Capture and imaging of a prehairpin fusion
intermediate of the paramyxovirus PIV5
Yong Ho Kima, Jason E. Donaldb, Gevorg Grigoryanb,1, George P. Leserc, Alexander Y. Fadeevd,
Robert A. Lambc,2, and William F. DeGradoa,b,2,3
aDepartment of Chemistry, University of Pennsylvania, Philadelphia, PA 19104;
Philadelphia, PA 19104;
dDepartment of Chemistry and Biochemistry, Seton Hall University, South Orange, NJ 07079
bDepartment of Biochemistry and Biophysics, University of Pennsylvania,
cHoward Hughes Medical Institute and Department of Molecular Biosciences, Northwestern University, Evanston, IL 60201; and
Contributed by William F. DeGrado, October 11, 2011 (sent for review August 26, 2011)
During cell entry, enveloped viruses fuse their viral membrane with
a cellular membrane in a process driven by energetically favor-
able, large-scale conformational rearrangements of their fusion
proteins. Structures of the pre- and postfusion states of the fu-
sion proteins including paramyxovirus PIV5 F and influenza virus
hemagglutinin suggest that this occurs via two intermediates. Fol-
lowing formation of an initial complex, the proteins structurally
elongate, driving a hydrophobic N-terminal “fusion peptide” away
from the protein surface into the target membrane. Paradoxically,
this first conformation change moves the viral and cellular bilayers
further apart. Next, the fusion proteins form a hairpin that drives
the two membranes into close opposition. While the pre- and post-
fusion hairpin forms have been characterized crystallographically,
the transiently extended prehairpin intermediate has not been
visualized. To provide evidence for this extended intermediate
we measured the interbilayer spacing of a paramyxovirus trapped
in the process of fusing with solid-supported bilayers. A gold-
labeled peptide that binds the prehairpin intermediate was used
to stabilize and specifically image F-proteins in the prehairpin inter-
mediate. The interbilayer spacing is precisely that predicted from a
computational model of the prehairpin, providing strong evidence
for its structure and functional role. Moreover, the F-proteins in the
prehairpin conformation preferentially localize to a patch between
the target and viral membranes, consistent with the fact that the
formation of the prehairpin is triggered by local contacts between
F- and neighboring viral receptor-binding proteins (HN) only when
HN binds lipids in its target membrane.
fusion protein refolding ∣ membrane fusion ∣ electron microscopy
class I fusion proteins that mediate coalescence of the viral and
target membranes (1–6). Homotrimeric class I fusion proteins are
synthesized as biologically inactive precursors that are activated
by proteolytic cleavage, which generates a new N-terminal hydro-
phobic sequence known as the fusion peptide. The crystal struc-
tures of the paramyxovirus F-protein in its metastable prefusion
form and the human parainfluenza virus 3 (hPIV3) F-protein
in its very stable postfusion form, and also the influenza virus HA
in pre- and postfusion conformations, reveal a unique protein
architecture that undergoes large-scale, irreversible refolding
during membrane fusion (7–10).
These structures provided snapshots of the locations of the
critical fusion peptide in the initial and final states: In the prefu-
sion form, the fusion peptide is located along the protein surface,
near the virus membrane; while in the postfusion state it is posi-
tioned to penetrate into the viral membrane through the forma-
tion of a hairpin conformation. Nevertheless, photochemical
labeling (11) indicates that the fusion peptide engages with the
target membrane in an intermediate state known as the “prehair-
pin” state. This prehairpin state has been hypothesized to be a
highly extended parallel coiled coil, which drives the fusion pep-
tide beyond the surface ofthe virus toallow it toinsert deeply into
nveloped viruses such as influenza virus, human immunode-
ficiency virus (HIV), and parainfluenza virus 5 (PIV5) encode
the target membrane. Interestingly, this conformational change
would be expected to drive the membranes further apart than
in the initial complex, due to the proposed elongated nature of
the prehairpin. Next, the formation of the final hairpin form is
hypothesized to drive the close apposition of the bilayers required
for fusion to proceed. The hypothetical extended prehairpin form
has served as a remarkably robust model for the design of peptide
drugs that inhibit HIV and PIV5 fusion such as T-20 and C1
(12–14). However, to date, the existence of this prehairpin inter-
mediate has been shown only indirectly (15–17). Direct observa-
tion of this state would be a valuable test of current models of
viral fusion and the mode of action of entry inhibitors.
Here we characterize the prehairpin intermediate of a para-
myxovirus, which fuses with the cytoplasmic membrane and does
not require endocytosis or low pH-triggering (2). The initial bind-
ing occurs via transient interactions between its receptor-binding
protein known as hemagglutinin-neuraminidase (HN) and sialic
acid moieties on the cell surface. The engagement of HN with its
substrate leads to conformational changes in neighboring F-pro-
teins on the viral envelope, which then convert to the prehairpin
form. We built a computational model for F in this metastable
intermediate, which predicted a relatively long bilayer-to-bilayer
spacing of 21 nm. We used electron microscopy (EM) to charac-
terize the fusion of virions with nano-bead supported bilayers
(18), which were expected to be less deformable than liposomes
or cells and hence slower to fuse with virus. A previously char-
acterized prehairpin-binding peptide was used to further trap this
intermediate and enable immuno-gold labeling. Direct measure-
ment of the interbilayer distance shows that the separation closely
matches that predicted from the computational model for the
prehairpin intermediate. Additionally, the F-proteins in the pre-
hairpin state on the surface of the virus specifically localize to
the contact area between the virus and the target membrane.
These experiments provide direct evidence for this prehairpin
intermediate in an intact cell, and suggest that the fusion protein
does travel through this large, extended intermediate state on the
path to fusion.
Results and Discussion
A Computational Modeling of the Prehairpin Intermediate State. An
atomic-scale model was constructed to determine the overall
Author contributions: Y.H.K., J.E.D., A.Y.F., R.A.L., and W.F.D. designed research; Y.H.K.,
J.E.D., and G.G. performed research; Y.H.K., J.E.D., G.G., and A.Y.F. contributed new
reagents/analytic tools; Y.H.K., J.E.D., G.G., G.P.L., R.A.L., and W.F.D. analyzed data; and
Y.H.K., J.E.D., R.A.L., and W.F.D. wrote the paper.
The authors declare no conflict of interest.
1Present address: Department of Computer Science, Dartmouth College, Hanover,
2To whom correspondence may be addressed. E-mail: firstname.lastname@example.org or
3Present address: Department of Pharmaceutical Chemistry, University of California, San
Francisco, San Francisco, CA 94143.
This article contains supporting information online at www.pnas.org/lookup/suppl/
20992–20997 ∣ PNAS ∣ December 27, 2011 ∣ vol. 108 ∣ no. 52 www.pnas.org/cgi/doi/10.1073/pnas.1116034108
length of the aqueous domain, and hence the expected interbi-
layer distance of the intermediate. Type I fusion proteins, includ-
ing PIV5 F, influenza virus HA, and HIV Env, have two helical
repeat (HRA and HRB) regions that form coiled coils during
different stages of fusion. In the prefusion state of PIV5, HRA
lies along the protein surface, while HRB forms a trimeric coiled
coil stalk adjacent to the viral membrane (7). In the prefusion
state, HRA is believed to detach from the protein surface and
refold into a long, extended three-stranded coiled coil projecting
towards the target bilayer, while HRB remains essentially intact
and anchored near the virus surface. During formation of the
hairpin intermediate, the HRA trimer remains constant, while
the original HRB bundle dissociates and its individual chains bind
along the exterior of the HRA trimer, pulling the viral membrane
toward the target bilayer until the postfusion trimer of helical
hairpins is fully formed (8, 19). Thus, the prehairpin intermediate
shows hybrid character, having HRA and HRB in the starting
and postfusion conformations, respectively. One domain of F
(residues 290–357) is largely invariant in the pre- and postfusion
structures; it appears to form a hinged hub about which HRA and
HRB pivot during these conformational rearrangements. Thus,
by superposing this domain in the pre- and postfusion states, it
was possible to define the positions of HRA and HRB in the pre-
hairpin intermediate (see Methods and Dataset S1). The end-
to-end distance of the resulting model (Fig. 1 A and B) is 21 nm
with an associated error of approximately 1 nm, primarily reflect-
ing the error in predicting where the transmembrane (TM) and
fusion peptides enter the membranes.
Bilayer-Supported Silica and Polystyrene Nanobeads. For micro-
scopic studies, lipid bilayers were supported on nanosilica orpoly-
styrene (PS) nanobeads. The silica-supported bilayers maintain,
by tethering the bilayer to the nanobead with poly(ethylene gly-
col) (PEG), an internal space that allows for incorporation of
membrane proteins containing cytoplasmic domains. For experi-
ments requiring thin-sectioning we used the softer polystyrene
(PS) as the support. An electron micrograph (Fig. S1) shows the
structure of silica and PS nanobeads stained using a mixture of
2% uranyl acetate and 2% osmium tetroxide. The membrane
surface of the silica-supported nanobeads is approximately 50 Å
from the nanobead core because of the flexible PEG linkages, as
compared to the 40 Å that would be observed if the membrane
was directly on the surface of the solid support. To control the size
homogeneity, curvature of membranes, and structural properties
of the nanobeads, 100–150 nm silica and ≈120 nm polystyrene
beads (average diameter) were selected.
To induce fusion of PIV5 particles (virons) with nanobead-
supported bilayers, the bilayer was modified with sialoganglio-
sides to serve as a receptor for binding by the PIV5 attachment
protein, HN. PIV5 HN binding to sialic acid is believed to lead to
activation of F-protein, priming the conformational transition
from the prefusion state to intermediates along the fusion reac-
tion pathway (20, 21).
Fusion Between Viruses and Nanobead-Supported Bilayers. For para-
myxoviruses, binding of virus to surface sialic acids at 37°C to
42°C triggers the conformational changes leading to fusion (22,
23). As expected, fusion was observed at 37° and 42 °C by EM
bound by the fusion inhibitor (right) show the distance expected to be observed in EM. The fusion peptide is shown in cyan, heptad repeat A in red, the
globular head domain in green, heptad repeat B in dark blue (cartoon), and the TM domain in magenta. The fusion-inhibitory peptide, C1 (12), is shown
in dark blue (filled cylinders). (B) Schematic diagram illustrating the free energy changes in the system. The canonical pathway moves from prefusion, to
the intermediate state, through a high barrier to the postfusion state (8). Introduction of the fusion-inhibitory peptide, C1, traps the protein in the inter-
mediate state conformation (12).
Structures of the fusion protein in fusion states and their thermodynamics. (A) Models of the prefusion (7, 27) (left) and the prehairpin intermediate
EM images of the samples. (A) At 4°C, nanobead-supported bilayers made up
of silica nanobeads were combined with purified PIV5 particles that express
the fusion (F) protein spike on their surface. At this temperature there is little
association between the virus and the nanobead-supported bilayers. Dark
spheres represent silica nanobead and light vesicles are viruses. Silica-sup-
ported bilayers appear darker under these conditions. (B) After warming to
37°C for 30 min viral particles fuse to nanobead-supported bilayers. Samples
in (A and B) were stained with 4% uranyl acetate. (C) EM gallery of fused
virus/nanobead-supported bilayers pairs in higher magnification. PIV5 virions
with a concentration of 1.0 × 1010plaque forming units ðPFUÞ∕mL were
mixed with silica nanobeads with an approximate concentration of 5 × 109
particles∕mL into 100 μL total volumes and incubated at the desired tem-
perature. The concentration of silica nanobeads was calculated from weight,
assuming a 120 nm mean diameter of silica and its density of 2.2 g∕cm3.
Temperature triggers viral fusion with nanobead-supported bilayers.
Kim et al.PNAS
December 27, 2011
after 30 min incubation (Fig. 2B), but not at 4°C and 25 °C
(Fig. 2A). Darker, nanobead-supported bilayers are readily dis-
tinguished from the virus particles.
Measurement of the Spacing Between Viral Membranes and Nano-
bead-Supported Bilayers. The computational model predicts a
bilayer separation of 21 nm in the prehairpin intermediate
(Fig. 1A). Using thin-sectioning EM with staining allowed the
surface spike glycoproteins of PIV5 to be observed. To trap the
F-protein in the prehairpin conformation, we included an inhibi-
tory peptide, C-1, derived from the HRB region of F (12). The
C-1 peptide binds specifically to the prehairpin form of F, and is
believed to inhibit PIV5 fusion by trapping the protein in this
state and preventing it from proceeding to the final hairpin form
(Fig. 3A). More than 150 images of virus/nano-particle mimic
pairs were analyzed. In both the presence and absence of the
fusion inhibitor, the distance between the viral and nanobead-
supported bilayer was measured (Fig. 3 A and D). Unlike unsec-
tioned silica nanobead-supported samples (Fig. 2), the contrast
of sectioned samples (Fig. 3) resulted in the viruses being more
darkly stained due to the internal ribonucleoprotein core. In the
absence of the fusion-inhibitory peptide, fusion was observed
(Fig. 3D) and most virus/nanobead-supported bilayer pairs are
separated by a short distance (Fig. 3C). In the presence of the
fusion-inhibitory peptide, the distribution of interparticle distances
peaked at approximately 20 nm (Fig. 3C), very closely matching
the value predicted by the computational model (Fig. 1A).
The conformation of the trapped prehairpin intermediate is
suggested by significant changes of proteins adjacent to the na-
nobead-supported bilayer surface (Fig. 3 A and B). While regions
of the virus far from the interface show densely packed protein
(presumably a mixture of HN and F in its prefusion conforma-
tion, blue arrows, Fig. 3B), regions close to the interface appear
to be less densely packed, (yellow arrows, Fig. 3B) consistent
with a transition between the prefusion conformation and the
prehairpin intermediate (Fig. 3B).
Immuno-Gold Labeling Shows Localization of Inhibited PIV5 F-Protein
to the Interfacial Region. We next localized the proteins that had
reached the prehairpin intermediate by using a fusion-inhibitory
peptide coupled to colloidal gold and visualized by EM. The na-
nobead-supported bilayers and PIV5 virions were first incubated
in the presence of the C-1 peptide that binds to the prehairpin
intermediate state (12). To allow gold-labeling a biotin was incor-
porated attached to the N terminus of the peptide via a short
PEG linker. Next, streptavidin-labeled gold particles were added
to label the location of the C-1 peptides and the F-protein pre-
hairpin intermediate. The nanobead-supported bilayers, virus
particles, and gold particles were then displayed by thin-section-
ing and EM (Fig. 4A).
Approximately 300 virus/nanobead-supported bilayer images
were then overlaid to show the position of the gold particles
relative to the virions and nanobead-supported bilayers (Fig. 4C).
To allow quantitative comparison of the distributions, all EM
images were superimposed along a line between the center of the
viral particle ellipse and the center of the PS nanobead-supported
bilayers; the midpoint of the virus-nanobead vector defines the
origin of the superimposed structures. The location of the gold
particles relative to the virus particles and nanobead-supported
bilayers are shown in red and blue circles, respectively (Fig. 4C).
The position of the gold particles clearly localize to the contact
area between the virus and the nanobead-supported bilayer
(Fig. 4A), indicating the accumulation of F-proteins in the pre-
hairpin intermediate state at the interface. The angle distribution
of gold particles around the particles (Fig. 4D) supports this con-
clusion, and shows the particles are localized in a narrow range
within 35° of the line between the centers of the two particles. The
counts within this angular peak are 5 to 20-fold greater than at
other angles, which represent locations outside of the contact
polystyrene beads from thin sectioned samples (A) at 42°C in the presence of the fusion-inhibitory peptide C-1 and (B) in the absence of the fusion-inhibitory
peptide, C-1, under the same condition as in (A). Under these staining conditions viral particles are darker than the PS nanobeads. (B) Blue arrows show
the putative prefusion state in the absence of the fusion-inhibitory peptide C-1, while yellow arrows show putative prehairpin intermediates. (C) Distance
distribution between the edge of the viral particle and nanobead with and without fusion inhibitor at 42°C. Note the peak near 20 nm for the inhibited state.
(D) EM gallery of viral particles fusing with nanobead-supported bilayers at 42 °C in the absence of fusion-inhibitory peptides C-1. PIV5 virions with a con-
centration of 1.0 × 1010plaque forming units∕mL were mixed with amino polystyrene nanobeads with a concentration of 5 × 1010particles∕mL. The faint
appearance of sections in EM images is an artifact arising from four independent quadrants in the camera used for image capture.
Observation of the prehairpin intermediate by thin-sectioning and EM. EM gallery of viral particles and PS nanobead-supported bilayers made up of
www.pnas.org/cgi/doi/10.1073/pnas.1116034108Kim et al.
zone. Thus, the prehairpin conformation segregates within the
contact zone, consistent with previous findings that formation
of this intermediate is triggered by contact between the virus
and the target at permissive temperatures.
Here we confirm the existence of the viral fusion prehairpin
intermediate and provide direct images of the bilayer spacing
and distribution of the population of F-protein in this intermedi-
ate state. Previous evidence for the existence of the prehairpin
intermediate in this and other viruses has been indirect. For ex-
ample, peptides with the sequence of HRB strongly inhibit fusion
despite the occlusion of its postfusion binding partner, HRA, in
the prefusion state (12–14). While this result is consistent with
an extended prehairpin intermediate that exposes HRA during
the fusion pathway, the direct measurements made here provide
strong new support for this hypothesis. Combined with computa-
state. (A) Gallery of thin sectioned EM images of viral particles and nanobead-supported bilayers labeled using immuno-gold. The small black spheres are 5 nm
gold particles. “V” labels viral particles, and “N” labels nanobeads. The thin sections containing immune-gold are approximately 70 nm thick, and produced a
silver interference color. (B) Illustration of how the angle θ is calculated for a virus and nanobead pair. (C) Superimposition of the positions of all gold particles
around interacting viral particles and nanobead-supported bilayers. The average virus ellipse is in red and the average nanobead-supported bilayer ellipse is in
blue. Gold particles are shown as yellow dark circles. (D) The distribution of the angle θ. Note the large peak at 0° the angle at which a gold particle is in the
space between the virus and the nanobead-supported bilayer. All scale bars in (A) are 100 nm. The faint appearance of sections in EM images is an artifact
arising from four independent quadrants in the camera used for image capture.
Immuno-gold labeling of the fusion inhibitor shows it is concentrated between the viral particles and nanobead-supported bilayers in the trapped
Kim et al.PNAS
December 27, 2011
tional modeling, these images also bring us much closer to an
atomic level structure that could be used in the design of future
Materials and Methods
Computational Modeling of the Prehairpin Intermediate. Computational mod-
eling of the prehairpin intermediate involved primarily four steps. First, the
crystal structures of the F prefusion state of PIV5 and F postfusion states of
hPIV3 were structurally aligned to create an extended model that contains
both HRA and HRB as trimeric coiled coils. Second, the hPIV3 sequence was
replaced by that of PIV5 and repacked. Third, the TM and fusion peptide
domains were added to the model. Finally, the combined model was mini-
mized to remove clashes.
To estimate the interbilayer spacing of the prehairpin intermediate we
built a detailed atomic model of this structure (24, 25), in which the N-term-
inal region (residues 103–289) has extended into a trimeric coiled coil project-
ing towards the target bilayer, and the C-terminal residue (residues 358–511)
undergoes its conformational change primarily at the late stages of fusion
(going from the prehairpin intermediate to the postfusion states). This
hypothesis is consistent with the fusion peptide from the N-terminal region
inserting during the early stages of fusion, prior to the zippering of HRB that
helps bring the bilayers together. The crystal structures of the prefusion (7)
and postfusion (8) states were aligned using a region of the globular head
domain, residues 290–357 (PIV5 numbering). This region is largely constant
within each chain between the two structures (Cα rmsd ¼ 1.7 Å when com-
paring individual chains) and presumably is also constrained in the prehairpin
intermediate. While the rmsd islarger when all chains are considered because
of relative rotation of the chains (Cα rmsd ¼ 4.4 Å), a structural alignment
of Cα positions places residues 357 of the postfusion structure and 358 of
the prefusion structure within close proximity, allowing connection of the
two regions. Therefore, the N-terminal region was taken from the postfusion
structure, and the C-terminal region taken from the prefusion structure.
Second, the sequence of PIV5 was threaded onto the postfusion structure
using the Rosetta Design software package to repack the new side chains
(26). Repacking occurred using the default parameters. Third, for the TM
domain adjacent to HRB, the model was taken from that obtained from dis-
ulfide cross linking (27). For the fusion peptide adjacent to HRA, sequence
conservation data (28) supports a coiled coil structure. The coiled coil from
HRA was extended into the fusion peptide to model this region, placing the
most conserved residues at the core of the coiled coil. The structure was then
briefly minimized (200 steps of steepest decent) in Gromacs (29) using the
all-atom OPLS force field (30, 31) to relax any steric clashes in the new struc-
ture. Minimization using other force fields, such as CHARMM (32), gave very
Preparation of Nanobead-Supported Bilayers. Small unilamellar vesicles (SUVs)
containing a 4∶4∶2 molar ratio of 1-palmitoyl-2-oleyl-sn-glycero-3-phospho-
choline (POPC), 1,2-dioleoyl-sn-glycero-3-[phosphor-L-serine] (DOPC) and
cholesterol (Avanti Polar Lipids) doped with 1 M% bovine brain disialogan-
glioside GD1a(Sigma Aldrich) at a lipid concentration of 5 mM were prepared
by tip sonication of vacuum dried lipid films in PBS buffer (pH 7.5, 100 mM
NaCl) for 30 min. PEG-tethered silica nanobeads were prepared by the
reaction of spherical bare silicas with ðC2H5OÞ3SiðCH2Þ3O½OCH2CH2?8–12-OH
(Gelest) in water (33). The silica particles were synthesized by Stöber’s method
(34). The mean particle diameter ∼120 nm was assessed by EM and its surface
area was determined by BET method (35). Overnight incubation of SUVs
with silica and polystyrene nanobeads formed lipid-bilayers on the hydrophi-
lic surface of silica and polystyrene nanobeads. Silica nanobeads were treated
with lipid concentrations sufficient to provide coverage of ∼2.5 lipids∕nm2
of the silica surfaces, which corresponds to a bilayer. Coverage was confirmed
by adsorption isotherms (33), and verified by phosphorus analysis and fluor-
escence of NBD-doped lipids. Polystyrene nanobeads (Polyscience, amino PSs
used with 5.68 × 1012particles∕mL in concentration) combined with 1 mL of
5 mM SUV stock solution formed lipid-bilayers on the hydrophilic surface of
nanobeads. After encapsulation by bilayers, the silica nanobeads were rinsed
at least three times with PBS buffer followed by vortexing and centrifugation
at 10;000 rpm for 2 min and the supernatant was discarded. The resulting
lipid-coated nanobeads were reconstituted in PBS buffer (pH 7.5). Additional
sonication of the lipid-coated nanobeads was avoided so as to not disrupt the
nanobead lipid bilayer. Silica nanobeads with a mean diameter of 120 nm
were used. Each tube of lipid-coated silica nanobeads contains 2.8 mg of
120 nm silica nanobeads in PBS buffer. Nanobeads were vortexed or soni-
cated until a homogeneous suspension formed. PS nanobeads were used
for the advanced EM measurements of the interparticle distance. Organic
beads such as PS provide excellent contrast levels and better brightness
to highlight proteins in EM. In addition, the softness of PS provides allows
sectioning. For the EM measurements, aqueous suspensions were stained
using 2% uranyl acetate (20) and observed using a 80 kV FEI-Tecnai T12 to
display the lipid membrane on the nanobeads.
Production and Purification of PIV5 Virions. PIV5 was grown in Madin-Darby
bovine kidney cells as described previously (36, 37). Virus was purified essen-
tially as described (38) on 15–60% sucrose/NTE (100 mM NaCl, 10 mM Tris
pH 7.4 and 1 mM EDTA) gradients by ultracentrifugation (24;000 rpm for
2 h at 4°C) in a Beckman SW 32 rotor. The virus band was collected, diluted
in NTE buffer, and virus pelleted at 100;000 × g for 1 h in a Beckman Ti60
rotor. The viral pellet was resuspended in NTE buffer and Dounce homoge-
nized. Purified PIV5 virions were aliquoted in with a concentration of 1.0 ×
1010plaque forming units (PFU) per mL and stored at −80°C.
Electron Microscopy in Combination with Sectioning and Staining Techniques.
Electron microscopy. EM images were taken on a JEOL 2010 microscope
operating at 120 kV or 200 kV for sectioned samples and a FEI-Tecnai G12
operating at 80 kV for silica nanobead-coated bilayers. The imaging device
was a Gatan image filter, and the Gatan Digital Micrograph software was
used to record the images. In the case of lipid-coated silica nanobeads (Fig. 2),
droplets of an aqueous suspension of a mixture of silica nanobead-supported
bilayers and PIV5 viruses were dropcast onto a carbon grid (Electron Micro-
scopy Sciences, 400-mesh) and visualized on an FEI-Tecnai G12 Transmission
electron microscope at 80 kV.
Negative/positive staining and sectioning protocols. The standard process of
sample fixation with paraformaldehyde or glutaraldehyde caused artifacts
in the EM examination after sectioning and was not used. Samples were fixed
in 2% osmium tetroxide (OsO4) with 1.5% potassium ferricyanide in 0.1 M
sodium cacodylate buffer for 1 h at room temperature. OsO4is known to
stabilize many proteins by transforming them into gels without destroying
structural features (39). Proteins that are stabilized by OsO4are not aggre-
gated by alcohols during dehydration. After washes in dH2O, samples were
stained en bloc with 2% aqueous uranyl acetate (20) prior to dehydration in a
graded ethanol series. Samples were embedded in a PolyBed 812 bed (Poly-
sciences Inc.). The embedded samples were sectioned on an ultramicrotome.
Thin sections were transferred to coated 300 mesh grids and were addition-
ally stained with uranyl acetate, OsO4, tannic acid, phosphotungstic acid
(PTA), and bismuth subnitrite (40). Staining times varied from 5 min
to 30 min.
Optimizing contrast of proteins and membranes against other compartments in
PIV5 virus. The sectioning and staining protocols were optimized to increase
the visibility of proteins on the viral surface (Fig. S2). OsO4embeds into mem-
branes, creating a high secondary electron emission without the need for
coating the membrane with a layer of metal which could obscure details
of the cell membrane (41). In staining the viral surface spike proteins (F
and HN), tannic acid-UA is commonly used as both a negative and positive
stain to improve resolution (42). The positively stained EM presents protein
compartments and nucleocapsids when combined with tannic acid and UA,
thus creating contrast. Depending on the presence of immuno-gold labels,
the staining conditions were found to be optimal at condition A (no immu-
no-gold): 1% tannic acid ð30 minÞ∕2% UA(30 min) (Fig. 3 A and D) or con-
dition B (with immuno-gold): 2% OsO4ð30 minÞ∕1% tannic acidð30 minÞ∕
2% UA(10 min) (Fig. 3B).
Synthesis and Purification of Inhibitory Fusion Peptide. Previous studies have
demonstrated that the inhibitory peptide C-1, derived from the heptad
repeat region B (HRB) of PIV5 F-protein, displays antiviral activity (43). The
sequence is: KLESSQILSIDPLDISQNLAAVNKSLSDALQHLAQSDTYLSAI. The C-1
was synthesized by solid-phase synthesis using Fmoc chemistry with HBTU
a coupling agent; biotin was attached via a flexible linker by coupling first
Fmoc-8-amino-3,6-dioxaoctanoic acid (PEPTIDES Internationals) then a single
cysteine residue to biotin (5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno [3,4-d]
imidazol-4-yl] pentanoic acid, Sigma-Aldrich). The products were cleaved
from the rink amide MBHA resin (Novabiochem, substitution level of
0.56 mmole∕g) in a mixture of trifluoroacetic acid (TFA)/triisopropyl silane∕
H2O (95∶2.5∶2.5 vol∕vol) at room temperature under N2flow for 2 h. The
crude peptides, precipitated with cold diethyl ether (Aldrich), were dried
in vacuo. Purification proceeded by preparative reverse phase high perfor-
mance liquid chromatography (HPLC, Varian ProStar 210) using a preparative
C4 column (Vydac) and a linear gradient of buffer A (99.9% H2O and 0.1%
trifluoroacetic acid) and buffer B (90% acetonitrile, 9.9% H2O and 0.1%
www.pnas.org/cgi/doi/10.1073/pnas.1116034108 Kim et al.
trifluoroacetic acid). Purity was assessed using an HP1100 analytical HPLC
system (Hewlett Packard) with a C4 column (Vydac). Molecular mass of all
peptides was confirmed by matrix-assisted laser desorption/ionization-time
of flight (MALDI-TOF) mass spectrometry using a Bruker Microflex 3.1.
Immuno-Gold Assay for Tracking the Prehairpin Intermediate State. Immuno-
gold labeling EM was performed using a JEOL 2010 microscope operated
at an acceleration voltage of 120 kV. The PIV5 virus particles were incubated
with lipid-coated PS nanobeads at 42°C in the presence of biotinylated
C-1 peptide using varied incubation time from 2 min to 1 h, followed by in-
cubation with streptavidin-coated 5 nm gold particles (Ted Pella) for 10 min.
The posttreatment with 5 nm gold particles labels the inhibitory peptide to
show the location of the inhibited fusion proteins. The same thin-sectioning
protocols were followed as in the other fusion assay. In the case of the
gold labels, the staining protocol allows for higher contrast against the light
membrane of the viruses and nanobeads.
Angular Distribution of Gold Particle Binding to Fusion Proteins in Immuno-Gold
Assay. For the localization of the prehairpin intermediate, EM images of viral
particle/PS nanobead pairs including immuno-gold were analyzed using a
JEOL 2010 microscope with over 200 images (about 300 pairs) in five sec-
tioned samples. In order to present a full distribution of gold particles bind-
ing to fusion protein in PIV5 virus, all EM images were superimposed such
that the midpoint of the line between the center of the viral particle ellipse
and the center of the PS nanobead-supported bilayers ellipse are at the ori-
gin, the viral particle is on the negative x axis, and the nanobead-supported
bilayer is on the positive x axis (Fig. 4C). The alignments were made using the
Matlab computing language. The angles of gold particles within 200 nm of
the nanobead-supported bilayer center were then quantified (Fig. 4D).
ACKNOWLEDGMENTS. We thank Dong Kuyn Ko and Taejong Baik for EM
measurements. We also thank Raymond Meade and Biao Zuo in BioMedical
Imaging Core at University of Pennsylvania for preparation of all sectioned
and stained samples for EM analysis. This work was supported by National
Institutes of Health (NIH) grants (GM54616, AI-23173) and the Materials Re-
search Science and Engineering Centers (MRSEC) program of the National
Science Foundation (NSF). G.P.L. is a Research Specialist and R.A.L. is an Inves-
tigator of the Howard Hughes Medical Institute.
1. Harrison SC (2008) Viral membrane fusion. Nat Struct Mol Biol 15:690–698.
2. Lamb RA, Jardetzky TS (2007) Structural basis of viral invasion: lessons from paramyx-
ovirus F. Curr Opin Struct Biol 17:427–436.
3. Colman PM, Lawrence MC (2003) The structural biology of type I viral membrane
fusion. Nat Rev Mol Cell Biol 4:309–319.
4. Weissenhorn W, et al. (1999) Structural basis for membrane fusion by enveloped
viruses. Mol Membr Biol 16:3–9.
5. Cross KJ, Burleigh LM, Steinhauer DA (2001) Mechanisms of cell entry by influenza
virus. Expert Reviews in Molecular Medicine 3:1–18.
6. White JM, Delos SE, Brecher M, Schornberg K (2008) Structures and mechanisms
of viral membrane fusion proteins: multiple variations on a common theme. Crit Rev
Biochem Mol Biol 43:189–219.
7. Yin HS, Wen X, Paterson RG, Lamb RA, Jardetzky TS (2006) Structure of the parain-
fluenza virus 5 F protein in its metastable, prefusion conformation. Nature 439:38–44.
8. Yin HS, Paterson RG, Wen X, Lamb RA, Jardetzky TS (2005) Structure of the uncleaved
ectodomain of the paramyxovirus (hPIV3) fusion protein. Proc Natl Acad Sci USA
9. Wilson IA, Skehel JJ, Wiley DC (1981) Structure of the haemagglutinin membrane
glycoprotein of influenza virus at 3 A resolution. Nature 289:366–373.
10. Bullough PA, Hughson FM, Skehel JJ, Wiley DC (1994) Structure of influenza haemag-
glutinin at the pH of membrane fusion. Nature 371:37–43.
11. Brunner J, Zugliani C, Mischler R (1991) Fusion activity of influenza virus PR8/34 cor-
relates with a temperature-induced conformational change within the hemagglutinin
ectodomain detected by photochemical labeling. Biochemistry 30:2432–2438.
12. Russell CJ, Jardetzky TS, Lamb RA (2001) Membrane fusion machines of paramyxo-
viruses: capture of intermediates of fusion. EMBO J 20:4024–4034.
13. Chan DC, Kim PS (1998) HIV entry and its inhibition. Cell 93:681–684.
14. Jiang S, Lin K, Strick N, Neurath AR (1993) HIV-1 inhibition by a peptide. Nature
15. Lee KK (2010) Architecture of a nascent viral fusion pore. EMBO J 29:1299–1311.
16. Damico RL, Crane J, Bates P (1998) Receptor-triggered membrane association of a
model retroviral glycoprotein. Proc Natl Acad Sci USA 95:2580–2585.
17. Furuta RA, Wild CT, Weng Y, Weiss CD (1998) Capture of an early fusion-active
conformation of HIV-1 gp41. Nat Struct Biol 5:276–279.
18. Tanaka M, Sackmann E (2005) Polymer-supported membranes as models of the cell
surface. Nature 437:656–663.
19. Baker KA, Dutch RE, Lamb RA, Jardetzky TS (1999) Structural basis for paramyxovirus-
mediated membrane fusion. Mol Cell 3:309–319.
20. Yuan P, et al. (2005) Structural studies of the parainfluenza virus 5 hemagglutinin-
neuraminidase tetramer in complex with its receptor, sialyllactose. Structure
21. Crennell S, Takimoto T, Portner A, Taylor G (2000) Crystal structure of the multifunc-
tional paramyxovirus hemagglutinin-neuraminidase. Nat Struct Biol 7:1068–1074.
22. Lamb RA, Parks GD (2007) Fields Virology, eds DM Knipe and PM Howley, pp
23. Paterson RG, Russell CJ, Lamb RA (2000) Fusion protein of the paramyxovirus SV5:
destabilizing and stabilizing mutants of fusion activation. Virology 270:17–30.
24. Lamb RA (1993) Paramyxovirus fusion: a hypothesis for changes. Virology 197:1–11.
25. Carr CM, Kim PS (1993) A spring-loaded mechanism for the conformational change
of influenza hemagglutinin. Cell 73:823–832.
26. Rohl CA, Strauss CE, Misura KM, Baker D (2004) Protein structure prediction using
Rosetta. Methods Enzymol 383:66–93.
27. Bissonnette ML, Donald JE, DeGrado WF, Jardetzky TS, Lamb RA (2009) Functional
analysis of the transmembrane domain in paramyxovirus F protein-mediated
membrane fusion. J Mol Biol 386:14–36.
28. Donald JE, et al. From the cover: transmembrane orientation and possible role of
the fusogenic peptide from parainfluenza virus 5 (PIV5) in promoting fusion. Proc Natl
Acad Sci USA 108:3958–3963.
29. Van Der Spoel D, et al. (2005) GROMACS: fast, flexible, and free. J Comput Chem
30. Jorgensen WL, Maxwell DS, Tirado-Rives J (1996) Development and testing of the
OPLS all-atom force field on conformational energetics and properties of organic
liquids. J Am Chem Soc 118:11225–11236.
31. Kaminski GA, Friesner RA, Tirado-Rives J, Jorgensen WL (2001) Evaluation and repar-
ametrization of the OPLS-AA force field for proteins via comparison with accurate
quantum chemical calculations on peptides. J Phys Chem B 105:6474–6487.
32. MacKerell AD, Jr, et al. (1998) All-atom empirical potential for molecular modeling
and dynamics studies of proteins. J Phys Chem B 102:3586–3616.
33. Fadeev AY, DeGrado WF (2011) Lipid membranes supported on optically transparent
nanosilicas: synthesis and application in characterization of protein-membrane inter-
actions. J Colloid Interface Sci 355:265–268.
34. Stöber W, Fink A, Bohn E (1968) Controlled growth of monodisperse silica spheres
in the micron size range. J Colloid Interf Sci 26:62–69.
35. Gregg SJ, Sing KSW (1982) Adsorption, Surface Area, and Porosity (Academic Press,
36. Peluso RW, Lamb RA, Choppin PW (1977) Polypeptide synthesis in simian virus
5-infected cells. J Virol 23:177–187.
37. Paterson RG, Harris TJ, Lamb RA (1984) Analysis and gene assignment of mRNAs
of a paramyxovirus, simian virus 5. Virology 138:310–323.
38. Lamb RA, Mahy BW, Choppin PW (1976) The synthesis of sendai virus polypeptides
in infected cells. Virology 69:116–131.
39. Hayat MA (2000) Principles and Techniques of Electron Microscopy: Biological Appli-
cations (Cambridge University Press, Cambridge United Kingdom; New York).
40. Yamaguchi K, Suzuki K, Tanaka K (2010) Examination of electron stains as a substitute
for uranyl acetatefor the ultrathinsectionsof bacterial cells. J ElectronMicrosc (Tokyo)
41. Bozzola JJ, Russell LD (1999) Electron Microscopy: Principles and Techniques for Biol-
ogists (Jones and Bartlett, Sudbury, Mass).
42. Haidar A, Ryder TA, Mobberley MA, Wigglesworth JS (1992) Two techniques for
electron opaque staining of elastic fibres using tannic acid in fresh and formalin fixed
tissue. J Clin Pathol 45:633–635.
43. Joshi SB, Dutch RE, Lamb RA (1998) A core trimer of the paramyxovirus fusion protein:
parallels to influenza virus hemagglutinin and HIV-1 gp41. Virology 248:20–34.
Kim et al. PNAS
December 27, 2011