|
Virus Structure group
|
Members
Ruigrok group
Rob Ruigrok - Professor (UJF)
Florence Baudin – CR1 (CNRS)
Guy Schoehn – CR2 (CNRS)
Inma Garcia Robles - postdoctoral fellow
Hatice Akarsu - predoctoral fellow (UJF)
Aurelie Albertini - predoctoral fellow (UJF)
Celine Fabry - DEA student (UJF)
Monique Perrissin - research assistant (UJF)
Introduction
In 2003 the group has become fully
French with the last students remaining under EMBL employment finishing their
work, leaving only people employed by the CNRS or the UJF. The group was
allowed to stay at EMBL and we are grateful for this hospitality. All approvals
for the construction of a new building that will house the future IVMS
(including the groups of Ruigrok, Seigneurin/Burmeister, Drouet plus groups to
be attracted from elsewhere) and the PSB (a partnership between the
international institutes on the site plus the IBS; http://psb.esrf.fr/ ). The
building will start spring 2004 and is expected to be finished before summer
2005.
During 2003 the group has developed
further into two major directions. First, Florence Baudin and her students have
developed the biochemistry of the interactions of viral components inside the
cell, in particular interactions necessary for the transport of influenza viral
nucleocapsids into and out of the nucleus of the infected cell. Apart from the
biochemistry, this work is developing into the determination of crystal
structures of the components involved in this transport and will hopefully
result in crystal structures of complexes between viral and cellular proteins.
These interactions will also be studied in living cells. The second direction
of the group is into the determination of high resolution EM structures of
viral components, in particular of measles virus nucleocapsids. This work,
driven by the activity of Guy Schoehn in collaboration with our colleagues
James Conway and Dick Wade at the “Institut de Biologie Structurale, IBS” is
not yet published and will be described next year. Our expertise in EM has led
to many collaborations, in particular with the group of Wisia Chroboczek at the
IBS and with Winfried Weissenhorn from the EMBL that will be described here.
Finally, the work on the structure of the cofactor of the Sendai virus
polymerase was continued through a collaboration between all the structural
biology institutes on the site with a final model for the C-terminal domain of
this molecule.
Nuclear transport of
influenza virus ribonucleoprotein
Florence Baudin, Hatice Akarsu and Inma Garcia Robles. During influenza virus infection,
viral ribonucleoproteins (vRNPs) are replicated in the nucleus and must be
exported to the cytoplasm before assembling into mature viral particles. There
is considerable evidence that the 14.5 kDa protein NEP (“nuclear export
protein”, formerly called NS2),behaves as an adapter protein analogous to HIV-1
Rev by mediating the association of Crm1 with vRNPs. NEP associates with the matrix protein M1,
which binds vRNPs with high affinity. The presence of M in the
nucleus is essential for vRNP export.
Despite
numerous trials, we were not able to obtain crystals of full-length NEP. Limited
proteolysis allowed us to define a stable C-terminal domain ranging from
residue 54 to 116, able to crystallize. The 2.6 Å crystal structure (Figure 1) of the C-terminal domain reveals an
amphipathic helical hairpin that dimerizes as a four-helix bundle. The two
helices share extensive contacts over their entire length, the majority of
which are hydrophobic in nature. The C-terminal domain is a highly compact unit
with little flexibility, explaining its resistance to proteolysis.
Figure 1. Three-dimensional structure of NEP C-terminal domain. Crystals belong to space group P3221, with cell parameters a=b=82.7 Å, c=61.1 Å, and two molecules per asymmetric unit. The final model includes residues 63 to 116 and 48 water molecules. Residues 59-62 are disordered. The C-terminal NEP domain forms a helical hairpin with approximate dimensions 40 Å x 25 Å x 15 Å. Residues 64-85 and 94-115 comprise helices C1 and C2, respectively, while residues 86-93 form the interhelical turn. The two helices are equal in length and nearly perfectly anti-parallel, giving the structure a remarkably flat appearance.
NEP appears to be a modular protein composed of
a protease-sensitive N-terminal domain which mediates RanGTP-dependent Crm1
binding, and a protease-resistant C-terminal domain chiefly responsible for M1
binding. Using pull-down experiments,we have shown that NEP-M1 interaction
involves two critical epitopes: an exposed tryptophan, Trp78, surrounded by a
cluster of glutamate residues on NEP, and the basic nuclear localization signal
(NLS) of M1. This NLS motif, 101RKLKR105, also mediates
binding to negatively charged liposomes (Baudin et al., 2001). Masking of the NLS motif of M1 by NEP may ensure that newly exported progeny
vRNPs destined for virus assembly at the plasma membrane are not re-imported
into the nucleus.
The efficient
passage of particularly large molecules through the nuclear pore complex has
been shown to require the recruitment of more than one nuclear transport
receptor. Indirect evidence suggesting that several Crm1 molecules are required
for Rev-mediated export of unspliced HIV mRNA is the observation that export
critically depends on the co-operative assembly of Rev multimers on the Rev
response element (RRE). For example, stem-loop IIB within the RRE has high
affinity for Rev monomers but is unable to direct Rev-mediated RNA export,
while Rev variants or mutants unable to form multimeric complexes are defective
in the nuclear export of unspliced RRE-containing mRNAs. For the moment,
similar evidence is lacking for the export of influenza vRNPs, but their large
size (average of ~5 MDa) suggests that they may also recruit several Crm1 molecules.
Although the precise M1:vRNP binding stoichiometry during export remains
undetermined, the results from a number of studies suggest that an individual
vRNP can bind several M1 molecules. Association of NEP with more than one of
these would thus permit a single vRNP to recruit multiple Crm1 molecules.
Experiments to test this hypothesis are currently underway.
Guy Schoehn and Celine Fabry in collaboration with Pascal Fender,
Anne-Laure Favier, Larissa Balakireva and Wisia Chroboczek (IBS Grenoble). Some of this work was published
in 2002: Favier et al.Virology 293, 75-85.
Adenoviruses, icosahedral double stranded DNA viruses, are important human pathogens but also one of the viruses most often used for gene therapy. We want to understand the entry of virus into cells and how individual adenovirus proteins or protein complexes could be used for gene therapy. To study the mechanism of human adenovirus (Ad) 2 entry, we used alveolar adenocarcinoma A549 cells, which have retained the ability of alveolar epithelial type II cells to synthesize the major component of pulmonary surfactant, disaturated phosphatidylcholine. Stimulation of phosphatidylcholine secretion by a calcium ionophore or phorbol ester augmented the susceptibility of these cells to Ad. Both Ad infection and recombinant-Ad-mediated transfection increased in the presence of dipalmitoyl phosphatidylcholine (DPPC) liposomes in culture medium. Importantly, in the presence of DPPC liposomes, virus penetrates the cells independently of virus-specific protein receptors. DPPC vesicles bind Ad and are efficiently incorporated by A549 lung cells, serving as a virus vehicle during Ad penetration. To identify the viral protein(s) mediating Ad binding, flotation of liposomes preincubated with structural viral proteins was employed, showing that the only Ad protein that bound to DPPC vesicles was the hexon protein. The hexon preserved its phospholipid-binding properties upon purification, confirming its involvement in virus binding to the phospholipid. This was also confirmed by electron microscopy. Given that disaturated phosphatidylcholine not only covers the inner surface of alveoli in the lungs but also re-enters alveolar epithelium during lung surfactant turnover, Ad binding to this phospholipid may provide a pathway for virus entry into the alveolar epithelium in vivo.
Enteric adenoviruses of serotypes 40
and 41 possess two fibers of different lengths and primary sequences. The
production protocol of this notoriously-difficult-to-grow virus has been
improved and reasonable amounts of particles could be obtained. Structural Ad41
proteins were analysed by biochemical methods, mass spectrometry, and electron
microscopy (EM), in order to identify and localize them on polyacrylamide SDS
gels and to assess the proportion of short and long fibers in the virion.
Surprisingly, the three proteins that make up the short and long pentons of the
virus did not totally enter the denaturing polyacrylamide gels, which is
probably due in part to their high pI. The pentons were separately purified and
their dimensions were estimated from EM data. The EM images suggest that there
are the same amounts of short and long fibers in each virion (see figure 2).
Figure
2: Negative stained image of one human
adenovirus 41 particle. The top arrow depicts a short fiber and the bottom one
a long fiber. On the right side, at the same magnification, are the
corresponding isolated pentons.
Guy Schoehn and Rob Ruigrok in collaboration with Winfried Weissenhorn
(EMBL Grenoble) The matrix protein VP40 from Ebola virus plays an
important role in the assembly process of virus particles by interacting with
cellular factors, cellular membranes, and the ribonucleoprotein complex. The
N-terminal domain of VP40 was expressed in E. coli and we
have been able to show that it folds into a mixture of two different oligomeric
states in vitro, namely hexameric and octameric ring-like structures, as
detected by gel filtration chromatography, chemical cross-linking, and electron
microscopy (see Fig. 3). The two forms bind differently to the carbon when
prepared for negative staining (the octamer bind preferentially in top views
whereas the hexameric form bind on the side). Octamer formation depends largely
on the interaction with nucleic acids, which, in turn, confers in vitro SDS resistance to the complex. Refolding experiments
with a nucleic acid-free N-terminal domain preparation reveal a mostly dimeric
form of VP40, which is transformed into an SDS-resistant octamer upon
incubation with E. coli nucleic acids. In
addition, we showed that the N-terminal domain of Marburg virus VP40 also folds
into ring-like structures, similar to Ebola virus VP40. Interestingly, Marburg
virus VP40 rings reveal a high tendency to polymerise into rods composed of
stacked rings. These results may suggest distinct roles for different oligomeric
forms of VP40 in the filovirus life cycle.
Figure
3: Comparison of the octameric (left) and the
hexameric form of the N-term part of expressed Ebola virus VP40. The octameric
form shows more top views where as the hexameric one shows more side views. The
top views look either round (left) or triangular (right) as showed by the
circle and the average image in the bottom left part of each image.
Another part of the work concerning
VP40 is the study of the role that it plays in virus assembly and budding at
the plasma membrane of infected cells. For efficient budding, a intact amino
terminus of VP40 is required, which includes a PPXY and a PT/SAP motif, both of
which have been proposed to interact with cellular proteins. We have shown that
Ebola VP40 can interact with cellular factors Nedd4 and Tsg101 in vitro
and that WW domain 3 of human Nedd4 is necessary and sufficient for binding to
the PPXY motif of VP40, which requires an oligomeric conformation of VP40.
Single particle electron microscopy reconstructions indicate by difference
imaging with VP40 rings alone that WW3 of Nedd4 is in close contact with the
N-terminal domain of hexameric VP40 (Fig 4). In contrast, the ubiquitin enzyme
variant domain of Tsg101 was sufficient for binding to the PT/SAP motif of
VP40, regardless of the oligomeric state of the matrix protein. These results
suggest that hNedd4 and Tsg101 may play complimentary roles at a late stage of
the assembly process, by recruiting cellular factors of two independent
pathways to the site of budding at the plasma membrane.
Figure
4: Complex between VP40 and WW23. Right: negative
stained image of the complex. The arrows show extra density fixed to the VP40
rings. Left: comparison between the reconstruction of the hexameric ring of
VP40 (left) and the complex between the hexameric ring and WW23 (center). The
extra densities are in blue and green. On the right is the interpretation of
the extra density based on the VP40-WW23 structure and on the VP40-WW34
structure (no shown).
Structure of the C-terminal domain
of the phosphoprotein of Sendai virus.
Rob Ruigrok in
collaboration with Laurence Blanchard, Martin Blackledge and Dominique Marion
from the IBS, Peter Timmins from the ILL (Institut Laue Langevin) and with Wim
Burmeister and Nicolas Tarbouriech, formerly ESRF but now respectively at the
UJF and EMBL. This work is in press: Blanchard, L., Tarbouriech, N., Blackledge, M.,
Timmins, P., Burmeister, W.P., Ruigrok, R.W.H. and Marion, D. (2004). Structure and dynamics of the nucleocapsid-binding domain
of the Sendai virus phosphoprotein in solution. Virology.
P
has 568 amino acid residues. The N-terminal half
of P (see Figure 5) is largely unstructured and the only functional part in
this half of the protein is a short predicted helix that is needed when P binds
to another form of N (N°), that is not yet bound to RNA. Bruno Canard, Sonia
Longhi and co-workers from UMR 6098 CNRS Marseille
have done extensive analyses on several sequences of paramyxovirus P proteins
(Sendai virus is the prototype paramyxovirus) and have suggested that the
N-terminal half of the protein is structurally disordered. The C-terminal half
from residue 344 is structured and is, by its own, active in supporting
transcription activity by L (most of the functional work on P has been done by
Joe Curran and Dan Kolakofsky, dept. Microbiology, University of Geneva). Using
limited proteolysis we were able to define a number of subdomains of the
C-terminal half of P. The first part of the
C-terminal half consists of the oligomerisation domain of P. For
Sendai virus we have shown before that P forms a tetramer and that it is this
tetramer that binds L to the nucleocapsid. The atomic structure of the
tetramerization domain was determined by Nicolas Tarbouriech and Wim Burmeister
when they were worked for the ESRF (Figure 2) (Tarbouriech et al., Nature
Struct. Biol. 7, 777-781 (2000).
The most C-terminal subdomain of P,
PX, represents the N-RNA binding domain. This part of the protein is also
expressed by itself in infected cells since the mRNA for P contains an internal
ribosome initiation site that allows translation of this domain. After
Figure 5.
Schematic representation of the Sendai virus phosphoprotein and the domains for
which atomic structures have been determined.
After extensive
crystallisation trials without success, we decided to approach the NMR group at
the IBS in order to solve the structure of this part of P. From the NMR
results, it appeared that PX itself consisted again of two parts, a very
flexible N-terminal part and a structured C-terminal part making up a
three-helical bundle that forms the actual N-RNA binding domain. Even though
the N-terminal part of PX was seen to be flexible by NMR, it did not have the
characteristics of an unfolded protein.
The intact C-terminal half could not
be crystallised, probably because many parts are flexible, but was studied by
small angle scattering using both neutron (SANS) and X-ray radiation (SAXS).
From SANS experiments on the intact C-terminal domain, on the tetramerization
domain and on intact PX including the flexible domain, combined with the X-ray
structure of the tetramerization domain and the NMR structure of the C-terminal
part of PX, we could build a model for the C-terminal domain, as shown in
Figure 6.
Figure 6. Model for the C-terminal
domain of Sendai virus phosphoprotein. The domains that binds to the
nucleocapsid (N) and to the polymerase (L) are indicated. The cylinders
indicate the size and the relative position of parts of the structure for which
no atomic structure information exists.
The phosphoprotein is an extremely flexible protein with only a few well structured domains. Most if not all of its N-terminal domain seems to be unstructured and can be deleted without losing the transcription support activity. The oligomerisation domain forms a four helix coiled coil but the part of this structure that binds to L is also particularly flexible (Tarbouriech et al., Nature Struct. Biol. 7, 777-781, 2000). C-terminal of this coiled coil is a domain with extreme flexibility which is followed by the nucleocapsid binding domain PX that is itself so small that it is at the size-limit of a stable domain. Only a combination of techniques such as used here will give full information on such a protein. The presence of the European synchrotron (ESRF), a neutron reactor (ILL) and the instruments present and developed at IBS and EMBL makes Grenoble a unique place in the world allowing this kind of research.
Web page of the EMBL virus group
Home page
mise à jour 19.1.04