Immunity, Vol. 8, 167–175, February, 1998, Copyright 1998 by Cell Press
Massive Expansion of Antigen-Specific
CD81 T Cells during an Acute Virus Infection
Eric A. Butz and Michael J. Bevan*
Howard Hughes Medical Institute
Department of Immunology
University of Washington
Seattle, Washington 98195
During LCMV infection, CD81 T cells expand greatly.
Bystander activation has been thought to play a role
because few cells score as LCMV specific in limiting
dilution analysis. In contrast, we find that at least a
quarter of the CD8 1 cells secrete IFNg specifically in
response to LCMV peptides at the peak of the response.
Moreover, by analyzing the expansion of adoptively
transferred LCMV-specific, TCR-transgenic CD81 T
cells in congenic hosts, we have determined that most
of the CD8 1 cell expansion is virus specific. Analysis
of the effect of the monospecific TCR-transgenic T
cells on the host response to three LCMV epitopes
suggests that CTL precursors compete for sites on
the APC in an epitope-specific fashion and that this
competition determines the specificity of the response.
Specific immune responses evolve following exposure
to foreign antigens. According to the clonal selection
theory, an antigen will stimulate cells that bear receptors
specific for the antigen, resulting in their proliferation
and functional activation. Virus infection can result in
an intense activation of the immune system, and especially of CD81 T cells.
Although CD81 T cells are normally present in the
spleen and lymph nodes of mice in lower numbers than
CD41 T cells, during many virus infections they are disproportionately expanded and may outnumber CD41 T
cells (Buchmeier et al., 1980; Rubin et al., 1981; Cauda
et al., 1987; Tripp et al., 1995; Callan et al., 1996). During
lymphocytic choriomeningitis virus (LCMV) infection of
B6 mice, CD81 T cells increase about 5-fold (Razvi et
al., 1995), and in humans there is a large increase in
circulating CD81 T cells during cytomegalovirus, Epstein-Barr virus, and varicella-zoster virus infections
(Rubin et al., 1981; Cauda et al., 1987; Callan et al.,
1996). As the total pool of CD81 cells expands during
the virus infection, the number of antigen-specific cytotoxic T lymphocytes (CTL) also rises (Lau et al., 1994;
Selin et al., 1994; Razvi et al., 1995; Tripp et al., 1995)
as judged by limiting dilution analysis (LDA). However,
LDA indicates that only a small number of the CD81 T
cells (1 in 4000 to 1 in 50, depending on the virus) are
actually specific for the infecting virus (Lau et al., 1994;
Selin et al., 1994; Razvi et al., 1995; Tripp et al., 1995).
The specificities of the remainder of the CD81 T cells
* To whom correspondence should be addressed (e-mail:[email protected]
have not been determined, although it has been shown
that during LCMV infection there is an increase in the
number of alloreactive and antigen cross-reactive CTL,
not all of which recognize LCMV antigens (Yang and
Welsh, 1986; Nahill and Welsh, 1993). In addition to their
expanded numbers, most CD81 T cells during the acute
immune response to LCMV show signs of activation.
They are enlarged; have elevated surface expression
of CD11a, CD11b, CD44, CD49d, and the interleukin-2
receptor; and show reduced expression of CD62L
(Lynch et al., 1989; McFarland et al., 1992; Andersson
et al., 1995).
Because of the low frequency of virus-specific CTL,
the large numbers of apparently activated cells, and the
appearance of alloreactive and cross-reactive CTL, it
has been assumed that the bulk of the CD81 cells have
been activated in a bystander fashion involving cytokines (Yang and Welsh, 1986; Tough and Sprent, 1996).
In support of this idea, it has been demonstrated that
cytokines can drive antigen-independent activation of
naive and memory phenotype T cells in vitro (Unutmaz
et al., 1994). There is also evidence that CD81 T cells
of memory phenotype may be more prone to bystander
activation than naive cells (Tough and Sprent, 1994;
Tripp et al., 1995), which has led to the suggestion that
this is a mechanism for the maintenance of CD81 T cell
memory (Beverley, 1996). It also has been suggested
that the cytokines expressed during the antiviral immune
response may act to amplify other coincidentally active
T cell responses that are not virus specific (Strang and
Rickinson, 1987). Similarly, it has been proposed that
bystander activation may play a role in the initiation and
maintenance of autoimmune diseases by expanding the
numbers of autoreactive T cells and breaking anergic
tolerance (Rott et al., 1995; Mueller and Jenkins, 1997).
In this study we have used LCMV infection of B6 mice
as a model for studying CD81 T cell proliferation during
the immune response to virus infection. By analyzing
the frequency of cells secreting interferon-g (IFNg) in
response to LCMV antigens, we have correlated CTL
activity with antigen-specific cell frequency and have
found that the number of LCMV-specific CD81 T cells
was much higher than initially expected (at least 24%
of CD81 cells). By adoptively transferring T cell receptor
(TCR)–transgenic CD81 T cells specific for an epitope of
LCMV into host mice and extrapolating their proliferation
during LCMV infection to the proliferation of host cells,
we have determined that at least half of the CD81 T cells
present are specific for LCMV. Although a significant
expansion of non–virus-specific CD81 T cells may still
occur during LCMV infection, the immune response is
much more narrowly focused on the virus than has previously been believed. These results are more in keeping
with the predictions of the clonal selection hypothesis.
CD8 1 T Cells Expand Dramatically during
A group of three C57Bl/6 mice was intravenously infected with 105 plaque-forming units (pfu) of LCMV. On
Figure 1. LCMV Infection of C57Bl/6 Mice Induces Massive Expansion of CD81 T Cells
Representative CD4 versus CD8 flow cytometry profiles of splenocytes from uninfected (A) or day 8 LCMV-infected B6 mice (B). The
numbers beside the gating boxes indicates the percentage of livegated cells. d8, day 8.
day 8 of infection, when the immune response to LCMV
peaks (Buchmeier et al., 1980; our unpublished data),
these mice and a group of three uninfected control mice
were sacrificed and splenocytes were prepared. For the
uninfected control mice there were 7.5 6 0.5 3 107 cells
per spleen (mean 6 SEM); for the infected mice there
were 1.5 6 0.2 3 108 cells per spleen. These cells were
analyzed for CD41 and CD81 cell content by two-color
flow cytometry (Figure 1). The uninfected spleen cells
contained 16% 6 1% CD81 cells and 22% 6 3% CD41
cells; infected spleens contained 50% 6 4% CD81 cells
and 5% 6 2% CD41 cells. In absolute cell numbers the
uninfected control mice contained an average of 1.2 6
0.8 3 107 CD81 cells and 1.7 6 0.3 3 107 CD41 cells.
Infected spleens contained 8 6 1 3 107 CD81 cells and
8 6 1 3 106 CD41 cells—an approximately 6-fold increase in the number of CD81 cells and a 2-fold decrease
in the number of CD41 cells. A similar increase in total
cell number and CD81 cell content was also observed
in cells pooled from the mesenteric, brachial, axilliary,
and inguinal lymph nodes of each mouse (data not
Antigen-Specific Lytic Activity and Frequency
of IFNg Secretion by Primary Antiviral CD8 1
T Cells Are Proportional to One Another
The B6 CTL response to LCMV is dominated by three
D b-restricted epitopes: gp33, gp276, and np396 (Gairin
et al., 1995). The data presented in Figure 2 show that
the responses to gp33 and np396 are stronger than the
response to gp276. Pooled spleen cells from pairs of
mice were prepared from uninfected and day 8 LCMVinfected mice for use as effector cells in an 8 hr 51Crrelease assay against peptide-coated EL4 target cells
(Figure 2A) and in an 18 hr IFNg ELISPOT assay against
peptide-coated EL4 cells (Figure 2B). Cells from the uninfected mice displayed no lytic activity, and the data
were omitted from Figure 2A for the sake of clarity. Cells
from the infected mice were highly lytic for gp33- and
np396-coated EL4 cells and less active against gp276coated targets. Lytic units calculated at 30% lysis were
5-fold lower for gp276 than for gp33 and np396.
These data correlate with the data from the ELISPOT
Figure 2. Antigen-Specific Cytolytic Activity Is Proportional to the
Frequency of IFNg-Secreting Cells
(A) Erythrocyte-depleted spleen cells were prepared from a pair of
day 8 LCMV-infected B6 mice and were used to determine primary
ex vivo CTL activity in an 8 hr 51Cr-release assay against EL4 target
cells uncoated or coated with the indicated LCMV peptides.
(B) The same pool of effector cells was used in an ELISPOT assay
to determine the frequency of cells secreting IFNg in response to
LCMV epitope peptides. The number of CD81 cells was determined
by FACS analysis. Error bars indicate the standard deviation of
triplicate samples. Three additional experiments yielded similar results. d8, day 8.
(C) The titration of spots per well versus number of LCMV day 8
splenocytes per well is shown for ELISPOT data from a separate
experiment. Splenocytes were stimulated in triplicate samples with
either control EL4 cells (open circles) or with gp33-coated EL4 cells
(filled squares). Stimulation with gp276- or np396-coated EL4 cells
resulted in similarly linear responses (data not shown).
assay of the frequency of cells secreting IFNg in response to each of the three LCMV epitopes: 10.6% 6
1.7% of CD81 cells secreted IFNg in response to gp33,
10.0% 6 1.4% in response to np396, and 2.9% 6 1.2%
in response to gp276. The ELISPOT response to EL4
alone was 0.5% 6 0.3% of CD81 cells, and fewer than
0.2% 6 0.04% of the CD81 cells from uninfected B6
mice secreted IFNg in this assay with any peptide. When
splenocytes from day 8 LCMV-infected mice were stimulated overnight with LCMV peptide-coated EL4 cells in
the presence of Brefeldin A and then analyzed by fluorescence-activated cell sorting (FACS), greater than
90% of the cells staining positive for IFNg were also
CD81 (data not shown). For this reason, and because
the IFNg secretion is revealed only in the presence of
the major histocompatibility (MHC) class I–binding peptides, we have presented these ELISPOT data as the
percentage of CD81 cells secreting IFNg. Also, because
a plot of the number of spots per well versus the number
of splenocytes per well is linear (Figure 2C), the assay
measures only antigen-specific cells and not cytokineinduced bystanders. From data of this type we conclude
that the relative frequency of CD81 T cells specific for
each LCMV epitope can be determined by titration of
their ability to lyse peptide coated target cells.
The aggregate number of cells secreting IFNg in response to the known LCMV epitopes (gp33 1 gp276 1
np396) is 24% 6 4% of CD81 T cells in the experiment
shown in Figure 2B. To estimate the efficiency of the
ELISPOT assay, we performed control experiments with
cloned anti-LCMV CTL. We were able to detect 24%–
100% of the CTL in ELISPOT assays, depending on the
particular clone and the time during its restimulation
cycle at which it was tested (data not shown). This suggests that, although 24% of splenic CD81 T cells from
Expansion of Antiviral CD81 Cells during Infection
VVflu-np–infected mice, 1.3% of the CD81 cells were
donor cells (Figure 3C), and in LCMV-infected mice,
0.8% of the CD81 cells were OT-1 cells (Figure 3D). The
lower representation of the OT-1 cells in the LCMVinfected chimeras probably reflects a greater dilution of
these cells by the larger overall expansion of CD81 cells
that occurs during LCMV infection than occurs during
VV infection. In the VVova-infected mice, 27% of the
CD81 cells were Thy1.21 (Figure 3E), confirming that the
donor cells were capable of proliferating in the host
mice. Not only did the OT-1 cells fail to proliferate significantly during VVflu-np or LCMV infection; there also
was no ex vivo anti-OVA257–264 cytolytic activity following
VVflu-np or LCMV infections, although such activity was
readily apparent following VVova infection (data not
Figure 3. Adoptively Transferred TCR-Transgenic CD81 T Cells Do
Not Expand Nonspecifically during VV or LCMV Infection
Flow cytometry profiles of pooled spleen cells from pairs of B6.PL
control mice (A) and B6.PL host mice that received 107 purified OT-1
CD81 T cells and were injected the next day with PBS (B), VVflu-np
(C), LCMV (D), or VVova (E). Splenocytes were analyzed on day 6
(A–C and E) or day 8 (D) of infection. The numbers beside the gating
boxes indicate the number of donor CD81 cells as a percentage
of total CD81 cells. A repetition of the experiment yielded similar
LCMV-infected mice were detected by ELISPOT, the
true number of anti-LCMV CTL could be substantially
TCR-Transgenic CD8 1 T Cells of Irrelevant
Specificity Do Not Expand Significantly
during Virus Infections.
OT-1 mice are transgenic for a TCR that recognizes
ovalbumin 257–264 (OVA257-264) in the context of H-2Kb
(Hogquist et al., 1994). To see if we could emulate the
bystander proliferation of non–virus-specific CD81 T
cells during infection, we transferred 107 purified OT-1
CD81 T cells into age- and sex-matched Thy1 congenic
B6.PL mice (PL/OT-1 chimeras). Mice were infected the
next day with LCMV, VVflu-np (a recombinant vaccinia
virus [VV] expressing the H-2Kd-restricted influenza nucleoprotein epitope, NP147–155 [Yewdell et al., 1985]),
or VVova (which expresses the full-length ovalbumin
protein [Bacik et al., 1994]). On day 6 of VV infection
and day 8 of LCMV infection, spleen cells were analyzed
for donor (CD81Thy1.21) cell content (Figure 3). As expected, there were no CD81Thy1.21 cells in uninfected
control B6.PL mice (Figure 3A), whereas 7 days after
transfer of OT-1 cells, 3% of the CD81 cells in uninfected
PL/OT-1 mice were of donor origin (Figure 3B). In the
TCR-Transgenic gp33-Specific CD81 T Cells
Expand Dramatically during LCMV Infection
As an independent method of tracking the number of
LCMV-specific CD81 cells at the peak of the immune
response to LCMV, we adoptively transferred unstimulated, LCMV-specific TCR-transgenic T cells into host
mice and infected them with LCMV. Similar methods
have been used to follow the in vivo behavior of both
CD41 and CD81 T cells (Zimmerman et al., 1996; Kedl
and Mescher, 1997; Pape et al., 1997). P14 mice are
transgenic for a TCR that recognizes LCMV gp33/Db
(Pircher et al., 1990). We transferred 105 P14 spleen cells
containing 30% CD81 T cells into Thy1 congenic B6.PL
host mice. These PL/P14 chimeras and B6.PL control
mice (three per group) were infected with LCMV. Spleen
cells were prepared and analyzed by flow cytometry for
T cell types and donor cell content on day 8 of infection.
The expansion of CD81 cells for both the B6.PL and PL/
P14 mice (Figure 4) was identical to that observed in B6
mice (Figure 1). In control, uninfected B6.PL mice that
received 105 P14 spleen cells, 0.2% of CD81 cells were
of donor origin 9 days after transfer (equivalent to day
8 of LCMV infection; data not shown). However, in the
LCMV-infected PL/P14 chimeras the donor cells were
greatly expanded, reaching 23% of the CD81 cell pool
in these mice (Figure 4D). Although the total CD81 T cell
expansion appeared similar in the two sets of infected
mice, we wondered whether the artificially high precursor frequency of gp33-specific CTL had substantially
altered the overall response of the chimeras to the virus.
Adoptive Transfer of P14 Splenocytes Does
Not Significantly Alter the Overall
CTL Response to LCMV
To assess the effect of donor P14 cells on the host
response to LCMV, we studied B6.PL mice and PL/P14
chimeras that had received 103, 105, or 106 P14 spleen
cells. One day after cell transfer the mice were infected
with LCMV. Spleen cells were prepared on day 8 of
infection, used as effector cells in a primary CTL assay
against peptide-pulsed EL4 target cells (Figure 5), and
analyzed by FACS. Uninfected chimeras that received
105 P14 cells did not respond to LCMV peptides. The
CTL responses of the infected PL/P14 chimeras to each
of the three LCMV epitopes were essentially identical
Figure 4. GP33-Specific TCR-Transgenic CD81 T Cells Expand Dramatically during LCMV Infection
Control B6.PL mice (A and C) or B6.PL mice that had received 10 5
P14 spleen cells (z3 3 104 transgenic CD81 cells) 1 day earlier (B
and D) were infected with LCMV. After 8 days spleens were removed
and assayed for CD4 versus CD8 staining (A and B) and donor
(CD81Thy1.21) CTL content (C and D). The numbers beside the
gating boxes in (A) and (B) indicate the percentage of live gated
cells; in (C) and (D) the numbers indicate the number of Thy1.21
donor cells as a percentage of CD81 cells. Results are from representative individual mice from groups of three. In uninfected chimeric mice the donor cells were less than 0.5% of CD81 cells (data
not shown). d8, day 8.
to those of the infected B6.PL mice, although following
transfer of 106 P14 cells the response to gp276 was
reduced (Figure 5C). In this experiment, the donor CD81
cells expanded to 1.5%, 26%, and 28% of CD81 cells
in the chimeras that received 103, 105, and 106 P14 cells,
As another way of looking at the effect of transferring
gp33-specific T cells into naive hosts, we compared
the responses to the three epitopes of LCMV in IFNg
ELISPOT assays. In individual LCMV-infected B6 and
B6.PL mice, the ratio of np396-stimulated IFNg-secreting cells to gp33-stimulated IFNg-secreting cells varied
from 0.7 to 1.7, and the ratio of gp276- to gp33-stimulated IFNg-secreting cells from 0.1 to 0.3 (data not
shown). In an additional experiment, groups of three PL/
P14 mice received 102, 103, 104 , 105, or 106 P14 spleen
cells and were infected with LCMV the next day. The
np396 to gp33 IFNg-secreting ratios were 0.76 6 0.02,
1.0 6 0.5, 1.5 6 0.4, 1.3 6 0.3, and 1.6 6 0.5, respectively.
The gp276 to gp33 IFNg ratios were 0.09 6 0.03, 0.13
6 0.04, 0.16 6 0.05, 0.11 6 0.04, and 0.09 6 0.04. Therefore, addition of 30–3 3 105 unstimulated gp33-specific
CD81 T cells does not alter the choice or hierarchy of
epitopes targeted in the anti-LCMV CTL response.
The FACS data from the PL/P14 mice that received
105 P14 cells indicated that at least 23%–26% of the
CD81 cells were gp33 specific. By comparing the lysis
of np396- and gp276-coated targets to gp33-specific
lysis, we calculated that another 23%–26% of the CD81
cells were np396 specific and that 5% of the CD81 cells
Figure 5. Adoptive Transfer of a Limited Number of gp33-Specific
CD81 T Cells Does Not Significantly Alter the Overall Response to
gp33, gp276, or np396 during the Immune Response to LCMV
Primary CTL activity of splenocytes from uninfected B6.PL mice
that received 10 5 P14 cells (open circles) and day 8 LCMV-infected
mice (filled symbols) that received no P14 cells (squares), 10 3 P14
cells (diamonds), 10 5 P14 cells (triangles), or 106 P14 cells (crosses)
1 day prior to infection. Lysis was determined on day 8 of infection
by 51Cr-release assay using EL4 targets without additional peptide
(A) and EL4 pulsed with gp33 (B), gp276 (C), or np396 (D) peptides.
Data are presented as the average specific lysis of targets by effectors from three individual mice per group, with standard deviation
indicated by error bars (some of which fall within the symbols). Two
repetitions of the experiment yielded similar results.
were gp276 specific. Therefore, a total of 54% of the
CD81 cells were LCMV specific on day 8 of infection.
Donor P14 and Host T Cells Are Functionally
Active in PL/P14 Chimeric Mice
To exclude the possibility that the TCR-transgenic donor
cells were proliferating but not functional within the infected chimeras, control B6, B6.PL, and PL/P14 mice
that had received 104 P14 cells 1 day earlier were infected with LCMV, and, 8 days later, spleen cells were
prepared and treated with rabbit complement alone,
anti-Thy1.1 and complement, or anti-Thy1.2 and complement and used as effectors in a 51Cr-release assay
(Figure 6). By flow cytometry, 13% of the CD81 cells
in the infected chimeras were of donor origin prior to
antibody treatment, and donor cells were reduced to
fewer than 0.5% of the CD81 cells by treatment with
anti-Thy1.2 and complement. As expected, the cytolytic
activity of B6 splenocytes (Figure 6A) was unaffected
by anti-Thy1.1 treatment (Figure 6D) but was ablated
by treatment with anti-Thy1.2 (Figure 6G). Conversely,
B6.PL activity (Figure 6B) was abolished by anti-Thy1.1
treatment (Figure 6E) and unaffected by anti-Thy1.2
treatment (Figure 6H). The overall pattern of CTL activity
of the PL/P14 chimeras (Figure 6C) was similar to the
activity of the B6 and B6.PL mice (Figures 6A and 6B).
Following treatment with anti-Thy1.1 and complement,
all activity against gp276 and np396 was lost, while
activity against gp33, the target of the donor cells, was
Expansion of Antiviral CD81 Cells during Infection
Figure 6. Both Adoptively Transferred P14
and Host CD81 T Cells Are Functionally Active in LCMV-Infected PL/P14 Chimeras
A group of three B6.PL mice received 104 P14
spleen cells. The next day the PL/P14 chimeras (C, F, and I) and groups of three B6 (A, D,
and G) and B6.PL (B, E, and H) mice were
infected with LCMV. On day 8 of infection,
splenocytes pooled for each group were prepared and treated with rabbit complement
alone (A–C), with anti-Thy1.1 (anti-host) plus
complement (D–F), or with anti-Thy1.2 (antidonor) plus complement (G–I), and the ability
of the surviving cells to lyse uncoated EL4
targets and EL4 cells coated with the indicated LCMV peptides was determined in a
reduced approximately 50% (Figure 6F). CTL activity of
the splenocytes from the chimeras against gp276 and
np396 was unaffected by treatment with anti-Thy1.2 and
complement, while activity against gp33 was reduced
approximately 50%. It thus appears that host CTL develop normally against gp276 and np396 in the chimeras
and that the CTL activity against gp33 is a 50:50 composite of host-derived and donor P14-derived CTL.
Since half of the CTL activity against gp33 is contained
in the donor cells and they make up 13% of the CD81
cells, we infer that 26% of the CD81 cells are gp33
specific. Therefore, by correlating effector cell numbers
with the peptide-specific CTL activity, we propose that
another 26% of the CD81 cells are np396 specific and
5% of the CD81 cells are gp276-specific, for a total of
57% of the CD81 cells.
The Extent of Antigen-Specific versus
Following Armstrong LCMV infection of B6 mice, there
was a large increase in the number of splenocytes and
in the representation of CD81 cells within the population
(Figures 1 and 4). It has been determined by LDA that
0.5%–2% of these CD81 cells are specific for LCMV
(Lau et al., 1994; Razvi et al., 1995) and, by inference,
that the bulk of the proliferating cells are bystanders
responding to cytokines produced during the immune
response (Yang et al., 1989; Tough et al., 1996).
We have found that the primary cytotoxic activity directed against LCMV epitopes and the frequency of
CD81 cells that secrete IFNg in response to these epitopes correlate well with each other (Figure 2). Although
we cannot show strictly that the same cells have both
activities, they both are clearly antigen-specific responses of the CD81 T cells to viral epitopes. At the
peak of the immune response to LCMV, when CD81 T
cells constitute half of the spleen (Figures 1 and 4), we
found in ELISPOT assays that a total of 24% of the CD81
T cells secrete IFNg in response to the three principle
H-2 b–restricted epitopes of LCMV: gp33, gp276, and
np396 (Figure 2). Therefore, since the ELISPOT assay
may not be 100% efficient, at least a quarter of the CD81
T cells present are responding specifically to the virus.
In light of the earlier published reports indicating much
lower CTL precursor frequencies, we became interested
in the nature of bystander activation and its relation to
virus specific CTL proliferation.
To see whether we could induce bystander proliferation of CD81 T cells of known but irrelevant specificity,
we adoptively transferred TCR-transgenic OT-1 cells
into Thy1 congenic host mice, which we then infected
with vaccina virus or LCMV. The OT-1 CD81 T cells are
specific for Kb/OVA 257–264 and do not have measurable
cytotoxic, cytokine, or proliferative responses to any of
the three LCMV epitopes in vitro, nor do they respond
to virus-infected cells (data not shown). We found that
the representation of the donor cells was reduced in
adoptively transferred chimeras infected with LCMV or
VVflu-np, a recombinant virus that expresses an irrelevant influenza epitope, and was greatly increased after
infection with VVova, which expresses ovalbumin (Figure 3). In agreement with our results, it has also been
reported that H-Y–specific CD81 T cells do not become
activated when adoptively transferred into hosts responding to LCMV (Zarozinski and Welsh, 1997) and
that vaccina virus infection of mice transgenic for an
LCMV gp33-specific TCR does not lead to substantial
activation of the CD81 cells bearing the transgenic receptor (Ehl et al., 1997).
Since the number of nonspecific CD81 T cells did not
appear to be significantly expanded in mice responding
to virus, we next sought to determine the degree to
which virus-specific CTL would proliferate in chimeric
mice that had received various numbers of gp33-specific P14 cells (PL/P14 chimeras). When chimeras that
received 105 P14 spleen cells (z3 3 104 gp33-specific
CD81 cells) were infected with LCMV, the number of
donor cells was greatly increased by day 8 of infection,
to about 24.5% of total CD81 cells. This occurred without significant alteration to the total expansion of CD81
cells or the overall CD8:CD4 ratio from that observed
in LCMV-infected B6 or B6.PL mice (Figures 1 and 4).
Moreover, the proliferation of the donor-derived gp33specific cells did not alter the targeting of the other two
epitopes in CTL assays (Figure 5). Since the cytolytic
activity is proportional to the frequency of the antigenspecific cells (Figure 2), we can deduce that the number
of host-derived np396-specific cells must be similar to
the number of gp33-specific cells and that there must
be, in addition, approximately one fifth as many hostderived gp276-specific cells. Therefore, the total number of CD81 T cells in these mice responding to the
three LCMV epitopes on day 8 of infection must be at
least 54% of the total CD81 cells. It also has been reported that half of the CD81 T cells express cytoplasmic
granules containing serine proteinase-1 (MTSP-1) at the
peak of LCMV infection (Kramer et al., 1989).
In a group of PL/P14 chimeric mice that initially received 104 P14 spleen cells (approximately 3000 transgenic CD81 T cells), the donor cells increased to 13%
of the CD81 cells by day 8 of infection. Since the total
anti-gp33 activity that develops during the immune response is the same regardless of the number of donor
cells transferred into the host (Figure 5), the total number
of anti-gp33 CTL (host 1 donor) must also be the same.
Following antibody plus complement depletion of host
or donor CTL from the day 8 spleen cells of PL/P14
chimeras that received 104 P14 (Figure 6), a comparison
of the reduction in gp33-specific cytolytic activity shows
that about 50% of the CTL activity was derived from
Thy1.11 host cells and 50% of the activity was derived
from Thy1.21 donor cells. The total number of gp33specific CTL, therefore, was 2 3 13%, or 26%. This
implies that another 26% of the cells must be np396
specific and that 5% of the CD81 cells are gp276 specific. Thus, 57% of the CD81 cells are LCMV specific,
a figure in close agreement with the results obtained
after transfer of 105 P14 cells (Figure 5).
Although we estimate on the basis of our adoptive
transfer studies that about half of the CD81 T cells on
day 8 of infection are specific for LCMV, only a quarter
of CD81 cells were detected as secreting IFNg in response to the three LCMV peptides in our ELISPOT
assays. It is possible that only half of the anti-LCMV
CD81 cells are able to secrete IFNg in response to viral
antigens or that the ELISPOT assay is only 50% efficient
at detecting IFNg-secreting cells.
If the non–LCMV-specific CD81 T cells of the naive
repertoire remain largely within the spleen during LCMV
infection, then, together with the half of the CD81 cells
that are LCMV specific, they constitute 63% of the CD81
cells within the spleen on day 8 of infection. This means
that the upper limit of bystander-expanded CD81 cells
is 37%, or about 2.5 times the original number of splenic
CD81 cells. It has been reported that the proliferation
of low-affinity CD81 T cells can be stimulated by much
lower peptide densities than are required for functional
responses such as in vivo protection, cytotoxicity, or
IFNg secretion (Speiser et al., 1992; Kageyama et al.,
1995; Valitutti et al., 1996). Thus, it may be that some
of the CD81 cells in the day 8 infected mice that we
cannot account for are, in fact, low-affinity anti-LCMV
CD81 T cells rather than cells expanding as a result of
bystander effects. This would further reduce the significance of bystander-induced proliferation during LCMV
Indirect evidence for the expansion of similarly large
numbers of antiviral CD81 T cells can be found in the
expansion of particular TCR Vb–bearing CD81 populations during certain infections (Cose et al., 1997). In
some individuals responding to Epstein-Barr virus infection (Callan et al., 1996) or to human immunodeficiency
virus infection (Pantaleo et al., 1994), a transient, oligoclonal expansion of CD81 T cells expressing particular TCR Vb chains to as much as 40% of CD81 T cells
has been observed. Because these expansions parallel
the rise and fall of circulating virus in the blood, because they are oligoclonal, and because different TCR
Vb are expanded in different individuals, the expansions
do not appear to be superantigen driven. In one case,
virus-specific CTL activity could also be found in the
expanded cell population (Pantaleo et al., 1994), suggesting that the cells were antigen specific. The significance of these expansions is unclear because they are
seen in only some individuals, but it may be that in the
usual response to infection there is an equally large
expansion of a polyclonal repertoire of antiviral CD81 T
cells. Our data support this model and suggest that large
expansions of antiviral CD81 T cells may be a common
feature of antiviral immune responses.
An Estimate of Precursor Frequency
The precursor frequency of CTL prior to antigen exposure has been determined for some antigens by LDA;
for LCMV it has been estimated to be 1 in 560,000 (Selin
et al., 1994). Because the gp33-specific CTL response
of PL/P14 chimeras that receive 104 P14 spleen cells
(z3000 CD81 cells) is evenly divided between host- and
donor-derived cells, we believe that there are approximately 3000 gp33-specific CTL precursors in naive
B6.PL mouse. Since we estimate that there are approximately 3 3 107 CD81 T cells in a naive mouse, the precursor frequency of gp33-specific CTL is about 1024. We
presume that the frequency of np396- and gp276-specific precursors is roughly similar. By day 8 of LCMV
infection we estimate that the total number of CD81 T
cells rises to about 1.5 3 108 (E. A. B., unpublished
data); 55%, or 8.3 3 107, are LCMV specific. This would
necessitate about 15 divisions over the 8 days of infection, or 1 division every 13 hr. While rapid, this rate of
proliferation is well below the 7 hr dividing time that has
been observed for B cells in germinal centers (Liu et al.,
1991). Since some of the P14 donor cells are probably
lost during the adoptive transfer, this is likely to be an
Expansion of Antiviral CD81 Cells during Infection
Figure 7. Competition for MHC Class I–Peptide Complexes during CTL Priming
(Left) A virus-infected APC displays a large number of MHC class I–peptide complexes, indicated by the different numbers on the surface of
(Center) When CTL precursors specific for epitopes 1 and 2 (T1 and T2 ) interact with the APC, they aggregate their respective target MHC–peptide
complexes at the zones of contact.
(Right) When enough T1 interact with the APC, they reduce the available MHC–peptide 1 complexes to the point where other T1 cells cannot
be primed, although cells of other specificities (T 2) can still interact productively with the APC.
overestimate of the CTL precursor frequency. There is
also a delay following infection of the mice as the viral
proteins are generated and subsequently processed
and presented by antigen-presenting cells (APC) to T
cells, so the time during which the T cells expand will
be shorter than 8 days. For these reasons the real rate
of T cell division will be greater than our estimates.
Epitope-Specific Competition for the APC
Since the TCR-transgenic donor cells make up a smaller
fraction of the total CD81 cells in the LCMV-infected
chimeras when fewer P14 cells are initially transferred,
it is clear that precursor frequency plays a large role in
deciding which clones respond to a given epitope in the
response to the virus. Thus, as greater numbers of P14
cells are transferred they come to dominate the antigp33 response. However, since we can vary the precursor frequency of gp33-specific CD81 T cells by more
than 100-fold without affecting the number of gp33specific, np396-specific, or gp276-specific CTL that ultimately develop, precursor frequency does not dictate
epitope dominance in this system. Rather, it must be
that antigen presentation is limiting in the priming and
expansion of the CD81 T cells.
When 103–105 P14 spleen cells (representing 3 3
102–3 3 104 CD81 cells) are transferred into naive B6.PL
hosts, there is no effect on the overall CTL response to
the three LCMV epitopes. At all of these doses, the level
of host CTL response to np396 and gp276 is unchanged
from the response of control B6.PL mice. The total hostplus-donor T cell response to gp33 also remains constant, but the composition of this response changes.
For example, at 103 cells transferred, only 1.5%–1.7%
of the CD81 T cells are donor derived, and these make
up a small fraction of the total CTL activity against gp33.
At 104 cells transferred, donor-derived cells make up
13%–16% of total CD81 cells and account for about
50% of the activity against gp33. At 105 cells transferred,
donor CD81 cells make up 23%–26% of total CD81 cells
and dominate the anti-gp33 activity to the point of suppressing host T cells of this specificity. It is noteworthy
that, as we increase the number of donor cells transferred from 104 to 105 , the contribution of gp33-specific
donor CD81 cells to the response does not increase
10-fold. Thus, an excess of gp33-specific precursors
holds the host and donor precursor T cells of the same
specificity in check while, at the same time, the host
response to the other two epitopes is unaffected. From
this observation we conclude that there is an epitopespecific regulation of the response.
The APC that present the LCMV peptides to CTL precursors are either LCMV-infected themselves or have
phagocytosed other infected cells and processed their
proteins for presentation by MHC class I (Bevan, 1995).
Since it is inconceivable that separate APC present each
of the three LCMV epitopes, we must conclude that
CD81 T cells compete for epitopes on the same APC.
Figure 7 illustrates this notion by supposing that, following viral infection, an APC presents a set of viral epitopes
on its surface. Due to competition with self peptides
that bind MHC class I, the number of copies of each
epitope expressed per APC is likely to be 103 or fewer
(Rammensee et al., 1993). CD81 T cells specific for any
epitope will take up space on the APC surface, but
in addition to this epitope-nonspecific competition, we
propose that T cells occupying the APC will sequester
their own target epitopes in a specific manner (Figure
7, middle). An excess of CD81 T cells specific for epitope
1 will sequester this epitope, reducing the available density from 103 per APC to a point where other T cells of
the same specificity cannot recognize the APC (Figure
7, right). At the same time, the APC still presents other
epitopes at 103 available copies per cell, so CD81 T cells
specific for these epitopes can continue to bind the APC
and become activated.
What is the nature of the sequestration of MHC class
I peptide epitopes on the APC? It may be that the TCR
and its ligand aggregate at the zone of contact of the
two cells. This aggregation of MHC molecules could
lead to their endocytosis by the APC. In this way the
class I–peptide complexes would be turned over just as
TCR complexes are turned over during T cell activation
(Valitutti et al., 1995). Even without TCR-induced endocytosis of MHC class I, however, it is clear that TCR
engagement may hide epitopes in a specific way. Whatever the mechanism, some degree of concentration or
aggregation of specific class I epitopes on the APC into
the zone of contact with the T cell would be required to
explain the epitope-specific regulation of the response.
This model has important implications for our understanding of how the epitope profile of a CD81 T cell
response is controlled and suggests that the nature of
T cell priming works to diversify the epitopes targeted
by CTL responses.
C57Bl/6 (B6) mice were purchased from Taconic Farms (Germantown, PA). Thy1 congenic B6.PL-Thy1a /Cy (B6.PL) mice were purchased from Jackson Laboratories (Bar Harbor, ME), and anti-Db/
LCMV GP-1 (33–41) TCR-transgenic P14 mice on a B6 background
(Pircher et al., 1990) were purchased from Jackson Laboratories
and then bred in the University of Washington specific pathogen–
free animal facilities. Anti-Kb/chicken OVA257–264 TCR-transgenic OT-1
mice on a B6 background have been described elsewhere (Hogquist
et al., 1994) and were bred in our specific pathogen–free animal
facilities. All mice used in these studies were 1- to 4-month-old
EL4 (ATCC TIB-41) are a B6-derived (H-2b), MHC class II2 thymoma
cell line and were maintained in RP10 (RPMI 1640 supplemented
with 10% heat-inactivated fetal calf serum, 2 mM L-glutamine, and
The Armstrong 53b strain of LCMV was originally obtained from
Peter J. Southern (University of Minnesota, Minneapolis, MN) and
was grown on BHK-21 cells (ATCC CCL-10) and titered on Vero
Cl008 cells (ATCC CRL-1586) as described (Ahmed et al., 1984).
Mice were infected by intravenous injection of 1 3 10 5 pfu of LCMV.
Recombinant VV were obtained from Jonathan Yewdell (National
Institutes of Health, Bethesda, MD). VVova expresses a full-length
cDNA for OVA (Bacik et al., 1994), and VVflu-np expresses a partial
influenza nucleoprotein cDNA (Yewdell et al., 1985). VV were grown
on HeLa S3 cells and titered on Vero cells according to standard
protocols (Mackett et al., 1985). Mice were infected by intravenous
injection of 5 3 10 6 pfu of VV.
The H-2D b-binding LCMV peptides LCMV gp33 (KAVYNFATC),
gp276 (SGVENPGGYCL), and np396 (FQPQNGQFI) and the K b-binding OVA 257–264 (SIINFEKL) were synthesized using an Applied Biosystems Synergy (Foster City, CA) peptide synthesizer. Peptide concentrations were determined using the BCA assay (Pierce Chemical,
Rockford, IL). The gp33 and gp276 peptides were dissolved in acidified RPMI-1640 with 1 mM 2-mercaptoethanol to prevent cysteine
Antibodies and Flow Cytometry
For flow cytometry we used directly conjugated anti-CD4–
fluorescein isothiocyanate (FITC), anti-CD8-phycoerythrin, antiThy1.1-FITC, and anti-TCR Va2-FITC (Pharmingen, San Diego, CA)
and anti-Thy1.2-biotin and streptavidin-Tricolor (CalTag, South San
Francisco, CA). All staining also included 2% normal mouse serum
and the anti-FcgRII antibody 24G2 (Pharmingen) to reduce nonspecific and Fc receptor–mediated binding. Analysis was done on a
Becton-Dickenson FACScan with Lysis-II software (Becton-Dickinson, Mountain View, CA).
The anti-IFNg antibody R4-6A2 (ATCC HB-170) was protein G
purified from tissue culture supernatants. Biotinylated XMG-1.2,
which recognizes a different epitope of murine IFNg, was purchased
from Pharmingen. The 19E12 monoclonal antibody, specific for
Thy1.1, and the 30H12 anti-Thy1.2 monoclonal antibody were produced as ascites.
IFNg ELISPOT Assays
Cells secreting IFNg in an antigen-specific manner were detected
using a standard ELISPOT assay (Miyahira et al., 1995). In brief, EL4
target cells were incubated in phosphate-buffered saline (PBS) with
or without 1 mM peptide as indicated, washed several times, and
added at 105 per well to graded numbers of erythrocyte-depleted
effector cells in 96-well Multiscreen-HA plates (Millipore, Bedford
MA) that had been precoated with protein G–purified R4–6A2. After
20–24 hr, cells were removed; the plates were extensively washed;
and the plates were developed by incubation with XMG-1.2-biotin,
followed by streptavidin–horseradish peroxidase and diaminobenzidine (Sigma, St. Louis, MO).
Primary Ex Vivo Chromium Release Assays
Target cells were prepared by incubation for 1–2 hr with or without
peptide in the presence of sodium 51Cr-chromate, washed three
times in PBS and resuspended in RP10. For the assay, 104 target
cells were added to 96-well round-bottom plates along with different
numbers of erythrocyte-depleted effector cells in a total volume of
200 ml. After 8 hr, 100 ml of supernatant was removed and counted in
a Wallac 1470 Wizard g-counter (Wallac Oy, Turku, Finland). Specific
lysis was calculated as ([experimental release 2 spontaneous release]/[maximum release 2 spontaneous release]) 3 100%. Spontaneous release was determined for target cells in medium alone, and
maximum release was determined by incubating target cells in 1%
Triton X-100. Spontaneous release was typically 10%–20%. Lytic
units were calculated as the number of effector cells required to
achieve 30% lysis of target cells.
In some experiments donor or host effector T cells were depleted
by incubating spleen cell suspensions with medium alone, with 5
mg/ml anti-Thy1.1 antibody (19E12), or with 10 mg/ml anti-Thy1.2
(30H12) on ice for 30 min followed by addition of rabbit complement
(Low-Tox M, Cedarlane, Westbury, NY) and incubation for 30 min
We thank Marianne Zollman and Ethan Ojala for technical assistance. We thank Stefan Martin, Brad Cookson, Jacqueline Kirchner,
Ananda Goldrath, and Laurel Lenz for their comments on the manuscript.
Received November 20, 1997; revised January 9, 1998.
Ahmed, R., Salmi, A., Butler, L.D., Chiller, J.M., and Oldstone, M.B. A.
(1984). Selection of genetic variants of lymphocytic choriomeningitis
virus in spleens of persistently infected mice. Role in suppression
of cytotoxic T lymphocyte response and viral persistence. J Exp
Med 160, 521–540.
Andersson, E.C., Christensen, J.P., Scheynius, A., Marker, O., and
Thomsen, A.R. (1995). Lymphocytic choriomeningitis virus infection
is associated with long-standing perturbation of LFA-1 expression
on CD81 T cells. Scand J Immunol 42, 110–118.
Bacik, I., Cox, J.H., Anderson, R., Yewdell, J.W., and Bennink, J.R.
(1994). TAP (transporter associated with antigen processing)-independent presentation of endogenously synthesized peptides is enhanced by endoplasmic reticulum insertion sequences located at
the amino- but not carboxyl-terminus of the peptide. J Immunol 152,
Bevan, M.J. (1995). Antigen presentation to cytotoxic T lymphocytes
in vivo. J Exp Med 182, 639–641.
Beverley, P.C. L. (1996). Generation of T-cell memory. Current Opinion in Immunology 8, 327–330.
Buchmeier, M.J., Welsh, R.M., Dutko, F.J., and Oldstone, M.B. A.
(1980). The virology and immunobiology of lymphocytic choriomeningitis virus infection. Adv Immunol 30, 275–331.
Expansion of Antiviral CD81 Cells during Infection
Callan, M.F., Steven, N., Krausa, P., Wilson, J.D., Moss, P.A., Gillespie, G.M., Bell, J.I., Rickinson, A.B., and McMichael, A.J. (1996).
Large clonal expansions of CD81 T cells in acute infectious mononucleosis. Nat Med 2, 906–911.
Rammensee, H.G., Falk, K., and Rotzschke, O. (1993). Peptides
naturally presented by MHC class I molecules. Annu Rev Immunol
Cauda, R., Prasthofer, E.F., Tilden, A.B., Whitley, R.J., and Grossi,
C. E. (1987). T-cell imbalances and NK activity in varicella-zoster
virus infections. Viral Immunol 1, 145–152.
Razvi, E.S., Welsh, R.M., and McFarland, H.I. (1995). In vivo state
of antiviral CTL precursors. Characterization of a cycling cell population containing CTL precursors in immune mice. J Immunol 154,
Cose, S.C., Jones, C.M., Wallace, M.E., Heath, W.R., and Carbone,
F.R. (1997). Antigen-specific CD81 T cell subset distribution in lymph
nodes draining the site of herpes simplex virus infection. Eur. J.
Immunol. 27, 2310–2316.
Rott, O., Mignon, G.K., Fleischer, B., Charreire, J., and Cash, E.
(1995). Superantigens induce primary T cell responses to soluble
autoantigens by a non-V beta-specific mechanism of bystander activation. Cell Immunol 161, 158–165.
Ehl, S., Hombach, J., Aichele, P., Hengartner, H., and Zinkernagel,
R. M. (1997). Bystander activation of cytotoxic T cells: studies on
the mechanism and evaluation of in vivo significance in a transgenic
mouse model. J Exp Med 185, 1241–1251.
Rubin, R.H., Carney, W.P., Schooley, R.T., Colvin, R.B., Burton, R.
C., Hoffman, R.A., Hansen, W.P., Cosimi, A.B., Russell, P.S., and
Hirsch, M.S. (1981). The effect of infection on T lymphocyte subpopulations: a preliminary report. Int J Immunopharmacol 3, 307–312.
Gairin, J.E., Mazarguil, H., Hudrisier, D., and Oldstone, M.B. (1995).
Optimal lymphocytic choriomeningitis virus sequences restricted
by H-2Db major histocompatibility complex class I molecules and
presented to cytotoxic T lymphocytes. J Virol 69, 2297–2305.
Selin, L.K., Nahill, S.R., and Welsh, R.M. (1994). Cross-reactivities
in memory cytotoxic T lymphocyte recognition of heterologous viruses. J Exp Med 179, 1933–1943.
Hogquist, K.A., Jameson, S.C., Heath, W.R., Howard, J.L., Bevan,
M. J., and Carbone, F.R. (1994). T cell receptor antagonist peptides
induce positive selection. Cell 76, 17–27.
Kageyama, S., Tsomides, T.J., Sykulev, Y., and Eisen, H.N. (1995).
Variations in the number of peptide-MHC class I complexes required
to activate cytotoxic T cell responses. J Immunol 154, 567–576.
Kedl, R.M., and Mescher, M.F. (1997). Migration and activation of
antigen-specific CD81 T cells upon in vivo stimulation with allogeneic tumor. J Immunol 159, 650–663.
Kramer, M.D., Fruth, U., Simon, H.G., and Simon, M.M. (1989). Expression of cytoplasmic granules with T cell-associated serine proteinase-1 activity in Ly-21(CD81) T lymphocytes responding to
lymphocytic choriomeningitis virus in vivo. Eur J Immunol 19,
Lau, L.L., Jamieson, B.D., Somasundaram, T., and Ahmed, R. (1994).
Cytotoxic T-cell memory without antigen. Nature 369, 648–652.
Speiser, D.E., Kyburz, D., Stubi, U., Hengartner, H., and Zinkernagel,
R. M. (1992). Discrepancy between in vitro measurable and in vivo
virus neutralizing cytotoxic T cell reactivities: Low T cell receptor
specificity and avidity sufficient for in vitro proliferation or cytotoxicity to peptide-coated target cells but not for in vivo protection. J
Immunol 149, 972–980.
Strang, G., and Rickinson, A.B. (1987). In vitro expansion of EpsteinBarr virus-specific HLA-restricted cytotoxic T cells direct from the
blood of infectious mononucleosis patients. Immunology 62,
Tough, D.F., and Sprent, J. (1994). Turnover of naive- and memoryphenotype T cells. J Exp Med 179, 1127–1135.
Tough, D.F., and Sprent, J. (1996). Viruses and T cell turnover: evidence for bystander proliferation. Immunol. Rev. 150, 129–142.
Tough, D.F., Borrow, P., and Sprent, J. (1996). Induction of bystander
T cell proliferation by viruses and type I interferon in vivo. Science
Liu, Y.J., Zhang, J., Lane, P.J., Chan, E.Y., and MacLennan, I.C.
(1991). Sites of specific B cell activation in primary and secondary
responses to T cell-dependent and T cell-independent antigens. Eur
J Immunol 21, 2951–2962.
Tripp, R.A., Hou, S., McMickle, A., Houston, J., and Doherty, P.C.
(1995). Recruitment and proliferation of CD81 T cells in respiratory
virus infections. J Immunol 154, 6013–6021.
Lynch, F., Doherty, P.C., and Ceredig, R. (1989). Phenotypic and
functional analysis of the cellular response in regional lymphoid
tissue during an acute virus infection. J Immunol 142, 3592–3598.
Unutmaz, D., Pileri, P., and Abrignani, S. (1994). Antigen-independent activation of naive and memory resting T cells by a cytokine
combination. J Exp Med 180, 1159–1164.
Mackett, M., Smith, G.L., and Moss, B. (1985). The construction and
characterization of vaccinia virus recombinants expressing foreign
genes. In DNA cloning: a practical approach, D.M. Glover, eds.
(Oxford: IRL), pp. 191–211.
Valitutti, S., Muller, S., Cella, M., Padovan, E., and Lanzavecchia, A.
(1995). Serial triggering of many T-cell receptors by a few peptideMHC complexes. Nature 375, 148–151.
McFarland, H.I., Nahill, S.R., Maciaszek, J.W., and Welsh, R.M.
(1992). CD11b (Mac-1): a marker for CD81 cytotoxic T cell activation
and memory in virus infection. J Immunol 149, 1326–1333.
Miyahira, Y., Murata, K., Rodriguez, D., Rodriguez, J.R., Esteban,
M., Rodriguez, M.M., and Zarala, F. (1995). Quantification of antigen
specific CD81 T cells using an ELISPOT assay. J. Immunol. Meth.
Mueller, D.L., and Jenkins, M.K. (1997). Autoimmunity: when selftolerance breaks down. Curr Biol 7, 255–257.
Nahill, S.R., and Welsh, R.M. (1993). High frequency of cross-reactive cytotoxic T lymphocytes elicited during the virus-induced polyclonal cytotoxic T lymphocyte response. J Exp Med 177, 317–327.
Pantaleo, G., Demarest, J.F., Soudeyns, H., Graziosi, C., Denis, F.,
Adelsberger, J.W., Borrow, P., Saag, M.S., Shaw, G.M., Sekaly, R.P.,
and et, a. l. (1994). Major expansion of CD81 T cells with a predominant V beta usage during the primary immune response to HIV.
Nature 370, 463–467.
Pape, K.A., Kearney, E.R., Khoruts, A., Mondino, A., Merica, R.,
Chen, Z. M., Ingulli, E., White, J., Johnson, J.G., and Jenkins, M.K.
(1997). Use of adoptive transfer of T-cell-antigen-receptortransgenic T cell for the study of T-cell activation in vivo. Immunol
Rev 156, 67–78.
Pircher, H., Moskophidis, D., Rohrer, U., Burki, K., Hengartner, H.,
and Zinkernagel, R.M. (1990). Viral escape by selection of cytotoxic
T cell-resistant virus variants in vivo. Nature 346, 629–633.
Valitutti, S., Muller, S., Dessing, M., and Lanzavecchia, A. (1996).
Different responses are elicited in cytotoxic T lymphocytes by different levels of T cell receptor occupancy. J Exp Med 183, 1917–1921.
Yang, H., and Welsh, R.M. (1986). Induction of alloreactive cytotoxic
T cells by acute virus infection of mice. J Immunol 136, 1186–1193.
Yang, H.Y., Dundon, P.L., Nahill, S.R., and Welsh, R.M. (1989). Virusinduced polyclonal cytotoxic T lymphocyte stimulation. J Immunol
Yewdell, J.W., Bennink, J.R., Smith, G.L., and Moss, B. (1985). Influenza A virus nucleoprotein is a major target antigen for cross-reactive anti-influenza A virus cytotoxic T lymphocytes. Proc Natl Acad
Sci U S A 82, 1785–1789.
Zarozinski, C.C., and Welsh, R.M. (1997). Minimal bystander activation of CD8 T cells during the virus-induced polyclonal T cell response. J Exp Med 185, 1629–1639.
Zimmerman, C., Brduscha, R.K., Blaser, C., Zinkernagel, R.M., and
Pircher, H. (1996). Visualization, characterization, and turnover of
CD81 memory T cells in virus-infected hosts. J Exp Med 183, 1367–