Download Recognition of measles virus-infected cells by CD8MT cells

Journal
of General Virology (1998), 79, 2583–2591. Printed in Great Britain
...................................................................................................................................................................................................................................................................................
Recognition of measles virus-infected cells by CD8MT cells
depends on the H-2 molecule
Claudia Neumeister and Stefan Niewiesk
Institute of Virology and Immunobiology, University of Wu$ rzburg, Versbacher Str. 7, 97078 Wu$ rzburg, Germany
H-2d mice are resistant to measles virus-induced
encephalitis (MVE) and develop Ld-restricted CD8M
T cells which lyse target cells infected with measles
virus or with a vaccinia virus recombinant expressing
the nucleocapsid protein of measles virus (vvN). In
contrast, H-2k mice are susceptible to MVE and
generate CD8MT cells which lyse target cells infected
with vvN, but not those infected with MV. We were
able to demonstrate that this difference is not due
to a defect in the antigen processing machinery, but
that Kk molecules require 100-fold more peptide to
sensitize target cells for lysis by CTL. vvN replicates
well in target cells and therefore enhances the level
Introduction
Infection with measles virus (MV) leads to acute measles
which is characterized by fever and the typical rash. Usually,
the disease is overcome within two weeks, but often
complications such as encephalitis, pneumonia and diarrhoea,
as well as secondary infections due to the marked immune
suppressive effect of MV arise (for a review, see Katz, 1995).
During infection, B and T cell responses arise. Observations in
patients with agammaglobulinaemia indicate that B cell
responses against MV are dispensable (Bruton, 1953 ; Good &
Zak, 1956). In contrast, patients with defects in the cellular
immune response are not able to clear the virus (Nahmias et al.,
1967). Although CD4+ as well as CD8+ T cells are generated
during acute infection, their contribution to overcoming
infection remains to be elucidated. The frequency of CD4+ T
cells and their highly lytic potential has been emphasized
(Long & Jacobson, 1989). MV is lymphotropic and, at least in
tissue culture, very cytopathic. Therefore, the generation of
CD8+ T cells with MV-infected stimulator cells has been
problematic, although with recent technical advances, a number
Author for correspondence : Stefan Niewiesk.
Fax ­49 931 201 3934. e-mail viro025!mail.uni-wuerzburg.de
0001-5677 # 1998 SGM
of epitope peptide available for CTL recognition. In
contrast, MV infection is abortive in mouse cells and
low levels of epitope peptide are produced. As Ld
requires 100-fold less peptide than Kk to sensitize
target cells for lysis, the low level of epitope peptide
is enough to induce lysis by CD8MT cells, whereas
for recognition via Kk, increased synthesis of protein
is required. We propose that the differences in
peptide binding between the two H-2 molecules
will have consequences for the kinetics of the
generation of CD8M T cells as well as the
absolute numbers of CD8MT cells generated.
of reports have described CD8+ T cell responses (Nanan et al.,
1995 ; Uytdehaag et al., 1994). However, the contribution of
CD4+ versus CD8+ T cells in overcoming disease is not yet
known. Indirect evidence for the importance of CD8+ T cells is
the fact that a high frequency of CD8+ T cells has been found
in children recovering from acute measles (van Binnendijk et al.,
1990). In the mouse model of MV-induced encephalitis (MVE),
susceptibility and resistance of various inbred and H-2
congenic mouse strains depend on their MHC haplotype
(Niewiesk et al., 1993). In resistant H-2d mice (like BALB}c),
the CD8+ T cell response is restricted by the Ld molecule and
in susceptible H-2k mice (like C3H) this response is restricted
by the Kk molecule. BALB}c mice generate good CD8+ T cell
responses which result in lysis of target cells infected with MV
as well as cells infected with a vaccinia virus recombinant
expressing the nucleocapsid (N) protein of MV (vvN). In
contrast, infection of C3H mice with MV and stimulation of
spleen cells with MV-infected stimulators give CD8+ T cells
restricted by Kk which do not recognize MV-infected target
cells, but readily lyse target cells infected with vvN (Niewiesk
et al., 1993). CD8+ T cells from both mouse strains recognize
target cells pulsed with the respective peptide epitopes [aa
281–289 on the N protein for Ld (Beauverger et al., 1993) ; aa
52–59 and aa 81–88 for Kk (Beauverger et al., 1994)]. In this
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CFID
C. Neumeister and S. Niewiesk
paper, we demonstrate that recognition of MV via Kkrestricted CTL can be achieved if target cells allow replication,
thus increasing the number of N protein molecules for
processing. In comparison to Ld, Kk requires 100-fold more of
its epitope peptide to sensitize target cells for lysis by CD8+
T cells.
Methods
+ Mice. Mice were bought from Harlan Winkelmann (Borchem) and
were specific pathogen-free (company specification). Every 6–8 months,
the animals were checked for pathogens by serological examination.
Animals were kept in a barrier system with light negative pressure
(150 Pa) and a 12 h day (artificial light), and were fed and watered ad
libitum. The room temperature (24³2 °C) and the humidity (70³5 %)
were regulated by air conditioning. Mice were used between the ages of
6–18 weeks.
+ Infection of mice. For the generation of CTL, mice were infected
intraperitoneally (i.p.) with 8¬10' p.f.u. MV (Edmonston strain) or
2¬10' p.f.u. of a vaccinia virus recombinant (vvR) expressing the
nucleoprotein (NP) of influenza A virus strain PR8.
+ Cells, antibodies and peptides. RMA-S cells that were stably
transfected with H2-Kk were obtained from Gu$ nther Ha$ mmerling,
Heidelberg, Germany and cells that were stably transfected with Ld were
from Hans-Georg Rammensee, Tu$ bingen, Germany. P815 cells that were
stably transfected with H2-Kk were from Denis Gerlier and Chantale
Rabourdin-Combe, Lyon, France, and C1R-Kk cells were from Peter
Cresswell, Yale, USA.
All cell lines were cultured in RPMI}10 % FCS. L929, L-Ld and NS20Y
cells were cultured in MEM}10 % FCS.
The MAb 11-4.1, which is specific for the Kk molecule, was purchased
from Pharmingen. Hybridoma F227, produced at the Institute of Virology
and Immunobiology, secretes MV N protein-specific antibody (Liebert
et al., 1990).
Peptides LDRLVRLI (from MV Edmonston strain N protein aa 52–59,
Kk-restricted), VESPGQLI (MV N protein aa 81–88, Kk-restricted),
YPALGLHEF (MV N protein aa 281–289, Ld-restricted) and SDYEGRLI
(influenza A virus NP aa 50–57, Kk-restricted) were synthesized and
purified (" 90 %) by Jerini Biotools.
+ Generation and culture of CTL. For the generation of CD8+ T
cells, spleen cells from mice were irradiated (300 Gy) and either infected
with MV (Edmonston strain) at an m.o.i. of 1 or incubated with peptide
(50 µg}10( cells) at 37 °C for 2 h. Infected stimulators (3¬10') were
cultivated in an upright 50 ml flask containing 15 ml RPMI}10 % FCS
and 1±5¬10( spleen cells from infected animals. After 7 days, living cells
were separated on a Percoll gradient (1±083 g}ml) and plated at a density
of 10' cells}ml in tissue culture plates. For restimulation, spleen cells (10'
cells}ml) infected with virus or incubated with peptide (see above) and
2 % rat spleen Concanavalin A supernatant were added. T cells were
tested for their cytolytic activity 5 days after restimulation.
+ Cytotoxic assay. Cells (10') were labelled with 3±7 MBq Na&"CrO
%
(DuPont) for 80 min at 37 °C and washed twice. If used, peptide was
added during the labelling step (50 µg}10' cells). Labelled target cells
(10%) in a volume of 50 µl were added to various numbers of T cells in
100 µl volumes in U-bottomed microtitre plates. After 5 h incubation at
37 °C, 75 µl supernatant was harvested and counted. The percentage
lysis was calculated as follows : 100¬[(experimental®spontaneous
release)}(total®spontaneous release)]. If not otherwise indicated, target
CFIE
cells were infected with MV strain Edmonston (m.o.i. of 10) overnight
and with vvR expressing the NP of influenza A virus strain PR8 (vvNP)
or the N protein of MV strain Edmonston (vvN) (m.o.i. of 5) for 5 h.
+ Recombinant vaccinia viruses. Recombinant vaccinia viruses
(vvR) were made by transfecting vaccinia virus Copenhagen DNA and
the shuttle vectors psc11.30R.2 or psc11L­ (Elliott et al., 1995)
containing the various epitopes into Tk-143 cells infected with vaccinia
virus ts7 as described previously (Bankamp et al., 1991). Vaccinia virus
was propagated in Tk-143 cells and isolated by standard procedures.
Plasmids psc11.30R.2 and psc11L­ both contain a Kozak sequence and
were kindly provided by Tim Elliott, Oxford, UK. In addition, psc11L­
contains the HA-1 leader sequence to ensure peptide transport
independent of the antigen presentation transporter (TAP). Oligonucleotides encoding either LDRLVRLI (from MV Edmonston strain N
protein aa 52–59), VESPGQLI (MV N protein aa 81–88) or YPALGLHEF
(MV N protein aa 281–289) were constructed with a 5« end NcoI site
(encoding Met at position 1) and a BglII site at the 3« end. These were
cloned into NcoI}BglII-cut psc11.30R.2 or psc11L­ plasmids and the
sequences confirmed by dideoxynucleotide sequencing. vvN and vvNP
have been described previously (Niewiesk et al., 1993).
+ Competition and sensitization assay. The &"Cr-release assay
was used to assess peptide competition as previously described (Cao et
al., 1996). RMA-S-Kk target cells (10%) were incubated with various
concentrations of competitor peptide for 15 min at room temperature.
Target peptide at a concentration of 10-) M was added for another
15 min at room temperature. For the sensitization assay, target peptide,
but no competitor peptide, was added in various concentrations. CTL
specific for the target peptide were then added at an effector to target
(E : T) ratio of 5 : 1 for the competition assay and 20 : 1 for the sensitization
assay in a final volume of 100 µl. After 5 h incubation at 37 °C, 50 µl
supernatant was harvested and counted.
+ Western blotting and immune precipitation. Western blotting
and immune precipitation were done as previously described (Reich et al.,
1992 ; Niewiesk et al., 1997). Briefly, L929 cells were infected with MV
Edmonston strain (m.o.i. of 10) overnight and with vvN (m.o.i. of 5) for
5 h. For Western blotting, cell lysates were separated on SDS–PAGE,
blotted onto nylon membranes and stained with a human hyperimmune
serum. For immune precipitation, proteins from cells labelled with [$&S]methionine and -[$&S]cysteine (Pro-mix ; Amersham) were precipitated using a human hyperimmune serum and separated by SDS–PAGE.
+ Flow cytometry. Cells were fixed with 3±7 % formaldehyde for 5
min, permeabilized with a Triton X-100 solution (0±25 % in PBS) for 3 min
and stained with the N protein-specific antibody F227 and a donkey antimouse serum labelled with FITC (Dianova). Cells were subsequently
analysed by flow cytometry (Facscan ; Becton Dickinson).
Results
General influence on antigen processing ?
We have verified that MV-infected H-2k mice generate
CD8+ T cells which lyse vvN-infected target cells, but not
MV-infected cells (Fig. 2 a). Some large DNA viruses interfere
by various mechanisms with the recognition of antigen
irrespective of its origin. Adenovirus, for example, inhibits
CTL recognition of influenza A virus after co-infection of
target cells (Yewdell et al., 1988). We tested whether MV was
able to inhibit lysis by influenza A virus-specific CTL after coinfection of target cells with influenza A virus and MV. MV
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CTL in MV encephalitis
(a)
(b)
Fig. 1. Kk-expression of L929 cells infected with MV or treated with IFN-γ. (a) Cell surface expression of the Kk molecule on
L929 cells was determined by flow cytometry. Cells were infected with MV (m.o.i. of 10) for 18 h, treated with 100 µg/ml
IFN-γ for 48 h or left untreated and were stained with Kk-specific antibody 11-4.1 or an isotype control. (b) To test whether the
Kk molecule was retained during MV infection, peptide 52–59 was added to uninfected cells and cells were infected with vvNP
or MV (Edmonston strain). Recognition of target cells was tested using CD8+ T cells specific for the 52–59 peptide. The same
results were obtained with peptide 81–88. +, E : T ratio 1 : 1 ; , E : T ratio 30 : 1.
infection did not influence the activity of influenza A virusspecific CTL (data not shown) and co-infection did not lead to
enhanced recognition of MV by CTL (data not shown).
Sometimes, faulty recognition of target cells due to an
inhibitory effect of a virus on antigen processing can be
overcome by the addition of IFN-γ. This cytokine changes the
recruitment of different subunits of the proteasome and
enhances the level of expression of the TAP transporters and
the MHC class I and II molecules (Hengel et al., 1994 ; Sibille et
al., 1992). Incubation of MV-infected cells with IFN-γ led to an
enhanced level of Kk expression (Fig. 1 a), but not to
recognition of MV-infected cells by CD8+ T cells (data not
shown). We investigated whether the Kk molecule might be
retained in the cell as a result of MV infection. Staining with
the Kk-specific MAb 11-4.1 and subsequent analysis by flow
cytometry indicated that MV-infected cells expressed more Kk
than non-infected cells (Fig. 1 a). In addition, MV-infected cells
could be rendered susceptible to lysis by addition of peptides
52–59 or 81–88 (Fig. 1 b) indicating that there is no retention
of the Kk molecule. These data confirm that the nonrecognition of MV-infected target cells by Kk-restricted CD8+
T cells is specific for MV and not an overall effect on the
cellular machinery.
Impairment of antigen processing or presentation ?
The lack of recognition could in principle be due to a defect
in antigen presentation (due to Ld or Kk) or processing (due to
various genes). In BALB}c mice, which are resistant to MVE,
Ld-restricted CD8+ T cells are generated which lyse MV-
infected P815 cells well (Niewiesk et al., 1993). To test whether
the generation of CD8+ T cells was dependent on the genes of
the BALB}c mouse background, we assayed CTL activity in
BaC3F1 and BALB}k mice. In BaC3F1 mice, every molecule is
encoded by an allele of each parent. In BALB}k mice (congenic
to BALB}c), H-2 is of the k haplotype whereas all other
molecules are imprinted by BALB}c genes. Interestingly
enough, the CTL response seemed to be governed by the
H-2 molecule only. Kk-restricted CD8+ T cells (either derived
from BALB}k or BaC3F1 mice) lysed target cells (L929)
infected with vvN, but did not recognize MV-infected target
cells (data not shown). In contrast, Ld-restricted CTL (also
derived from BaC3F1 mice) lysed MV-infected cells (P815)
(Fig. 2 b).
To investigate the importance of cell type on CTL
recognition, we tested Ld-restricted CTL (from BALB}c mice)
on L929 cells (H-2k) transfected with Ld (L-Ld) (Fig. 2 b). These
cells were lysed after infection with MV, although never quite
as efficiently as P815 cells. In the reverse experiment, CD8+ T
cells from C3H mice specific for either Kk-restricted epitope
were tested on P815 cells transfected with Kk. Infection with
vvN but not MV led to recognition of target cells (Fig. 2 c).
These data indicate that the non-recognition of target cells is
either due to differences in presentation by Ld and Kk or
differences in processing of the respective epitopes.
No defect in antigen processing
As the level of protein content might differ between vvNand MV-infected target cells, we estimated the level of protein
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CFIF
C. Neumeister and S. Niewiesk
be defined as the cause of the non-recognition of MV-infected
cells.
(a)
Protein synthesis is important for recognition by CD8M
T cells
So far the only difference in antigen recognition was the
difference in lysis between MV- and vvN-infected cells. A
major difference between vaccinia virus and MV is that
vaccinia virus replicates well in murine cells whereas MV
replicates abortively. Although MV is taken up by L929 cells
and its presence can be detected by Western blotting, immune
precipitation showed that N was produced in large quantities
only in vaccinia virus-infected cells (Fig. 3 b). To investigate the
effect of MV replication on lysis by CD8+ T cells, we used the
semi-permissive mouse neuroblastoma cell line NS20Y and a
fully permissive human B cell line (C1R) transfected with the
Kk molecule. After infection of NS20Y cells, no more than 15 %
of cells expressed the N protein (as measured by flow
cytometry) and in persistently infected NS20Y, no more than
45 % were N protein-positive (data not shown). The extent of
N protein expression correlated with lysis seen by CD8+ T
cells (Fig. 5 a). Similarly, the percentage of N protein-positive
C1R-Kk cells after infection correlated with lysis (Fig. 5 b).
These data showed that a higher amount of N protein resulted
in lysis.
(b)
(c)
Good binding of peptides to the Kk molecule
Fig. 2. Expression of the Kk or Ld molecule rather than the nature of the
target cell determines recognition of MV-infected cells. (a) CD8+ T cells
from C3H mice infected with MV were tested on L929 cells infected with
vvN (+), MV (_) and vvNP ( ). (b) CD8+ T cells from BALB/c mice
were tested on cells expressing the Ld molecule (P815, E : T ratio 30 : 1 ;
L929-Ld, E : T ratio 6 : 1). +, vvNP ; 8, MV ; , vvN. (c) CD8+ T cells from
C3H mice were tested on cells expressing the Kk molecule (L929, E : T
ratio 30 : 1 ; P815-Kk, E : T ratio 6 : 1). +, vvNP ; 8, MV ; , vvN.
present in infected cells. After infection under standard
conditions, protein was present in both cell samples (as seen by
Western blotting ; Fig. 3 a). Therefore, both samples should be
recognized by Kk- and Ld-restricted CD8+ T cells. As Kkrestricted epitopes are not recognized, antigen processing of
these epitopes might be impaired. We tried to address specific
steps in the antigen processing machinery by expressing the
full-length N protein, the Kk-restricted epitopes alone (to
circumvent faulty protein degradation by the proteasome), and
the epitopes with the HA-1 signal peptide (to circumvent
putatively inefficient transport by the TAP molecules) in vvR
(Fig. 4). Kk-restricted CD8+ T cells recognized all constructs
equally well so that no specific step in antigen processing could
CFIG
If more protein (and therefore more peptide epitope) is
needed to sensitize target cells for lysis then this indicates a
low binding affinity of the epitope peptides to the Kk molecule.
We tested the ability of MV-derived epitope peptides (52–59
and 81–88) to sensitize RMA-S cells transfected with Kk for
lysis by CD8+ T cells in a competition assay (Feltkamp et al.,
1995). RMA-S cells are not able to transport peptide from the
cytosol to the endoplasmic reticulum due to expression of a
truncated TAP 2 molecule (Powis et al., 1992). Therefore,
addition of peptide from the outside is the only possibility for
stabilization and expression of MHC class I molecules on the
cell surface. As a positive control, we chose a Kk-restricted
peptide epitope from the NP of influenza A virus (aa 50–57)
(Brown et al., 1994). In competition experiments, the two
peptides 52–59 and 81–88, derived from MV N protein,
bound at least as well as the peptide derived from influenza A
virus NP (Fig. 6). In a different approach, we tested the off-rate
(i.e. the rate of dissociation of the epitope peptide from the
H-2 molecule) of MV-derived peptides. Twenty hours after
peptide-pulsing, target cells were compared to freshly pulsed
targets in a &"Cr-release assay in order to estimate the stability
of peptide binding to Kk. Target cells pulsed with MV peptides
52–59 and 81–88 were killed more efficiently than cells loaded
with influenza A virus peptide 50–57 (data not shown). This
indicates a low off-rate and good binding of MV epitope
peptides to the Kk molecule.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CTL in MV encephalitis
(a)
(b)
1
2
3
1
2
3
4
60 kDa
56 kDa
60 kDa
Fig. 3. MV N protein is present in MV-infected cells, but protein synthesis occurs after infection with vvN. (a) To test for the
presence of MV N protein in L929 cells after infection by Western blotting, cell lysates from uninfected cells (lane 1), cells
infected with vvN (lane 2) and MV (lane 3) were blotted onto nylon membranes and stained with a human serum containing
MV-specific antibodies. (b) To test for synthesis of MV N protein by immune precipitation, labelled cell lysates of uninfected
cells (lane 1), cells infected with vvNP (from influenza A virus, 56 kDa ; lane 2), vvN (from MV, 60 kDa ; lane 3) or MV (lane 4)
were precipitated with a human serum containing antibodies specific for MV and influenza A virus and separated by PAGE.
Discussion
Fig. 4. No difference in recognition of vvR expressing the full-length N
protein, the peptide epitope or the epitope bound to the HA-1 leader
peptide. L929 cells were infected with vvNP, MV, vvN (expressing the fulllength N protein), vv52–59 (expressing the 52–59 peptide) and
vvL­52–59 (expressing the 52–59 peptide with a leader peptide).
Recognition of the different viruses was determined with CTL specific for
the 52–59 peptide. The same set of experiments was performed with the
81–88 peptide and the same results were obtained. +, E : T ratio 3 : 1 ; 8,
E : T ratio 30 : 1.
Kk requires more peptide than Ld in a sensitization
assay
As the binding of Kk-restricted peptides seemed to be
comparable to that of the influenza A virus epitope (which is
well recognized after infection), we compared the ability of the
Kk- and Ld-restricted peptides to sensitize their respective
target cells for lysis. A 100-fold more Kk-restricted peptide was
required to sensitize L929 cells for lysis than Ld-restricted
peptide was required for P815 cells (Fig. 7 a). In order to ensure
that this effect was not due to differences in cell type, we
compared sensitization efficiencies on RMA-S cells transfected
with either Kk or Ld. The results were the same and could be
reproduced with different T cell lines (Fig. 7 b).
In most virus infections, CD8+ T cells play a dominant role
in clearing virus infection and protecting against disease
(Koszinowski et al., 1991 ; Kagi & Hengartner, 1996). For MVE,
it has been reported that, in the resistant BALB}c mouse,
depletion of CD4+ T cells leads to breakdown of resistance
(Finke & Liebert, 1994). In the light of recent findings, these
data have to be re-evaluated. Firstly, it has been shown that
without CD4+ T cell help, the clonal burst of CD8+ T cells
specific for poorly replicating viruses (like MV in the mouse) is
greatly diminished (Zimmermann et al., 1997) and secondly,
even primed CD8+ T cells are not able to clear virus from the
brain without CD4+ T cell help (Stohlman et al., 1998). In MVE,
CTL activity correlates with resistance and susceptibility of
different mouse strains (Niewiesk et al., 1993). Ld-restricted
CD8+ T cells (from resistant BALB}c mice) recognize both
target cells infected with MV and target cells infected with
vvN. Kk-restricted CTL (from susceptible C3H mice) recognize
the N protein after expression via a vvR but not after infection
of target cells with MV. Infection with adenovirus leads to
non-recognition of target cells co-infected with influenza A
virus (Yewdell et al., 1988). After co-infection of target cells
with influenza A virus and MV, recognition by influenza A
virus-specific CTL is not impaired. In addition, IFN-γ, which
helps to overcome inefficient antigen processing in cells
infected with murine cytomegalovirus (Hengel et al., 1994),
does not affect the recognition of MV-infected cells. These
data indicate that there is no general effect of MV gene
products on the antigen processing machinery of the cell. As
CD8+ T cells restricted by Ld lyse MV-infected cells [L-Ld,
P815 (Niewiesk et al., 1993) or C1R-Ld (C. Neumeister & S.
Niewiesk unpublished)], this phenomenon seems to be due to
either the Kk-molecule itself or to the processing of the Kk
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CFIH
C. Neumeister and S. Niewiesk
(b)
(a)
1
2
3
4
Fig. 5. The level of recognition correlates with the number of N protein-synthesizing cells. (a) Uninfected (1), persistently
infected (2) and NS20Y cells infected with MV (m.o.i. of 5) for 24 (3) or 48 h (4) were tested for recognition by peptide
81–88-specific CD8+ T cells in a 51Cr-release assay. Similar data were obtained for peptide 52–59-specific CD8+ T cells. E : T
ratio was 20 : 1. (b) C1R-Kk cells were infected with MV (m.o.i. of 0±1) for 24 or 48 h. The percentage of N protein-positive
cells was determined by flow cytometry using the N protein-specific antibody F227. The percentage of specific lysis was
measured in a 51Cr-release assay using CD8+ T cells specific for the 52–59 peptide and CD8+ T cells specific for the 81–88
peptide. E : T ratio was 40 : 1. E, N protein-expressing C1R-Kk cells ; *, 52–59 peptide-specific CD8+ T cells ; ^, 81–88
peptide-specific CD8+ T cells.
(b)
(a)
Fig. 6. Efficient binding of 52–59 and 81–88 peptides to Kk. A competition assay was performed using RMA-S-Kk cells. CTL
specific for the 81–88 (a) and 52–59 (b) peptides were used as effector cells (E : T ratio 5 : 1). Peptide 50–57 from the NP of
influenza A virus strain PR8 or the other MV-derived peptide were used as competitor peptides. (a) E, Peptide 50–57 from
influenza A virus ; ^, 52–59 peptide. (b) E, Peptide 50–57 from influenza A virus ; *, 81–88 peptide. lg, Log scale.
peptide epitopes. In contrast to herpesvirus (Hill et al., 1995 ;
Del Val et al., 1992 ; Wiertz et al., 1996 ; Ahn et al., 1996) and
adenovirus (Jeffries & Burgert, 1990) infection, our investigations have shown that the transport of peptide and binding
to the Kk molecule is not impaired and that the Kk molecule is
not retained within the cell after MV infection. Some viruses,
e.g. adenovirus 12, impair the proteasome activity and the
TAP-dependent transport of peptides (Rotem-Yehudar et al.,
1996). However, if the MV N protein is expressed by a vvR it
is recognized by Kk-restricted CTL. The recognition is the
same whether the full-length protein, an epitope or an epitope
with a leader peptide enabling transport into the endoplasmic
reticulum is expressed. This means that there is no specific
CFII
defect in antigen processing. The important difference between
MV and vvN infection of mouse cells is the poor replication of
MV. It is known that MV usually does not replicate in murine
cells (Horvat et al., 1996 ; Niewiesk et al., 1997). Also primary
mouse B cells, which express the human MV receptor CD46
via a transgene, replicate MV in vitro after mitogen activation
but not in vivo (Horvat et al., 1996). In agreement with these
data, we found that infection of L929 cells with MV leads to
uptake of virus (as detected by Western blot), but not to
replication (as shown by immune precipitation). In contrast,
vvN does replicate in L929 cells and is recognized by CD8+ T
cells. In the MV semi-permissive NS20Y mouse neuroblastoma
cell line, CTL recognition is partly restored. In human C1R cells
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CTL in MV encephalitis
(b)
(a)
Fig. 7. Kk requires 100-fold more peptide than Ld to sensitize target cells for lysis. Target cells were incubated with different
peptide concentrations and the percentage of specific lysis was measured in a 51Cr-release assay using peptide-specific CD8+
T cells. (a) P815 cells (naturally expressing the Ld molecule) were pulsed with the Ld-restricted 281–289 peptide (*), L929
cells (naturally expressing the Kk molecule) were pulsed with the Kk-restricted 52–59 (E) or the 81–88 (_) peptides.
(b) RMA-S cells, transfected with the Ld molecule, were pulsed with the 281–289 peptide (*), RMA-S cells, transfected with
the Kk molecule, were pulsed with the 52–59 (E) or the 81–88 (_) peptides. lg, Log scale.
transfected with Kk, MV replicates well and these cells are
readily lysed by CTL. It therefore seems that the amount of
newly synthesized protein (and therefore epitope peptide
produced), rather than the processing, limits CTL recognition.
This does not completely rule out variation in the processing
efficiency of the Kk and Ld epitopes as has been described for
other pathogens (Anto! n et al., 1997 ; Sijts et al., 1996). Our data
demonstrate that the binding affinity of the Kk-restricted
epitopes is at least as good as the affinity of the Kk-restricted
influenza A virus epitope. In addition, the amount of peptide
required to sensitize Ld-expressing cells correlates with binding
studies of peptide to Ld (Kageyama et al., 1995). However, in
comparison the amount of epitope peptide required to sensitize
cells expressing Ld for CTL recognition is 100-fold less than
the amount required to sensitize cells expressing Kk. It should
be noted that, although the number of MHC–peptide complexes as well as T cell receptor affinity may influence the
competition and sensitization assay, these systems seem to
correlate well with peptide binding (Feltkamp et al., 1995).
In mice, MV replicates in brain tissue after intracerebral (i.c.)
infection, whereas i.p. administration leads to abortive replication. Still, CD8+ T cells develop after i.c. as well as i.p.
infection and can be stimulated with MV-infected stimulator
cells in vitro (Niewiesk et al., 1993). MV replicates in brain
tissue of H-2k mice (Fennelly et al., 1995) and therefore CD8+
T cells should be able to recognize it and protect against
encephalitis. In a previous report, it has been demonstrated
that the lytic ability of CD8+ T cells in vitro correlates with
their protective capacity in vivo (Del Val et al., 1991). Similarly,
we find that Kk-restricted CD8+ T cells have poor lytic ability
and this might explain their lack of protection in vivo. Another
important point might be the kinetics of infection and
development of CD8+ T cells. C3H mice die after i.c. infection
with MV after 5–9 days. In the lymphocytic chorio-meningitis
virus system, it has been shown that CD8+ T cells against a
major epitope develop after 6 days with peak activity on day
8, whereas CD8+ T cells against a minor epitope were seen on
day 8 with peak activity on day 10 (Weidt et al., 1998).
Therefore, a delayed CD8+ T cell response might occur too
late to exert a protective effect.
In summary, we have shown that the difference in the CTL
response in C3H mice susceptible to MVE and BALB}c mice
resistant to MVE depends on the amount of newly synthesized
viral protein and the binding abilities of the H-2 molecules.
We are grateful for the generous gift of cells from Gu$ nther
Ha$ mmerling (Heidelberg, Germany), Hans-Georg Rammensee
(Tu$ bingen, Germany), Chantale Rabourdin-Combe (Lyon, France) and
Peter Cresswell (Yale, USA) and plasmids from Tim Elliott (Oxford, UK).
S. N. was supported by Stipendienprogramm Infektionsforschung
(Bundesministerium fu$ r Bildung, Wissenschaft, Forschung und Technologie). This work was in part supported by Deutsche Forschungsgemeinschaft, Bundesministerium fu$ r Bildung, Wissenschaft, Forschung
und Technologie, Pfleger-Stiftung and the World Health Organization.
References
Ahn, K., Angulo, A., Ghazal, P., Peterson, P. A., Yang, Y. & Fruh, K.
(1996). Human cytomegalovirus inhibits antigen presentation by a
sequential multistep process. Proceedings of the National Academy of
Sciences, USA 93, 10990–10995.
Anto! n, L. C., Yewdell, J. W. & Bennink, J. R. (1997). MHC class Iassociated peptides produced from endogenous gene products with
vastly different efficiencies. Journal of Immunology 158, 2535–2542.
Bankamp, B., Brinckmann, U. G., Reich, A., Niewiesk, S., ter Meulen, V.
& Liebert, U. G. (1991). Measles virus nucleocapsid protein protects rats
from encephalitis. Journal of Virology 65, 1695–1700.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CFIJ
C. Neumeister and S. Niewiesk
Beauverger, P., Buckland, R. & Wild, F. T. (1993). Measles virus
antigens induce both type-specific and canine distemper virus crossreactive cytotoxic T lymphocytes in mice : localization of a common Ldrestricted nucleoprotein epitope. Journal of General Virology 74,
2357–2363.
Beauverger, P., Chadwick, J., Buckland, R. & Wild, T. F. (1994).
Serotype-specific and canine distemper virus cross-reactive H-2Kkrestricted cytotoxic T lymphocyte epitopes in the measles virus
nucleoprotein. Virology 203, 172–177.
Brown, E. L., Wooters, J. L., Frerenz, C. R., O’Brien, C. M., Hewick,
R. M. & Herrmann, S. H. (1994). Characterization of peptide binding to
the murine MHC class I H-2Kk molecule. Journal of Immunology 153,
3079–3092.
Bruton, O. C. (1953). Agammaglobulinemia. Pediatrics 9, 722–728.
Cao, W., Myers-Powell, B. A. & Braciale, T. J. (1996). The weak CD8+
CTL response to an influenza hemagglutinin epitope reflects limited T cell
availability. Journal of Immunology 157, 505–511.
Kagi, D. & Hengartner, H. (1996). Different roles for cytotoxic T cells
in the control of infections with cytopathic versus noncytopathic viruses.
Current Opinions in Immunology 8, 472–477.
Katz, M. (1995). Clinical spectrum of measles. In Measles Virus, 1–12.
Edited by M. Billeter & V. ter Meulen. Berlin : Springer.
Koszinowski, U. H., Reddehase, M. J. & Jonjic, S. (1991). The role of
CD4 and CD8 T cells in viral infections. Current Opinions in Immunology
3, 471–475.
Liebert, U. G., Schneider-Schaulies, S., Baczko, K. & ter Meulen, V.
(1990). Antibody-induced restriction of viral gene expression in measles
encephalitis in rats. Journal of Virology 64, 706–713.
Long, E. O. & Jacobson, S. (1989). Pathways of viral antigen processing
for presentation by MHC class I molecules depends on its neighboring
residues in the protein. Cell 66, 1145–1153.
and presentation to CTL : defined by the mode of entry? Immunology
Today 10, 45–48.
Nahmias, A. J., Griffith, D., Salsbury, C. & Yoshida, K. (1967). Thymic
aplasia with lymphopenia, plasma cells, and normal immunoglobulins.
Journal of the American Medical Association 201, 729–734.
Nanan, R., Carstens, C. & Kreth, H. W. (1995). Demonstration of virusspecific CD8+ memory T cells in measles-seropositive individuals by in
vitro peptide stimulation. Clinical and Experimental Immunology 102,
40–45.
Del Val, M., Hengel, H., Ha$ cker, H., Hartlaub, U., Ruppert, T., Lucin, P.
& Koszinowski, U. (1992). Cytomegalovirus prevents antigen pres-
Niewiesk, S., Brinckmann, U., Bankamp, B., Sirak, S., Liebert, U. G. &
ter Meulen, V. (1993). Susceptibility to measles virus-induced en-
entation by blocking the transport of peptide-loaded major histocompatibility complex I molecules into the medial-Golgi. Journal of
Experimental Medicine 176, 729–738.
Elliott, T., Willis, A., Cerundolo, V. & Townsend, A. (1995). Processing
of major histocompatibility class I-restricted antigens in the endoplasmic
reticulum. Journal of Experimental Medicine 181, 1481–1491.
cephalitis in mice correlates with impaired antigen presentation to
cytotoxic T lymphocytes. Journal of Virology 67, 75–81.
Del Val, M., Schlicht, H.-J., Ruppert, T., Reddehase, M. J. &
Koszinowski, U. H. (1991). Efficient processing of an antigenic sequence
Feltkamp, M. C. W., Vierboom, M. P. M., Toes, R. E. M., Ossendorp, F.,
ter Schegget, J., Melief, C. J. M. & Kast, W. M. (1995). Competition
Niewiesk, S., Schneider-Schaulies, J., Ohnimus, H., Jassoy, C.,
Schneider-Schaulies, S., Diamond, L., Logan, J. S. & ter Meulen, V.
(1997). CD46 expression does not overcome the intracellular block of
measles virus replication in transgenic rats. Journal of Virology 71,
7969–7973.
Powis, S. J., Deverson, E. V., Coadwell, J. W., Ciruela, A., Huskisson,
N. S., Smith, H., Butcher, G. W. & Howard, J. C. (1992). Effect of
inhibition of cytotoxic T-lymphocyte (CTL) lysis, a more sensitive
method to identify candidate CTL epitopes than induction of antibodydetected MHC class I stabilization. Immunology Letters 47, 1–8.
polymorphism of an MHC-linked transporter on the peptides assembled
in a class I molecule. Nature 357, 211–215.
Fennelly, G. J., Flynn, J. A. L., ter Meulen, V., Liebert, U. G. & Bloom,
B. R. (1995). Recombinant Bacille Calmette Guerin priming against
Reich, A., Erlwein, O., Niewiesk, S., ter Meulen, V. & Liebert, U. G.
(1992). CD4+ T cells control measles virus infection of the central
measles. Journal of Infectious Diseases 172, 698–705.
Finke, D. & Liebert, U. G. (1994). CD4+ T cells are essential in
overcoming experimental murine measles encephalitis. Immunology 83,
184–189.
Good, R. A. & Zak, S. J. (1956). Disturbances in gamma globulin
synthesis as ‘ experiments of nature ’. Pediatrics 18, 109–149.
nervous system. Immunology 76, 185–191.
Rotem-Yehudar, R., Groettrup, M., Soza, A., Kloetzel, P. M. & Ehrlich,
R. (1996). LMP-associated proteolytic activities and TAP-dependent
evade host immunity. Nature 375, 411–415.
peptide transport for class I MHC molecules are suppressed in cell lines
transformed by the highly oncogenic adenovirus 12. Journal of Experimental Medicine 183, 499–514.
Sibille, C., Gould, K., Ha$ mmerling, G. & Townsend, A. (1992). A defect
in the presentation of intracellular viral antigens is restored by interferonγ in cell lines with impaired major histocompatibility complex class I
assembly. European Journal of Immunology 22, 433–440.
Sijts, A. J. A. M., Neisig, A., Neefjes, J. & Pamer, E. G. (1996). Two
Listeria monocytogenes CTL epitopes are processed from the same antigen
with different efficiencies. Journal of Immunology 156, 685–692.
Horvat, B., Rivailler, P., Varior-Krishnan, G., Cardoso, A., Gerlier, D. &
Rabourdin-Combe, C. (1996). Transgenic mice expressing human
Stohlman, S. A., Bergmann, C. C., Lin, M. T., Cua, M. T. & Hinton, D. R.
(1998). CTL effector function within the central nervous system requires
Hengel, H., Lucin, P., Jonjic, S., Ruppert, T. & Koszinowski, U. H.
(1994). Restoration of cytomegalovirus antigen presentation by gamma
interferon combats viral escape. Journal of Virology 68, 289–297.
Hill, A., Jugovic, P., York, I., Russ, G., Bennink, J., Yewdell, J., Ploegh,
H. & Johnson, D. (1995). Herpes simplex virus turns off the TAP to
measles virus (MV) receptor CD46 provide cells exhibiting different
permissivities to MV infection. Journal of Virology 70, 6673–6681.
Jeffries, W. A. & Burgert, H.-G. (1990). E3}19K from adenovirus 2 is an
immunosubversive protein that binds to a structural motif regulating the
intracellular transport of major histocompatibility complex class I
proteins. Journal of Experimental Medicine 172, 1653–1664.
CD4+ T cells. Journal of Immunology 160, 2896–2904.
Kageyama, S., Tsomides, T. J., Sykulev, Y. & Eisen, H. N. (1995).
CD8 T cells after infection with measles virus suggests a role for CD8 class
I MHC-restricted cytotoxic T lymphocytes (CTL) in recovery from
measles. Journal of Immunology 144, 2394–2399.
Variations in the number of peptide–MHC class I complexes required to
activate cytotoxic T cell responses. Journal of Immunology 154, 567–576.
CFJA
Uytdehaag, F. G., van Binnendijk, R. S., Kenter, M. J. & Osterhaus,
A. D. (1994). Cytotoxic T lymphocyte responses against measles virus.
Current Topics in Microbiology and Immunology 189, 151–167.
van Binnendijk, R. S., Poelen, M. C. M., Kuijpers, K. C., Osterhaus,
A. D. M. E. & Uytdehaag, F. G. C. M. (1990). The predominance of
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CTL in MV encephalitis
Weidt, G., Utermo$ hlen, O., Heukeshoven, J., Lehmann-Grube, F. &
Deppert, W. (1998). Relationship among immunodominance of single
CD8+ T cell epitopes, virus load, and kinetics of primary antiviral CTL
response. Journal of Immunology 160, 2923–2931.
Wiertz, E. J. H. J., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J. &
Ploegh, H. L. (1996). The human cytomegalovirus US11 gene product
CTL recognition of adenovirus-transformed cells infected with influenza
virus : lysis by anti-influenza CTL parallels adenovirus-12-induced
suppression of class I MHC molecules. Virology 162, 236–238.
Zimmermann, C., Seiler, P., Lane, P. & Zinkernagel, R. M. (1997).
Antiviral immune responses in CTLA4 transgenic mice. Journal of
Virology 71, 1802–1807.
dislocates MHC class I heavy chains from the endoplasmic reticulum to
the cytosol. Cell 84, 769–779.
Yewdell, J. W., Bennink, J. R., Eager, K. B. & Riccardi, R. P. (1988).
Received 8 May 1998 ; Accepted 6 July 1998
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 28 Apr 2017 21:59:54
CFJB