JOURNAL OF VIROLOGY, Nov. 2011, p. 12067–12072
Copyright © 2011, American Society for Microbiology. All Rights Reserved.
Vol. 85, No. 22
The Infectivity of prM-Containing Partially Mature West Nile Virus
Does Not Require the Activity of Cellular Furin-Like Proteases?
Swati Mukherjee, Tsai-Yu Lin, Kimberly A. Dowd, Carolyn J. Manhart, and Theodore C. Pierson*
Viral Pathogenesis Section, Laboratory of Viral Diseases, National Institute of Allergy and
Infectious Disease, National Institutes of Health, Bethesda, Maryland
Received 29 June 2011/Accepted 22 August 2011
Cleavage of the flavivirus prM protein by a cellular furin-like protease is a hallmark of virion maturation.
While this cleavage is a required step in the viral life cycle, it can be inefficient. Virions that retain uncleaved
prM may be infectious. We investigated whether cleavage by furin of prM on partially mature West Nile virus
(WNV) during virus entry contributes to infectivity. Using quantitative assays of WNV infection, we found that
virions incorporating considerable amounts of uncleaved prM protein were insensitive to treatment of cells
with a potent inhibitor of furin activity. Thus, partially mature WNV does not require furin-like proteases for
Flaviviruses are a group of positive-stranded RNA viruses
responsible for considerable global morbidity and mortality.
West Nile virus (WNV) is a mosquito-borne neurotropic mem-
ber of this genus that is an emerging pathogen (16). Flavivi-
ruses are small spherical particles that incorporate two viral
glycoproteins that orchestrate virus entry into cells (reviewed
in reference 18). The envelope (E) protein is an approximately
53-kDa protein that mediates virus attachment to cells and
drives the pH-dependent fusion of viral and cellular mem-
branes in the endosomes of infected cells. On immature viri-
ons, the 20-kDa premembrane (prM) protein interacts with the
E protein to form 60 heterotrimeric spikes and functions to
prevent adventitious fusion between viral and cellular mem-
branes during egress through acidic compartments of the se-
cretory pathway (9, 30, 31). Transit through these compart-
ments catalyzes a pH-dependent rearrangement of prM and E
proteins on the virion that exposes a recognition motif on prM
for cellular furin-like serine proteases (15, 27, 28). Cleavage of
prM results in the formation of an approximately 14-kDa “pr”
protein and a small membrane-anchored M peptide. Release
of the virion into the neutral environment of the extracellular
space promotes the release of the pr protein and the formation
of mature virus particles (28).
While the cleavage of prM is a critical step in the flavivirus
life cycle (5), the extent of cleavage required for virus infec-
tivity is not clear. Biochemical studies of the prM content of
mosquito-borne flaviviruses released from cells suggest that
cleavage can be inefficient (6, 9, 10, 13, 14, 17). The demon-
stration that more than 90% of dengue virus (DENV) particles
can be precipitated with a prM-reactive antibody suggests that
“partially mature” virions may be a significant component of
the population of virus particles released from cells (12). Sev-
eral lines of evidence support the idea that uncleaved prM is
present on infectious virions (3, 8). The presence of prM on
virions has been shown to increase the sensitivity of virus
particles to neutralization by some E-specific antibodies (1,
19). Furthermore, antibodies specific for prM enhance virus
infectivity in vitro and in vivo (2, 4, 11, 25, 29). Enhancing
understanding of the stoichiometry of prM cleavage required
for infectivity would facilitate a deeper conception of the com-
plexity arising from heterogeneity in populations of virions
released from infected cells. Complicating this analysis is the
intriguing possibility that prM on virus particles is also cleaved
by furin in acidic compartments of the endosome during virus
entry. In support of this possibility, antibody-dependent en-
hancement (ADE) of prM-containing “immature” DENV vi-
rions (with no detectable cleavage of prM) produced in furin-
deficient Lovo cells was found to be blocked by treatment with
an inhibitor of furin-like proteases (25).
To investigate the requirement for cleavage of prM by a
furin-like protease on WNV during the virus entry process, we
performed a series of infection-inhibition studies using com-
mercially available inhibitors of furin. Control experiments
performed using recombinant furin and a well-characterized
fluorogenic substrate revealed dose-dependent furin-like pro-
tease activity that could be inhibited by the inhibitor Dec-
RVKR-CMK (Fig. 1A and B) (7). Furin-like protease activity
was also measured in a Raji B cell line that stably expresses the
C-type lectin DC-SIGNR (Fig. 1C); these cells are highly per-
missive to WNV infection (3). Furin-like protease activity was
also measured in a variety of other cell lines (Vero, HEK-293T,
BHK-21, and K562) (data not shown). In each instance, treat-
ment of 5 ? 104cells, corresponding to the number of cells
used in the infectivity studies described below, with furin in-
hibitor (FI) resulted in a reduction of substrate cleavage to
background levels (Fig. 1D). Two additional inhibitors tested
(hexa-D-arginine and anthrax lethal factor protease inhibitor)
were capable of inhibiting the activity of recombinant furin to
various degrees but were considerably less potent in assays
with cells and were not used in subsequent studies (data not
We next investigated the impact of inhibition of the activity
* Corresponding author. Mailing address: Viral Pathogenesis Sec-
tion, Laboratory of Viral Diseases, National Institutes of Health, 33
North Drive, Building 33, Room 2E19A.2, Bethesda, MD 20892.
Phone: (301) 451-7977. Fax: (301) 451-7978. E-mail: piersontc@mail
?Published ahead of print on 31 August 2011.
of furin-like proteases on the infectivity of WNV. These stud-
ies were performed using an infectious clone of WNV lineage
I (NY99 strain) engineered to express green fluorescent pro-
tein (GFP) by the use of a modification of a previously de-
scribed molecular clone strategy (T. Y. Lin and T. Pierson,
submitted for publication) (26). Raji-DC-SIGNR cells were
preincubated with FI (50 ?M) and infected with serial 2-fold
dilutions of virus stocks produced in mammalian (HEK-293T,
Vero, or BHK) or insect (C6/36) cells. Virus infectivity was
measured 16 h postinfection using flow cytometry; this time
point was selected because it reflects a single round of infection
with the GFP-expressing virus (Fig. 2A to D). Overall, the
efficiencies of infection in the presence and absence of FI at
each dilution of virus were similar; the largest mean reduction
in infection of FI-treated cells was approximately 13% (n ? 3).
While FI treatment was not overtly toxic to cells (data not
shown), it did result in a reduction of approximately 10% in
GFP expression, as measured by mean channel fluorescence of
GFP in infected Raji-DC-SIGNR cells, identifying a potential
indirect effect of inhibiting furin-like enzymes on viral replica-
tion. Because prM-containing virions may represent a modest
proportion of infectious WNV virions released from cells,
analysis of the infectivity of the entire population of virions by
the use of an inhibitor of furin may be an insensitive test.
Therefore, we used two additional approaches to focus our
analysis on those virions that exhibited incorporation of prM.
In this context, a significant impact of FI would be expected if
cleavage of prM during virus entry were a required event in the
life cycle of partially mature viruses.
First, we measured the infectivity of WNV lineage II strain
956 in the presence or absence of FI. This WNV strain lacks an
asparagine-linked carbohydrate on the E protein but is glyco-
sylated on prM (10). Interactions between WNV and Raji-DC-
SIGNR cells are mediated by carbohydrates and therefore
require the presence of prM on the WNV 956 virion; use of
Raji-DC-SIGNR cells in infectivity studies preferentially se-
lects for WNV 956 virions that retain uncleaved prM (3). A
variant of WNV 956 that encodes GFP was produced from a
molecular clone by the use of HEK-293T cells as described
previously (23). Raji-DC-SIGNR cells were incubated in the
presence or absence of FI and infected with serial 2-fold dilu-
tions of virus. Infectivity was monitored at 16 h postinfection
using flow cytometry. Treatment of cells with FI did not sig-
nificantly reduce the infectivity of WNV 956 (Fig. 2E; n ? 2).
Because attachment to and infection of Raji-DC-SIGNR cells
by this strain is restricted to the subset of virus particles that
retain uncleaved prM, these results demonstrate that complete
cleavage of prM is not an absolute requirement for infectivity.
We next performed similar studies using populations of vi-
rions that vary markedly with respect to the efficiency of prM
cleavage. WNV reporter virus particles (RVPs) are pseudoin-
fectious virions produced by genetic complementation of a
subgenomic replicon that carries a reporter gene (24). Ap-
proaches to modulate the efficiency of maturation of WNV
FIG. 1. Detection and inhibition of furin-like protease activity. (A) The indicated mass of recombinant furin (R&D Systems) was diluted in
assay buffer (0.5% Triton X-100, 0.5 mM CaCl2, 100 mM HEPES, pH 7.0) and incubated at 30°C in the presence or absence of ERTKR-AMC
fluorogenic substrate (R&D Systems) (100 ?M). Fluorescence was measured at 460 nM every 45 s for 11 h using a SpectraMax M5 reader
(Molecular Devices) (excitation wavelength of 380 nM). Data represent relative fluorescence units (RFU) obtained from experiments using
triplicate wells; error bars reflect standard errors of the mean. (B) Recombinant furin was incubated in the presence of a previously described furin
inhibitor (FI), Dec-RVKR-CMK (Enzo Life Sciences) (50 ?M). Data represent mean fluorescence measured using triplicate wells; error bars
reflect standard errors of the means. (C) The activity of furin-like proteases in Raji-DC-SIGNR cells was measured as described for panel A.
Fluorescence was monitored as described above. Data represent averages of results determined using triplicate wells. (D) The potency of the FI
in Raji-DC-SIGNR cells was monitored as described for panels A and B. Cells were incubated in the presence of FI for 30 min, lysed in assay buffer,
added to the substrate, and incubated at 30°C. A total of 50,000 cells were selected for study, because that was the number of cells used in all the
functional experiments described here. The data presented in each panel are representative of the results of three independent experiments.
12068 NOTESJ. VIROL.
RVPs have been previously described (19); ectopic expression
of human furin in RVP-producing cells markedly increases the
efficiency of prM cleavage, whereas treatment of producer cells
with NH4Cl reduces the extent of maturation. WNV RVPs
were produced in HEK-293T cells in the presence of furin
overexpression or NH4Cl. A marked difference in the efficien-
cies of prM cleavage when RVPs were produced using these
methods was verified by Western blot analysis (Fig. 3A). Sim-
ilar results were obtained using WNV RVPs produced using a
variant of the prM protein that encodes a V5 tag (V5 RVP)
(Fig. 3C). Both stocks of RVPs were subjected to titration on
Raji-DC-SIGNR cells in the presence or absence of FI (Fig. 3B
and D). As expected, the specific infectivity of RVPs produced
in the presence of NH4Cl was markedly reduced in comparison
to that seen with the more mature preparation (19). Notably,
no significant reduction of infectivity was observed when target
Raji-DC-SIGNR cells were incubated in the presence of FI,
despite the considerable prM content of each NH4Cl prepara-
tion (n ? 4 and 6 for WNV RVP and WNV VS RVP, respec-
The neutralization potency of many E protein-specific anti-
bodies is strongly influenced by the maturation state of the
virion. Several E protein-specific monoclonal antibodies
(MAbs) bind virions more efficiently in the presence of un-
cleaved prM (1, 19) and are useful functional probes for de-
tection of the presence of prM on infectious virions. To con-
firm that the significant uncleaved prM detected in NH4Cl
preparations was present on infectious virions, neutralization
studies were performed with maturation state-sensitive anti-
bodies. MAb E16 (domain III) was capable of equivalently
FIG. 2. WNV infection of Raji-DC-SIGNR cells does not require the activity of a furin-like protease. (A to D) Raji-DC-SIGNR cells were
incubated in the presence or absence of 50 ?M furin inhibitor (FI) and then infected with serial 2-fold dilutions of WNV lineage I (strain NY99)
produced in HEK-293T, Vero, BHK, or C6/36 cells. Infection was monitored as a function of GFP expression at 16 h postinfection by flow
cytometry. The WNV encoding GFP used in these studies is described elsewhere (T.-Y. Lin and T. C. Pierson, submitted for publication). The
averages of the results of duplicate infections are shown for each dilution; error bars represent standard errors of the means. The data presented
are representative of the results of three independent experiments. (E) Raji-DC-SIGNR cells were incubated in the presence or absence of 50 ?M
FI and infected with serial 2-fold dilutions of lineage II (strain 956) WNV that encodes GFP and was generated in HEK-293T cells. The averages
of the results of duplicate infections are shown for each dilution; error bars represent standard errors of the means. The data presented are
representative of the results of two independent experiments.
VOL. 85, 2011NOTES12069
neutralizing furin and NH4Cl RVPs (Fig. 3E), as expected
(P ? 0.17; n ? 7) (19–21), whereas furin RVPs were consid-
erably more resistant to neutralization by the maturation-sen-
sitive MAbs E53 (domain II) and E121 (domain I) than the
NH4Cl RVP preparations (19, 22) (Fig. 3F and G). The neu-
tralization potency (i.e., the 50% effective concentration
[EC50]) of E16 (P ? 0.41; n ? 7), E53 (P ? 0.87; n ? 7), and
E121 (P ? 0.41; n ? 3) was not impacted significantly by FI
treatment. Altogether, these studies demonstrate that infectiv-
ity of RVPs that retain significant uncleaved prM does not
require processing by furin during virus entry.
Antibody-dependent enhancement (ADE) of immature
DENV infection has been shown to be sensitive to treatment
with FI (25). To investigate whether the activity of a furin-like
protease is required during Fc-? receptor-mediated entry of
WNV into cells, ADE experiments were performed using furin
and NH4Cl RVPs and a panel of six maturation-sensitive an-
tibodies. RVPs were incubated with serial dilutions of antibody
for 1 h and then added to Fc-?RIIA-expressing K562 cells in
the presence or absence of FI. All MAbs were capable of
enhancing infection of both furin and NH4Cl RVPs. The levels
of impact of FI treatment were minimal and equivalent for the
FIG. 3. Decreasing the extent of virion maturation does not confer sensitivity of virus infection as a result of furin inhibitor treatment. WNV RVPs
were produced in HEK-293T cells in the presence of the weak base NH4Cl or of a plasmid that ectopically expresses furin. (A) The efficiency of prM
cleavage in RVPs released from transfected cells was evaluated. RVPs were partially purified by centrifugation through a 20% sucrose cushion, followed
by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting with the anti-E MAb 4G2 and anti-prM polyclonal
antibody (Imgenex). (B) The sensitivity of two independent preparations of furin and NH4Cl RVPs to furin inhibitor (FI) was evaluated following the
infection of Raji-DC-SIGNR cells. Serial 2-fold dilutions of RVPs were added to cells in the presence or absence of FI. Infectivity was monitored as a
function of GFP expression as measured by flow cytometry 2 days postinfection. Infections were performed in duplicate. The infectivity of each stock is
expressed as a function of the RNA content of the preparation. The data represent the results of four independent assays performed using two
independent stocks of furin and NH4Cl RVPs. (C) RVPs were produced using a modified structural gene construct that encodes a V5 tag immediately
after the signal cleavage site of prM. The tag was introduced into the prM gene by the use of overlap extension PCR and conventional cloning
methodologies. The efficiency of maturation of V5 RVP stocks was established as described for panel A using a commercially available anti-V5 antibody
(Invitrogen). (D) The infectivity of V5 RVPs in the presence and absence of FI was studied as described for panel B. The data presented are
representative of the results of a total of six independent experiments performed with three independent stocks of furin and NH4Cl-V5 RVPs. (E to G)
Antibody dose-response studies were performed using furin and NH4Cl RVPs. Duplicate serial 4-fold dilutions of MAb E16 (panel E; E-DIII lateral
ridge) or the maturation state-sensitive MAbs E53 (panel F; E-DII fusion loop) and E121 (panel G; E-DI lateral ridge) were incubated with RVPs for
1 h at room temperature prior to the addition of Raji-DC-SIGNR cells incubated with or without 50 ?M FI. Virus infection was scored as a function
of GFP expression 2 days later using flow cytometry. Data were analyzed using nonlinear regression with a variable slope. The data presented are
representative of the results of seven, seven, and three independent experiments performed with E16, E53, and E121, respectively.
12070 NOTESJ. VIROL.
prM-containing (NH4Cl) and prM-deficient (furin) RVP prep-
arations (Fig. 4A to F). To explore whether the enhancement
mediated by antibodies that recognize prM was sensitive to FI
treatment, experiments were performed using V5 RVPs and a
commercial anti-V5 antibody. Significant enhancement was
observed with RVPs produced in the presence of NH4Cl but
not with exogenously expressed furin, as expected, and was not
sensitive to the presence of FI in K562 target cells (Fig. 4G).
Flavivirus-infected cells release a heterogeneous population
of virions that differ with respect to the extent of virion mat-
uration. While uncleaved prM is readily apparent in biochem-
ical studies of flaviviruses, the distribution of uncleaved prM
molecules among infectious virions is not known. In this study,
we investigated whether the infectivity of incompletely cleaved
virions requires processing of prM by furin-like proteases dur-
ing entry. Our studies failed to identify a subset of virions
whose infectivity was sensitive to inhibition of furin-like pro-
teases, even when the efficiency of virion maturation was sig-
nificantly reduced. A previous study of immature DENV pro-
duced in Lovo cells suggested a requirement for cleavage of
prM during antibody-mediated entry of virus into K562 cells
that was not apparent in our studies of WNV (25). Several
factors may account for the differences between these two
studies. One interesting possibility is that the stoichiometric
requirements for cleavage of prM may differ between DENV
and WNV. While inhibitors of furin markedly decrease cleav-
age of WNV prM during virus egress (data not shown), it is
possible that as-yet-unidentified proteases that are not sensi-
tive to inhibitors of furin-like proteases may cleave partially
mature WNV during entry. The sensitivity of DENV infectivity
to FI was established principally by using experiments that
measured virus release (25). A quantitative assessment of the
size and contribution of the FI-sensitive population of DENV
by the use of indirect assays may be difficult, as small changes
in the number of infection events could translate into large
differences in virus output. This is particularly true if FI-treat-
ment has a subtle impact on the viral life cycle that does not
relate directly to prM cleavage. While we cannot rule out the
possibility that virions containing essentially unprocessed prM
may be cleaved by furin during virus entry, our data suggest
that particles at this extreme of the maturation spectrum rep-
resent a very small fraction of the prM-containing partially
mature WNV released from cells. Taken together, our studies
suggest that the infectivity of prM-containing partially mature
WNV is not dependent on cleavage of prM during virus entry.
This work was supported by the Intramural Research Program of the
NIH, National Institute of Allergy and Infectious Diseases (NIAID).
We are grateful to Michael Diamond for his thoughtful comments
on our manuscript. We thank members of our laboratory for useful
discussions and their comments on the manuscript.
1. Cherrier, M. V., et al. 2009. Structural basis for the preferential recognition
of immature flaviviruses by a fusion-loop antibody. EMBO J. 28:3269–3276.
2. Colpitts, T. M., et al. 22 June 2011, posting date. prM-antibody renders
immature West Nile virus infectious in vivo. J. Gen. Virol. doi:10.1099/
vir.0.031427-0. [Epub ahead of print.]
3. Davis, C. W., et al. 2006. West Nile virus discriminates between DC-SIGN
and DC-SIGNR for cellular attachment and infection. J. Virol. 80:1290–
4. Dejnirattisai, W., et al. 2010. Cross-reacting antibodies enhance dengue
virus infection in humans. Science 328:745–748.
5. Elshuber, S., S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Cleavage of
protein prM is necessary for infection of BHK-21 cells by tick-borne enceph-
alitis virus. J. Gen. Virol. 84:183–191.
6. Elshuber, S., and C. W. Mandl. 2005. Resuscitating mutations in a furin
cleavage-deficient mutant of the flavivirus tick-borne encephalitis virus.
J. Virol. 79:11813–11823.
7. Garten, W., et al. 1994. Processing of viral glycoproteins by the subtilisin-like
endoprotease furin and its inhibition by specific peptidylchloroalkylketones.
8. Guirakhoo, F., R. A. Bolin, and J. T. Roehrig. 1992. The Murray Valley
encephalitis virus prM protein confers acid resistance to virus particles and
alters the expression of epitopes within the R2 domain of E glycoprotein.
9. Guirakhoo, F., F. X. Heinz, C. W. Mandl, H. Holzmann, and C. Kunz. 1991.
Fusion activity of flaviviruses: comparison of mature and immature (prM-
containing) tick-borne encephalitis virions. J. Gen. Virol. 72(Pt. 6):1323–
10. Hanna, S. L., et al. 2005. N-linked glycosylation of West Nile virus envelope
proteins influences particle assembly and infectivity. J. Virol. 79:13262–
FIG. 4. Antibody-dependent enhancement of prM-containing virions is not blocked by furin inhibitor treatment. (A to G) Antibody dose-
response studies were performed using furin and NH4Cl RVPs produced and characterized as part of the experiments described for Fig. 3.
Duplicate serial 4-fold dilutions of MAbs E121 (A), E53 (B), E28 (C), E113 (D), E111 (E), and E34 (F) and of anti-V5 (G) were incubated with
RVPs for 1 h at room temperature prior to the addition of K562 cells in the presence or absence of 50 ?M furin inhibitor (FI). The WNV E-specific
antibodies used in these studies have been described previously (21, 22). Virus infection was scored as a function of GFP expression on day 2 using
flow cytometry. Data are presented as the severalfold increases in infection relative to the levels of infectivity observed in the absence of antibody
and are representative of the results of two independent experiments.
VOL. 85, 2011NOTES12071
11. Huang, K. J., et al. 2006. The dual-specific binding of dengue virus and target Download full-text
cells for the antibody-dependent enhancement of dengue virus infection.
J. Immunol. 176:2825–2832.
12. Junjhon, J., et al. 2010. Influence of pr-M cleavage on the heterogeneity of
extracellular dengue virus particles. J. Virol. 84:8353–8358.
13. Khromykh, A. A., A. N. Varnavski, and E. G. Westaway. 1998. Encapsidation
of the flavivirus kunjin replicon RNA by using a complementation system
providing Kunjin virus structural proteins in trans. J. Virol. 72:5967–5977.
14. Kimura, T., and A. Ohyama. 1988. Association between the pH-dependent
conformational change of West Nile flavivirus E protein and virus-mediated
membrane fusion. J. Gen. Virol. 69(Pt. 6):1247–1254.
15. Li, L., et al. 2008. The flavivirus precursor membrane-envelope protein
complex: structure and maturation. Science 319:1830–1834.
16. Mackenzie, J. S., D. J. Gubler, and L. R. Petersen. 2004. Emerging flavivi-
ruses: the spread and resurgence of Japanese encephalitis, West Nile and
dengue viruses. Nat. Med. 10:S98–S109.
17. Moesker, B., I. A. Rodenhuis-Zybert, T. Meijerhof, J. Wilschut, and J. M.
Smit. 2010. Characterization of the functional requirements of West Nile
virus membrane fusion. J. Gen. Virol. 91:389–393.
18. Mukhopadhyay, S., R. J. Kuhn, and M. G. Rossmann. 2005. A structural
perspective of the flavivirus life cycle. Nat. Rev. Microbiol. 3:13–22.
19. Nelson, S., et al. 2008. Maturation of West Nile virus modulates sensitivity to
antibody-mediated neutralization. PLoS Pathog. 4:e1000060.
20. Nybakken, G. E., et al. 2005. Structural basis of West Nile virus neutraliza-
tion by a therapeutic antibody. Nature 437:764–769.
21. Oliphant, T., et al. 2005. Development of a humanized monoclonal antibody
with therapeutic potential against West Nile virus. Nat. Med. 11:522–530.
22. Oliphant, T., et al. 2006. Antibody recognition and neutralization determi-
nants on domains I and II of West Nile virus envelope protein. J. Virol.
23. Pierson, T. C., et al. 2005. An infectious West Nile virus that expresses a GFP
reporter gene. Virology 334:28–40.
24. Pierson, T. C., et al. 2006. A rapid and quantitative assay for measuring
antibody-mediated neutralization of West Nile virus infection. Virology 346:
25. Rodenhuis-Zybert, I. A., et al. 2010. Immature dengue virus: a veiled patho-
gen? PLoS Pathog. 6:e1000718.
26. Shustov, A. V., P. W. Mason, and I. Frolov. 2007. Production of pseudoin-
fectious yellow fever virus with a two-component genome. J. Virol. 81:11737–
27. Stadler, K., S. L. Allison, J. Schalich, and F. X. Heinz. 1997. Proteolytic
activation of tick-borne encephalitis virus by furin. J. Virol. 71:8475–8481.
28. Yu, I. M., et al. 2008. Structure of the immature dengue virus at low pH
primes proteolytic maturation. Science 319:1834–1837.
29. Zellweger, R. M., T. R. Prestwood, and S. Shresta. 2010. Enhanced infection
of liver sinusoidal endothelial cells in a mouse model of antibody-induced
severe dengue disease. Cell Host Microbe 7:128–139.
30. Zhang, Y., et al. 2003. Structures of immature flavivirus particles. EMBO J.
31. Zhang, Y., B. Kaufmann, P. R. Chipman, R. J. Kuhn, and M. G. Rossmann.
2007. Structure of immature West Nile virus. J. Virol. 81:6141–6145.
32. Reference deleted.