AL.
49. Saad, J. S., J. Miller, J. Tai, A. Kim, R. H. Ghanam, and M. F. Summers.
2006. Structural basis for targeting HIV-1 Gag proteins to the plasma membrane for virus assembly. Proc. Natl. Acad. Sci. USA 103:11364–11369.
50. Shaw, A. S. 2006. Lipid rafts: now you see them, now you don’t. Nat.
Immunol. 7:1139–1142.
51. Sherer, N. M., M. J. Lehmann, L. F. Jimenez-Soto, A. Ingmundson, S. M.
Horner, G. Cicchetti, P. G. Allen, M. Pypaert, J. M. Cunningham, and W.
Mothes. 2003. Visualization of retroviral replication in living cells reveals
budding into multivesicular bodies. Traffic 4:785–801.
52. Shui, G., A. K. Bendt, K. Pethe, T. Dick, and M. R. Wenk. 2007. Sensitive
profiling of chemically diverse bioactive lipids. J. Lipid Res. 48:1976–1984.
53. Stolz, D. B., and B. S. Jacobson. 1992. Examination of transcellular membrane protein polarity of bovine aortic endothelial cells in vitro using the
cationic colloidal silica microbead membrane-isolation procedure. J. Cell
Sci. 103:39–51.
54. Trajkovic, K., C. Hsu, S. Chiantia, L. Rajendran, D. Wenzel, F. Wieland, P.
Schwille, B. Brugger, and M. Simons. 2008. Ceramide triggers budding of
exosome vesicles into multivesicular endosomes. Science 319:1244–1247.
55. Trubey, C. M., E. Chertova, L. V. Coren, J. M. Hilburn, C. V. Hixson, K.
Nagashima, J. D. Lifson, and D. E. Ott. 2003. Quantitation of HLA class II
protein incorporated into human immunodeficiency type 1 virions purified
by anti-CD45 immunoaffinity depletion of microvesicles. J. Virol. 77:12699–
12709.
56. van Meer, G. 2005. Cellular lipidomics. EMBO J. 24:3159–3165.
J. VIROL.
57. Welsch, S., O. T. Keppler, A. Habermann, I. Allespach, J. Krijnse-Locker,
and H. G. Krausslich. 2007. HIV-1 buds predominantly at the plasma membrane of primary human macrophages. PLoS Pathog. 3:e36.
58. Wenk, M. R., L. Lucast, G. Di Paolo, A. J. Romanelli, S. F. Suchy, R. L.
Nussbaum, G. W. Cline, G. I. Shulman, W. McMurray, and P. De Camilli.
2003. Phosphoinositide profiling in complex lipid mixtures using electrospray
ionization mass spectrometry. Nat. Biotechnol. 21:813–817.
59. Wubbolts, R., R. S. Leckie, P. T. Veenhuizen, G. Schwarzmann, W. Mobius,
J. Hoernschemeyer, J. W. Slot, H. J. Geuze, and W. Stoorvogel. 2003. Proteomic and biochemical analyses of human B cell-derived exosomes. Potential implications for their function and multivesicular body formation. J. Biol.
Chem. 278:10963–10972.
60. Yeung, T., M. Terebiznik, L. Yu, J. Silvius, W. M. Abidi, M. Philips, T.
Levine, A. Kapus, and S. Grinstein. 2006. Receptor activation alters inner
surface potential during phagocytosis. Science 313:347–351.
61. Yu, C., M. Alterman, and R. T. Dobrowsky. 2005. Ceramide displaces cholesterol from lipid rafts and decreases the association of the cholesterol
binding protein caveolin-1. J. Lipid Res. 46:1678–1691.
62. Zheng, Y. H., A. Plemenitas, T. Linnemann, O. T. Fackler, and B. M.
Peterlin. 2001. Nef increases infectivity of HIV via lipid rafts. Curr. Biol.
11:875–879.
63. Zimmerberg, J., and M. M. Kozlov. 2006. How proteins produce cellular
membrane curvature. Nat. Rev. Mol. Cell Biol. 7:9–19.