JOU~'~'I. °J~P.~1'~era/, ,~T!r()~Ig~. ( !.9 ? 6).~.7 7.'.2,6~ 5Z2 6 9 9:.,,, iP~int~d,,!n..Grear. ,Brita!n
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
SHORT COMMUNICATION
The ectodomain of measles virus envelope glycoprotein does
not gain access to the cytosol and MHC class I presentation
pathway following virus-cell fusion
Alicia I. Cardoso, 1 Denis Gerlier, z T. Fabian Wild 3 and Chantal R a b o u r d i n - C o m b e 1
11mmunobiologie Mol6culaire, CNRS, UMR 49, I~cole Normale Sup6rieure de Lyon, 46 Allee d'ltalie, 69364 Lyon Cedex 07, France
2 Immunit6 et Infections Virales, IVMC, CNRS-UCBL, UMR 5537, Faculte de M6decine Lyon R. T. H. Laennec, 69372 Lyon Cedex 08,
France
3 INSERM U 404 'Immunity and Vaccination', Institut Pasteur de Lyon, Avenue Tony-Gamier, 69365 Lyon Cedex 07, France
To unravel the intracellular fate of measles virus
(MV) haemagglutinin (H) following fusion of the
virus envelope with the cell membrane, its presentation by MHC molecules to T cells was explored.
After MV infection, murine cells expressing CD46
were lysed by MHC class I-restricted CD8 CTLs
specific for the ectodomain of H. In contrast, when
sensitized with UV-inactivated MV, they were not
lysed by these effectors, but were recognized by Hspecific and class II-restricted CD4 CTLs. Thus, after
MV binding and fusion, H becomes associated with
plasma membrane and its ectodomain can reach
the endosomal MHC-II but not the cytosolic MHC-I
antigen presentation pathway. From these data
and a reappraisal of previous reports, it appears
that the ectodomains of both MV haemagglutinin
fusion proteins, having undergone the fusion step,
are not translocated into the cytosol and end up
in the endosomes.
Measles virus (MV) attaches to human host cells by binding
to the CD46 receptor (Naniche eta[., 1993) and the virus
envelope fuses with the cell membrane at neutral pH (see
Gerlier eta]., 1995, for review). The haemagglutinin (H) binds
to CD46 and the fusion protein (F) triggers the fusion process.
This requires a precise scaffolding between H, F and CD46 so
that the nuc[eocapsid can be injected into the cell without
extracellular leakage of cytosol. An appropriate H and F
structural pairing governs the efficiency of fusion (Wild et aI.,
1994) (see Gerlier et al., 1995, for review), and the distance
between the H binding site on CD46 and its membrane anchor
must be kept within an optimum interval (Buchholz et aI.,
Authorfor correspondence:ChantalRabourdin-Combe.
Fax + 33 72 72 86 86. e-mail [email protected]
0001-4079 © 1996 SGM
1996). Moreover, fusion probably involves insertion of a
hydrophobic fusion peptide implying a conformational change
in the F protein (Lamb, 1993), as modelled from the aciddependent fusion induced by influenza virus haemagglutinin
(White, 1994). Little is known as to the fate of H and F
glycoproteins after fusion. Together with other viral proteins
H and F are targets of the CD4- and CD8-dependent immune
response. Knowledge of how the individual MV antigens are
routed intracellularly in order to be presented by MHC class I
and II molecules to CD8 and CD4 T cells would help in
defining the requirements a vaccine must meet in order to
stimulate the corresponding T cells. Indeed, MHC class Irestricted presentation involves antigen processing in the
cytosol and/or endoplasmic reticulum (ER) whereas MHC-IIrestricted presentation uses an endosomal pathway (Germain,
1994). Failure to stimulate CD8 CTL and/or CD4 ThI/Th2
subsets may result in a severe immunopathology after infection
with the wild-type virus, as documented with formalininactivated vaccine against MV and respiratory syncytial virus
(Cardoso et at., I995; Connors eta/., I994; Fulginiti et aI.,
I967). To follow the intracellular fate of the ectodomain of MV
H glycoprotein after CD46-mediated MV binding and fusion,
the efficiency of H antigen presentation by MHC-I and MHCII molecules was investigated.
Polyclonal H-specific MHC class I-restricted CTLs were
obtained by immunization of H-2 a mice with a recombinant
vaccinia virus encoding MV H (rVV-H) (Wild et al., 1992) and
in vitro secondary stimulation with MHC-I+II- P815-H cells,
which constitutively express MV H (see Beauverger eta[.,
1994, for details). The polyclonal CTLs were Ld-restricted and
recognized two H epitopes, amino acid sequence 343-351 and
544-552, which both belong to the ectodomain of the protein
(Beauverger et aI., 1994). Murine MHC-I+II- P815 cells
expressing (Cardoso et al., 1995) or not expressing CD46 were
incubated overnight with infectious MV (Hall6 strain) at
0"45 p.f.u, per cell, and then tested for their sensitivity to lysis
by the polyclonal CTLs. When infected with MV, P815-CD46,
but not parental P815 cells, were readily lysed by H-specific
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Jul 2017 16:31:04
_~69.~
100
(a)
.o/{.o.°,°°,..°°°'°°°°"°'/~
80
60
20
oi
100
I
10
100
(b)
80
~
6o
~
4o
2
0
0
~
I
10
E:T ratio
100
Fig. 1. CD46 expression is required for presentation of infectious NIV by
murine cells and H from UV-inactivated MV is not presented by P81 5CD46 cells to H-specific CTLs. Specific cytolysis (%) by H-specific CTLs
of: (a) MV-infected P81 5 cells ( t l ) , MV-infected P81 5-CD46 cells (111),
P81 5-H (A), uninfected P81 5-CD46 cells (rq) and P81 5-NP ( 0 ) ; (b)
UV-MV-loaded P81 5 cells ( 0 ) , UV-MV-loaded P81 5-CD46 cells (11) and
unloaded P81 5-CD46 cells (C1).
CTLs (Fig. la). These CTLs were specific for H since they lysed
P815-H but not P815-NP cells, which expressed the MV
nucleoprotein (Beauverger et al., 1994) (Fig. la), and, as
expected, they belong to the CD8 lymphocyte subset (see Fig.
3a). Thus, expression of CD46 was both necessary and
sufficient to allow murine P815 mastocytoma cells infected
with MV to present H to specific MHC class I-restricted CTLs.
To determine whether the source of H peptides was the H
brought by the infecting virus particles or the neosynthesized
H, UV-inactivated virus was used as a source of H antigen. UV
inactivation of MV leaves the binding and fusion abilities
intact (Cardoso et al., 1995). P815-CD46 cells incubated with
UV-inactivated MV were not recognized by H-specific and
MHC-I-restricted CTLs (Fig. i b). To verify that UV-inactivated
MV could still bind to CD46-expressing cells and provide
them with MV antigens, the same preparation was incubated
with the B cell line MHC-I+II + M12-CD46 (Naniche eta].,
1993), and the sensitivity of these targets to lysis by H-specific
and MHC-II-restricted CTLs was tested. H-specific and class IIrestricted CTLs were derived by in vitro stimulation of spleen
cells from rVV-H immune mice with UV-inactivated MV.
!696
These CTLs were able to lyse class I+II + M12-CD46 cells that
were either infected with MV or sensitized with UVinactivated MV (Fig. 2 a, b). The class II restriction of this CTL
activity was ascertained by showing that (i) these effectors
were unable to kill MHC class I+II - P815-H or P815-CD46
cells infected with MV or incubated with UV-inactivated MV
(Fig. 2 c, d); (ii) killing of M12-CD46 cells sensitized with MV
was abolished in the presence of anti-CD4 but not anti-CD8
antibodies (Fig. 3 b, c); (iii) killing was also inhibited when
antigen processing was blocked by treatment (50 I~M)with the
lysosomotropic agent chloroquine during antigen loading of
the targets and the CTL assay according to a procedure
reported by Lombard-Platet et al. (1993) (Fig. 3 c); and (iv) they
lack any detectable NK activity since no killing of YAC cells
was observed (Fig. 2 c). These results show that UV-inactivated
MV can provide the CD46-expressing antigen-presenting cells
(APC) with H antigen available for class II- but not for class Irestricted presentation to T cells.
Thus, expression of the MV receptor, human CD46, in
mouse cells is an absolute requirement for efficient presentation
of MV H to MHC class I-restricted specific CTLs. This effect is
a result of the ability of CD46-transfected cells to capture MV
particles, allow fusion, genome replication and de novo
synthesis of viral proteins (Gerlier et al., I994; Naniche et aI.,
1993; D. Gerlier, unpublished). Indeed, H synthesized in the
target cells after infection with MV was presented by class I
molecules, since P815-CD46 cells infected with MV were
lysed by H-specific CTLs as were the P815-H cells (transfected
with H gene). With UV-inactivated MV, H glycoprotein could
not be presented by MHC class I molecules since P815-CD46
cells, which expressed only class I molecules, were not
recognized by H-specific CTL when sensitized with UVinactivated MV. The lack of class I-restricted presentation of H
from UV-MV was not due to the lack of H association with the
target cells because H can be processed and presented by MHC
class II molecules, as shown by the recognition of UV-MVsensitized M12-CD46 cells by H-specific CD4 CTLs (see Fig.
2 a, b) and by a class II I-Ea-restricted CD4 T cell hybridoma
(Gerlier et a]., 1994). It should be stressed that the UV-MV was
still able to fuse with the cell plasma membrane as the NP
component was presented by the cytosolic MHC class I
antigen presentation pathway (Cardoso et aI., 1995).
Why is there class II- but not class I-restricted presentation
of H from UV-inactivated MV? In H-2 d mice, both class I(Beauverger et al., 1994) and class II- (P. Devaux & D. Gerlier,
unpublished) restricted epitopes mapped to the ectodomain
(luminal domain) of H. On the one hand, the lack of class Irestricted presentation of the luminal part of H provided to the
APC by fusion of the envelope from UV-inactivated MV
indicates that it cannot gain access to any cellular compartment
relevant for appropriate MHC-I-restricted antigen processing,
including the cytosol. Indeed, four pathways leading to the
formation of peptide-class I complexes from transmembrane
antigens have been identified so far (see Siliciano & Soloski,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Jul 2017 16:31:04
i iiiiiiii i i i iiii i i iiii ii ii iiii iii i i iiiii i iiiii!iii!iiii iiiiiiiiiiiiii iii
60
60
40
~" 40
03
co
~" 20
0
20
1
60
~
o!
10
Fig. 2. H from infectious or UVinactivated MV is presented by M12CD46 but not by P81 5-CD46 cells to
CD4+CD8 - class N-restricted CTLs.
Specific cytolysis (%) by H-specific CTLs
induced after in vitro stimulation with UVM V of; (a) MY-infected M12-CD46 ceils
(11) and unloaded M12-CD46 ( R ) ; (b)
UV-MV-loaded M12-CD46 cells (11) and
unloaded M12-CD46 (FI) ; (c) P815-H
cells (A), P815-NP cells (O) and YAC-1
cells (I-]) ; (d) MV-infected P815-CD46
cells (11), UV-NV-loaded P815-CD46
cells ( 0 ) and unloaded P815-CD46 cells
(D).
....................c~ ..................... D
10
60 - (d)
(c)
4o
~'4o
-
~" 20 -
0
0
10
E:T ratio
'
1
E:T ratio
(a)
10
(c)
100
T
=
75
.~
5o
%
•
%
#
#'
= 25
%
P
#"
%
#
,
%
,#
•
#
,
,
%
J
•
0, I . R m l l J i ~
P815-CD46
+inf. MV
M12-CD46
+ inf. MV
M12-CD46
+ UV-MV
Fig. 3. Characterization of CTL effectors. (a) H-specific polyclonal CTLs from mice immunized with rVV-H and restimulated in
vitro with P81 5-H cells are CD8+CD4 -. Spleen cells were tested for their cytolytic activity at an E/T ratio of 3:1 in the
presence or absence of anti-CD8 H35.17 or anti-CD4 H129,19 MAbs added to the effectors prior to incubation of the P81 5CD46 targets with MV. Results are expressed as percentage inhibition of cytolysis. (b, c) H-specific polyclonal CTLs from mice
immunized with rW-H and restimulated in vitro with UV-irradiated N V are CD4+CD8 -. Spleen cells were tested for their
cytolytic activity at E/T = 3/1 in the presence or absence of anti-CD8 or anti-CD4 NAbs added to the effectors prior to Ioadin 9
of the M12-CD46 targets with (b) infectious or (c) UV-irradiated MV. Other tests were done in the presence of 50 pMchloroquine added to the APC prior to the virus and maintained throughout the bioassay (CHQ column).
1995, for review). One involves antigen degradation into the
ER and peptide association with the nascent heavy chain of
MHC-I. In the second, some misfolded proteins fail to be
translocated into the ER and are degraded into peptides by
cytosolic proteasomes. The peptides are then transported
through the TAP1/TAP2 heterodimers to the ER where they
can associate with nascent MHC-I molecules. Alternatively,
the transmembrane antigen is degraded within the ER into
long peptides that are able to cross the ER membrane and reach
the cytosol. Here, they can be further degraded into smaller
peptides able to return to the ER via the TAP1/TAP2
transporter. In a fourth and unusual pathway, the transmembrane antigen is internalized by endocytosis where it is
degraded. The peptides became available (by 'regurgitation'?)
for association with (recycling?) cell-surface MHC-I molecules.
On the other hand, the class II-restricted presentation of the
luminal part of H indicates that it has access to the cellular
compartment relevant for appropriate MHC-II-restricted antigen processing, i.e. the endosomal compartment (Germain,
1994). As a newly incorporated plasma membrane glycoprotein, H from the fused viral envelope has easy access to the
endosomes as it follows the usual endocytic fate of plasma
membrane proteins. In order to reach the lumen of the
endosomes, the extracellular domain of H will have to remain
oriented towards the outside of the cell, as demonstrated in the
case of the human immunodeficiency virus glycoprotein
(Callahan et aI., 1993). From these data, we infer that during
fusion of the MV envelope with the plasma membrane of
target cells, the ectodomain of H (or at least the domain of H
encompassing the H 343-351 and 544-552 class I epitopes)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Jul 2017 16:31:04
!69:
does not cross the plasma membrane and remains oriented
towards the outside of the cell.
What is the orientation of the F glycoprotein during fusion?
A partial answer come from our reappraisal of class I-restricted
presentation of F in humans as reported by Van Binnendijk et
al. (I992, I993). F provided to the APC as UV-inactivated MV
was also found not to be presented by M H C class I whereas F
from infectious M V was processed and presented by M H C
class I molecules. One of the CTL clones (clone F40) recognized
the amino acid sequence 438-446 of F localized in the
extracellular domain of the F1 subunit. These data imply that
during the fusion process, the extracellular domain of F1 (at
least the domain encompassing the F438-446 epitope) remains
oriented outside the cell and is not translocated to the cytosolic
side, although its C-terminal part, which contains the fusion
peptide (Fl16-147), is thought to be inserted into the plasma
membrane during fusion (Paterson & Lamb, 198 7). Altogether,
this favours the probability that during MV-to-cell fusion, both
envelope H and F glycoproteins keep their extracellular domain
oriented towards the cell exterior (i.e. they are not translocated
into the cytosol) and end up into endosomal compartments.
We predict that this topology of the envelope glycoproteins
after fusion will be confirmed by other means such as immunoelectron microscopy. Indeed, during membrane fusion with the
influenza virus envelope there is no merger of the inner and
outer leaflets of the two fusing membranes, so that transverse
membrane asymmetry is maintained (Klotz et aI., 1996). In
contrast, when H and F are endogenously synthesized by cells,
their luminal portion can be presented by M H C class I
molecules but it has to be processed prior to its expression at
the cell surface. These observations should be taken into
account when engineering and evaluating new vaccines against
M V in order to obtain an adequate stimulation of CD8 a n d / o r
CD4 T h l / T h 2 subsets and to avoid the imbalance of the
immune response which may be responsible for the exacerbated atypical disease as seen with inactivated vaccines
(Cardoso et al., 1995; Fulginiti et al., 1967).
The authors are indebted to K. Belorizky, I. Fugier-Vivier, C. Gimenez,
B. Horvat, S. Krantic-Moyse, J. Marvel, P. Rivailler, Mo-C. TrescolBi6mont and G. Varior-Krishnan for helpful discussions and comments.
This work was supported in part by grants from the Association pour la
Recherche contre le Cancer (CRC 6108) and the World Health
Organization/UNPP Programme for Vaccine Development (PVD)
(B.H.). A. Cardoso was supported by Fondation M6rieux and WHO.
References
Beauverger, P., Buckland, R. & Wild, F. (1994). Measles virus
hemagglutinin induces an La-restricted CD8+ cytotoxic T lymphocyte
response to two specific epitopes. Virology 200, 281-283.
Buchholz, C., Schneider, U., Devaux, P., Gerlier, D. & Cattaneo, R.
(1996). Cell entry by measles virus: long hybrid receptors uncouple
binding from membrane fusion. Journal of Virology 70, 3716--3723.
Callahan, K.M., Rowell, J.F., Soloski, M.J., Machamer, C.E. &
Siliciano, R. F. (1993). HIV-1 envelope protein is expressed on the
_~69~
surface of infected cells before its processing and presentation to class IIrestricted T lymphocytes. Journal of Immunology 151, 2928-2942.
Cardoso, A. I., Beauverger , P., Gerlier, O., Wild, T. F. & RabourdinCombe, C. (1995). Formaldehyde inactivation of measles virus abolishes
CD46-dependent presentation of nucleoprotein to murine class Irestricted CTLs but not to class II-restricted helper T cells. Virology 212,
255-258.
Connors, M., Giese, N. A., Kulkarni, A. B., Firestone, C.-Y., Morse H. C.~
III& Murphy,B. R. (1994). Enhanced pulmonary histopathology induced
by respiratory syncytial virus (RSV) challenge of formalin-inactivated
RSV-immunized BALB/c mice is abrogated by depletion of interleukin4 (]L-4) and IL-I0. Journal of Virology 68, 5321-5325.
Fulginiti, V.A., Eller, I.J., Downie, A.W. & Kempe, C. H. (1967).
Altered reactivity to measles virus: atypical measles in children
previously immunized with inactivated measles vaccines. Journal of the
American Medical Association 202, 1075-1080.
Gerlier, D., TrescoI-Bi4mont, M. C., Varior-Krishnan, G., Naniche, D.,
Fugier-Vivier, I. & Rabourdin-Combe, C. (1994). Efficient MHC classII-
restricted presentation of measles virus relies on hemagglutinin-mediated
targeting to its cellular receptor human CD46 expressed by murine B
cells. Journal of Experimental Medicine 179, 353-358.
Gerlier, D., Varior-Krishnan, V. & Devaux, P. (1995). CD46-mediated
measles virus entry: a first key to host-range specificity. Trends in
Microbiology 3, 338-345.
Germain, R. N. (1994). MHC-dependent antigen processing and peptide
presentation: providing ligands for T lymphocyte activation. Cell 76,
287-299.
Klotz, K.-H., Bartoldus, I. & Stegman, T. (1996). Membrane asymmetry
is maintained during influenza-induced fusion. Journal of Biological
Chemistry 271, 2383-2386.
Lamb, R.A. (1993). Paramyxovirus fusion: a hypothesis for changes.
Virology 197, 1-11.
Lombard-Platet, S., Bertolino, P., Deng, H. A., Gerlier, D. & RabourdinCombe, C. (1993). Inhibition by chloroquine of the class II major
histocompatibility complex-restricted presentation of endogenous antigens varies according to the cellular origin of the antigen presenting cells,
the nature of the T cell epitope, and the responding T cell. Immunology
80, 566-573.
Naniche, D., Varior-Krishnan, G., Cervoni, F., Wild, T. F., Rossi, C.,
Rabourdin-Combe, C. & Gerller, D. (1993). Human membrane cofactor
protein (CD46) acts as cellular receptor for measles virus. Journal of
Virology 67, 6025--6032.
Paterson, R.G. & Lamb, R.A. (1987). Ability of the hydrophobic
fusion-related external domain of a paramyxovirus F protein to act as a
membrane anchor. Cell 48, 441-452.
Siliciano, R.F. & Soloski, M.J. (1995). MHC class I-restricted
processing of transmembrane proteins - mechanism and biologic significance - commentary. Journal of Immunology 155, 2-5.
Van Binnendijk, R. S., Van Baalen, C. A., Poelen, M. C. M., Devries, P.,
Boes, J., Cerundolo, V., Osterhaus, A. D. & Uytedehaag, F. G. (1992).
Measles virus transmembrane fusion protein synthesized de novo or
presented in immunostimulating complexes is endogenously processed
for HLA class 1- and class II-restricted cytotoxic T cell recognition. Journal
of Experimental Medicine 176, 119-128.
Van Binnendijk, R.S., Versteegvanosten, J.P., Poelen, M.C.M.,
Brugghe, H. F., Hoogerhout, P., Osterhaus, A. D. & Uytedehaag, F. G.
(1993). Human HLA class I- and HLA class II-restricted cloned cytotoxic
T lymphocytes identify a cluster of epitopes on the measles virus fusion
protein. Journal of Virology 67, 2276-2284.
White, J. (1994). Fusion of influenza virus in endosomes: role of the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Jul 2017 16:31:04
iiiiiiiii iii i i!iii! iiii iii i iii iii iiii iii i ii iiiiiii iii iii iii iii i i iiiiiiiiiiiiiiiiii
hemagglutinin. In Cellular Receptors for Animal Viruses, pp. 281-302.
Edited by E. Wimmer. New York: Cold Spring Harbor Laboratory.
Wild, T. F., Bernard, A., Spehner, D. & Drillien, R. (1992). Construction
of vaccinia virus recombinants expressing several measles virus proteins
and analysis of their efficacy in vaccination of mice. Journal of General
Virology 73, 359-367.
Wild, T. F., Fayolle, J., Beauverger, P. & Buckland, R. (1994). Measles
virus fusion: role of the cysteine-rich region of the fusion glycoprotein.
Journal of Virology 68, 7546--7548.
Received 17 April 1996; Accepted 16 July 1996
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Jul 2017 16:31:04
~_69~