Lipids in Viral Fusion 61
From: Methods in Molecular Biology, vol. 199: Liposome Methods and Protocols
Edited by: S. Basu and M. Basu © Humana Press Inc., Totowa, NJ
Lipids in Viral Fusion
Anu Puri, Maite Paternostre, and Robert Blumenthal
1.1. Viral Fusion Process
Enveloped animal viruses infect host cells by fusion of viral and target
membranes. This crucial fusion event occurs either with the plasma membrane
of the host cells or with the endosomal membranes (1). Fusion is triggered by
specific glycoproteins in the virus membrane (2) and involves a range of steps
before the final merging of membranes occurs. These steps include molecular
processes, such as envelope protein conformational changes; aggregation of
envelope protein; lipid–envelope protein interactions; and fusion pore forma-
tion and pore widening (see Fig. 1) (3,4). The reader is referred to a number of
reviews on viral glycoprotein-mediated membrane fusion (5–9).
Because lipids are integral elements in the membrane fusion process, a num-
ber of key issues concerning the role of lipids deserve thorough examination.
1.2. Contribution of Physicochemical Properties
of Various Membrane Lipids to Promote Fusion
The ability of membrane lipids to bend into curved structures is a crucial
property in the fusion process. Chernomordik and co-workers (10) have shown
that addition of amphiphiles to the outer monolayer that promote (e.g., oleic
acid) or inhibit (e.g., lysophosphatidylcholine [lysoPC]) negative curvature
respectively, will promote or inhibit the formation of the initial lipid junction.
Conversely, amphiphiles added to the inner monolayer that promote positive
curvature (e.g., lysoPC) will promote the formation of the fusion pore.
62 Puri, Paternostre, and Blumenthal
1.3. Effects of Target Membrane Lipid Asymmetry on Viral Fusion
The human erythrocyte plasma membrane is the most highly characterized
system with respect to lipid asymmetry. In the human erythrocyte membrane,
the aminophospholipids phosphatidylserine (PS) and phosphatidylethanolamine
(PE) are preferentially located on the inner monolayer, while zwitterionic
lipids such as PC and sphingomyelin (SM) are on the outer layer. Human
erythrocyte ghosts can be manipulated such that their lipids are distributed
either asymmetrically or symmetrically between inner and outer monolayers
(11). We found that fusion of vesicular stomatitis virus (VSV) with lipid-
symmetric erythrocyte ghosts was rapid at low pH and 37°C, whereas little
or no fusion was observed with lipid-asymmetric ghosts. Biophysical studies
indicate that the susceptibility to VSV fusion is not dependent on any particular
phospholipid but is rather related to packing characteristics of the outer leaflet
of the target membrane (11).
1.4. Requirement of Specific Lipids as Viral Receptors/Fusogens
in Viral Fusion
A good example is Semliki Forest virus (SFV), which requires both choles-
terol (12) and sphingomyelin (13). Cholesterol is required for binding, whereas
SM acts as a cofactor, possibly through activation of the viral fusion protein
(14). Specific glycosphingolipids (GSLs) are also known to serve as receptors
for viruses such as influenza (15) and Sendai (16).
1.5. Modulation of Fusion by GSLs in the Target Membrane
Evidence for this concept is based on the fact that inhibitors of GSL
biosynthesis affect human immunodeficiency virus type 1 (HIV-1) infection
and fusion, and that fusion activity can be recovered following addition of
Fig. 1. A cartoon depicting various fusion intermediates following structural
changes in viral proteins. This cartoon taken from Blumenthal et al. (3) shows fusion
intermediates after virus binds to the target cells.
Lipids in Viral Fusion 63
purified GSLs to the fusion-impaired cells (17,18). Studies in reconstituted
monolayers of purified GSL at the air–water interface showed evidence for
CD4-induced interactions between HIV-1 gp120 and certain GSLs (Gb3 and
1.6. Methods to Study the Role of Membrane Lipids in Viral Fusion
The study of the role of lipids in fusion is not as straightforward as that
of viral envelope glycoproteins and their protein receptors. Once the proteins
have been cloned they can readily be expressed on the cell surface and their
structure and function can be studied in the context of the fusion reaction. To
examine the role of viral lipids more elaborate strategies have been developed.
1. Amphiphiles (e.g., lysoPC, fatty acids) canbe directly added to target membranes,
or target membranes can be modified with phospholipases to produce these
amphiphiles in situ (21). Although these methods have been successful in
studying the role of membrane curvature in viral fusion, issues of partitioning,
metabolism, and amount incorporated into the membrane have to be carefully
2. Liposomes with defined lipid (phospholipid, cholesterol, and/or sphingolipid)
composition can be used as targets for membrane fusion. However, in lipid
mixing assays, nonspecific effects (unrelated to the known behavior of the
viral envelope glycoprotein) have been noted (23). A number of low-pH fusing
viruses (such as influenza, SFV, VSV) have been shown to fuse efficiently with
liposomes with appropriate controls. A neutral pH fusing virus, Sendai fuses
with liposomes, but has also been shown to fuse with greater efficiency with
biological targets (24). Attempts to monitor viral envelope glycoprotein-specific
fusion of other neutral pH-fusing viruses such as murine leukemia virus (MuLV)
and HIV with liposomes have been unsuccessful.
3. Cholesterol levels in biological membranes have been altered by using agents
such as methyl-?-cyclodextrin (25). Cell lines have also been selected with defec-
tive mobilization of cholesterol from the plasma membrane to the endoplasmic
reticulum (26). Recent experiments from our laboratory indicate that cholesterol
removal from target membranes significantly reduces their susceptibility to
HIV-1 envelope glycoprotein-mediated fusion (M. Viard et al., unpublished
observations). It is also possible to alter target membrane GSL levels by treatment
of cells with inhibitors of GSL biosynthesis (18,37).
During the past few years, we have developed assays to monitor viral fusion
in lipid-modified biological membranes. We describe in detail the following
two methods of lipid addition to biological targets: (1) lipid incorporation into
plasma membrane vesicles (PMVs) by detergent solubilization methods; (2)
addition of GSL to nucleated cells by influenza hemagglutinin (HA)-mediated
64 Puri, Paternostre, and Blumenthal
fusion of liposomes containing specific GSLs. We also describe ways to monitor
fusion of lipid-modified targets with viral envelope-glycoprotein-expressing
cells. Finally we discuss the technical limitations encountered to incorporate
natural lipids to the membranes of cells and possible experimental approaches
to overcome these difficulties.
1. Fluorescent probes were from Molecular Probes (Eugene, OR).
2. Phospholipids were purchased from Avanti Polar Lipids (Alabaster, AL).
3. 10× Trypsin–EDTA: 0.5% Trypsin, 5.3 mM EDTA 4Na.
4. 100× Penn-Strep: 10,000 U/mL of penicillin G, 10,000 μg/mL of streptomycin
sulfate in 0.85% saline.
5. 100× Penn-Strep-Gln: 10,000 U/mL of penicillin G sodium,10,000 μg/mL of
streptomycin sulfate, and 29.2 mg of L-glutamine/mL in 0.85% saline.
6. Fetal bovine serum (FBS): Heat inactivated, mycoplasma tested (Life Technolo-
gies, cat. no. 16140-055).
7. RPMI: RPMI-1640 + L-glutamine (1×).
8. DMEM: Dulbecco’s modified Eagle medium with 4.5 g of glucose per liter,
9. Phosphate-buffered saline (PBS): 137 mMNaCl, 2.7 mMKCl, 8.1 mMNa2HPO4,
1.5 mM KH2PO4, pH 7.4 (Life Technologies, cat. no. 14190-144).
10. D-PBS: Dulbecco’s PBS containing 2 mM Ca2+ and 2 mM Mg2+.
11. D2: DMEM containing 2% FBS.
12. D10: DMEM supplemented with 10% heat-inactivated FBS, 100 U/mL of
penicillin, 100 μg/mL of streptomycin (Life Technologies, Custom Formula
13. D5: 500 mL of DMEM containing 10 mL of G418 sulfate solution (50 mg/mL
GENETICIN aqueous solution, Life Technologies cat. no. 10131-027), 25 mL
of FBS, and 5 mL of Penn-Strep-Gln.
14. R10: 500 mL of RPMI-1640 containing 50 mL of FBS and 5 mL of Penn-Strep
15. HEPES-NaCl buffer: 10 mM 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic
acid (HEPES), 145 mM NaCl, pH 7.4, filter sterile through a 0.2-μm filter.
16. 1× Trypsin–EDTA: 20 mL of trypsin–EDTA (10×) is diluted with 180 mL of
PBS. It is recommended to prepare the solution as needed but it can be stored
at 4°C for about 2 wk.
17. Homogenization buffer: The quantities are given for 1-L solution. 1.42 g (10 mM,
mol wt 142) Na2HPO4, 203 mg (1 mM, mol wt 203) MgCl2•6H2O, 1.753 g
(30 mM, mol wt 58.44) NaCl, 154 mg (1 mM, mol wt 154) dithiothreitol (DTT),
0.87 mg (0.005 mM, mol wt 174.2) phenylmethylsulfonyl fluoride (PMSF); 200 mg
(0.02%) NaN3. NaN3may be omitted depending on the experimental set up.
Addition of a few micrograms of DNase is recommended to avoid precipitation
Lipids in Viral Fusion 65
of DNA during the homogenization step. Homogenization buffer is filtered sterile
through a 0.22-μm filter and stored at 4°C.
18. Mink-CD4 cells (27), a gift from Dr. Paul Clapham, Chester Beatty Institute,
London, were grown in D5.
19. TF228.1.16 cells that constitutively express HIV-1 gp120-gp41 (28) (from
Zdenka L. Jonak, Smithkline Beecham Pharmaceuticals, PA) were grown in
20. Vero cells (an African green monkey kidney cell line) and HeLa cells (a human
cervix epitheloid carcinoma cell line) were obtained from American Type Culture
Collection (Rockville, MD) and grown in D10.
21. Purified VSV (Indiana serotype) was prepared by J. Brown and B. Newcomb at
the University of Virginia as described (29).
22. Influenza virus (A/PR8/34/H1N1 strain) was grown in 11-d embryonated eggs
as described previously (30). One-milliliter aliquots of influenza virus (allantoic
fluid) were stored at –70°C.
23. PD10 column (Pharmacia)
24. Trypsin: 1 mg/mL stock in PBS. Use tissue culture tested (Sigma).
3.1. Incorporation of Cardiolipin into the PMVs
3.1.1. Preparation of PMVs (35)
1. Vero cells are grown in D10 to nearly 90% confluency in six to eight T150
tissue culture flasks.
2. The cells are harvested using 1× trypsin–EDTA. Each flask is treated as follows:
culture medium is removed, and 15 mL of 1× trypsin–EDTA solution is added,
rinsed, and trypsin solution is aspirated. This step is repeated and 4 mL of 1×
trypsin–EDTA solution is added to the flask and incubated at 37°C for 8–12 min.
The cells should be detached from the flask as seen under a microscope. Sixteen
milliliters of D10 medium is added to the cells, mixed by pipetting several
times, and transferred to a 50-mL centrifuge tube. The cells are pelleted by
centrifugation at 800–1200 rpm for 5–10 min at 4°C.
3. The cell pellets are combined into a 50-mL centrifuge tube and washed by
resuspending in 30–40 mL of PBS, followed by centrifugation as described in
step 2. The supernatant is removed and this step is repeated again. The cell
numbers are counted at this step using a cell counting chamber.
4. The cells are centrifuged and the pellet is resuspended in cold homogenization
buffer at a concentration of 1–2 × 108cells/mL and homogenized using an ice
bath by means of a precooled Teflon-coated homogenizer (10–15-mL capacity).
Usually 10–30 strokes are sufficient for complete lysis of cells.
5. An aliquot is tested for efficient lysis of cells as follows: 20 μL of lysed cells
are mixed with an equal volume of Trypan blue (0.4% solution in sterile PBS,
66 Puri, Paternostre, and Blumenthal
stored at room temperature) and observed under the microscope. Maximum lysis
is determined by uptake of dye by the cells. (Any dye that can stain live/dead
cells can be used at this step.)
6. The nuclei are removed by centrifugation at 500g for 10 min at 4°C. The
supernatant (which contains PMV) is transferred to another tube, adjusting the
volume to 3 mL by addition of cold homogenization buffer.
7. One milliliter of 41% sucrose solution (prepared in sterile distilled water) is
placed in a 4-mL capacity centrifuge tube (for Beckman Rotor SW60), 3 mL of
supernatant (from step 6) is gently layered on the 1-mL 41% sucrose cushion,
and the solution centrifuged at 100,000g for 1 h at 4°C.
8. The tubes are carefully removed from the rotor buckets. A cloudy layer of PMV
should be visible above the 41% sucrose layer. The clear supernatant just above
the PMV layer is discarded using a pasteur pipet. The PMVs are collected and
carefully transferred to a 1-mL tube (volume of PMV at this step is approx
0.5 mL). The samples are dialyzed for 6–8 h at 4°C against 100–200 volumes
of cold PBS to remove most of the sucrose from the PMV preparation. The
protein content of PMV is determined using the Protein-BCA reagent kit (Pierce,
Rockford, IL). Typically, a yield of 2.0–2.5 mg of PMV protein is obtained. The
yield depends on the type of cells used for preparing the PMVs. At this step
PMVs can be aliquoted and stored at –20°C.
3.1.2. Solubilization and Reconstitution of PMV (Reconstituted PMV
[R-PMV] and Cardiolipin Incorporated Reconstituted PMV [CL-R-PMV])
The solubilization process by detergents consists in the progressive disrup-
tion of the membrane in favor of mixed micellar structure. Detergents are
soluble amphiphiles that form soluble aggregates (micelles) when their total
concentration is higher than their critical micellar concentration (CMC). The
solubilization process depends on choice of detergent, the temperature, ionic
strength of the solution, as well as nature and concentration of the lipids and the
proteins present in the membrane (31–33). Therefore, it is important to control
these parameters to determine the exact detergent concentration required for
solubilization. The simplest way to monitor the solubilization process is to
record the evolution of the turbidity of an initial membrane suspension on
detergent addition. Indeed, this process is characterized by a change in the
organization of the amphiphiles, which leads to important changes in the size
and shape of the amphiphile aggregates. Natural membranes such as PMV
consist of “big” aggregates of a few hundred nanometers whereas mixed
micelles of detergents, lipids, and proteins consist of “small” aggregates
of about few tens of nanometers. Therefore, these aggregates scatter light
differently (i.e., membranes scatter light much more than mixed micelles)
and the solubilization process can be visualized by a sharp drop in turbidity
Lipids in Viral Fusion 67
220.127.116.11. TURBIDITY MEASUREMENTS
1. PMV (from Subheading 3.1.1., step 8) is resuspended at a typical concentration
of 1.3 mg/mL in PBS for the protein, which corresponds to 1.0 mg/mL for the
lipids based on a published analysis of the plasma membrane fraction (35).
2. A 200 mM solution of octyl glucoside (OG, mol wt 292.4) is prepared in PBS.
OG solution can be stored at 4°C for at least a month.
3. For the turbidity experiment, any spectrophotometer ( UV-Visible or spectro-
fluorimeter) can be used to measure the optical density of the sample. To record
turbidity changes during solubilization the wavelengths are chosen in a range in
which no light absorption is recorded (e.g., at 450 nm for the PMV). A quartz
cuvette (1 cm optical length) is placed in the spectrofuorimeter (SLM8000)
ex/em 450 nm) and 0.4 mL of PMVs + 1.6 mL of PBS are added to the cuvette
to final PMV protein and lipid concentrations of approx 0.26 mg/mL and
0.2 mg/mL, respectively.
4. The sample is maintained at 25°C and stirred continuously with a Teflon-coated
stir bar (2 × 7 mm). Small aliquots (10 μL) of the OG solution are added to
the PMV suspension and after each OG addition, the signal is recorded until
it is completely stable. (This stabilization may take a few seconds or a few
minutes depending on the concentration.) In this experiment, 20–26 mM final
OG concentration was sufficient for complete solubilization of PMVs. It is
recommended to optimize the choice and the concentration of the detergent for
any protein–lipid mixture to be solubilized.
5. The solubilization curve is then obtained by plotting the turbidity of the suspen-
sion after each detergent addition and after stabilization as a function of the total
detergent concentration in the cuvette (Fig. 2).
6. To remove insoluble material, the sample is transferred from the quartz cuvette
to a centrifuge tube and centrifuged in a Beckman Ti50 rotor, 60 min, 4ºC at
7. After centrifugation, the supernatant is collected and saved for detergent
removal without or with addition of exogeneous lipid, cardiolipin (see below).
The turbidity of the supernatant after centrifugation is about 10 times lower
than the turbidity recorded at the end of the solubilization experiment, indicat-
ing that some insoluble materials have been eliminated from the sample by
18.104.22.168. GENERATION OF R-PMV BY DETERGENT REMOVAL USING BIO BEADS
The PMVs are reconstituted by removal of detergent. This can be easily
achieved by using Bio Beads SM2 (BB SM2, cat. no. 152-3920, Bio-Rad,
Hercules, CA). BB SM2 are small hydrophobic beads, with a high surface
capacity, that absorb any amphiphilic molecules, particularly detergents. One
must be careful, however, because some of the solubilized lipids and proteins
can also be removed by using this technique (31,32,36). Therefore, successive
68 Puri, Paternostre, and Blumenthal
steps using small quantities of BB SM2 are preferred for detergent elimination.
For each 200 μg vesicle lipid sample, the following steps are used:
1. Ten glass vials (3–5 mL capacity) with caps containing 2 × 5 mm stir bars are
prepared and labeled as 1–10. Twenty milligrams of BB SM2 (wet wt) are placed
in each vial and 1 mL of PBS is added.
2. PBS is removed from vial 1 (containing BB SM2) and 1 mL of solubilized PMV
supernatant (from Subheading 3.1.1., step 8) is added. The sample is incubated
with BB SM2 for 12 min at 25°C under gentle magnetic stirring. At the end of
incubation, stirring is stopped to allow the BB SM2 to sediment. The PBS from
vial 2 is removed and the sample from vial 1 is transferred to vial 2. This process
is repeated for the remaining vials for complete detergent removal.
22.214.171.124. INCORPORATION OF EXOGENEOUS LIPID CARDIOLIPIN
INTO PMV (CL-R-PMV)
1. Fifty micrograms of cardiolipin (CL, Avanti Polar Lipids, cat. no. 840012) is
dried in a 5-mL glass tube under a stream of N2and the tube is placed in a vacuum
desiccator for 12–24 h to ensure removal of traces of solvent.
2. One milliliter of solubilized PMV supernatant (from Subheading 3.1.1., step 8)
is added to the tube containing CL.
3. The samples are incubated at room temperaturewith gentle shaking for 30–60 min
to allow complete solubilization of CL.
Fig. 2. Determination of OG concentration to solubilize PMVs.
Lipids in Viral Fusion 69
We use a ratio of 50 μg CL for 200 μg plasma membrane lipid in our
reconstitution procedure. This ratio can be modified and any exogeneous lipid
can be reconstituted in PMVs at this step. Control samples are prepared in the
same manner except that CL is omitted. Removal of detergent using Bio Beads
is performed as described in Subheading 126.96.36.199.
188.8.131.52. LABELING OF PMV WITH OCTADECYL RHODAMINE. The method described
is this subheading is used for octadecyl rhodamine (R18) labeling of PMV,
R-PMV, or CL-R-PMV.
1. A stock solution of R18 (cat. no. 0-246, Molecular Probes, Eugene, OR) is
prepared in absolute alcohol in a glass vial at 1 mg/mL and stored at –20°C
in small aliquots.
2. Ten to fifty microliters of R18 is added to 1 mL of PMV preparation (containing
approx 200 μg of vesicle protein from Subheadings 3.1.1., step 8; 3.1.2., step 2;
or 3.1.2., step 3) using a Hamilton Syringe under constant vortex mixing and the
mixture is incubated for 30 min at room temperature in the dark.
3. R18-labeled PMVs are centrifuged on a 10–41% sucrose gradient at 100,000g
for 60 min at 4°C as described in Subheading 3.3.1.
4. R18-labeled PMVs are collected from the 41% sucrose layer. This step is crucial
as it eliminates any unincorprated R18 on the 10% sucrose layer which can lead
to nonspecific effects in fusion assays.
5. R18-labeled PMV are passed through a PD10 column using D-PBS as an eluant.
Fluorescent PMVs are collected in a total volume of 1 mL.
6. R18-labeled PMV obtained at this step are used as targets to monitor fusion with
VSV-G expressing cells (see Subheading 3.3.1.).
3.2. Incorporation of GSLs into the Membranes of Mink-CD4 Cells
We have described the method of GSL transfer for GSL extracted from
human (GSL-Hu) or bovine erythrocytes (GSL-Bov)(17). The method for GSL
transfer is applicable to a number of commercially available GSL (18,37).
3.2.1. Preparation of GSL-Containing Liposomes
1. The following components are mixed in a 7-mL glass tube: 3.0 mg of egg
PC (0.6 mL, from a 5 mg/mL stock in CHCl3, Avanti Polar Lipids, cat. no.
840051), 1.5 mg of egg PE (0.3 mL, from a 5 mg/mL stock in CHCl3, Avanti
Polar Lipids, cat. no. 840021), 1.0 mg of GSL (1 mL, from a 1 mg/mL stock
in CHCl3–methanol (2?1).
2. The solvents are removed under a stream of N2and a thin lipid film is formed
on the wall of the glass tube. The lipid film is dried in a vacuum desiccator
for 12–24 h.
3. Three milliliters of sterile PBS is added to the tube and vortex-mixed intermit-
tently for 5–10 min at room temperature. This step results in the formation of
multilamellar vesicles (MLVs).
70 Puri, Paternostre, and Blumenthal
4. An additional 3 mL of PBS is added to the MLVs, bringing the final volume
to 6.0 mL (0.9 mg of lipid/mL). The lipid composition is PC–PE–GSL
(4.5?2.25?1.5, by wt).
6. The MLVs are transferred to a 15-mL tube (round-bottom polypropylene, Falcon
cat. no. 2059) and subjected to at least five freeze–thaw cycles to generate a
homogeneous preparation of MLVs. At this step, MLVs can be aliquoted and
stored at –20°C until further use.
7. To generate unilamellar liposomes, MLVs are extruded through a 0.2-μm filter
using an Extruder (Lipex Biomembranes, Inc., Vancouver, BC) to yield vesicles
with a diameter of about 200 nm. The extrusion process is repeated at least three
to five times. Resulting liposomes stored at 4°C can be used within 24–36 h.
3.2.2. Infl uenza (PR/8) Infection of Mink-CD4 Cells and Fusion
Efficiency of influenza infection depends on the cell type and the strain of
influenza used. Therefore, it is recommended to determine efficiency of influenza
infection when a new virus preparation or a different cell line is used.
1. Mink-CD4 cells are plated on microwells (105 per dish, 14 mm coverslips no. 0,
MatTek Corp., Ashland, MA) overnight at 37°C.
2. The influenza virus is thawed, diluted (50 μL of PR/8 allantoic fluid/mL of D2)
and vortex-mixed briefly to remove any virus aggregates.
3. Diluted influenza virus is added to the coverslip (0.5 mL/dish) and incubated
1–2 h at 37°C. Unbound virus is removed, 1 mL of D5 is added, and the mixture
incubated 5–6 h at 37°C.
184.108.40.206. ACTIVATION OF INFLUENZA HAO TO FUSION ACTIVE HA
BY TRYPSIN TREATMENT
Because HA is expressed on the surface of cells as an uncleaved precursor
(HAo), the cells are treated with 5 μg/mL of trypsin for 5 min at room
temperature to convert HAo to fusogenic HA. This procedure involves the
following steps (38):
1. Trypsin, 1 mg/mL stock in PBS (or water),is prepared and stored in small aliquots
at –20°C. A working dilution 1?200 in D-PBS is made to a final concentration
of 5 μg of trypsin/mL.
2. Each dish is washed three times with D-PBS (1 mL each time) to completely
remove D5 from the wells.
3. One milliliter of diluted trypsin is added to each dish and the dish is incubated
5 min at room temperature. The cells are checked after 3 min; if they look healthy,
incubation is continued 2 min more; otherwise trypsin is removed at this point and
1 mL of D5 is added. Control wells are treated in parallel except trypsin is omitted
from the incubation mixture. Controls are designated as HA0 controls.
Lipids in Viral Fusion 71
220.127.116.11. LIPOSOME BINDING
1. Wheat germ agglutinin (WGA ) stock at 1 mg/mL in PBS is prepared and stored at
–20°C in small aliquots. A working solution is prepared by adding 160 μL of WGA
solution to 4 mL with D-PBS to a final concentration of 40 μg/mL of WGA.
2. Medium is removed from the wells. Twenty-five microliters of diluted WGA
solution is layered into the central well of each microwell.
3. Two hundred microliters of unilamellar liposomes (see Subheading 3.2.1.)
are added to the center of the microwell and incubated 20–30 min at room
temperature to promote binding of liposomes to the cells.
4. Liposome solution is aspirated from the central well, 1 mL of prewarmed
PBS (preadjusted to pH 5.1) is added, and the solution is incubated at 37°C
for only 1–2 min.
5. The PBS, pH 5.1, is aspirated immediately and replaced with 1 mL of D10, pH 7.4.
Incubation is continued in the culture medium at pH 7.4 for 20–30 min at room
temperature. GSL-supplemented cells are used for fusion with TF228 cells (see
3.3. Fusion of Lipid-Supplemented Biological Targets
with Viral Envelope Glycoprotein Expressing Cells
3.3.1. Kinetics of Fusion of R18-Labeled PMVs with VSV-G Expressing
18.104.22.168. Infection of HeLa cells with VSV
1. 3 × 106HeLa cells are plated in a T75 tissue culture flask 1 d prior to infection
2. VSV is diluted it into 2.0 mL of D-PBS (at 6 multiplicity of infection) and
3. Medium is removed from the flask and diluted VSV is added to HeLa cells on
T75 flasks. Virus and cells are incubated for 30–45 min at 37°C.
4. Unbound virus is removed, 10 mL of D10 added, and incubation continued at
37°C for 6 h. Infection for 6–8 h results in high expression of VSV-G on the
cell surface. At the end of the incubation, the cells are scraped off the flask, and
suspended cells are transferred to a 15-mL polypropylene centrifuge tube.
5. The cells are pelleted by centrifugation at 800–1200 rpm for 5 min. The pellet is
resuspended in 0.3 mL of D-PBS (approx 106cells/100 μL), and 100 μL aliquots
are transferred in to a 15-mL tube (×3).
22.214.171.124. BINDING OF R18-LABELED PMV TO VSV-G EXPRESSING CELLS
1. One milliliter of R18-labeled PMV preparation (approx 100 μg of vesicle protein,
fromSubheading 126.96.36.199) is added to 100 μL of cells (from Subheading 188.8.131.52.,
step 5) and the cells–PMV mixture is placed on ice and incubated for 30–40 min
with shaking in the dark.
72 Puri, Paternostre, and Blumenthal
2. After incubation, 10 mL of PBS is added to the tubes and the cells are pelleted.
The supernatant is discarded and cell pellet is resuspended in 100 μL of PBS.
184.108.40.206. KINETICS OF FUSION
1. Kinetics of fusion is monitored with an SLM8000 spectrofluorimeter. A cuvette
(all sides clear, suitable for fluorescence spectroscopy) containing a 2 × 7 mm
stir bar is placed into the fluorimeter; 2 mL of HEPES-NaCl buffer, pH 7.4 or at
desired preadjusted pH value, is added; the solution is equilibrated at the desired
temperature using a circulating water bath equipped with the fluorimenter (39).
2. Twenty to fifty microliters of vesicle-cell suspension is added to the cuvette
and fluorescence is measured at ex/em 560/590 nm with a 560 nm cutoff filter
at the emission.
3. For preset pH values, measurement is continued for the desired time.
4. For samples at pH 7.4, 10–100 μL of 0.5 M 4-morpholino ethanesulfonic acid
(MES) is added using a Hamilton syringe at 20 s to lower the pH. The amount of
MES required to obtain a desired pH is calibrated beforehand.
5. At the end of the run, 10 μL of an aqueous Triton X-100 (10% w/v solution
is added and fluorescence measurement is continued to until a steady signal is
achieved (usually 20–50 s).
6. Percentage of dequenching is calculated according to:
% dequenching = 100 × (F – Fo)/(Ft – Fo), (1)
where F, Fo, and Ftare fluorescence values at a given time, at zero time, and
after total dequenching by addition of Triton X-100 (0.05% final concentration),
Figure 3 (top panel) shows fusion of R-PMV with VSV-G-expressing HeLa
cells at 37°C. Fusion was monitored at the preset pH 6.66 for the indicated
time period and additional MES was added to lower the pH to 5.6 (indicated by
arrow). As can be seen, VSV-G-induced fusion kinetics of R18-labeled R-PMV
is similar to that previously observed with R18-labeled PMV (39). A slight
decrease in fusion observed with the samples preincubated at pH 6.66 (250 s)
was consistent with previously observed inactivation of fusion activity of intact
VSV (40). Therefore, treatment of PMV with OG does not impair fusion with
VSV-G-expressing cells. R-PMV display characteristics similar to intact cells
as targets for VSV-G mediated fusion. To ascertain incorporation of additional
components in OG-solubilized PMV, we tested the ability of a fusogenic lipid
CL to enhance fusion of R-PMV. Figure 3 (lower panel) shows fusion of
R18-labeled CL-R-PMV with VSV-G-expressing HeLa cells. As PMV fuse
efficiently with VSV-G-expressing cells at 37°C, pH 6.3 (39), we monitored
fusion of CL-R-PMV at the subthreshold temperature 32°C. Data presented in
Lipids in Viral Fusion 73
Fig. 3. Fusion of VSV-G expressing HeLa cells with R18-labeled PMVs. Cells
were infected with VSV at 6 m.o.i. for 6 h. R18-labeled PMVs were bound to cells at
4°C at pH 7.4 and unbound PMVs were removed by centrifugation. The vesicle-cell
suspension was added to preequilibrated buffer at various pH values. At the end of the
incubation, Triton X-100 was added and percentage fusion was calculated as described
in Subheading 3. (Top) Fusion of VSV-G expressing HeLa cells with R18-labeled
R-PMV at 37°C. The vesicle-cell suspension was added to preequilibrated buffer at pH
6.66. 0.5 M MES (50–70 μL) was added (indicated by arrow) to bring the pH of the
samples to 5.6. (Bottom) Fusion of VSV-G expressing HeLa cells with R18-labeled
CL-R-PMV at 32°C at pH 6.3.
74 Puri, Paternostre, and Blumenthal
Fig. 3(lower panel) show that incorporation of CL into R-PMV enhanced fusion
of PMV with VSV-G-expressing cells. Observed fluorescence dequenching
was specific, as we did not see any dequenching at pH 6.8 and neutral pH
(data not shown).
3.3.2. Fusion of GSL-Supplemented Mink-CD4 Cells with HIV-1 Envelope
Glycoprotein Expression Cells
220.127.116.11. LABELING OF MINK-CD4 CELLS WITH CMFDA
1. The cytoplasmic fluorescent probe 5-chloromethylfluorescein diacetate (CMFDA;
Molecular Probes cat. no. C-7025, ex/em 492/516 nm) is solubilized at 10 mM
concentration in dimethyl sulfoxide (DMSO), aliquoted into a 10- or 20-μL
volume, and stored at –20°C. A working dilution of CMFDA (1?500) is prepared
2. Medium is removed from GSL-containing mink-CD4 cells on microwells (from
Subheading 3.2.2.), 1 mL of diluted CMFDA solution is layered, and samples
are incubated at 37°C for 45–60 min. Dye solution is replaced with 1 mL of D10
and incubation continued for additional 15–30 min.
18.104.22.168. LABELING OF TF228 CELLS WITH 1,1′-DIOCTADECYL-DII
1. DiI (Molecular Probes cat. no. D282, ex/em 550/565 nm) is solubilized in 10 mM
DMSO, divided into small aliquots and stored at –20°C. Five microliters of
3,3,3′,3′-tetramethylindo-carbocyanine perchlorate (DiI) solution is added to
0.1 mL of Diluent C (Sigma, cat. no. CGL-DIL) in a 1.5-mL Eppendorf tube and
vortex-mixed (the dye solution should be clear solution without any insoluble
2. TF228 cells (5 × 106) are pelleted by centrifugation at 1000 rpm, and the
pellet resuspended in 10 mL of PBS and centrifuged again. The cell pellet is
resuspended in 0.5 mL of PBS.
3. One-half milliliter of cells (from step 2) are added to the DiI solution and mixed
quickly. The mixture is transferred to a fresh Eppendorf tube and incubated for
2–5 min at room temperature.
4. One milliliter of R10 is added to the mixture and the sample is centrifuged for
30 s in a microfuge. The supernatant is discarded and the cells resuspended in 5 mL
of R10 and incubated at room temperature for 5 min.
5. The cells are pelleted and washed three times with PBS (5 mL each wash), then
resuspended at 105/mL in R10.
22.214.171.124. CELL–CELL FUSION
1. Two milliliters of DiI-labeled TF228 cells (from Subheading 126.96.36.199., step 5)
are added to CMFDA-labeled mink-CD4 cells and the two cell populations
incubated for 3–5 h at 37°C.
Lipids in Viral Fusion 75
2. At the end of incubations, medium is replaced with 1 mL of D-PBS and the
phase and fluorescence images are acquired at room temperature using an
Olympus IX70 inverted microscope with a 40× oil immersion UPlanApo
objective (1.0 NA). We routinely use the cooled charge coupled device (CCD)
camera (Princeton Instruments, Trenton, NJ), and the Metamorph image analysis
software package for image acquisition (Universal Imaging, West Chester, PA)
which allows average intensities to be determined within user-defined regions
of an image.
U-MNG filter cube (530–550 nm ex, 570 nm dichroic mirror, 590 nm high pass
em) is used for DiI observation and the U-MNIBA filter cube (470–490 nm ex,
505 nm dichroic mirror, 515–550 nm em) to visualize CMFDA fluorescence.
Images are collected randomly from 6–10 different selected fields for each
3. The data are analyzed using Metamorph software (Universal Imaging Inc.)
by overlaying and counting the images. The total number of cells positive for
CMFDA are counted. Then the number of cells positive for both fluorescent
probes are scored. Bright field images are used to distinguish false-positives
where labeled cells were lying over one another but had not actually fused.
4. Percent fusion is calculated as
% fusion = 100 × [number of cells positive for both dyes]
% fusion = 100 × ————————————————
[total number of target cells]
Figure 4 shows fusion of GSL-supplemented mink-CD4 cells with TF228
cells. The data presented in Fig. 4 show that addition of only human GSL to
mink-CD4 cells resulted in subsequent fusion with TF228 cells. Lack of fusion
with bovine GSL-supplemented mink-CD4 showed that recovery of fusion was
specific. Percentage fusion from one such experiment was as follows: GSL-
Hu-mink-CD4 cells, 35–40%; GSL-Bov-mink-CD4 cells, 8–10%; mink-CD4
cells without GSL addition, 8–10%.
We have described here two methods to incorporate lipids into biological
targets. We will briefly discuss the rationale behind the two approaches and
compare the limitations and benefits of these two methods.
Previous experiments from our laboratory have shown that time-resolved
kinetics of cell-to-cell fusion based on R18 dequenching can be successfully
monitored for viruses that utilize erythrocytes as targets (41,42). Erythrocytes
possess the advantage that their membranes can be labeled with R18 at
quenched concentrations without internalization of the dye, in contrast to other
biological membranes (viz. cultured cells). In an attempt to develop a versatile
assay to monitor the fusion kinetics of viral envelope proteins expressed in