Virology 294, 60–69 (2002)

doi:10.1006/viro.2001.1303, available online at http://www.idealibrary.com on

Hantaan Virus Enters Cells by Clathrin-Dependent Receptor-Mediated Endocytosis

Mirim Jin,* Junghyun Park,* Sungwook Lee,* Boyoun Park,* Jinwook Shin,* Ki-Joon Song,† Tae-In Ahn,‡
Sue-Yun Hwang,§ Byung-Yoon Ahn,* and Kwangseog Ahn* ,1

*Division of Life Science and Graduate School of Biotechnology and †Department of Microbiology, College of Medicine, Korea University,
Seoul 136-701, Korea; ‡Division of Life Science, Seoul National University, Seoul, Korea; and §Research Institute of Immunobiology,
Catholic Institutes of Medical Science, The Catholic University of Korea, Seoul, Korea

Received September 19, 2001; returned to author for revision October 26, 2001; accepted November 15, 2001

The cellular entry of Hantaan virus (HTN) occurs through interactions with ␤ 3 integrins as cellular receptors. However, the
process of HTN infection following attachment to the cell surface is not well understood. Our data indicate that overexpression
of a dominant-negative mutant dynamin inhibits HTN internalization and that compounds that block clathrin- but not caveolae-
dependent endocytosis also reduce HTN infectivity. In addition, we show that HTN colocalizes with the clathrin heavy chain but
not with caveolae. At the early phase of infection HTN colocalizes with EEA-1, an early endosome marker, and later, HTN
colocalizes with LAMP-1, a lysosome marker. Cells treated with lysosomotropic agents are largely resistant to infection, suggesting
that a low-pH-dependent step is required for HTN infection. These findings demonstrate that HTN enters cells via the clathrin-
coated pit pathway and uses low-pH-dependent intracellular compartments for infectious entry. © 2002 Elsevier Science (USA)
Key Words: Hantaan virus; entry; clathrin-dependent endocytosis; endosome.

INTRODUCTION HTN must penetrate the plasma membrane and target its
genome to the proper cellular compartments.
Hantaan virus (HTN), a prototypic member of the Han- At present, at least two distinct internalization path-
tavirus genus of the Bunyaviridae family, is a causative ways for receptor-mediated virus endocytosis are sug-
agent for Korean hemorrhagic fever with renal syndrome, gested: one pathway via clathrin-coated pits and the
which is characterized by fever, vascular hemorrhage, other pathway via caveolae. It remains unclear whether
and renal dysfunction (Ko, 1992; Lee et al., 1982; Lee et HTN enters cells by a clathrin-dependent pathway or by
al., 1983). This disease is one of the most severe hem- a caveola-dependent pathway. Clathrin coats are known
orrhagic diseases worldwide, an important public health to be involved in receptor-mediated and fluid-phase en-
concern because of its high mortality rate. HTN is known docytosis from plasma membranes to early endosomes
to spread by aerosolized excreta of specific rodents, and as well as transport from the trans-Golgi network to
human infections occur upon close contact with the endosomes (Schmid, 1997). Enveloped or nonenveloped
natural host, the field mouse Apodemus agraricus. HTN viruses, such as influenza virus, adenovirus, canine par-
is enveloped and contains a tripartite negative-stranded vovirus, and human polyomavirus JC virus, seemingly
RNA genome. The gene segments L, M, and S encode use the clathrin-dependent pathway for infection (Bartlett
the viral RNA polymerase, two integral membrane sur- et al., 2000; Parker and Parrish, 2000; Pho et al., 2000;
face glycoproteins (G1 and G2), and the nucleocapsid Roy et al., 2000). Caveolae are small flask-shaped invagi-
protein (N), respectively (Arikawa et al., 1990; Schmal- nations of the plasma membrane, characterized by high
john and Dalrymple, 1983). In both animals and humans, levels of cholesterol and glycosphingolipids and also by
hantavirus predominantly infects and replicates in endo- the presence of caveolin, an integral membrane protein
thelial cells, macrophages, kidney glomerula, and epi- of 20 to 24 kDa. One example of a virus that enters cells
thelial cells (Pensiero et al., 1992). Recent studies have via caveolae is Simian virus 40 (Pelkmans et al., 2001).
shown that early entry of HTN into cells involves its The fission and budding of both clathrin- and caveolae-
attachment to the cell surface receptor, the ␤ 3 integrin coated vesicles requires the GTPase activity of dynamin
(Gavrilovskaya et al., 1998, 1999). Following attachment, (Oh et al., 1998; van der Bliek et al., 1993), a cytosolic
protein of ⬃100 kDa. The expression of dominant-nega-
tive dynamin mutants, formed by a substitution mutation
1
of lysine to alanine (K44A) in the GTP binding site, inhib-
To whom correspondence and reprints should be addressed at the
Graduate School of Biotechnology, Korea University, 1, 5-ka, Anam-
its the formation of clathrin-coated pits and the budding
Dong, Sungbuk-Gu, Seoul 136-701, Korea. Fax: ⫹82-2-927-9028. E-mail: of caveolae-coated vesicles (Baba et al., 1995; Damke et
[email protected]. al., 1994, 1995; Oh et al., 1998).
0042-6822/02 $35.00
© 2002 Elsevier Science (USA) 60
All rights reserved.
 ENTRY OF HANTAAN VIRUS 61

Viruses that are taken into cells by receptor-

mediated endocytosis are trafficked to either acidic or
nonacidic compartments (Pelkmans et al., 2001). The
acidic condition in endosomes triggers the reactions
that lead to fusion between the viral and the endoso-
mal membranes (Helenius and Marsh, 1982). For ex-
ample, influenza virus and Semliki Forest virus require
an endosomal pH of less than 6.0 to cause their
glycoprotein spike complexes to undergo the confor-
mational changes that are needed for fusion of the
viral envelope with the cellular membrane (Edwards
and Brown, 1991; Schmid et al., 1989; Yoshimura et al.,
1982). Internalization through the endocytic pathway
usually results in the delivery of ligands to the hydro-
lytic compartment of lysosomes, an environment in
which ligands are rapidly inactivated and degraded FIG. 1. HTN enters cells by dynamin-required endocytosis. (A) West-
(Marsh et al., 1982). Here we use a variety of ap- ern blot analysis of dynamin expression. HeLa cells stably transfected
proaches to define the endocytosis and infection path- with the HA epitope-tagged wild-type or K44A dynamin mutant were
way of HTN into cells. Our results indicate that cyto- cultured in the presence (⫹) or absence (⫺) of tetracycline for 48 h. (B)
solic GTPase-active dynamin is critical for the entry of Dynamin dependence of HTN infection. Wild-type and K44A dynamin
stable cells grown with or without tetracycline for 48 h were inoculated
HTN. Agents that disrupt clathrin-dependent, receptor- with HTN, and the percentage of infected cells was assayed after 30 h
mediated endocytosis inhibit infection of cells by HTN. by anti-nucleocapsid staining and immunofluorescence microscopy.
In contrast, agents that inhibit the caveolae-dependent The results are expressed as the percentage of cells infected in the
endocytosis pathway have no effect on entry of HTN. induced culture relative to the percentage of cells infected in unin-
In addition, on the basis of the observed colocalization duced cultures. The average and standard deviation of three indepen-
dent experiments are shown.
of HTN with markers, we demonstrate that shortly after
uptake, HTN particles are delivered to endosomes and
later to lysosomes, suggesting that these intracellular
compartments participate in the life cycle of HTN.
surface (data not shown). These results indicate that
RESULTS dynamin K44A inhibited dynamin-mediated endocytosis
Dynamin is required for HTN entry in the induced cells. Next, we compared the efficiency of
HTN infection of dynamin K44A-expressing cells to the
Two major internalization pathways, clathrin-mediated efficiency of infection in uninduced and induced wild-
endocytosis and caveolae-dependent endocytosis, re- type dynamin-expressing cells. At 30 h postinfection,
quire dynamin (Damke et al., 1994; Oh et al., 1998). We
HTN-infected HeLa cells were stained with anti-nucleo-
used a dynamin mutant to examine the involvement of
capsid antibody and then evaluated for infection by
dynamin in HTN entry. For these experiments, we used
counting the HTN-positive cells. For each case, about
HeLa cells that had been stably transfected with either
900 cells were scored. We found that between 45 and
wild-type or dominant-negative (K44A) dynamin (van der
60% of the uninduced K44A cells became infected, while
Bliek et al., 1993). For both dynamin-expressing cell lines,
dynamin is tagged with an influenza virus HA epitope, the dynamin K44A-expressing cultures had 60% fewer
and expression of dynamin is controlled by a tetracy- infected cells on average than the uninduced cultures
cline-inducible promoter (Gossen and Bujard, 1992). (Fig. 1B), similar to what has been reported for adenovi-
Upon removal of tetracycline from the growth medium, rus infection of HeLa cells (Wang et al., 1998). Taking into
cells are induced to express wild-type or mutant dy- account the possibility that some of cell population
namin (Fig. 1A), both of which can be detected via the would be in the S phase of the cell cycle in the unin-
HA tag. duced cells, which would reduce the number of cells
We examined whether blocking the endocytic pathway permissive for viral replication, we estimate that a 60%
in HeLa cells by expression of dynamin K44A would reduction is still significant. In contrast, there was no
prevent infection by HTN. Transferrin uptake was used significant difference in the infection rate of induced and
as a marker of dynamin-mediated endocytosis (Damke et uninduced wild-type dynamin-expressing cells. Notable
al., 1994). After 10 min at 37°C, Texas red-conjugated is that at best 70% of the HeLa cells became infected
transferrin was predominantly observed in the perinu- even at a multiplicity of infection (m.o.i.) of 10. These
clear region in uninduced cells, whereas in the dynamin results suggest that HTN enters cells via dynamin-medi-
K44A-expressing cells it remained mostly at the cell ated endocytosis.
62 JIN ET AL.

HTN entry is mediated by clathrin-dependent

endocytosis but not by caveolae-mediated
endocytosis
To determine whether HTN enters cells by a clathrin-
or a caveolae-dependent pathway, Vero E6 cells were
treated with drugs or subjected to conditions that selec-
tively inhibit each of these pathways. Both chlorproma-
zine and hypertonicity have been shown to inhibit clath-
rin-dependent endocytosis (Okamoto et al., 2000; Pho et
al., 2000). Since the internalization of transferrin by clath-
rin-dependent receptor-mediated endocytosis is well
known, as a control, we analyzed the uptake of fluores-
cein isothiocyanate (FITC)-conjugated transferrin into
Vero E6 cells with either chlorpromazine or a high con-
centration of sucrose. Internalization of transferrin re-
sults in a characteristic punctate pattern of cytoplasmic
staining (Killisch et al., 1992). As expected, treatment with
chlorpromazine (5 ␮g/ml for 30 min) or sucrose (0.45 M
for 10 min) inhibited the uptake of FITC-conjugated trans-
ferrin in the cells (Fig. 2A, b and d, respectively). Under
the same condition that disrupts the clathrin-dependent
pathway, cells were infected with HTN, and 30 h postin-
fection, expression of the HTN nucleocapsid antigen
was analyzed by immunofluorescence microscopy.
Treatment of cells with chlorpromazine or sucrose sig-
nificantly reduced the number of infected cells (Fig. 2B).
Moreover, we observed that addition of chlorpromazine
or sucrose 30 min postinfection did not affect the number
of HTN-infected cells, indicating that the blocking effects
of these drugs occur at an early step in the entry pathway
(data not shown). In the cases of both chlorpromazine
and sucrose, however, the inhibition was not complete.
To confirm these results and to provide direct and quan- FIG. 2. HTN entry is mediated by the clathrin-dependent pathway. (A)
titative evidence for the involvement of clathrin in HTN Clathrin-dependent internalization of FITC-conjugated transferrin was
endocytosis, we compared the total amount of cellular visualized in Vero E6 cells that were either untreated (a, c) or treated
with chlorpromazine (5 ␮g/ml) (b) or with 0.45 M sucrose (d). (B) HTN
nucleocapsid proteins in infected cells in the absence
entry is inhibited by chlorpromazine or sucrose. Vero E6 cells, un-
and presence of the drugs. Quantification of the proteins treated or pretreated with chlorpromazine (CPM) or sucrose, were
by Western blot analysis revealed that the total amount of infected with HTN at 37°C for 45 min. The remaining viruses were
nucleocapsid proteins obtained from the infected cells in neutralized by the addition of rat anti-HTN antiserum. At 30 h postin-
the presence of drugs was three- to fivefold less than the fection, infected cells were identified by immunofluorescence by direct
inspection of the coverslips. The percentage of infected cells was
amount obtained from cells without drug treatment (Fig. calculated by scoring infected/uninfected; in most cases, at least 900
2C, top). Endogenous ␤-actin was included as an inter- cells were scored per experimental point. (C) Quantitative analysis of
nal loading control for the Western blot (bottom). To- inhibition of HTN entry. Under the same conditions as described for B,
gether with the results in Fig. 2B, these data suggest that Vero E6 cells were lysed and the nucleocapsid protein of HTN was
clathrin-mediated endocytosis might be the predominant detected by Western blot analysis. The expression of endogenous
␤-actin was used as the internal loading control.
pathway for the entry of HTN, although we cannot ex-
clude other clathrin-independent mechanisms.
Treatment of Vero E6 cells with phorbol 12-myristate
13-acetate (PMA) and filipin, both known inhibitors of PMA or filipin (1 ␮g/ml) for 45 min (Fig. 3A, b and d,
caveola-dependent endocytosis (Anderson et al., 1996; respectively). However, using these same conditions for
Schnitzer et al., 1994), had no significant effect on infec- drug treatment, we could detect no significant difference
tion by HTN. Consistent with the findings of a previous in HTN internalization between treated and untreated
study (Skretting et al., 1999), the entry of the cholera toxin cells (Fig. 3B). Also, as shown in Fig. 3C, the total
B subunit, a well-known marker of the caveolae-depen- amounts of nucleocapsid proteins that could be detected
dent pathway, was blocked in the presence of 10 ␮M in untreated cells and cells treated with PMA or filipin did
 ENTRY OF HANTAAN VIRUS 63

tion with HTN showed any FITC staining. However, upon

incubation with HTN, colocalization of the FITC-conju-
gated nucleocapsid protein and the Texas red-conju-
gated clathrin heavy chain was observed, with colocal-
ization appearing yellow by digital overlay of both images
(Fig. 4A, bottom).
To further clarify the involvement of the clathrin-medi-
ated entry pathway for HTN, we infected Vero E6 cells
with a high titer of HTN (m.o.i. of 100 to 1,000), added
Texas red-conjugated transferrin to the medium, and
examined the localization of virus and transferrin. Both
FITC-conjugated virus particles and Texas red-conju-
gated transferrin could be seen in a small punctate
pattern of staining within the cytoplasm. An overlay of the
red and green channels showed that HTN colocalized
with transferrin (Fig. 4B), suggesting that HTN and trans-
ferrin coexist in the same endocytic compartments. In
contrast, no colocalization was found between HTN and
cholera toxin B subunit, a marker of the caveloae-depen-
dent pathway (Fig. 4C). These results provide additional
support that HTN entry into cells takes place through
clathrin-coated pits.

HTN requires passage through distinct acidic

compartments for infectious entry
It is generally accepted that internalized virions pene-
trate the cytoplasm by fusing with the membranes of
endosomes or lysosomes (Gruenberg and Howell, 1989).
To determine whether internalized HTN is also targeted
to endosomes, we infected Vero E6 cells with HTN. At 90
min after infection, we double-immunostained HTN par-
ticles and the endosome marker EEA-1 with FITC and
Texas red, respectively (Wilson et al., 2000), and ana-
lyzed cells by confocal microscopy. Both FITC-stained
FIG. 3. HTN entry is not mediated by the caveolae-dependent path- HTN and Texas red-stained EEA-1 exhibited a punctate
way. Caveolae-dependent endocytosis of FITC-conjugated cholera staining pattern. An overlay of the green and red chan-
toxin was visualized in Vero E6 cells that were either untreated (a, c) or
treated with filipin (1 ␮g/ml) (b) or with PMA (10 ␮M) (d). (B) HTN entry
nels indicated that HTN substantially colocalized with
is not inhibited by filipin or PMA. The same procedures were followed EEA-1 in endosomes (Fig. 5A).
as described for Fig. 2. (C) Quantitative analysis of inhibition of HTN In an effort to reveal the intracellular trafficking of HTN
entry by filipin or PMA. after entry, we examined whether the virions that fused
with endosomes are targeted to lysosomes, which is the
case for some viruses (Miyazawa et al., 2001). HTN-
not differ. These data suggest that HTN does not use a infected Vero E6 epithelial cells were double-labeled
caveolae-dependent pathway for entry. with anti-HTN nucleocapsid and anti-LAMP-1 antibodies,
late endosome or lysosome markers (Tabuchi et al.,
Colocalization of HTN and the clathrin-dependent
2000). At 4 h postinfection HTN-infected Vero E6 cells
endocytic pathway
labeled with the FITC-stained anti-nucleocapsid antibody
To confirm the involvement of clathrin-coated pits in showed a perinuclear punctate pattern (Fig. 5B, top),
HTN entry, we examined the colocalization of HTN and whereas no specific fluorescence was detected in mock-
the clathrin heavy chain by double-immunofluorescence infected cells (data not shown). Cytoplasmic staining
staining and confocal microscopy. We incubated HTN with a distinct punctate pattern was observed for Texas
with cells at 4°C for 1 h, a condition in which virus binds red-labeled LAMP-1. The superposition of the two im-
to the cell surface but does not enter the cells. Neither ages showed that HTN nucleocapsid proteins colocal-
the uninfected cells (Fig. 4A, top) nor the cells (data not ized with LAMP-1 at 4 h postinfection (Fig. 5B, top). We
shown) stained only with secondary antibody after infec- believe that the nucleocapsid proteins detected at this
64 JIN ET AL.

and the nucleocapsid proteins were exclusively concen-

trated on lysosomes in perinuclear regions (Fig. 5C). In
view of the expression profile of nucleocapsid proteins
as shown in Fig. 5D, we believe that the nucleocapsid
species detected by immunostaining at 24 h postinfec-
tion were synthesized after viral replication. These data
suggest that at late times of infection, the newly synthe-
sized nucleocapsid proteins also localize to late endo-
somes or lysosomes.

FIG. 4. (A) Colocalization of HTN and the clathrin heavy chain.

Double-immunofluorescence labeling of clathrin heavy chain and HTN
in Vero E6 cells after incubation with HTN at 4°C for 1 h. Confocal
microscopy photographs of HTN (left), clathrin (middle), and combined
FITC and Texas red channels (right). Arrowheads mark the colocaliza-
tion of HTN and clathrin in the plasma membrane. Colocalization of
HTN and transferrin (B) or cholera toxin B subunit (C). For direct
evaluation of endocytosis of HTN, 10 4 Vero E6 cells were grown on
coverslips and incubated with 10 6 to 10 7 HTN in the presence of Texas
red-labeled transferrin or Texas red-labeled cholera toxin B subunit at
4°C for 1 h. Endocytosis was initiated by shifting cells to 37°C for 30
min. Arrows mark the colocalization of HTN and transferrin within the
same endosomal compartment.

time point were derived from internalized HTN particles

but not from synthesis after infection. A time-course
study of the synthesis of nucleocapsid proteins clearly
showed that radiolabeled nucleocapsid proteins could
not be detected until 6 h after infection (Fig. 5D), which
generally agrees with a previous study showing that the
synthesis of viral proteins of HTN does not occur until
approximately 6 h after infection (Elliott et al., 1984). To
confirm that late endosome or lysosomal targeting of FIG. 5. (A) Localization of HTN in early endosomes. The intracellular
internalized HTN also occurs in other cell types, we compartment to which HTN is targeted after uptake was directly de-
performed essentially the same experiments using hu- termined by simultaneously probing cells with anti-HTN rat antiserum
man umbilical vein endothelial cells (HUVECs), a primary and anti-EEA-1 rabbit antiserum at 90 min postinfection. Bound anti-
body was traced with FITC-conjugated goat anti-rat antibody (left) or
endothelial cell line that is also permissive to HTN infec-
Texas red-conjugated goat anti-rabbit antibody (middle). (B) Colocal-
tion (Pensiero et al., 1992). Similar to our results in Vero ization of HTN and LAMP-1 in Vero E6 cells (top) or in HUVECs
E6 epithelial cells, the nucleocapsid protein and LAMP-1 (bottom). At 4 h postinfection, the cells were probed with anti-HTN
lysosome marker were colocalized in lysosomes (Fig. 5B, nucleocapsid antibody and anti-LAMP-1 antibody followed by appro-
top), suggesting that the route of HTN trafficking to late priate secondary antibodies. (C) Colocalization of the expressed HTN
nucleocapsid and LAMP-1 at 24 h postinfection. (D) Time-dependent
endosomes or lysosomes is not cell-type specific but
expression profile of HTN nucleocapsid proteins. At the indicated time
rather is conserved among cell types. Noteworthy, at points postinfection, HTN-infected Vero E6 cells were labeled with
24 h postinfection, the late endosomal or lysosomal [ 35S]methionine for 1 h postinfection. Cell lysates were immunoprecipi-
staining pattern was largely modified in size and shape, tated with the anti-HTN nucleocapsid antibody.
 ENTRY OF HANTAAN VIRUS 65

TABLE 1 Using pharmacological agents and colocalization as-

Effect of NH 4Cl and Chloroquine on HTN Infection says, we asked whether HTN infects cells by clathrin- or
by caveolae-dependent endocytosis. We used the drug
Drug % HTN nucleocapsid-positive cells chlorpromazine and high concentrations of sucrose to
inhibit clathrin-dependent endocytosis, and we used
No treatment 28.8
PMA and filipin to inhibit caveolae-dependent endocyto-
NH 4Cl (50 ␮M) 9.8
NH 4Cl (100 ␮M) 3.6 sis. Chlorpromazine inhibits the clathrin-dependent path-
Chloroquine (100 ␮M) 0 way by reducing the number of coated-pit-associated
receptors at the cell surface and by causing the accu-
Note. The effects of lysosomotropic agents were quantified as the mulation of clathrin and AP-2 in an endosomal compart-
percentage of cells expressing nucleocapsid protein at 30 h postinfec-
ment (Sofer and Futerman, 1995). Hypertonicity induced
tion. Cells were pretreated with NH 4Cl and chloroquine for 30 min and
infected with HTN in the presence or absence of agents. The results by sucrose dissociates clathrin by depleting the pool of
are expressed as the percentage of nucleocapsid protein-positive K ⫹ ions (Heuser and Anderson, 1989). Under these drug
cells. treatments, HTN infection was inhibited, and at 30 h
postinfection, expression of the nucleocapsid protein
was not observed, supporting the idea that HTN entry
To explore the biological importance of HTN targeting
takes place through the clathrin-mediated route. In con-
to late endosomes or lysosomes, we examined whether
trast, neither PMA nor filipin inhibited infectious entry of
involvement of an acidic compartment is required for
HTN. We do not exclude the possibility that many other
HTN infection by pretreating Vero E6 cells with NH 4Cl or
effects on the cell could occur during drug treatment. The
chloroquine. These drugs, within several minutes of
contrasting effects of drugs on the HTN entry, however,
treating cells, are known to raise the pH of intracellular
validate the selectivity of their action. Therefore, we have
organelles and to inhibit low-pH-dependent endosomal
concluded that entry of HTN proceeds by clathrin-depen-
fusion. To maintain the neutralized pH of the endosome,
dent but not by caveolae-dependent endocytosis. This
we infected cells in the presence of these compounds for
conclusion is further supported by the finding that HTN
45 min at 37°C. Any remaining virus was neutralized with
largely colocalizes with clathrin but not with caveolae.
antibodies, and the cells were further incubated at 37°C
Noncolocalized HTN particles might exist as vesicles
for 30 h. Cells were assayed for productive entry by
without coated clathrin in the inner part of cytoplasm.
counting the number of nucleocapsid-expressing cells.
Since only a few minutes are required for the clathrin
Treatment with 100 ␮M NH 4Cl inhibited HTN infection by
coat to wrap around vesicles and pinch off, it looks as if
⬎90%, and inhibition was concentration-dependent (Ta-
every particle cannot be colocalized with clathrin.
ble 1). In particular, treatment with 100 ␮M chloroquine
As a putative HTN receptor, the ␤3 integrin has the
completely blocked the expression of the nucleocapsid
tyrosine-based motif YXX⌽ at amino acid positions 763
protein in Vero E6 epithelial cells. Treatment of cells with
to 766 (where X is any amino acid and ⌽ is an amino acid
these concentrations did not affect the integrity or viabil-
with a bulky hydrophobic residue), which has been
ity of the cells (data not shown). In total, these results
shown to be necessary for localizing certain receptors to
demonstrate the exclusive localization of HTN in endo-
coated pits (Mellman, 1996; Trowbridge et al., 1993). This
somes or lysosomes and requirement of intracellular
motif might mediate the direct association of the AP-2
acidic compartments for HTN infection.
adaptor complexes, which in turn are connected to the
clathrin-based triskelion. In support of this concept, ad-
DISCUSSION
enovirus, coxsackievirus A9, and human parenchovirus
In this study we have defined the previously unchar- 1, which are known to be internalized via the clathrin-
acterized mechanisms of HTN entry and its subsequent dependent pathway (Huang et al., 1996; Roivainen et al.,
intracellular endocytic pathways. We demonstrated that 1994), use the ␣ v␤ 3 integrin for entry (Duan et al., 1999;
uptake of HTN from the plasma membrane occurs via Joki-Korpela et al., 2001; Wang et al., 1998). Although
dynamin-dependent endocytosis, as shown by the strong further studies are necessary to determine if this ty-
inhibition of HTN uptake by overexpression of the dom- rosine-based motif in the ␤3 integrin mediates HTN up-
inant-negative dynamin K44A mutant. Our results also take into clathrin-coated pits, our findings indicate that
showed that HTN rapidly enters into an antibody neutral- HTN enters cells by the receptor-mediated and clathrin-
ization-resistant compartment within 30 min postadsorp- dependent coated pit pathway.
tion. Rapid uptake of ligands is characteristic of clathrin- After attachment to the receptor, the virus enters the
mediated endocytosis (Mellman, 1996). The fast internal- host cell, and uncoating and release of the viral genome
ization kinetics in conjunction with dynamin-dependent follow. Early endosomes are a major sorting compart-
entry strongly support that HTN is internalized by recep- ment from which ligands may be released and trans-
tor-mediated endocytosis but not by direct fusion with ported to lysosomes and receptors recycled to the
the plasma membrane. plasma membrane or from which ligand–receptor com-
66 JIN ET AL.

plexes are transported to lysosomes (Luzio et al., 2000). late stages of the entry process, at least beyond early
A ligand, dissociated by low endosomal pH, may be endosomes.
confined to the main body of endosomes and follow a
nonselective default pathway to late endosomes and MATERIALS AND METHODS
lysosomes. The underlying mechanisms of intracellular
Cells and viruses
fusion and the intracellular site for HTN replication cur-
rently remain elusive. The only membrane fusion prop- The monkey kidney cell line Vero E6 was grown in
erties to be studied were of the La Crosse virus, also a Dulbecco’s modified eagle medium (DMEM) (Life Tech-
member of the family Bunyaviridae. After exposure to an nologies, Rockville, MD) supplemented with 10% heat-
acidic environment in vitro (Pekosz and Gonzalez- inactivated fetal bovine serum (HyClone, Logan, UT),
Scarano, 1996), the G1 protein of the La Crosse virus penicillin (50 U/ml), and streptomycin (50 ␮g/ml).
undergoes a conformational change, and the small gly- HUVECs were purchased from BioWhittaker (Walkers-
coprotein G2 becomes more susceptible to cleavage by ville, MD) and maintained in EBM-2 (BioWhittaker) me-
proteases, a process that is reminiscent of an endoso- dium containing 0.1% endothelial cell growth factor. HeLa
mal occurrence. EEA-1 is a 180-kDa peripheral mem- cells expressing the tetracycline-inducible dynamin mu-
brane protein of early endosomes (Sorkina et al., 1999). tant (K44A) were generously supplied from Sandra
Our data show that HTN completely colocalizes with Schmid (Scripps Research Institute at San Diego, CA)
EEA-1 at 90 min postinfection, indicating that the virus These cells were grown in DMEM containing puromycin
targets early endosomes. Distinct EEA-1 colocalization (400 ng/ml) and tetracycline (2 ␮g/ml) and maintained in
at 90 min has also been observed with parvovirus and a 5% CO 2 incubator at 37°C. For all experiments, Han-
certain adenovirus subtypes (Miyazawa et al., 2001; taan virus was cultivated in biosafety level 3 facilities.
Parker and Parrish, 2000). However, this is much longer Hantaan virus strain 78-116 (HTN) was propagated in
time than colocalization with endosomal marker would Vero E6 cells, which had been infected with HTN at a
be expected for cellular ligands such as transferrin or m.o.i. of 0.5. After 7 days, cultured viral supernatants
epidermal growth factor (Kornilova et al., 1996; Puertol- were harvested, and debris was removed by centrifuga-
lano et al., 2001), implying a specific, slow endocytic tion at 7900 g for 30 min. Viruses were concentrated
trafficking event for these viruses. This suggests that using Centricon Plus-80 (Millipore, Bedford, MA) filter
although HTN rapidly enters cells within vesicles, the cups according to the manufacturer’s instructions.
majority of particles remain within the endosomal com- Briefly, we used a sample filter cup to centrifuge 50 to 80
partment for up to 90 min after uptake and may be only ml of cultured virus supernatant at 3500 g for 20 min and
gradually trafficked to another compartment. Consider- then centrifuged the inverted unit again at 1000 g for 5
min. Concentrated virus aliquots were stored at ⫺70°C
ing the colocalization of HTN with the lysosomal marker
until use.
LAMP-1 at 4 h postinfection, we conclude that HTN
finally reaches late endosomes or lysosomes. This ob-
Antibodies
servation is consistent with a mode of entry similar to
that described for enveloped Semliki Forest virus, in HTN nucleocapsid-specific antibody and rat anti-HTN
which the virus is endocytosed in endosomes. Exposure antisera were generated as described previously (Kang
to the mildly acidic environment of endosomes causes et al., 1999). Rabbit anti-EEA-1 antibody was kindly pro-
conformational changes within the spike protein, trigger- vided by Harald Stenmark (The Norwegian Radium Hos-
ing a fusion of viral membranes and endosomal mem- pital, Oslo, Norway). Mouse monoclonal anti-human
branes, which releases the nucleocapsid into the cyto- LAMP-1 antibody was purchased from DBSH Hybridoma
plasm (Garoff et al., 1994; Kielian, 1995). Thus, we hy- Center (University of Iowa). Clathrin heavy chain antibody
pothesize that HTN also undergoes the conformational was purchased from Transduction Laboratories (Lexing-
changes associated with membrane fusion activity at ton, KY) and FITC- or Texas red-conjugated secondary
endosomes or lysosomes, which might be an essential antibodies were purchased from Jackson ImmunoRe-
process during the infectious life cycle of HTN. In fact, search Lab. ␤-Actin antibody, FITC- or Texas red-conju-
we found that the targeting of HTN to these acidic com- gated cholera toxin B subunit, chlorpromazine, filipin,
partments is necessary for HTN infection since lysoso- PMA, and FITC- or Texas red-conjugated transferrin were
motropic agents such as NH 4Cl and chloroquine com- purchased from Sigma (St. Louis, MO) or Molecular
pletely block viral protein synthesis. Although we do not Probes (Eugene, OR).
provide information as to where the release of the RNA
Infection and immunofluorescence
genome could take place, these data suggest that inter-
nalization of HTN capsids in early endosomes is not HeLa cells grown on glass coverslips were infected
sufficient for the initiation of a productive replication with HTN at a m.o.i. of 10. For inducing expression of the
cycle and that the viral genome may be released at the wild-type or K44A mutant dynamin, HeLa cells were
 ENTRY OF HANTAAN VIRUS 67

cultured without tetracycline for 48 h. Cells were plated starved for 30 min in a methionine-free medium, labeled
on coverslips to 50% confluency, washed twice with for 1 h with 0.1 mCi/ml [ 35S]methionine (Trans-Label;
serum-free medium, incubated with HTN for 90 min at Amersham, Arlington Heights, IL), and harvested at the
37°C, and washed once with serum-free medium. Cells indicated time points postinfection. For Western blot
were incubated for 30 h after infection. At 40 min before analysis, cells were washed twice with cold PBS and
terminating the virus growth period, 100 ␮l of 35 ␮g/ml lysed with 1% Nonidet P-40 in PBS containing 1 ␮g/ml
Texas red-conjugated transferrin was added to the me- leupeptin and 1 mM phenylmethylsulfonyl fluoride. Su-
dium to label cells for their endocytosis phenotype. Cells pernatants were obtained by centrifugation at 10,000 g.
were washed three times in cold phosphate-buffered For Western blot analysis, we separated samples that
saline (PBS) and fixed with 4% (w/v) paraformaldehyde contained equivalent amounts of protein by 12% SDS–
for 1 h, and residual aldehyde was quenched by treat- PAGE and transferred samples onto nitrocellulose mem-
ment with 50 mM NH 4Cl for 5 min. The fixed cells were branes. These membranes were incubated for 1 h with a
permeabilized with 0.02% saponin and reacted with anti- 1 to 1000 dilution of the HA-specific antibody (12CA5
HTN nucleocapsid antibody for 1 h at room temperature. mAb), a 1 to 3000 dilution of the anti-nucleocapsid anti-
The coverslips were washed in PBS and then incubated body, or a 1 to 2000 dilution of the anti-␤-actin antibody.
with FITC-conjugated goat anti-rabbit antibody. After a After washing with PBS/0.1% Tween 20 three times, we
1-h incubation, the slides were washed in PBS at least incubated membranes with the corresponding horserad-
three times and mounted in Antifade Reagent (Bio-Rad, ish peroxidase-conjugated secondary antibodies for 1 h
Hercules, CA). Data from the experiment with wild-type at 37°C, washed excess antibody three times with PBS/
and mutant dynamin were obtained by direct inspection 0.1% Tween 20, and used the Super signal chemilumi-
of the coverslips. Expression of the wild-type HA-tagged nescence substrate (Pierce, Rockford, IL) to visualize
dynamin or the K44A mutant HA-tagged dynamin was proteins.
identified by immunofluorescence labeling and then
evaluated for infection by counting the HTN-positive Lysosomotropic drug treatment
cells. For each experimental point, at least 900 cells
Ammonium chloride and chloroquine (Sigma), both
were scored.
known to neutralize low-pH organelles, were used to
study the importance of organelle acidity for HTN infec-
Inhibition of entry
tion. Cells were pretreated with 50 or 100 mM ammonium
Vero E6 cells growing on coverslips were incubated at chloride or 100 ␮M chloroquine for 30 min and then, in
37°C with chlorpromazine (5 ␮g/ml) for 30 min or su- the presence of the respective drug, cells were infected
crose (0.45 M) for 10 min to inhibit the formation of with HTN for 45 min at 37°C. At 30 h postinfection,
clathrin-coated pits (Heuser and Anderson, 1989; Oka- immunofluorescence-positive cells were scored and
moto et al., 2000). To block the caveolae-dependent path- data were represented by percentage infectivity.
way for entry, we added PMA (10 ␮M in dimethyl sulfox-
ide) or filipin (1 ␮g/ml in dimethyl sulfoxide) to cells for 45 Confocal microscopy
min at 37°C (Henley et al., 1998; Oh et al., 1998). Control Vero E6 cells and HUVECs were plated at 30% conflu-
cells were incubated in medium with or without dimethyl ency onto coverslips in 12-well plates. HTN was allowed
sulfoxide. The cells were incubated in the continued to bind to cells at 4°C for 1 h, and the cells were infected
presence of these drugs with HTN (m.o.i. of 10) and/or (m.o.i. of 100 to 1000) at a shift to 37°C. The cellular
FITC-conjugated transferrin or FITC-conjugated cholera localization of proteins was analyzed after fixing cells for
toxin for 45 min at 37°C. The cells were then washed 30 min with paraformaldehyde (4.0%). The fixed cells
three times in serum-free medium and incubated with a were permeabilized with 0.1% Triton X-100 and pro-
neutralizing concentration of rat anti-HTN serum. The cessed for immunofluorescence microscopy by a 1-h
efficiency of virus entry was analyzed by counting the incubation at room temperature with the anti-HTN nu-
number of positive cells by immunofluorescence micros- cleocapsid antibody and either anti-EEA-1 or anti-
copy at 30 h postinfection. Data were represented by the LAMP-1, followed by incubation with the appropriate sec-
percentage of infection, which was calculated as the ondary antibodies. We used a Zeiss Axiophot micro-
number of infected cells divided by the total number of scope equipped with Bio-Rad confocal optics to obtain
cells counted multiplied by 100. all confocal images.
Metabolic labeling, immunoprecipitation, and Western
ACKNOWLEDGMENTS
blot analysis
This work was supported by a grant from KOSEF (R01-1999-00143).
We metabolically labeled cells with methionine using We thank S. Schmid, Harald Stenmark, and the Developmental Studies
previously described procedures (Lee et al., 2000). Hybridoma Bank at the University of Iowa for the generous gift of cell
Briefly, after HTN infection, cells were methionine- lines and antibodies.
68 JIN ET AL.

REFERENCES Killisch, I., Steinlein, P., Romisch, K., Hollinshead, R., Beug, H., and
Griffiths, G. (1992). Characterization of early and late endocytic com-
Anderson, H. A., Chen, Y., and Norkin, L. C. (1996). Bound simian virus partments of the transferrin cycle. Transferrin receptor antibody
40 translocates to caveolin-enriched membrane domains, and its blocks erythroid differentiation by trapping the receptor in the early
entry is inhibited by drugs that selectively disrupt caveolae. Mol. Biol. endosome. J. Cell Sci. 103, 211–232.
Cell. 7, 1825–1834. Ko, K. W. (1992). Haemorrhagic fever with renal syndrome: Clinical
Arikawa, J., Lapenotiere, H. F., Iacono-Connors, L., Wang, M. L., and aspects. Pediatr. Nephrol. 6, 197–200.
Schmaljohn, C. S. (1990). Coding properties of the S and the M Kornilova, E., Sorkina, T., Beguinot, L., and Sorkin, A. (1996). Lysosomal
genome segments of Sapporo rat virus: Comparison to other caus- targeting of epidermal growth factor receptors via a kinase-depen-
ative agents of hemorrhagic fever with renal syndrome. Virology 176, dent pathway is mediated by the receptor carboxyl-terminal residues
114–125. 1022–1123. J. Biol. Chem. 271, 30340–30346.
Baba, T., Damke, H., Hinshaw, J. E., Ikeda, K., Schmid, S. L., and Lee, H. W., Baek, L. J., and Johnson, K. M. (1982). Isolation of Hantaan
Warnock, D. E. (1995). Role of dynamin in clathrin-coated vesicle virus, the etiologic agent of Korean hemorrhagic fever, from wild
formation. Cold Spring Harbor Symp. Quant. Biol. 60, 235–242. urban rats. J. Infect. Dis. 146, 638–644.
Bartlett, J. S., Wilcher, R., and Samulski, R. J. (2000). Infectious entry Lee, M., Lee, J. S., and Kim, B. K. (1983). Disseminated intravascular
pathway of adeno-associated virus and adeno-associated virus vec- coagulation in Korean hemorrhagic fever. Bibl. Haematol. 49, 181–
tors. J. Virol. 74, 2777–2785. 199.
Damke, H., Baba, T., Warnock, D. E., and Schmid, S. L. (1994). Induction Lee, S., Yoon, J., Park, B., Jun, Y., Jin, M., Sung, H. C., Kim, I. H., Kang, S.,
of mutant dynamin specifically blocks endocytic coated vesicle for- Choi, E. J., Ahn, B. Y., and Ahn, K. (2000). Structural and functional
mation. J. Cell Biol. 127, 915–934. dissection of human cytomegalovirus US3 in binding major histo-
Damke, H., Gossen, M., Freundlieb, S., Bujard, H., and Schmid, S. L. compatibility complex class I molecules. J. Virol. 74, 11262–11269.
(1995). Tightly regulated and inducible expression of dominant inter- Luzio, J. P., Rous, B. A., Bright, N. A., Pryor, P. R., Mullock, B. M., and
fering dynamin mutant in stably transformed HeLa cells. Methods Piper, R. C. (2000). Lysosome–endosome fusion and lysosome bio-
Enzymol. 257, 209–220. genesis. J. Cell Sci. 113, 1515–1524.
Duan, D., Li, Q., Kao, A. W., Yue, Y., Pessin, J. E., and Engelhardt, J. F. Marsh, M., Matlin, K., Simons, K., Reggio, H., White, J., Kartenbeck, J.,
(1999). Dynamin is required for recombinant adeno-associated virus and Helenius, A. (1982). Are lysosomes a site of enveloped-virus
type 2 infection. J. Virol. 73, 10371–10376. penetration? Cold Spring Harbor Symp. Quant. Biol. 46, 835–843.
Edwards, J., and Brown, D. T. (1991). Sindbis virus infection of a Chinese Mellman, I. (1996). Endocytosis and molecular sorting. Annu. Rev. Cell.
hamster ovary cell mutant defective in the acidification of endo- Dev. Biol. 12, 575–625.
somes. Virology 182, 28–33. Miyazawa, N., Crystal, R. G., and Leopold, P. L. (2001). Adenovirus
Elliott, L. H., Kiley, M. P., and Mccormick, J. B. (1984). Hantaan virus:
serotype 7 retention in a late endosomal compartment prior to
Identification of virion proteins. J. Gen. Virol. 65, 1285–1293.
cytosol escape is modulated by fiber protein. J. Virol. 75, 1387–1400.
Garoff, H., Wilschut, J., Liljestrom, P., Wahlberg, J. M., Bron, R., Suoma-
Oh, P., Mcintosh, D. P., and Schnitzer, J. E. (1998). Dynamin at the neck
lainen, M., Smyth, J., Salminen, A., Barth, B. U., Zhao, H., et al. (1994).
of caveolae mediates their budding to form transport vesicles by
Assembly and entry mechanisms of Semliki Forest virus. Arch. Virol.
GTP-driven fission from the plasma membrane of endothelium. J. Cell
Suppl. 9, 329–338.
Biol. 141, 101–114.
Gavrilovskaya, I. N., Brown, E. J., Ginsberg, M. H., and Mackow, E. R.
Okamoto, Y., Ninomiya, H., Miwa, S., and Masaki, T. (2000). Cholesterol
(1999). Cellular entry of hantaviruses which cause hemorrhagic fever
oxidation switches the internalization pathway of endothelin receptor
with renal syndrome is mediated by beta3 integrins. J. Virol. 73,
type A from caveolae to clathrin-coated pits in Chinese hamster
3951–3959.
ovary cells. J. Biol. Chem. 275, 6439–6446.
Gavrilovskaya, I. N., Shepley, M., Shaw, R., Ginsberg, M. H., and
Parker, J. S., and Parrish, C. R. (2000). Cellular uptake and infection by
Mackow, E. R. (1998). Beta3 integrins mediate the cellular entry of
canine parvovirus involves rapid dynamin-regulated clathrin-medi-
hantaviruses that cause respiratory failure. Proc. Natl. Acad. Sci. USA
95, 7074–7079. ated endocytosis, followed by slower intracellular trafficking. J. Virol.
Gossen, M., and Bujard, H. (1992). Tight control of gene expression in 74, 1919–1930.
mammalian cells by tetracycline-responsive promoters. Proc. Natl. Pekosz, A., and Gonzalez-Scarano, F. (1996). The extracellular domain
Acad. Sci. USA 89, 5547–5551. of La Crosse virus G1 forms oligomers and undergoes pH-depen-
Gruenberg, J., and Howell, K. E. (1989). Membrane traffic in endocyto- dent conformational changes. Virology 225, 243–247.
sis: Insights from cell-free assays. Annu. Rev. Cell Biol. 5, 453–481. Pelkmans, L., Kartenbeck, J., and Helenius, A. (2001). Caveolar endo-
Helenius, A., and Marsh, M. (1982). Endocytosis of enveloped animal cytosis of simian virus 40 reveals a new two-step vesicular-transport
viruses. Ciba Found. Symp. 92, 59–76. pathway to the ER. Nat. Cell Biol. 3, 473–483.
Henley, J. R., Krueger, E. W., Oswald, B. J., and Mcniven, M. A. (1998). Pensiero, M. N., Sharefkin, J. B., Dieffenbach, C. W., and Hay, J. (1992).
Dynamin-mediated internalization of caveolae. J. Cell Biol. 141, 85– Hantaan virus infection of human endothelial cells. J. Virol. 66, 5929–
99. 5936.
Heuser, J. E., and Anderson, R. G. (1989). Hypertonic media inhibit Pho, M. T., Ashok, A., and Atwood, W. J. (2000). JC virus enters human
receptor-mediated endocytosis by blocking clathrin-coated pit for- glial cells by clathrin-dependent receptor-mediated endocytosis.
mation. J. Cell Biol. 108, 389–400. J. Virol. 74, 2288–2292.
Huang, S., Kamata, T., Takada, Y., Ruggeri, Z. M., and Nemerow, G. R. Puertollano, R., Martinez-Menarguez, J. A., Batista, A., Ballesta, J., and
(1996). Adenovirus interaction with distinct integrins mediates sepa- Alonso, M. A. (2001). An intact dilysine-like motif in the carboxyl
rate events in cell entry and gene delivery to hematopoietic cells. terminus of MAL is required for normal apical transport of the
J. Virol. 70, 4502–4508. influenza virus hemagglutinin cargo protein in epithelial Madin–
Joki-Korpela, P., Marjomaki, V., Krogerus, C., Heino, J., and Hyypia, T. Darby canine kidney cells. Mol. Biol. Cell 12, 1869–1883.
(2001). Entry of human parechovirus 1. J. Virol. 75, 1958–1967. Roivainen, M., Piirainen, L., Hovi, T., Virtanen, I., Riikonen, T., Heino, J.,
Kang, J. I., Park, S. H., Lee, P. W., and Ahn, B. Y. (1999). Apoptosis is and Hyypia, T. (1994). Entry of coxsackievirus A9 into host cells:
induced by hantaviruses in cultured cells. Virology 264, 99–105. Specific interactions with alpha v beta 3 integrin, the vitronectin
Kielian, M. (1995). Membrane fusion and the alphavirus life cycle. Adv. receptor. Virology 203, 357–365.
Virus Res. 45, 113–151. Roy, A. M., Parker, J. S., Parrish, C. R., and Whittaker, G. R. (2000). Early
 ENTRY OF HANTAAN VIRUS 69

stages of influenza virus entry into Mv-1 lung cells: Involvement of and eps15 in endosomes containing activated epidermal growth
dynamin. Virology 267, 17–28. factor receptors. J. Cell Sci. 112, 317–327.
Schmaljohn, C. S., and Dalrymple, J. M. (1983). Analysis of Hantaan Tabuchi, M., Yoshimori, T., Yamaguchi, K., Yoshida, T., and Kishi, F.
virus RNA: Evidence for a new genus of Bunyaviridae. Virology 131, (2000). Human NRAMP2/DMT1, which mediates iron transport
482–491. across endosomal membranes, is localized to late endosomes and
Schmid, S., Fuchs, R., Kielian, M., Helenius, A., and Mellman, I. (1989). lysosomes in HEp-2 cells. J. Biol. Chem. 275, 22220–22228.
Acidification of endosome subpopulations in wild-type Chinese ham- Trowbridge, I. S., Collawn, J. F., and Hopkins, C. R. (1993). Signal-
ster ovary cells and temperature-sensitive acidification-defective dependent membrane protein trafficking in the endocytic pathway.
mutants. J. Cell Biol. 108, 1291–1300. Annu. Rev. Cell Biol. 9, 129–161.
Van Der Bliek, A. M., Redelmeier, T. E., Damke, H., Tisdale, E. J.,
Schmid, S. L. (1997). Clathrin-coated vesicle formation and protein
Meyerowitz, E. M., and Schmid, S. L. (1993). Mutations in human
sorting: An integrated process. Annu. Rev. Biochem. 66, 511–548.
dynamin block an intermediate stage in coated vesicle formation.
Schnitzer, J. E., Oh, P., Pinney, E., and Allard, J. (1994). Filipin-sensitive
J. Cell Biol. 122, 553–563.
caveolae-mediated transport in endothelium: Reduced transcytosis,
Wang, K., Huang, S., Kapoor-Munshi, A., and Nemerow, G. (1998).
scavenger endocytosis, and capillary permeability of select macro- Adenovirus internalization and infection require dynamin. J. Virol. 72,
molecules. J. Cell Biol. 127, 1217–1232. 3455–3458.
Skretting, G., Torgersen, M. L., Van Deurs, B., and Sandvig, K. (1999). Wilson, J. M., De Hoop, M., Zorzi, N., Toh, B. H., Dotti, C. G., and Parton,
Endocytic mechanisms responsible for uptake of GPI-linked diphthe- R. G. (2000). EEA1, a tethering protein of the early sorting endosome,
ria toxin receptor. J. Cell Sci. 112, 3899–3909. shows a polarized distribution in hippocampal neurons, epithelial
Sofer, A., and Futerman, A. H. (1995). Cationic amphiphilic drugs inhibit cells, and fibroblasts. Mol. Biol. Cell 11, 2657–2671.
the internalization of cholera toxin to the Golgi apparatus and the Yoshimura, A., Kuroda, K., Kawasaki, K., Yamashina, S., Maeda, T., and
subsequent elevation of cyclic AMP. J. Biol. Chem. 270, 12117–12122. Ohnishi, S. (1982). Infectious cell entry mechanism of influenza virus.
Sorkina, T., Bild, A., Tebar, F., and Sorkin, A. (1999). Clathrin, adaptors J. Virol. 43, 284–293.