Synthetic recombinant bat SARS-like coronavirus
is infectious in cultured cells and in mice
Michelle M. Becker
, Rachel L. Graham
, Eric F. Donaldson
, Barry Rockx
, Amy C. Sims
, Timothy Sheahan
Raymond J. Pickles
, Davide Corti
, Robert E. Johnston
, Ralph S. Baric
, and Mark R. Denison
Microbiology and Immunology, Vanderbilt University, Nashville, TN 37232; Departments of
Microbiology and Immunology,
Cystic Fibrosis/Pulmonary Research and Treatment Center and Department of Medicine, and
Institute, University of North Carolina, Chapel Hill, NC 27599; and
Institute for Research in Biomedicine, CH-6500 Bellinzona, Switzerland
Edited by Peter Palese, Mount Sinai School of Medicine, New York, NY, and approved October 14, 2008 (received for review August 18, 2008)
Deﬁning prospective pathways by which zoonoses evolve and
emerge as human pathogens is critical for anticipating and controlling
both natural and deliberate pandemics. However, predicting tenable
pathways of animal-to-human movement has been hindered by
challenges in identifying reservoir species, cultivating zoonotic or-
ganisms in culture, and isolating full-length genomes for cloning and
genetic studies. The ability to design and recover pathogens recon-
stituted from synthesized cDNAs has the potential to overcome these
obstacles by allowing studies of replication and pathogenesis with-
out identiﬁcation of reservoir species or cultivation of primary iso-
lates. Here, we report the design, synthesis, and recovery of the
largest synthetic replicating life form, a 29.7-kb bat severe acute
respiratory syndrome (SARS)-like coronavirus (Bat-SCoV), a likely
progenitor to the SARS-CoV epidemic. To test a possible route of
emergence from the noncultivable Bat-SCoV to human SARS-CoV, we
designed a consensus Bat-SCoV genome and replaced the Bat-SCoV
Spike receptor-binding domain (RBD) with the SARS-CoV RBD (Bat-
SRBD). Bat-SRBD was infectious in cell culture and in mice and was
efﬁciently neutralized by antibodies speciﬁc for both bat and human
CoV Spike proteins. Rational design, synthesis, and recovery of hy-
pothetical recombinant viruses can be used to investigate mecha-
nisms of transspecies movement of zoonoses and has great potential
to aid in rapid public health responses to known or predicted emerg-
ing microbial threats.
emerging pathogens 兩 synthetic biology 兩 vaccine development 兩 zoonoses
mergence of zoonotic-human pathogens is increasingly recog-
nized as a threat to public health (1). Human population growth
and environmental change s have created new opportunities for
contact between humans and zoonotic organisms that may result in
cross-species transmission and human disease (1, 2). Recent exam-
ples include RNA viruses such as severe acute respiratory syndrome
coronavirus (SARS-CoV), Ebola, Hanta, Nipah, and Chikungunya
viruse s (2). However, the determinants regulating succe ssful trans-
species movement remain poorly understood due to challenges in
identifying viral precursors and animal reservoirs, thereby compli-
cating vaccine and therapeutic design (3). SARS-CoV, which
exemplifies the challenges inherent in studying emerging patho-
gens, was the first 21st century emerging virus to exhibit efficient
human-to-human transmission and rapid global spread (4–6).
Although zoonotic SARS-CoV strains were isolated from civets
and raccoon dogs (7–9), the epidemic likely originated from strains
circulating in bats (Bat-SCoVs) (10–12). Bat CoVs cluster in both
major mammalian CoV taxonomic groups, raising the possibility
that Bat CoVs may be progenitors to all 4 known pathogenic human
CoVs (11, 13). Bats are also predicted to function as reservoirs for
other important emerging human and animal viruses (14, 15).
Although several complete Bat CoV genome sequence s are avail-
able, no Bat CoV has been successfully cultivated in cell culture or
in animals (11, 13), severely limiting the identification of determi-
nants of zoonotic CoV transspecies movement.
The SARS-CoV Spike is a 180-kDa type I membrane glycopro-
tein that contains a well-defined receptor-binding domain (RBD).
The SARS-CoV genome is likely a mosaic of sequence s derived
from multiple recombination events, although this hypothesis is
somewhat controversial (16). However, recombination within Spike
has been described often (17), suggesting that the RBDs may be
interchangeable between strains (18–20). During the SARS-CoV
outbreak, evolution in the Spike RBD allowed for more efficient
use of human angiotensin-converting enzyme 2 (hACE2) as a
receptor for entry (21, 22). Because future zoonoses are likely, it is
critical to identify strategies used by viruses to adapt in human
populations. In this study, we have combined phylogenetic and
bioinformatics analyses, large-scale cDNA synthesis, chimeric gene
design, and reverse genetics to generate a consensus Bat-SCoV.
Succe ssful recovery of the infectious chimeric virus, Bat-SRBD,
which includes the RBD within Spike from human SARS-CoV,
demonstrates the plasticity of the CoV type I glycoprotein. The
synthetic reconstruction and recovery of this novel chimeric virus
identifies a necessary genetic element for CoV cross-species trans-
mission, establishes a model system for testing experimental evo-
lution of zoonotic CoVs, and allows for testing of vaccine and
therapeutics against possible future zoonotic strains.
Consensus Bat-SCoV Sequence Design and Construction. When this
study was initiated, 4 Bat-SCoVs had been identified (HKU3–1,
HKU3–2, HKU3–3, and RP3) as the virus reservoir populations
from which SARS-CoV emerged (10–12). Because none had been
recovered in culture, the infectivity of the reported viral genomic
RNA sequences was hypothetical, having been derived from RT-
PCR sequencing of bat fecal or rectal swab samples. Sequence
database s have error frequencies from 1/500 to 1/10,000, making
viable genome reconstruction problematic with increasing size (23).
Therefore, we used the 4 reported Bat-SCoV sequences to establish
a putative consensus Bat-SCoV sequence (GenBank acce ssion no.
FJ211859) and designed cDNA fragments with junctions precisely
aligned to the existing SARS-CoV reverse genetics system [Fig. 1A;
supporting information (SI) Fig. S1] (24). The defined and func-
tional SARS-CoV 5⬘ UTR and transcriptional regulatory sequences
Author contributions: M.M.B., R.L.G., E.F.D., R.S.B., and M.R.D. designed research; M.M.B.,
R.L.G., E.F.D., B.R., and A.C.S. performed research; R.L.G., E.F.D., T.S., R.J.P., D.C., and R.E.J.
contributed new reagents/analytic tools; M.M.B., R.L.G., E.F.D., B.R., A.C.S., R.J.P., and R.S.B.
analyzed data; and M.M.B., R.L.G., R.S.B., and M.R.D. wrote the paper.
Conﬂict of interest statement: R.E.J. is a coinventor of the Venezuelan Equine Encephalitis
(VEE) expression vector technology and holds an equity interest in AlphaVax, Inc., the
company that has licensed this technology from the University of North Carolina.
This article is a PNAS Direct Submission.
Freely available online through the PNAS open access option.
Data deposition: The sequences reported in this paper have been deposited in the GenBank
database (accession nos. FJ211859 and FJ211860).
M.M.B and R.L.G. contributed equally to this work.
To whom correspondence may be addressed. E-mail: firstname.lastname@example.org or mark.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2008 by The National Academy of Sciences of the USA
December 16, 2008
no. 50 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0808116105
were used because the 5⬘ UTRs of the Bat-SCoVs were incomplete.
The genomic cDNA fragments were commercially synthesized,
inserted into vectors, assembled into a full-length cDNA, and
transcribed in vitro to yield genomic RNA. Initial attempts to
recover and passage infectious Bat-SCoV failed. Electroporated
cells contained high levels of genome and leader-containing sub-
genomic transcripts on day 2, but not day 5 postelectroporation
(p.e.) (Fig. 1C), indicating that the synthetic consensus Bat-SCoV
genome expressed a functional replicase. We did recover infectious
virus consisting of SARS-CoV genome fragments A–E and Bat
fragment F (Fig. 1 B and D). The resulting virus, Bat-F, encoded a
chimeric Spike. Thus, the amino-terminal two-thirds of SARS-CoV
Spike, including the RBD, and the fusion core contained within the
carboxyl-terminal third of Bat-SCoV Spike can successfully drive
productive infection. Also, because Bat-F contained Bat-SCoV
acce ssory and structural genes 3⬘ to the Spike gene, these down-
stream ORFs are clearly interchangeable.
Generation and Recovery of Chimeric Bat-SRBD. The ectodomain of
Spike can be exchanged among CoVs, altering host-range specific-
ity (25, 26). To test whether the RBDs of Bat-SCoV and SARS-
CoV were interchangeable, we replaced the Bat-SCoV RBD
(amino acid 323–505) with the SARS-CoV RBD (amino acid
319–518) (27, 28) (GenBank accession no. FJ211860), simulating a
theoretical recombination event that might occur during mixed
infection in vivo (Fig. 1B). After electroporation, Bat-SRBD ge-
nome RNA and leader-containing subgenomic mRNA transcripts
were detected (Fig. 1C), and progeny virions were detected by
plaque assay. After 2 additional passages, the population genome
sequence was identical to the Bat-SRBD molecular clone. How-
ever, 4 nucleotides exhibited dual peaks on the sequencing elec-
tropherograms, suggesting quasispecies variation at these positions
(Table S1). Recovery and passage of Bat-SRBD demonstrated the
functional interchangeability of human and animal SARS-CoV-like
The crystal structure of SARS-CoV RBD complexed with its
receptor, hACE2 (29), implicated 13 residues within the carboxyl
terminus of the RBD (amino acid 426R-518D) in ACE2 engage-
ment. Homology modeling indicated that this receptor-binding
motif (RBM) may be sufficient to allow ACE2 engagement, and
further predicts that inclusion of 6 residues amino-terminal to the
RBM (amino acid 388V-393D) may enhance ACE2 engagement by
functioning as a distal ‘‘hinge.’’ To test this possibility, chimeric
Bat-SCoV genomes were constructed containing either the SARS-
CoV RBM (Bat-SRBM) or the RBM plus the distal hinge residue s
(Bat-Hinge) (Fig. 1B). Electroporation yielded genome and sub-
genomic leader-containing transcripts at day 2, but not 5, p.e. (Fig.
1 C and D), and progeny virions could not be successfully passaged
Virus Replication in Primate and Murine Cells. We next infected Vero
cells, murine delayed brain tumor (DBT) cells, and DBT cells
expre ssing hACE2 or civet (c)ACE2 (DBT-hACE2 and DBT-
cACE2) (22) with Bat-SRBD or SARS-CoV at a multiplicity of
infection (MOI) of 0.01 or 1 plaque forming unit (PFU) per cell
(Fig. 2). Bat-SRBD and SARS-CoV exhibited productive growth in
Vero, DBT-hACE2, and DBT-cACE2 cells that was remarkably
similar in kinetics and peak titers (Fig. 2). In contrast, DBT cells
lacking ACE2 expression did not support growth of either SARS-
CoV or Bat-SRBD (data not shown). These data indicate that
Bat-SCoV expressing the SARS-CoV RBD is capable of entering
cells by using ACE2 from humans, nonhuman primates, or civets as
receptor, and replicating efficiently.
Detection of Bat-SRBD Replicase Proteins by SARS-CoV Antibodies.
Comparison of SARS-CoV and Bat-SRBD predicted high, but not
complete, identity of amino acid sequences across replicase proteins
(Table S2). Because antibody cross-reactivity is a potential tool for
detection and analysis of Bat-SCoVs, we tested whether antibodies
specific for SARS-CoV proteins (30) could also detect Bat-SRBD
homologues. Immunoblots were performed by using rabbit poly-
clonal antibodies (pAbs) specific for SARS-CoV nsp1, nsp8, nsp9,
22440 22988 25202
323 I 505 E 12431 (aa)
21492 (nt) 22446 23045 25259
319 I 518 D 12561 (aa)
Fig. 1. Schematic representation of SARS-CoV and Bat-SCoV variants. (A)
Schematic representation of SARS-CoV and Bat-SCoV (GenBank accession no.
FJ211859) genomes and reverse genetics system. (Top) Arrowheads indicate nsp
processing sites within the ORF1ab polyprotein (open arrowheads, papain-like
proteinase mediated; ﬁlled arrowheads, nsp5 [3C-like proteinase] mediated).
Immediately below are the fragments used inthe reversegenetics system, labeled
A through F. The fragments synthesized to generate Bat-SCoV exactly recapitu-
late the fragment junctions of SARS-CoV with the exception that the Bat-SCoV
has 2 fragments, Bat-E1 and Bat-E2,which correspondto the SARS-Efragment. (B)
Schematic representation showing organization of the SARS-CoV and Bat-SCoV
Spike proteins. The engineered Spike proteins are pictured below with the virus
name to the left. Bat-SRBD includes all of the Bat-SCoV Spike sequence except
that the Bat-SCoV RBD (Bat-SCoV amino acid 323–505) is replaced with the
SARS-CoV RBD (amino acid 319 –518) (GenBank accession no. FJ211860). Bat-
SRBD-MA includes the MA15 Spike RBD change atSARS-CoV aa Y436H. Bat-SRBM
includes the minimal 13 SARS-CoV residues critical for ACE2 contact, resulting in
a chimeric RBD of Bat-SCoV amino acid 323I-429T and SARS-CoV amino acid
426R-518D. Bat-Hinge is Bat-SRBM sequence, with Bat-SCoV amino acid 392L-
397E replaced with SARS-CoV amino acid 388V-393D. Bat-F includes nt 1–24057
of SARS-CoV (to Spike amino acid 855), with the remaining 3⬘ sequence from
Bat-SCoV. To the right of the schematic representations, observation of transcript
activity and approximate stock titers at passage 1 (P1) are indicated. ND indicates
no infectious virus detected by plaque assay. (C and D) Presence of genomic and
subgenomic transcripts after electroporation of in vitro transcribed viral RNA.
Band corresponding to mRNA1 indicates the presence of genomic RNA, either
electroporated genomic RNA or progeny genomic RNA, and the presence of a
band corresponding to mRNA9 indicates the presence of leader-containing sub-
genomic RNA, consistent with mRNA transcription.
Becker et al. PNAS
December 16, 2008
and nsp10 (Fig. S2 and SI Materials and Methods). Proteins of the
predicted size were detected in cells infected with SARS-CoV or
Bat-SRBD demonstrating cross-reactivity of the antibodies and
confirming expression of the cognate replicase proteins in Bat-
Neutralization of Bat-SRBD. To examine antigenic relatedness, the
Bat-SCoV Spike was cloned into Venezuelan Equine Encephalitis
viral replicon particle (VRP) vectors, and pools of mouse (m)pAbs
were tested for neutralization of Bat-SRBD infectivity. All Bat-
SCoV Spike-specific sera efficiently neutralized Bat-SRBD, with
50% neutralization titers ranging from 1/100 to 1/400 dilutions (Fig.
3A). In parallel, these sera did not neutralize SARS-CoV infectivity,
suggesting that the antibodies recognize epitopes in Spike outside
the RBD in Bat-SRBD. Because the RBD appears to be the
minimal motif required to alter host range of the Bat-SCoV
precursor, we also tested whether Bat-SRBD could be neutralized
with human monoclonal antibodies (hmAbs S109.8, S227.14, and
S230.15), which recognize unique epitopes within the SARS-CoV
RBD and cross-neutralize human and zoonotic SARS-CoV isolates
in vitro and in vivo (31). The hmAbs neutralized Bat-SRBD,
demonstrating the accessibility of the neutralizing epitopes of
SRBD in the background of the Bat-SCoV Spike (Fig. 3B).
Neutralization also functions as an important safety feature in
design and study of Bat-SRBD viruses.
Bat-SRBD Replicates in Human Airway Epithelial (HAE) Cell Cultures.
Because Bat-SRBD grew equivalently to SARS-CoV in culture, we
tested whether Bat-SRBD could replicate in primary HAEs, which
recapitulate the epithelium of the human conducting airway. We
have previously identified zoonotic SARS-CoV variants that rep-
licated efficiently in Vero cells, but not in HAE cultures (22, 32).
Therefore, HAE cultures provide a more relevant and stringent
measure of the replicative potential of chimeric recombinant zoo-
notic SARS-CoV viruses in the human host. HAE cultures were
inoculated by means of the apical surface and the media sampled
at different times postinfection (p.i.). Peak titers of SARS-CoV and
Bat-SRBD were similar, although Bat-SRBD growth was delayed,
compared with SARS-CoV (Fig. 4A). Because hACE2 is detected
primarily on the apical surface of ciliated cells in HAEs (33), we
looked for differences in targeting of Bat-SRBD and SARS-CoV
infection in HAEs. Histological sections from HAE at 144-h p.i.
were probed with Spike-specific antisera, and localization assessed
by indirect immunofluorescence. For both viruses, Spike was
detected predominantly on the apical surface of ciliated cells (Fig.
4 B and C), but was not detected in nonciliated cells.
SARS-CoV Mouse-Adapted Spike Mutation Enhances Bat-SRBD Repli-
cation in Mice.
SARS-CoV replicates in mouse lungs, but causes only
slight morbidity (34, 35). Replication and pathogenesis are en-
hanced in infections of BALB/c mice with MA15, a mouse-adapted
SARS-CoV containing 6 amino acid changes, including a Y436H
substitution in the Spike RBD (36). Although modeling predicts
that Y436H enhances RBD-mACE2 receptor engagement (Fig.
S3), both SARS-CoV and MA15 replicate efficiently in mouse
lungs, complicating assignment of Y436H contributions. To test the
Viral Titer (Log
Time Post-infection (h) Time Post-infection (h)
Viral Titer (Log
Time Post-infection (h)
500 10203040 50010203040 5010 20 30 400
Viral Titer (Log
Fig. 2. Growth of SARS-CoV and Bat-SRBD in 3 different cell types. (A–C) Vero cells (A), DBT-hACE2 cells (B), or DBT-cACE2 cells (C) were infected with SARS-CoV
at a MOI ⫽ 1(
)orMOI⫽ 0.01 (F), or Bat-SRBD at a MOI ⫽ 1(䊐)oraMOI⫽ 0.01 (
). Infected cultures were sampled, in triplicate, at the times indicated and
viral titer was quantiﬁed by plaque assay on Vero cells. Error bars indicate SD.
mAb Concentration (µ g/ml)
Fig. 3. Neutralization of the Bat-SRBD by mouse serum and human mAbs. (A)
Immune sera from 5 mice (1,
; and 5,
) vaccinated with Bat-SCoV
Spike were used to neutralize Bat-SRBD. Controls include prebleed serum (
Mouse 1 serum used to neutralize SARS-CoV (⽧). (B) Human mAbs S109.8 (
), and S230.15 (
) were used to neutralize Bat-SRBD. Results are
expressed as the percentage of neutralization. Error bars indicate SD.
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0808116105 Becker et al.
hypothesis that Y436H enhances interaction with mACE2, Bat-
SRBD was constructed with this substitution (Bat-SRBD-MA).
Electroporation with Bat-SRBD-MA genome RNA resulted in
production of infectious virus with titers similar to Bat-SRBD (Fig.
1B). Next, 14-month-old BALB/c mice were infected with Bat-
SRBD, Bat-SRBD-MA, or SARS-CoV. Mice were weighed and
monitored daily for morbidity, and on days 2 and 4 p.i., mice were
euthanized and lungs harvested (Fig. 5). Mice infected with chi-
meric viruse s did not exhibit significant weight loss or morbidity
after infection (Fig. 5A). However, although Bat-SRBD replicated
in infected lungs, Bat-SRBD-MA replicated ⬇1.5 logs more effi-
ciently at day 2 (Fig. 5B), providing support for the hypothesis that
the Y436H substitution in Bat-SRBD-MA may improve mACE2
Reverse genetic systems have revolutionized our understanding of
the molecular basis of viral replication, pathogenesis, and vaccine
design for many virus families. However, application of these
technologies to emerging pathogens has been limited by factors
involved in constructing and manipulating molecular clones and
characterizing recombinant viruses, particularly those with large
genomes. Also, standard approaches for development of infectious
clones have required availability of viral RNA or DNA, and
consequently have been mostly limited to viruses that replicate in
culture. Last, classical approaches to combat emerging or deliber-
ately introduced human pathogens may not allow responses in a
timeframe adequate to significantly reduce mortality or morbidity.
Time Post-infection (h)
2 8 24 48 72
Viral Titer (Log
Fig. 4. Efﬁcient replication of Bat-SRBD in human ciliated airway epithelial cells.
(A) Growth curves for SARS-CoV and Bat-SRBD were obtained from apical washes
of human ciliated airway epithelial cell cultures inoculated with either virus.
Samples were serially diluted and titers determined by plaque assay on Vero cells.
Titers are expressed as PFU per mL. Both SARS-CoV and Bat-SRBD replicated to
titers of ⬇10
, although Bat-SRBD growth was delayed compared with SARS-CoV.
All inoculations were performed in duplicate. SARS-CoV, F; Bat-SRBD, E.(B–D)
Representative histological sections of HAE 144 h p.i. with SARS-CoV (B), Bat-
SRBD (C), or vehicle alone (D) and probed with mouse polyclonal sera directed
against the Bat-CoV Spike and visualized with mouse-speciﬁc secondary antibod-
ies conjugated to AlexaFluor 488 (green). Detection of Spike immunoreactivity
was localized speciﬁcally to the apical surface of ciliated cells indicating that
SARS-CoV and Bat-SRBD both infect ciliated cells after apical inoculation. Note
that at 144 h p.i. cilial morphology shows considerable cytotoxicity. Spike immu-
noreactivity was not observed in nonciliated cell-types. (Scale bar, 5
Fig. 5. Weight loss and viral replication of Bat-SRBD, Bat-SRBD-MA, and
SARS-CoV in aged BALB/c mice. Ten 14-month-old female BALB/c mice were
infected intranasally with 10
PFU of the indicated virus or an equivalent volume
L) of PBS. (A) Weights ofall surviving mice per infection group wererecorded
each day, averaged, and plotted as a percentage of starting weight. Error bars
indicate SD. (B) On days 2 and 4 p.i., 5 mice per group were killed and lungs were
harvested. Lung homogenates were titered on Vero cells. Circles represent titers
of individual mouse lungs. Bars represent the average titer of each infection
group. BSRBD, Bat-SRBD; BSRM, Bat-SRBD-MA; SARS, SARS-CoV.
Becker et al. PNAS
December 16, 2008
Thus, new methodologies to rapidly recover and test emerging
zoonotic pathogens are critical. Recent studies have provided
crucial steps toward the goal of synthetic reconstruction of large
microbial genomes. The 7.5- and 5.6-kb genomes of poliovirus and
X174, respectively, were reconstructed from known sequences by
using commercially synthesized cDNA fragments and PCR assem-
bly (37, 38). Similarly, the segmented genome of the 1918 strain of
influenza was reconstructed, in part, by using synthetic design (39).
Human endogenous retroviruses were assembled by PCR-directed
assembly of synthetic oligonucleotides into a consensus provirus
(40), and recombinant SARS-CoVs bearing synthetic zoonotic
Spike sequences were derived by our group (22, 32). Last, the
cloning of the 580-kb microbial genome of Mycoplasma genitalium
was reported, although this work has not yet yielded a replicating
organism (41). To our knowledge, no studies to date have used a
synthetic approach to assess potential mechanisms of zoonotic
emergence of a noncultivable virus.
Because it is possible that even minor events of recombination
and mutation-driven evolution can alter CoV population structure
and promote emergence (22, 32, 42), CoVs may select for alter-
ations in discrete regions of Spike to achieve host-range expansion.
The CoV Spike, a type I fusion protein, contains 2 discrete regions,
a carboxyl-terminal S2 region that encodes fusion and heptad
repeat domains in an arrangement shared by other viral attachment
proteins possessing type I architecture (43), whereas the S1 region
encode s the RBD. Studies in our and other laboratories have
identified mutations in both S1 and S2 regions associated with CoV
host-range expansion (42, 44). These 2 regions may also indepen-
dently or cooperatively mediate transspecie s expansion and neu-
tralization escape, in that some mAbs that target the SARS-CoV S1
RBD select for escape mutants in the RBD, but also in the S2 region
(45). These data suggest that coordinated interactions between the
S1 RBD and select S2 domains may be important in epitope
presentation and Spike function.
We have identified several hmAbs that bind distinct, conserved
locations in the SARS-CoV RBD and neutralize strains that
originate from animal and human hosts (31). These hmAbs did
efficiently neutralize Bat-SRBD, an important finding as it was not
previously clear that mAbs targeting the SARS-CoV RBD would
neutralize virus in the context of a different Spike backbone. It is
also informative that antibodies specific for the Bat-SCoV Spike
protein neutralize Bat-SBRD, which expresses a chimeric Spike
protein, but not SARS-CoV, indicating the existence of neutralizing
antibodies that target portions of Spike outside the RBD. Impor-
tantly, these results sugge st that hmAbs specific for SARS-CoV, and
by inference the current panel of SARS-CoV vaccines, may provide
significant protection against other SARS-like CoVs that emerge
from zoonotic pools by natural recombination or are deliberately
designed to cross species. Second, both SARS-CoV RBD-specific
hmAbs and Bat-SCoV mpAbs specific for Spike epitopes outside of
the RBD are able to recognize and neutralize virus even in the
setting of a chimeric Spike, providing important safety features for
studies of emerging zoonotic CoVs.
It has been shown that the 5⬘ UTRs of CoVs can influence the
capacity of the virus to replicate in cells (46). Although the 5⬘ UTR
in the synthetic Bat-SCoV originated from SARS-CoV, the sole
difference between Bat-SCoV, which was not capable of amplifi-
cation in cell culture, and Bat-SRBD, which could be recovered, was
the RBD derived from SARS-CoV. Our re sults also confirm and
extend a previous report predicting that deletions and mutations
within the Bat-SCoV RBM ablate interaction with hACE2 and
cACE2 molecules (47). Thus, in this report, we show that the CoV
Spike RBD is interchangeable, is sufficient to confer efficient
growth and infectivity in cells from multiple species, and likely
represents a critical determinant of transspecies movement of
To protect against future emerging zoonotic pathogens, it is
crucial to develop cell culture and animal models to test vaccines
and therapeutics, ideally against entire families of organisms, such
as CoVs. Both SARS-CoV and Bat-SRBD replicated efficiently in
HAE cultures, providing a direct human airway model for com-
parison of existing and new antivirals. However, Bat-SRBD repli-
cated poorly in vivo, calling for additional modifications to facilitate
studies in mouse models. Robust structural information exists on
the RBD-ACE2 interaction (29), mutations affecting this interac-
tion have been identified (22, 31), and Rosetta-modeling of short
range RBD-mACE2 receptor interfaces can identify key residues
essential for retargeting the host specificity of Bat-SRBD (Fig. S3)
(48). Previous studies had identified a mutation in the RBD of the
MA15 strain, Y436H, but its exact role in vivo was not clear (36).
By using structural modeling algorithms, we predicted that the
Y436H substitution would enhance the interaction of Bat-SRBD
with mACE2. Bat-SRBD-MA did exhibit increased growth effi-
ciency in aged mice. However, this did not result in clinical disease,
suggesting the requirement for additional adaptive changes. For
example, SARS-CoV encodes at least 5 IFN antagonists, predicted
to function in virulence (49–52). Our model system will allow
mapping of the domains in the Bat-SCoV and SARS-CoV genetic
backgrounds involved in regulation of virulence in aged animals.
In this report, sequence and structural information was inte-
grated with synthetic genomics, reverse genetics, and protein design
to recover a zoonotic precursor virus from a hypothetical infectious
sequence. The resulting chimera exhibited cross-reactivity with
previously identified therapeutics and highlighted possible previ-
ously undescribed mechanisms for host-range expansion. Here, we
articulate a model to predict and directly test tenable emergence
pathways. Paired with a greater availability of reagents and thera-
peutics, our studies represent an approach for rapid recovery and
testing of newly identified pathogens, and which may improve
public health preparedness and intervention strategies against
natural or intentional zoonotic-human epidemics.
Materials and Methods
Cells and Viruses. VeroE6 cells (Vero) were maintained in MEM (Invitrogen), and
delayed brain tumor (DBT, murine astrocytoma) cells were maintained in Dul-
becco’s MEM (Invitrogen) containing 10% FBS. DBT-hACE2 and DBT-cACE2 cells
were cultured as described (22). HAE cells were plated and differentiated as
described (33). SARS-CoV Urbani strain (hereafter, SARS-CoV) and Bat-SCoV wild
type and chimeric viruses were propagated and assessed by plaque assay on Vero
cells. All studies with viable SARS-CoV and SARS-CoV-like viruses were performed
in certiﬁed BSL3 laboratories in biological safety cabinets, by using safety proto-
cols that were reviewed and approved by the Institutional Biosafety Committees
of Vanderbilt University and the University of North Carolina at Chapel Hill.
Determining the Consensus Sequence for Synthetic Bat-SCoV and Conceptual
Design of the Bat-SCoV Clone. HKU3–1 (DQ022305), HKU3–2 (DQ084200),
HKU3–3 (DQ084199), and RP3 (DQ071615) genomes were aligned by using
ClustalXv1.83 to determine a consensus sequence (Fig. S1 and Figs. S4 –S6) (Gen-
Bank accession no. FJ211859). The consensus Bat-SCoV sequence was designed to
ligate interchangeably with the SARS-CoV infectious clone (24). Notably, there
was no consensus at the 5⬘end of the Bat-SCoV genomes, so we used the 5⬘ most
region of SARS-CoV to append the T7 promoter site to the 5⬘ end of the Bat-A
Construction of Chimeric Spike Variants. Insertions of SARS-CoV sequence in
place of Bat-SCoV sequence were engineered by using PCR and the primers
shown in Table S3. PCR amplicons for Bat-SRBD (GenBank accession no. FJ211860)
and Bat-SRBM were generated by using fragments Bat-E2, and SARS-E. PCR
amplicons for Bat-Hinge (Bat-SRBM plus 6 additional residues from SARS-CoV
Spike) were generated by using Bat-SRBM as template. PCR-generated products
were cloned into the Bat-E2 plasmid by using unique 5⬘-BstBI and 3⬘-MscI sites.
Successful insertions of SARS-CoV sequence were conﬁrmed by restriction diges-
tion and nucleotide sequencing across the region of PCR ampliﬁcation.
Generation of SARS-CoV and Bat-SCoV Mutant Viruses. Viruses containing
PCR-generated insertions within the viral coding sequence were produced by
using the SARS-CoV assembly strategy (24, 33, 53) with the following modiﬁca-
tions. Brieﬂy, for Bat-F virus, full-length cDNA was constructed by ligating restric-
tion products from SARS-CoV fragments A–E and Bat-SCoV fragment F, which
required a BglI-NotI digestion. For Bat-SCoV and Bat-SRBD, Bat-SRBM, and Bat-
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0808116105 Becker et al.
Hinge, plasmids containing the 7 cDNA fragments of the Bat-SCoV genome were
digested by using BglI for Bat-A, Bat-B, Bat-C, and Bat-D, BglI and AﬂII for Bat-E1
and Bat-E2, and BglI and NotI for Bat-F. Digested, gel-puriﬁed fragments were
simultaneously ligated together. Transcription was driven by using a T7 mMes-
sage mMachine kit (Ambion), and RNAwas electroporated intoVero cells(24, 53).
Virus viability was determined by cytopathic effect and progeny viruses were
passaged at low MOI. RNA was recovered from infected cell monolayers by using
TRIzol (Invitrogen) according to the manufacturer’s instructions, and genome
origins were veriﬁed by RT-PCR and nucleotide sequencing.
Assay for Bat-SCoV Leader-Containing Transcripts in Electroporated Cells and
Mouse Lungs. At days 2 and 5 p.e., generation of leader-containing N (ORF9) and
genome (ORF1) transcripts was determined by RT-PCR. Brieﬂy, RT-PCR was per-
formed by using random hexamers (ABI) and SuperScript III (Invitrogen) to
generate ﬁrst-strand cDNA at an extension temperature of 55 °C for 1 h. Leader-
containing cDNAs were ampliﬁed by PCR by using Taq with Thermopol buffer
(NEB) and the followingprimers: 5⬘-CAGGAAAAGCCAACCAACCTTG (leader)and
5⬘-CGCTACGACCGAACTGAATGCC to detect Bat-SCoV genomic RNA; and leader
and 5⬘-GTGAGAGCTGTGAACCAAGACG to detect mRNA 9 transcripts. Presence
or absence of PCR products was assessed by electrophoresis on 1.5% agarose gels.
Viral Growth and Plaque Assays. Vero, DBT, DBT-hACE2, or DBT-cACE2 cells were
infected at a MOI of 1 or 0.01 PFU/cell. After1hat37°C,theinocula were
removed, cells were washed and samples were taken at different times p.i. To
determine viral titer, samples were diluted, inoculated onto Vero cell monolayers
in 6-well plates for 1 h, and overlaid with complete media plus 1% agar. Plaques
were visualized between 48 and 52 h p.i. by neutral red staining (Sigma).
Generation of Polyclonal Bat-SCoV Spike-Specific Sera and Neutralization Assay.
Murine pAbs speciﬁc for the Bat-SCoV Spike were generated as previously de-
scribed (33). Neutralizing titers for mpAbs and hmAbs S109.8, S227.14, and
S230.15 were determined by plaque reduction neutralization titer assay
(PRNT50%) (31). The percentage of neutralization was calculated as follows:
1-(number of plaques in the presence of antibody/number of plaques in the
absence of antibody) x 100%.
Infection of Aged BALB/c Mice. Ten each of aged (14 months) female BALB/c mice
(National Institute of Aging) were lightly anesthetized and infected intranasally
PFU of Bat-SRBD, Bat-SRBD-MA, or SARS-CoV; 5 additional mice were
inoculated with an equivalent volume (50
L) of PBS. Mice were weighed daily
through 4 days p.i., and on days 2 and 4 p.i., 5 mice of each group were killed and
lungs harvested for determination of viral titer. Lungs were weighed and ho-
mogenized in 500
L of PBS at 6,000 rpm for 60 s in a MagnaLyser (Roche).
Clariﬁed homogenates were then diluted serially, and titers were determined by
plaque assay on Vero cells.
ACKNOWLEDGMENTS. We thank XiaoTao Lu and Sunny Lee for technical assis-
tance, Susan Burkett for maintenance of the HAE cultures, Perry Myrick for
immunoﬂuorescence assays, and the University of North Carolina Cystic Fibrosis
Tissue Culture Core for HAE cells. M.M.B., R.L.G., R.S.B., and M.R.D. are supported
by the National Institute of Allergy and Infectious Diseases Public Health Service
Award P01 AI59943. Additional support was provided by Public Health Service
Award CA68485 to the Vanderbilt University DNA Sequencing Shared Resource
of the Vanderbilt–Ingram Cancer Center. The Baric laboratory is supported by the
Gillings Innovation Fund.
1. Weiss RA, McMichael AJ (2004) Social and environmental risk factors in the emergence
of infectious diseases. Nat Med 10:S70 –76.
2. Woolhouse ME, Gowtage-Sequeria S (2005) Host range and emerging and reemerging
pathogens. Emerg Infect Dis 11:1842–1847.
3. Webby R, Hoffmann E, Webster R (2004) Molecular constraints to interspecies trans-
mission of viral pathogens. Nat Med 10:S77– 81.
4. Drosten C, et al. (2003) Identiﬁcation of a novel coronavirus in patients with severe
acute respiratory syndrome. N Engl J Med 348:1967–1976.
5. Ksiazek TG, et al. (2003) A novel coronavirus associated with severe acute respiratory
syndrome. N Engl J Med 348:1953–1966.
6. Peiris JS, et al. (2003) Coronavirus as a possible cause of severe acute respiratory
syndrome. Lancet 361:1319 –1325.
7. Tu C, et al. (2004) Antibodies to SARS coronavirus in civets. Emerg Infect Dis 10:2244–2248.
8. Kan B, et al. (2005) Molecular evolution analysis and geographic investigation of severe
acute respiratory syndrome coronavirus-like virus in palm civets at an animal market
and on farms. J Virol 79:11892–11900.
9. Guan Y, et al. (2003) Isolation and characterization of viruses related to the SARS
coronavirus from animals in southern China. Science 302:276 –278.
10. LiW, et al. (2005) Bats are natural reservoirs of SARS-like coronaviruses. Science 310:676– 679.
11. Shi Z, Hu Z (2008) A review of studies on animal reservoirs of the SARS coronavirus. Virus
Res 133:74 –87.
12. Lau SK, et al. (2005) Severe acute respiratory syndrome coronavirus-like virus in Chinese
horseshoe bats. Proc Natl Acad Sci USA 102:14040 –14045.
13. Dominguez SR, O’Shea TJ, Oko LM, Holmes KV (2007) Detection of group 1 coronavi-
ruses in bats in North America. Emerg Infect Dis 13:1295–1300.
14. Halpin K, et al. (2007) Emerging viruses: Coming in on a wrinkled wing and a prayer.
Clin Infect Dis 44:711–717.
15. Calisher CH, et al. (2006) Bats: Important reservoir hosts of emerging viruses. Clin
Microbiol Rev 19:531–545.
16. Holmes EC, Rambaut A (2004) Viral evolution and the emergence of SARS coronavirus.
Philos Trans R Soc Lond B Biol Sci 359:1059 –1065.
17. Lai MM, Cavanagh D (1997) The molecular biology of coronaviruses. Adv Virus Res
18. Stavrinides J, Guttman DS (2004) Mosaic evolution of the severe acute respiratory
syndrome coronavirus. J Virol 78:76– 82.
19. Li W, et al. (2006) Animal origins of the severe acute respiratory syndrome coronavirus:
Insight from ACE2-S-protein interactions. J Virol 80:4211– 4219.
20. Li W, et al. (2007) The S proteins of human coronavirus NL63 and severe acute respiratory
syndrome coronavirus bind overlapping regions of ACE2. Virology 367:367–374.
21. Li W, et al. (2005) Receptor and viral determinants of SARS-coronavirus adaptation to
human ACE2. EMBO J 24:1634–1643.
22. Sheahan T, et al. (2008) Mechanisms of zoonotic severe acute respiratory syndrome
coronavirus host range expansion in human airway epithelium. J Virol 82:2274 –2285.
23. Pruitt KD, Tatusova T, Maglott DR (2003) NCBI Reference Sequence project: Update and
current status. Nucleic Acids Res 31:34 –37.
24. Yount B, et al. (2003) Reverse genetics with a full-length infectious cDNA of severe
acute respiratory syndrome coronavirus. Proc Natl Acad Sci USA 100:12995–13000.
25. Kuo L, et al. (2000) Retargeting of coronavirus by substitution of the spike glycoprotein
ectodomain: Crossing the host cell species barrier. J Virol 74:1393–1406.
26. Haijema BJ, Volders H, Rottier PJ (2003) Switching species tropism: An effective way to
manipulate the feline coronavirus genome. J Virol 77:4528 –4538.
27. Chakraborti S, Prabakaran P, Xiao X, Dimitrov DS (2005) The SARS coronavirus S glycoprotein
receptor binding domain: Fine mapping and functional characterization. Virol J 2:73.
28. Wong SK, et al. (2004) A 193-amino acid fragment of the SARS coronavirus S protein
efﬁciently binds angiotensin-converting enzyme 2. J Biol Chem 279:3197–3201.
29. Li F, Li W, Farzan M, Harrison SC (2005) Structure of SARS coronavirus spike receptor-
binding domain complexed with receptor. Science 309:1864 –1868.
30. Prentice E, et al. (2004) Identiﬁcation and characterization of severe acute respiratory
syndrome coronavirus replicase proteins. J Virol 78:9977–9986.
31. Rockx B, et al. (2008) Structural basis for potent cross-neutralizing human monoclonal
antibody protection against lethal human and zoonotic severe acute respiratory
syndrome coronavirus challenge. J Virol 82:3220 –3235.
32. Rockx B, et al. (2007) Synthetic reconstruction of zoonotic and early human severe
acute respiratory syndrome coronavirus isolates that produce fatal disease in aged
mice. J Virol 81:7410–7423.
33. Sims AC, et al. (2005) Severe acute respiratory syndrome coronavirus infection of
human ciliated airway epithelia: Role of ciliated cells in viral spread in the conducting
airways of the lungs. J Virol 79:15511–15524.
34. Subbarao K, Roberts A (2006) Is there an ideal animal model for SARS? Trends Microbiol
35. Roberts A, et al. (2008) Animal models and vaccines for SARS-CoV infection. Virus Res
36. Roberts A, et al. (2007) A mouse-adapted SARS-coronavirus causes disease and mor-
tality in BALB/c mice. PLoS Pathog 3:e5.
37. Smith HO, Hutchison CA, III, Pfannkoch C, Venter JC (2003) Generating a synthetic
genome by whole genome assembly: PhiX174 bacteriophage from synthetic oligonu-
cleotides. Proc Natl Acad Sci USA 100:15440 –15445.
38. Cello J, Paul AV, Wimmer E (2002) Chemical synthesis of poliovirus cDNA: Generation
of infectious virus in the absence of natural template. Science 297:1016 –1018.
39. Tumpey TM, et al. (2005) Characterization of the reconstructed 1918 Spanish inﬂuenza
pandemic virus. Science 310:77– 80.
40. Lee YN, Bieniasz PD (2007) Reconstitution of an infectious human endogenous retro-
virus. PLoS Pathog 3:e10.
41. Gibson DG, et al. (2008) Complete chemical synthesis, assembly, and cloning of a
Mycoplasma genitalium genome. Science 319:1215–1220.
42. McRoy WC, Baric RS (2008) Amino acid substitutions in the S2 subunit of mouse hepatitis
virus variant V51 encode determinants of host range expansion. J Virol 82:1414 –1424.
43. Lamb RA, Jardetzky TS (2007) Structural basis of viral invasion: Lessons from paramyxo-
virus F. Curr Opin Struct Biol 17:427– 436.
44. Schickli JH, Thackray LB, Sawicki SG, Holmes KV (2004) The N-terminal region of the
murine coronavirus spike glycoprotein is associated with the extended host range of
viruses from persistently infected murine cells. J Virol 78:9073–9083.
45. Mitsuki YY, et al. (2008) A single amino acid substitution in the S1 and S2 Spike protein
domains determines the neutralization escape phenotype of SARS-CoV. Microbes
Infect 10:908 –915.
46. Hofmann MA, Senanayake SD, Brian DA (1993) A translation-attenuating intraleader
open reading frame is selected on coronavirus mRNAs during persistent infection. Proc
Natl Acad Sci USA 90:11733–11737.
47. Ren W,et al. (2008) Difference in receptor usage between severe acute respiratory syndrome
(SARS) coronavirus and SARS-like coronavirus of bat origin. J Virol 82:1899 –1907.
48. Li W, et al. (2004) Efﬁcient replication of severe acute respiratory syndrome coronavirus in
mouse cells is limited by murine angiotensin-converting enzyme 2. J Virol 78:11429–11433.
49. Kopecky-Bromberg SA, et al. (2007) Severe acute respiratory syndrome coronavirus
open reading frame (ORF) 3b, ORF 6, and nucleocapsid proteins function as interferon
antagonists. J Virol 81:548–557.
50. Kopecky-Bromberg SA, Martinez-Sobrido L, Palese P (2006) 7a protein of severe acute
respiratory syndrome coronavirus inhibits cellular protein synthesis and activates p38
mitogen-activated protein kinase. J Virol 80:785–793.
51. Frieman M, et al. (2007) Severe acute respiratory syndrome coronavirus ORF6 antag-
onizes STAT1 function by sequestering nuclear import factors on the rough endoplas-
mic reticulum/Golgi membrane. J Virol 81:9812–9824.
52. Wathelet MG, Orr M, Frieman MB, Baric RS (2007) Severe acute respiratory syndrome
coronavirus evades antiviral signaling: Role of nsp1 and rational design of an attenu-
ated strain. J Virol 81:11620–11633.
53. Graham RL, et al. (2005) The nsp2 replicase proteins of murine hepatitis virus and severe acute
respiratory syndrome coronavirus are dispensable for viral replication. J Virol 79:13399 –13411.
Becker et al. PNAS
December 16, 2008