Download All Activated CD8 T Cells A Transgenic Mouse Model Genetically Tags

A Transgenic Mouse Model Genetically Tags
All Activated CD8 T Cells
This information is current as
of August 3, 2017.
Subscription
Permissions
Email Alerts
J Immunol 2003; 171:2393-2401; ;
doi: 10.4049/jimmunol.171.5.2393
http://www.jimmunol.org/content/171/5/2393
This article cites 54 articles, 21 of which you can access for free at:
http://www.jimmunol.org/content/171/5/2393.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2003 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
References
Charles H. Maris, Joseph D. Miller, John D. Altman and
Joshy Jacob
The Journal of Immunology
A Transgenic Mouse Model Genetically Tags All Activated
CD8 T Cells1
Charles H. Maris, Joseph D. Miller, John D. Altman, and Joshy Jacob2
Identifying and characterizing Ag-specific CD8ⴙ T cells are central to the study of immunological memory. Although powerful
strategies such as MHC tetramers and peptide-induced cytokine production assays exist for identifying Ag-specific CD8ⴙ T cells,
alternate strategies that are not dependent upon a priori knowledge of the immunodominant and subdominant antigenic epitopes,
as well as the MHC background of the animal are of obvious utility. In this study, we present a transgenic mouse model that uses
Cre-loxP recombination to permanently mark all activated CD8ⴙ T cells with ␤-galactosidase. We used the lymphocytic choriomeningitis virus infection model to track the dynamics of the antiviral CD8ⴙ T cell responses. We show that in this transgenic
mouse model system, all of the antiviral effector and memory CD8ⴙ T cells are contained within the ␤-gal-marked CD8ⴙ T cell
population. The Journal of Immunology, 2003, 171: 2393–2401.
Department of Microbiology and Immunology and Emory Vaccine Center, Emory
University, Atlanta, GA 30329
Received for publication February 21, 2003. Accepted for publication June 27, 2003.
the individual epitopes. Similarly, quantitation of virus-specific
CD8⫹ T cells by intracellular cytokine staining or ELISPOT assays is dependent upon a priori knowledge of immunogenic
peptides.
Although tetramers have been pivotal in dissecting Ag-specific
CD8⫹ T cell responses, several recent studies have described
novel types of Ag-specific CD8⫹ T cells that, despite being Ag
specific, fail to bind tetramers. Using a TCR transgenic model,
Blohm et al. (22) have shown that tumor-infiltrating TCR transgenic lymphocytes that are found in tumors expressing their cognate peptide failed to bind MHC/tetramer, exhibited impaired cytolytic activity, and failed to proliferate or elaborate IFN-␥ when
stimulated with specific peptide. Interestingly, PBL and splenocytes from these same tumor-bearing mice did bind tetramers, exhibited potent cytolytic activity, produced IFN-␥, and proliferated
following peptide stimulation. Spencer and Braciale (23) have described influenza-specific murine CD8⫹ T cells specific for a subdominant epitope of influenza virus that fails to bind influenza
tetramers directly ex vivo. Following repeated rounds of stimulation in vitro, these cells transition from a tetramer-negative to tetramer-positive phenotype. Schneck and colleagues (24) have
shown that in vitro activated CD8⫹ T cells from the 2C TCR
transgenic mouse bind peptide/MHC dimers 20- to 50-fold better
than naive TCR transgenic T cells from the same mouse. Finally,
in both mouse and human chronic viral infections, the appearance
and persistence of tetramer-positive CD8⫹ T cells that fail to elaborate effector functions have been described (25–28). Peptide/
MHC multimers may be of limited utility in certain situations.
Thus, there exists a need for model systems in which polyclonal
populations of activated CD8⫹ T cells specific for various Ags can
be identified and followed independent of tetramer binding and
peptide-induced cytokine production.
We have previously described a transgenic mouse model system
in which both effector and memory CD8⫹ T lymphocytes are
tagged via Cre-loxP recombination, with the marker protein, human placental alkaline phosphatase (PLAP)3 (29). The system is
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by a grant from the National Institutes of Health (AI
47253).
2
Address correspondence and reprint requests to Dr. Joshy Jacob, Emory University,
954 Gatewood Road, Atlanta, GA 30329. E-mail address: [email protected]
Copyright © 2003 by The American Association of Immunologists, Inc.
Abbreviations used in this paper: PLAP, placental alkaline phosphatase; ␤-gal,
␤-galactosidase; FDG, fluorescein di-␤-D galactopyranoside; GBC, granzyme B promoter driving expression of Cre recombinase; LCMV, lymphocytic choriomeningitis
virus; STOP, truncated yeast His3 gene and polyadenylation sequence.
3
0022-1767/03/$02.00
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
T
he immune system responds with enhanced vigor to Ags
encountered in the past. This exaggerated recall immune
response, or immune memory, is a central concept in immunology, and it forms the basis of vaccination against infectious
disease. Immune memory is mediated, in part, by memory CD4
and CD8⫹ T lymphocytes that persist in the host long after resolution of the antigenic insult or infection. Following viral infection,
APCs process viral Ags and present peptides to CD8⫹ T cells in
the context of class I MHC. Ag-specific CD8⫹ T cells that recognize the viral peptides presented by the APC undergo massive
proliferation and differentiation, a response that can be categorized
into three distinct phases: activation and expansion, death, and
memory (1–9). The expansion phase is quite spectacular; recent
studies in mice have shown that precursor CD8⫹ T cells specific
for individual viral epitopes are able to expand from below the
limits of detection (⬍1000 cells per spleen) to greater than 107
cells per spleen (10). Following the elimination of viral Ag (6) and
due to an inherent genetic program (11), there is an equally dramatic death phase, during which 90 –95% of the Ag-specific CD8⫹
T cells die. Virus-specific CD8⫹ T cells that survive the death
phase persist in the host as a stable pool of memory lymphocytes
that provides lifelong immunity (6, 8, 9, 12).
Technical advances such as multimers of viral peptide/MHC
and peptide-induced cytokine secretion assays (intracellular cytokine staining and ELISPOT) have allowed the quantitation of antiviral CD8⫹ T cell responses with great accuracy and precision in
a variety of model systems, including viruses, bacteria, and tumors, in both rodents and primates (6, 13–21). The advent of peptide/MHC multimers has revolutionized the study of CD8⫹ T cell
memory; however, the construction of MHC multimers is predicated on the assumption that immunodominant and subdominant
viral epitopes have been defined, as well as the MHC restriction of
2394
CD28 (clone 37.51; 2 ␮g ml⫺1), added ⬎24 h earlier, or with 10 ␮g ml⫺1
poly(I:C). Following stimulation, live cells were assayed by flow cytometry for expression of ␤-gal, CD8, and CD69.
Peptides
LCMV and control peptides were synthesized by B. Evavold at the Emory
University Department of Microbiology and Immunology Peptide Core
Facility using F-moc chemistry on a Symphony/Multiplex Peptide Synthesizer. The peptides were purified by HPLC (Rainin Instruments,
Woburn, MA).
Abs and staining for flow cytometry
At various time points following LCMV infection, GBC ⫻ ROSA26R
mice were euthanized by cervical dislocation, spleens were harvested and
RBC lysed, and single cell suspensions were made from spleen. To obtain
PBLs, mice were retro-orbitally bled into 4% sodium citrate, and lymphocytes were enriched using lymphocyte separation medium (Cellgro). For
detection of ␤-gal, cells were hypotonically loaded with FDG, as previously described (32). Following FDG loading, the cells were surface
stained with fluorochrome-conjugated or biotinylated Abs to CD8 (clone
53-6.7 from BD Biosciences), CD4, NK1.1, CD44, CD62L, CD43, CD25,
CD122, CD69, or CD11a (clones RM4-5, IM7.8.1, MEL-14, 1B11, PC61
5.3, TM-␤1, H1.2F3, and 2D7 from Caltag). Streptavidin-coupled PE and
allophycocyanin were purchased from Caltag. Abs to Thy-1.1 and Thy-1.2
were purchased from BD Biosciences. Allophycocyanin-conjugated MHC
tetramers of LCMV gp33– 41, NP396 – 404, or Qa-1b tetramers folded
around the Qdm peptide were generated, as described (15, 34). Stained
cells were kept on ice at all times in FACS buffer (1.5% BSA and 0.02%
sodium azide) and analyzed on a FACSCalibur (BD Biosciences) flow
cytometer running CellQuest software. Data were analyzed with FlowJo
(Treestar, San Carlos, CA) or CellQuest software.
Cell sorting
RBC-lysed, single cell suspensions of splenocytes from LCMV-infected or
naive mice were hypotonically loaded with FDG and stained with anti-CD8
PE and anti-CD4 allophycocyanin (Caltag and BD Biosciences) in PBS
containing 3% BSA (Sigma-Aldrich). ␤-gal⫹CD8⫹CD4⫺ and ␤-gal⫺
CD8⫹CD4⫺ populations were sorted on a high-speed Cytomation MoFlo
cell sorter to 95–98% purity (data not shown).
Cytotoxicity assay
C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor,
ME) and housed in specific pathogen-free conditions in an American Association of Laboratory Animal Care-accredited facility at Emory University. Granzyme B Cre, CD2-STOP-PLAP, and ROSA26R (29, 32) mice
were bred and maintained at Emory University following the guidelines of
the Institutional Animal Care and Use Committee. LCMV Armstrong and
LCMV clone 13 were triple plaque purified on VERO cells and amplified
one round on BHK-21 cells (American Type Culture Collection (ATCC),
Manassas, VA) (33). For viral infections, mice were injected once i.p. with
2 ⫻ 105 PFU LCMV Armstrong. For viral rechallenge assays, mice were
injected with 2 ⫻ 106 PFU LCMV clone 13 i.v. LCMV Armstrong and
LCMV clone 13 were kind gifts from R. Ahmed, Emory University.
CTL assays were performed, as previously described (6), with the following modifications. MC57G fibroblast cells (ATCC) were pulsed with
LCMV gp33– 41, LCMV NP396 – 404, or OVA SIINFEKL peptides at 10
␮g ml⫺1 for 90 min at 37oC, followed by labeling with 51Cr (PerkinElmer,
Boston, MA). Sorted ␤-gal⫹ and ␤-gal⫺CD8⫹ T cells were plated in triplicate in 96-well plates at various dilutions; 51Cr-loaded and peptide-pulsed
targets were added to achieve the desired E:T ratios, and incubated for 6 h
at 37oC. In control wells, medium without effectors were added to target
cells to calculate the spontaneous 51Cr release; 1% Triton X-100 was added
to targets, followed by a 6-h incubation to obtain maximal 51Cr release.
Following incubation, supernatants were collected and transferred to Wallac CTL plates (Wallac/PerkinElmer) and air dried overnight in a laminar
flow chemical hood. The plates were then read on a 1450 Microbeta liquid
scintillation counter (Wallac/PerkinElmer). Data are expressed as percentage of specific 51Cr release ((experimental release – spontaneous release)/
(maximum release – spontaneous release) ⫻ 100).
Stimulations
Adoptive transfers
Single cell suspensions from naive GBC ⫻ ROSA26R double-transgenic
mice were cleared of RBCs using RBC lysis buffer (Sigma-Aldrich, St.
Louis, MO). A total of 5 ⫻ 106 cells was plated in a volume of 1 ml of
complete RPMI (10% FCS, 2 mM L-glutamine, 1000 U penicillin G, 1000
U streptomycin, 0.25 ␮g ml⫺1 amphotericin B, and 10 mM HEPES (Cellgro, Herndon, VA)). Cells were stimulated with either PMA and ionomycin
or PHA and IL-2. PMA and ionomycin were added at a final concentration
of 1 ␮g ml⫺1 and 2 ␮g ml⫺1, respectively. Cells were incubated for 48 h
at 37oC in a humidified incubator with 5% CO2. PHA and IL-2 were added
to a final concentration of 10 ␮g ml⫺1 and 25 U ml⫺1, respectively. Cells
were periodically assayed by FACS after hypotonic loading with fluorescein di-␤-D galactopyranoside (FDG; Molecular Probes, Eugene, OR) and
cell surface staining for CD4, CD8, and CD69 (anti-CD8 from BD Biosciences, San Diego, CA; anti-CD4 and anti-CD69 from Caltag, Burlingame, CA). Poly(I:C) and anti-CD3/CD28 stimulations were conducted on
sorted populations of ␤-gal⫹ and ␤-gal⫺ CD8⫹ T cells. Sorted cells were
cultured with plate-bound Abs to CD3⑀ (clone 145-2C11; 5 ␮g ml⫺1) and
␤-gal⫹CD8⫹CD4⫺ and ␤-gal⫺CD8⫹CD4⫺ T lymphocytes were sorted on
a Cytomation MoFlo, as described above. A total of 8 ⫻ 105 sorted cells
from GBC ⫻ ROSA26R double-transgenic mice that had received 2 ⫻ 105
PFU LCMV Armstrong ⬎30 days previously was injected into congenic
Thy-1.1 mice. As a control, 8 ⫻ 105 unsorted cells were injected into
separate Thy-1.1 recipient mice. Twelve hours later, recipients were challenged with 2 ⫻ 106 PFU LCMV clone 13 i.v. Recipient mice were sacrificed 7 days later, and their spleens were harvested for FACS analysis,
and serum viral titers were determined by plaque assay.
Materials and Methods
Animals and viral infections
Plaque assay
Plaque assays to determine viral titers were performed exactly as described
(35). Briefly, serum samples were serially diluted in 1% RPMI and overlaid
on VERO (ATCC) cell monolayers in six-well plates (Invitrogen, Carlsbad, CA) and incubated for 90 min. After 90 min, a 1:1 mixture of 2⫻ 199
medium (Life Technologies/Invitrogen) and 1% agarose was gently overlaid on the VERO cells, and the plates were incubated in a humidified
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
comprised of two transgenic mouse lines; one has a truncated human granzyme B promoter driving the expression of Cre recombinase (GBC) in activated T cells, and the other has a T cellspecific human CD2 promoter driving the expression of a lox-Pflanked transcriptional truncated yeast His3 gene and
polyadenylation sequence (STOP) fragment, followed by PLAP
(CD2-STOP-PLAP). This mouse model was originally designed to
mark all activated CD8⫹ T cells, but only a small subset of the
activated effector CD8⫹ T cells is marked with PLAP. For instance, when mice doubly transgenic for CD2-STOP-PLAP and
GBC are infected with lymphocytic choriomeningitis virus
(LCMV), only a small subset (10%) of the CD8⫹ T cells express
surface PLAP at the peak of the anti-LCMV immune response,
even though LCMV peptide/MHC multimers, intracellular cytokine flow cytometric assays, and direct ex vivo cytotoxicity studies
have unequivocally demonstrated that 50 –70% of the splenic
CD8⫹ T cells are responding to LCMV (6, 29).
To complement MHC tetramers for the detection of Ag-specific
CD8⫹ T cells and to overcome the deficit of the previously described transgenic model, we sought to establish a transgenic
mouse model system that would permanently mark not just a subset, but all activated CD8⫹ T cells. In the experiments described in
this work, we chose bacterial ␤-galactosidase (␤-gal) as the reporter instead of PLAP by using the Cre reporter transgenic mice,
ROSA26R (30 –32). In ROSA26R mice, Cre-mediated excisional
recombination leads to permanent expression of ␤-gal from the
constitutively active ROSA promoter. We crossed ROSA26R mice
with the previously described GBC transgenic mice (29) and used
the double-transgenic offspring to probe CD8⫹ T cell activation
using the LCMV infection model. Data presented in this work
demonstrate that there is a high degree of efficacy in this transgenic
marking method; all activated CD8⫹ T cells are marked with
␤-gal. We show that following LCMV infection, the population of
␤-gal-marked CD8⫹ T cells generated contains all of the antiviral
effector and memory CD8⫹ T cells.
TRACKING ACTIVATED CD8 T CELLS
The Journal of Immunology
incubator with 5% CO2 for 5 days. On the fourth day, a 1:1 mixture of 2⫻
199 medium and 1% agarose with 0.25% Neutral Red (Sigma-Aldrich) was
added to visualize plaques.
Results
Rationale for choosing ␤-gal as a reporter to mark activated
CD8⫹ T cells
FIGURE 1. Schematic diagram of the GBC and ROSA26R transgenes.
a, In the GBC transgene, the truncated human granzyme B promoter drives
the expression of Cre recombinase in activated T cells. The human growth
hormone (hGH) polyadenylation signal sequence was included to enhance
mRNA stability. ROSA26R is a Cre-reporter strain in which a loxPflanked, phosphoglycerate kinase (pgk) promoter-driven neomycin phosphotransferase expression cassette and ␤-gal-encoding LacZ gene have
been engineered downstream of the ROSA promoter using homologous
recombination (30). b, Cre-expressing cells recombine to excise the neomycin cassette, leaving behind ␤-gal under the control of the ubiquitous
and constitutive ROSA promoter (31).
32). We analyzed activation-induced CD8⫹ T cell marking in double-transgenic offspring of the GBC mouse strain crossed to
ROSA26R. By examining double-transgenic GBC ⫻ ROSA26R
mice, we tested the fidelity of the granzyme B Cre transgene to
induce Cre recombination-mediated genetic tagging of all activated CD8⫹ T cells, and were thus able to characterize the size and
temporal dynamics of the entire pool of responding CD8⫹ T cells
in an LCMV infection model.
T cell marking in naive doubly transgenic mice
It is known that ⬃8 –10% of the CD8⫹ T cells in naive mice
exhibit an activated phenotype. These activated memory-phenotype CD8⫹ T cells, which are CD44high, arise due to endogenous
immune activation presumably by gut flora or food Ags (37–39).
We analyzed the spleen and peripheral blood CD8⫹ T cells of
naive GBC ⫻ ROSA26R double-transgenic mice for the presence
of ␤-gal-marked CD8⫹ T cells. As anticipated, 5–10% of CD8⫹ T
cells in naive mice expressed ␤-gal (Fig. 2). The surface phenotype
of the ␤-gal-marked CD8 T cells in the naive mice was further
examined to determine their activation status. ␤-gal⫹CD8⫹ T cells
were CD25⫺, Fas⫹, Fas ligand⫺, TCR␣␤⫹, CD71⫺, CD69⫺, and
biphasic for CD44 expression (Fig. 2, and data not shown). Of the
␤-gal⫹CD8⫹ T cells, 82% were CD62Lhigh (Fig. 2). The activation markers expressed by ␤-gal⫹CD8⫹ T cells are consistent with
a memory phenotype (40). In the naive animals, ␤-gal does not
appear to be a surrogate marker for CD44; as many as 50% of the
␤-gal⫹CD8⫹ T cells are CD44low (Fig. 2), indicating that the rules
FIGURE 2. The endogenously activated memory-phenotype CD8⫹ T
cells in GBC ⫻ ROSA26R double-transgenic mice express ␤-gal. Splenocytes from naive double-transgenic mice were hypotonically loaded with
the tracker dye FDG that fluoresces after cleavage by ␤-gal and stained
with fluorochrome-labeled Abs to CD8, CD25, CD44, and CD62L. a, Approximately 5–7% of CD8⫹ T cells in naive double-transgenic mice are
␤-gal⫹. b, Flow cytometry plots of ␤-gal⫹CD8⫹ T cells from naive double-transgenic mice stained for the activation markers CD25, CD44, and
CD62L are shown in a. c, ␤-gal expression in NK and CD4⫹ T cells of
naive double-transgenic mice is shown as histogram plots. More than 10
naive double-transgenic mice were analyzed. Representative plots are
shown.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
To better understand the population dynamics, kinetics, and ultimate fate of CD8 T cells responding to a viral infection, it is
important to track activated CD8 T cells throughout the course of
an immune response. To this end, we have previously described a
transgenic mouse model system to mark activated CD8 T cells by
generating mice doubly transgenic for GBC and CD2-STOPPLAP (29). These double-transgenic (GBC ⫻ CD2-STOP-PLAP)
mice, upon infection with LCMV, marked only a subset (10%) of
activated CD8⫹ T cells; the frequency of PLAP-marked CD8⫹ T
cells was considerably less than what others had reported using
MHC multimers and cytokine expression-based assays (6, 9, 29).
LCMV infection-elicited PLAP-marked CD8⫹ T cells apparently
represent a novel effector memory population, as they score positive in secondary bulk stimulation CTL assays, whereas PLAPnegative CD8⫹ T cells do not (29). The majority of the activated
CD8 T cells were PLAP⫺ because the PLAP protein gets degraded
in activated primary lymphocytes. Interestingly, the degradation of
PLAP protein upon activation is observed in primary lymphocytes,
but not in Jurkat T cell line that is stably transfected with PLAP (J.
Jacob, unpublished observations). In this study, we have used
␤-gal as a reporter instead of PLAP by using the ␤-gal-expressing
Cre reporter mouse strain, ROSA26R. The two transgenic mouse
strains used in this study are depicted in Fig. 1. The truncated
granzyme B promoter is sufficient for transcription in activated
CD8⫹ and CD4⫹ T lymphocytes (36). In GBC ⫻ ROSA26R double-transgenic mice, the Cre recombinase is expressed in activated
T lymphocytes, and it will induce an excisional recombination
juxtaposing the ␤-gal gene immediately downstream of the strong,
ubiquitous ROSA promoter (31). Cells in which this recombination event has occurred will permanently express ␤-gal. The intracellular accumulation of ␤-gal can then be easily detected by
flow cytometric analysis using the fluorescent substrate FDG (30,
2395
2396
TRACKING ACTIVATED CD8 T CELLS
cells. Taken together, these data demonstrate that the truncated
GBC and Cre-loxP recombination-mediated ␤-gal expression can
be induced in CD8⫹ T cells by TCR-dependent, TCR-independent, and cytokine-mediated stimuli.
TCR-dependent, TCR-independent, and cytokine stimuli lead to
␤-gal expression in CD8⫹ T cells
␤-gal marking of CD8⫹ T cells during an immune response to
LCMV
It was first necessary to determine the rules that govern the induction of Cre-mediated recombination and de novo expression of
␤-gal in CD8⫹ T cells from GBC ⫻ ROSA26R mice. Splenocytes
from naive GBC ⫻ ROSA26R double-transgenic mice, as well as
a single-transgenic littermate control mice, were stimulated with
TCR-dependent and TCR-independent stimuli in vitro. PMA/Ionomycin was used as TCR-independent stimuli, and PHA or antiCD3/CD28 was used as the TCR-dependent stimuli. Splenocytes
were stimulated for 48 h (Fig. 3a); all types of in vitro stimulation
resulted in increased ␤-gal expression. The frequency of ␤-gal⫹
CD8⫹ T cells after stimulation was 67, 51, and 70% for anti-CD3/
CD28, PHA, and PMA/ionomycin stimulations, respectively.
Thus, both TCR-dependent and TCR-independent stimuli lead to
granzyme B promoter-induced Cre recombination and ␤-gal expression in double-transgenic CD8⫹ T cells. Five to seven percent
of the input CD8⫹ T cells are endogenously ␤-gal⫹ before in vitro
stimulation. We ruled out the possibility that the mitogens were
selectively expanding the endogenously activated, pre-existing
␤-gal⫹CD8⫹ T cells by sorting ␤-gal⫹ and ␤-gal⫺CD8⫹ T cells
from naive double-transgenic mice and then stimulating the CD8⫹
T cell populations in culture. For stimulation, we used anti-CD3/
CD28 and poly(I:C). Poly(I:C), a potent stimulator of type I IFNs
(41, 42), was used for stimulation to determine whether cytokine
stimuli played a role in activating Cre recombination and ␤-gal
expression. Fig. 3b shows the stimulation of sorted cells with antiCD3/CD28 and poly(I:C). ␤-gal⫹CD8⫹ sorted T cells that already
expressed ␤-gal before stimulation continued to express ␤-gal after
48 h in vitro, confirming that ␤-gal was stable and not downregulated upon continued stimulation in activated CD8⫹ T cells.
Stimulation of purified ␤-gal⫺ naive CD8⫹ T cells with anti-CD3/
CD28 for 48 h led to ␤-gal expression in ⬎90% of the CD8⫹ T
cells, and poly(I:C) induced ␤-gal expression in 72% of CD8⫹ T
To study T cell activation during the course of an immune response to a viral infection, we infected cohorts of double-transgenic GBC ⫻ ROSA26R mice and single-transgenic littermate
control mice that are unable to recombine and express ␤-gal with
a single i.p. injection of 2 ⫻ 105 PFU of LCMV Armstrong.
LCMV is a natural mouse pathogen, and the Armstrong strain is
quickly cleared by the antiviral CD8 immune response that peaks
8 days postinfection (35, 43, 44). At various time points following
infection, CD8⫹ T cells from spleen and peripheral blood were
analyzed for expression of ␤-gal in conjunction with markers for
lymphocyte activation. Fig. 4a shows that at the peak of the immune response, ⬃80% of the CD8⫹ T cells in peripheral blood
express ␤-gal, and expression was equivalent in spleen (79.05% ⫾
2.76%, Fig. 4b). LCMV infection is quickly eradicated, and the
majority of the CD8⫹ T cells, as tracked by MHC tetramers and
viral peptide-elicited IFN-␥ expression, undergo apoptosis, such
that by day 30 postinfection, only 5–10% of the CD8⫹ T cells are
LCMV specific (6, 9). However, at day 30 postinfection, as many
as 74% of peripheral blood CD8⫹ T cells are ␤-gal⫹ (Fig. 4d). The
frequency of ␤-gal⫹CD8⫹ T cells eventually levels off to ⬃40%
by day 100 postinfection (Fig. 4, a and d). The kinetics of ␤-gal
expression in CD8⫹ T cells of the spleen mirrors that of peripheral
blood, except that the overall frequency is proportionally lower at
the same time points (Fig. 4b). Overall, the total number of ␤-galmarked splenic CD8⫹ T cells (Fig. 4c) displayed typical immune
response kinetics, i.e., expansion, followed by a death phase, followed by a stable memory phase.
The surface phenotype of ␤-gal⫹CD8⫹ lymphocytes was examined by flow cytometry at the peak (day 8) of the antiviral
immune response. At day 8, ⬃80% of the PBL CD8⫹ T cells are
␤-gal⫹, and we analyzed ␤-gal⫹CD8⫹ T cells for the coexpression
of T cell activation markers (Fig. 5). Of the ␤-gal⫹CD8⫹ T cells,
FIGURE 3. TCR-dependent and TCR-independent stimuli lead to Cre-mediated recombination and
expression of ␤-gal in CD8⫹ T cells. a, Splenocytes
from naive double-transgenic GBC ⫻ ROSA26R
mice were stimulated in vitro with either PHA PMA/
ionomycin or anti-CD3/CD28 Abs. Cells were harvested 48 h later, hypotonically loaded with FDG,
and assayed for CD8 and CD69 expression by flow
cytometry. All flow cytometry plots shown are gated
on live CD8⫹ T cells. Percentages denote the frequency of ␤-gal⫹CD69⫹ CD8 T cells. b, CD8⫹␤-gal⫹
and CD8⫹␤-gal⫺ T cells were sorted by FACS and
stimulated in vitro for 48 h with either anti-CD3/CD28
or poly(I:C) and assayed as before. Percentages denote
the frequency of gated CD8⫹ T cells that were CD69⫹
and expressed ␤-gal. These experiments were done
using ⬎12 doubly transgenic mice, and the experiments were repeated twice.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
governing CD44 and ␤-gal expression in naive mice are not the
same. In naive double-transgenic mice, ␤-gal expression was not
restricted to CD8⫹ T cells. Fifty percent of splenic NK cells and
12% of CD4⫹ T cells expressed ␤-gal (Fig. 2c).
The Journal of Immunology
2397
99.5% were CD44high, 94% were CD62Llow, 83% were CD43high,
99% were Ly-6Chigh, 99% were CD11ahigh, and 81% bound tetramers of Qa-1b, a reagent that binds to the surface activation
marker CD94/NKG2 (34, 45, 46). In contrast, ␤-gal⫺ CD8⫹ T
cells did not exhibit an activated phenotype (data not shown). Almost all (⬎95%) of the ␤-gal⫹CD8⫹ T cells were negative for the
early activation markers CD25, CD69, CD71, and CD43 at day 8
post-LCMV infection (Fig. 5, and data not shown). When splenocytes from immune GBC ⫻ ROSA26R mice were analyzed 30
days following infection (when stable memory CD8⫹ T cells have
formed), the ␤-gal-marked CD8⫹ T cells exhibited a memory phenotype. Greater than 80% of CD8⫹ ␤-gal⫹ T cells were CD44high,
68% were CD62Llow (32% were CD62Lhigh), 88% were Ly6Chigh, and 86% were CD94/NKG2⫹ (Fig. 6), similar to what is
FIGURE 5. ␤-gal-marked CD8⫹ T cells observed at
the peak of the antiviral immune response exhibit an
activated phenotype. Splenocytes were harvested from
GBC ⫻ ROSA26R mice infected i.p. 8 days earlier
with a single dose of 2 ⫻ 105 PFU LCMV Armstrong.
The cells were loaded with FDG to detect intracellular
␤-gal expression, and surface stained for activation
makers using Abs to CD44, CD62L, CD43, Ly-6C,
CD11a, CD69, CD25, or tetramers of Qa-1b to detect
CD94/NKG2. The histograms are gated on CD8⫹␤-gal⫹
T cells.
seen by other groups in immune C57BL/6J mice (6, 34, 46). ␤-galmarked memory CD8⫹ T cells were CD25⫺ and CD69⫺ (data not
shown).
The frequency and kinetics of ␤-gal-marked CD8⫹ T cells following LCMV infection mirror that shown by others who have
used LCMV peptide/MHC multimers to track virus-specific CD8⫹
T cells. However, the data are distinct on two important points.
One is that the overall activation of the mouse CD8⫹ T cell repertoire is greater than that predicted solely by reagents that track
Ag-specific CD8⫹ T cells. In the primary response, the frequency
of ␤-gal⫹CD8⫹ T cells is higher than the frequency (50 –70%) of
LCMV tetramer-binding CD8⫹ T cells, as reported by MuraliKrishna et al. (6). By summing together all known LCMV
epitopes, Murali-Krishna et al. (6) estimated the frequency of
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 4. ␤-gal-marked CD8⫹ T cells exhibit clonal dynamics of expansion, death, and stability in LCMV-infected GBC ⫻ ROSA26R mice. The
frequency of ␤-gal⫹CD8⫹ T cells in peripheral blood (a) and spleen (b) was monitored by FACS at various time points after LCMV infection. Total number
of ␤-gal⫹CD8⫹ T cells in spleen plotted over time is shown in c. Approximately 70% of the ␤-gal-marked cells in the spleen die after the peak of the
immune response between days 8 and 60. Both the frequency and number of marked CD8⫹ T cells reach a plateau by 50 days postinfection. The kinetics
of ␤-gal expression in peripheral blood CD8⫹ T cells of a representative LCMV-infected double-transgenic mouse is shown in d; plots are gated on CD8⫹
T cells. The same mouse was bled at various time points postinfection, and the frequency of ␤-gal⫹ CD8⫹ T cells is shown (d). Data presented are
representative of 3–15 mice at each time point. The top row of contour plots represents data from single-transgenic littermate control mice.
2398
TRACKING ACTIVATED CD8 T CELLS
FIGURE 6. ␤-gal⫹CD8⫹ T cells have a memory phenotype in immune mice. GBC ⫻ ROSA26R double-transgenic mice were infected with LCMV
Armstrong, and spleen cells were examined 30 days later. Splenocytes were hypotonically loaded with FDG; surface stained with Abs to CD44, CD62L,
CD43, Ly-6C, or tetramers of Qa-1b to detect surface CD94/NKG2; and analyzed by flow cytometry. All flow cytometry plots shown are gated on CD8⫹
T cells. Eight to ten mice were analyzed for each marker, and the experiment was repeated at least three times.
⫹
␤-gal-marked CD8 T cells exhibit LCMV-specific cytolytic
activity
Cytolytic killing of virus-infected or viral peptide-pulsed target
cells is a hallmark of virus-specific effector CD8⫹ T cells. To
determine the cytolytic capability of ␤-gal-marked CD8⫹ T cells,
CD8⫹CD4⫺␤-gal⫹ and CD8⫹CD4⫺␤-gal⫺ T cells were isolated
by cell sorting (purity ⬎95%) from the spleens of GBC ⫻
ROSA26R mice infected 8 days earlier with LCMV. The cytolytic
capability of sorted cells was tested by incubating them with 51Crlabeled target cells that were pulsed with LCMV gp33– 41,
NP396 – 404, or an irrelevant OVA-specific peptide (SIINFEKL).
Only the ␤-gal⫹CD8⫹ T cells (Fig. 8a) contained antiviral effectors capable of killing viral peptide-pulsed targets. Conversely, the
purified ␤-gal⫺CD8⫹ T cells did not contain any cytolytic effectors (Fig. 8b). These data further demonstrate that the antiviral
CTL effectors were contained within the ␤-gal-marked, activated
CD8⫹ T cell population at the peak of the immune response.
␤-gal-marked CD8⫹ T cells present in the immune phase
contain LCMV-specific memory CD8⫹ T cells
We have shown that ␤-gal⫹CD8⫹ T cells at the peak of the infection contain antiviral effectors capable of CTL activity; that
␤-gal⫺CD8⫹ T cells were incapable of cytolytic activity (Fig. 8);
and that the ␤-gal⫹ fraction also contained LCMV tetramer-binding CD8⫹ T cells (Fig. 7). Once the Cre-mediated excision has
FIGURE 7. LCMV-specific CD8⫹ T cells are contained within the
␤-gal-marked fraction. Double-transgenic GBC ⫻ ROSA26R and singly
transgenic littermate controls were infected with 2 ⫻ 105 LCMV Armstrong. a, PBLs were analyzed at various time points by flow cytometry for
␤-gal expression and binding of LCMV tetramers Dbgp33– 41 and
DbNP396 – 404. Representative flow cytometry plots from a littermate control and double-transgenic mice are shown. b, In the double-transgenic
mice, all of the tetramer⫹ CD8⫹ T cells were ␤-gal⫹. In memory doubletransgenic mice, the DbGP33– 41 and DbNP396 tetramer⫹ CD8⫹ T cells
are exclusively ␤-gal⫹CD44high (data for gp33– 41 not shown). ␤-gal⫹ and
␤-gal⫺CD8⫹ T cells were analyzed for expression of CD44 and
DbNP396 – 404 tetramer binding (b). All of the DbNP396 – 404 tetramerbinding CD8⫹ T cells were in the ␤-gal⫹, and not in the ␤-gal⫺CD44high
fraction. Gating strategy was as depicted.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
memory CD8⫹ T cells to be ⬃10%. In contrast, LCMV-immune
GBC ⫻ ROSA26R double-transgenic mice have substantially
more ␤-gal⫹CD8⫹ T cells in the memory phase (⬎day 30) (Fig.
4). Another striking difference is that the decrease or decay of
␤-gal-expressing CD8⫹ T cells following the peak of the immune
response is much more protracted and does not reach the same low
frequency, as demonstrated by tetramer studies. As the genetic
recombination that takes place in GBC ⫻ ROSA26R mice is permanent, these data directly demonstrate that activated CD8⫹ T
cells are dying more slowly or non-Ag-specific CD8⫹ T cells are
being activated by bystander activation. To better determine the
Ag specificity of ␤-gal⫹CD8⫹ T cells, we combined ␤-gal expression with LCMV tetramer staining.
Costains for ␤-gal and MHC tetramers were performed on
CD8⫹ T cells during the effector and memory phases of an antiLCMV immune response. We were interested to know whether the
␤-gal-marked CD8⫹ T cells contained the antiviral effectors.
GBC ⫻ ROSA26R mice are H-2b, and LCMV-specific CD8⫹ T
cells can be identified by MHC multimers using viral peptides.
When MHC tetramers were used to stain cells in GBC ⫻
ROSA26R mice that had been LCMV infected, we observed that
all tetramer-binding CD8⫹ T cells were all ␤-gal⫹. These analyses
were performed using LCMV Dbgp33– 41 and DbNP396 – 404 tetramers to track the antiviral response using two different LCMV
immunodominant peptides (Fig. 7a). Tetramer binding and ␤-gal
expression were compared in CD8⫹ T cells at days 8, 15, 30, and
60 post-CMV infection. At all time points analyzed, during both
the effector and memory phases, LCMV peptide/MHC tetramer
binding correlated with ␤-gal expression. In the memory phase, we
wished to determine whether the tetramer-binding LCMV-specific
CD8⫹ T cells resided in the ␤-gal⫹ or the ␤-gal⫺ fraction. ␤-gal⫹
and ␤-gal⫺CD8⫹ T cells were analyzed for expression of CD44
and tetramer binding (Fig. 7b). All of the LCMV-specific CD8⫹ T
cells were found in the ␤-gal⫹CD44high fraction, and not in the
␤-gal⫺ fraction (Fig. 7b). We also observed that 94% of the ␤-galmarked CD8⫹ T cells were CD44high; however, ⬃20% of the
␤-gal⫺CD8⫹ T cells were CD44high, indicating that the rules that
govern CD44 expression and ␤-gal marking are not the same (Fig.
7b). Overall, these data demonstrate that CD8⫹ T cells responding
to LCMV are contained within the ␤-gal-marked population.
The Journal of Immunology
2399
infection and did not inhibit the endogenous CD8⫹ T cells from
responding to LCMV infection (Fig. 9a). We then assessed the
ability of the transferred ␤-gal⫹ and ␤-gal⫺CD8⫹ T cells to control viral replication in the recipient mice. For challenge experiments, we used the virulent, clone 13 strain of LCMV. Recipients
that received ␤-gal⫹CD8⫹ T cells were able to limit the replication
of LCMV, and three of five mice had serum viral titers below the
limit of detection (Fig. 9b). In contrast, mice that received
␤-gal⫺CD8⫹ T cells had high viral titers, typically ⬎106 PFU
ml⫺1 serum (Fig. 9b). These mice had comparable serum viral
titers to naive mice that had been infected with LCMV clone 13
(data not shown). Taken together, these data clearly demonstrate
that antiviral memory CD8⫹ T cells were contained within the
␤-gal-marked CD8⫹ T cells.
Discussion
occurred, all lineal descendants of a recombined T cell will propagate the recombination event and continue to express ␤-gal. Consequently, we hypothesized that the ␤-gal-marked T cells would
contain LCMV-specific memory CD8⫹ T cells. A primary feature
of memory CD8⫹ T cells is their ability to rapidly expand following re-exposure to their cognate Ag. We used adoptive transfer to
determine the fraction (CD8⫹␤-gal⫹ or CD8⫹␤-gal⫺) in which
the LCMV-specific memory CD8⫹ T cells exist. We isolated, by
cell sorting, purified populations of CD8⫹␤-gal⫹CD4⫺ and
CD8⫹␤-gal⫺CD4⫺ T cells from the spleens of GBC ⫻ ROSA26R
mice that had been infected ⬎45 days earlier with LCMV Armstrong. The two cell populations were adoptively transferred into
naive recipient congenic Thy-1.1 mice. GBC ⫻ ROSA26R mice
express Thy-1.2 (CD90.2), and therefore donor T cells can be easily identified in recipient Thy-1.1 (CD90.1) mice. Following adoptive transfer, recipient mice were challenged with virulent LCMV
clone 13, and the capacity of the transferred cells to respond was
measured by their ability to proliferate and expand in response to
LCMV (Fig. 9a). The transferred ␤-gal⫹CD8⫹ T cells responded
to the LCMV infection, and they represented ⬃14% of the recipient mouse’s spleen 7 days postchallenge, and it constituted more
than half of the recipient mouse’s total CD8⫹ T cells. These
␤-gal⫹CD8⫹ T cells were able to bind LCMV tetramers,
Dbgp33– 41 and DbNP396 – 404, and continued to express ␤-gal
(Fig. 9a). The presence of the transferred ␤-gal⫹ memory CD8⫹ T
cells dampened the endogenous response to LCMV (Fig. 9a), a
phenomenon observed by others as well (10). In contrast, transferred ␤-gal⫺CD8⫹ T cells failed to expand in response to LCMV
FIGURE 9. ␤-gal-marked CD8⫹ T cells from immune mice contain
memory cells capable of proliferation and conferring protection upon viral
challenge. CD4⫺CD8⫹␤-gal⫹ and CD4⫺CD8⫹␤-gal⫺ T cells were sorted
from Thy-1.2⫹ immune GBC ⫻ ROSA26R mice, adoptively transferred
into congenic Thy-1.1⫹ recipients, and challenged with LCMV clone 13.
Seven days later, the recipient mice were sacrificed, and their splenocytes
were analyzed for expansion and tetramer binding. a, Mice that received
␤-gal⫺CD8⫹ T cells (top panel) showed proliferation of host NP396- and
gp33-specific recipient (Thy-1.1⫹) CD8⫹ T cells. No expansion of the
transferred Thy-1.2⫹ ␤-gal⫺CD8⫹ T cells was observed. In contrast, mice
that received Thy-1.2⫹ ␤-gal⫹CD8⫹ T cells underwent proliferation and
accounted for almost all of the NP396- and gp33-specific CD8⫹ T cells
(bottom panel). Expansion of the transferred ␤-gal⫹CD8⫹ T cells severely
dampened the recipient (Thy-1.1⫹ or Thy-1.2⫺) CD8⫹ T cell responses. b,
Adoptive transfer of ␤-gal⫹, and not ␤-gal⫺CD8⫹ T cells reduced viral
titers in recipient mouse sera. Seven days post-LCMV clone 13 challenge,
the sera of recipient mice were assayed for LCMV by plaque assay. The
serum viral titers of mice that received ␤-gal⫹ or ␤-gal⫺CD8⫹ T cells are
shown. Each point represents an individual Thy-1.1⫹ recipient.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 8. Antiviral cytotoxic effector cells are present exclusively
within the ␤-gal⫹CD8⫹ T cell fraction. GBC ⫻ ROSA26R mice were
infected with 2 ⫻ 105 PFU LCMV Armstrong, and splenocytes were harvested 8 days later. CD8⫹␤-gal⫹ and CD8⫹␤-gal⫺ T cells were purified by
high speed cell sorting, and the ability of these two fractions to kill LCMV
peptide-pulsed target cells was tested in a standard 51Cr release assay.
MHC-matched target cells were pulsed with immunodominant LCMV
(gp33– 41 or NP396 – 404) peptides or an irrelevant OVA control peptide
(SIINFEKL) and loaded with 51Cr. Sorted CD4⫺CD8⫹␤-gal⫹ (a) or
CD4⫺CD8⫹␤-gal⫺ (b) effectors were incubated with targets at various E:T
ratios for 6 h, and the amount of released 51Cr was measured. One percent
Triton X-100 and targets only (no effectors) were used to ascertain maximal and spontaneous 51Cr release, respectively. Percent specific lysis was
calculated as (experimental – spontaneous)/ (max – spontaneous).
In this work, we describe a transgenic mouse model that marks not
just a subset, but all activated CD8⫹ T lymphocytes with ␤-gal.
These mice are healthy and otherwise normal, and are able to clear
acute LCMV infection with a kinetics and magnitude indistinguishable from those of C57BL/6 mice. Interestingly, ␤-gal is expressed in the CD8⫹ T cells of naive double-transgenic mice, albeit at a low frequency and a relatively low number. This is not
unexpected, as there is a basal level of CD8⫹ T cell activation to
endogenous Ags such as food or gut flora, as described by others
who have shown that in naive mice, 8 –12% of splenic or PBL
CD8⫹ T cells are CD44high (47–50). This also raises the issue as
2400
have tried alternative ␤-gal substrates, but none of them have performed as well as FDG. In addition, FDG staining is cumbersome,
and intracellular cytokine staining cannot be done because permeabilization of the cells leads to loss of hypotonically loaded FDG.
Wurtz et al. (54) have recently developed an elegant CD2 promoter-driven T cell-specific Cre reporter transgenic line, pLU, in
which T cells before Cre recombination express human CD2 and
after recombination express Thy-1.1 on the cell surface. Generating GBC ⫻ pLU double-transgenic mice will enable CFSE labeling and cell division studies, and these breedings are currently in
progress. In addition, it will be possible to track Thy-1.1⫹-tagged
CD8⫹ T cells using in situ immunofluorescence. Analysis of
spleen sections from naive and LCMV-infected doubly transgenic
pLU ⫻ GBC mice will allow tracking of both cell-cell interactions
and spacial dynamics of activated CD8⫹ T cells in situ.
Powerful transgenic strategies such as the one presented in this
work will be beneficial in situations in which defined antigenic
peptides are unavailable and therefore MHC multimers have not
yet been created. Because the strength of the model is that it can
irreversibly tag all activated CD8⫹ T cells, we anticipate that this
mouse model system will be useful in monitoring the population
dynamics of activated T lymphocytes in infection, transplantation,
and autoimmunity models. In addition, the GBC transgenic mice
will also be useful in breeding to the various conditional gene
knockout mice to dissect the mechanisms involved in lymphocyte
activation, expansion, and memory generation.
Acknowledgments
We are grateful to Craig P. Chappell, Melinda Mackereth, and Dr. Guido
Silvestri for critically reading the manuscript and helpful suggestions,
Karin Smith and Lorra Miller for mouse colony management, Robert
Karaffa and Michael Hulsey for cell sorting, and Dr. Rafi Ahmed for encouragement and guidance.
References
1. Welsh, R. M., M. Y. Lin, B. L. Lohman, S. M. Varga, C. C. Zarozinski, and
L. K. Selin. 1997. ␣␤ and ␥␦ T-cell networks and their roles in natural resistance
to viral infections. Immunol. Rev. 159:79.
2. Kundig, T. M., M. F. Bachmann, P. S. Ohashi, H. Pircher, H. Hengartner, and
R. M. Zinkernagel. 1996. On T cell memory: arguments for antigen dependence.
Immunol. Rev. 150:63.
3. Tough, D. F., and J. Sprent. 1995. Life span of lymphocytes. Immunol. Res. 14:1.
4. Swain, S. L., M. Croft, C. Dubey, L. Haynes, P. Rogers, X. Zhang, and
L. M. Bradley. 1996. From naive to memory T cells. Immunol. Rev. 150:143.
5. Doherty, P. C., D. J. Topham, and R. A. Tripp. 1996. Establishment and persistence of virus-specific CD4⫹ and CD8⫹ T cell memory. Immunol. Rev. 150:23.
6. Murali-Krishna, K., J. D. Altman, M. Suresh, D. J. Sourdive, A. J. Zajac,
J. D. Miller, J. Slansky, and R. Ahmed. 1998. Counting antigen-specific CD8 T
cells: a reevaluation of bystander activation during viral infection. Immunity
8:177.
7. Bevan, M. J., and A. W. Goldrath. 2000. T-cell memory: you must remember
this. Curr. Biol. 10:R338.
8. Butz, E., and M. J. Bevan. 1998. Dynamics of the CD8⫹ T cell response during
acute LCMV infection. Adv. Exp. Med. Biol. 452:111.
9. Butz, E. A., and M. J. Bevan. 1998. Massive expansion of antigen-specific CD8⫹
T cells during an acute virus infection. Immunity 8:167.
10. Blattman, J. N., R. Antia, D. J. Sourdive, X. Wang, S. M. Kaech,
K. Murali-Krishna, J. D. Altman, and R. Ahmed. 2002. Estimating the precursor
frequency of naive antigen-specific CD8 T cells. J. Exp. Med. 195:657.
11. Badovinac, V. P., B. B. Porter, and J. T. Harty. 2002. Programmed contraction of
CD8⫹ T cells after infection. Nat. Immun. 3:619.
12. Kaech, S. M., E. J. Wherry, and R. Ahmed. 2002. Effector and memory T-cell
differentiation: implications for vaccine development. Nat. Rev. Immun. 2:251.
13. Kerksiek, K. M., and E. G. Pamer. 1999. T cell responses to bacterial infection.
Curr. Opin. Immunol. 11:400.
14. Busch, D. H., and E. G. Pamer. 1999. T lymphocyte dynamics during Listeria
monocytogenes infection. Immunol. Lett. 65:93.
15. Altman, J. D., P. A. Moss, P. J. Goulder, D. H. Barouch, M. G. McHeyzer-Williams,
J. I. Bell, A. J. McMichael, and M. M. Davis. 1996. Phenotypic analysis of antigenspecific T lymphocytes. Science 274:94.
16. Lee, P. P., C. Yee, P. A. Savage, L. Fong, D. Brockstedt, J. S. Weber, D. Johnson,
S. Swetter, J. Thompson, P. D. Greenberg, et al. 1999. Characterization of circulating T cells specific for tumor-associated antigens in melanoma patients. Nat.
Med. 5:677.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
to whether ␤-gal is just a surrogate marker for CD44. In the periphery of naive animals, ␤-gal does not appear to be a surrogate
marker for CD44 because only 50% of the ␤-gal⫹CD8⫹ T cells are
CD44high (Fig. 2). Following LCMV infection, there is considerable overlap between the CD44high and ␤-gal⫹CD8⫹ T cell populations; 94% of the ␤-gal⫹CD8⫹ T cells are CD44high. Interestingly, the tetramer⫹ LCMV-specific CD8⫹ T cells are in the
␤-gal⫹CD44⫹ fraction, and not in the ␤-gal⫺CD44⫹ fraction (Fig.
7b). Thus, although there is considerable overlap, the rules that
govern CD44 and ␤-gal expression are indeed not the same.
During the acute phase (day 8) of the LCMV response, there
is a good correlation between the numbers of tetramer⫹ and
␤-gal⫹CD8⫹ T cells. In the memory phase (⬎30 days), the best
estimates using MHC tetramers and cytokine elaboration assays
for all the known immunodominant and subdominant LCMV
epitopes place the frequency of LMCV memory CD8⫹ T cells at
5–10% (6, 9). However, in double-transgenic mice, 100 days postLCMV infection, ⬃40% of CD8⫹ T cells express CD44 and ␤-gal
(Fig. 4). Thus, there is a substantial gap between the number of
memory CD8⫹ T cells accounted by tetramer staining and the
memory phenotype CD8⫹ T cells, as determined by CD44 or ␤-gal
expression. Our data show that in the memory phase, apoptosis of
the ␤-gal⫹CD8⫹ T cells is more reduced when compared with
primary effectors visualized by the LCMV tetramers. Recently,
Grayson et al. (51) demonstrated that there is a differential sensitivity of naive and memory CD8⫹ T cells to apoptosis in vivo.
They showed that upon restimulation, memory CD8⫹ T cells undergo rapid expansion, and the secondary effector T cells generated from these memory CD8⫹ T cells underwent much less contraction or apoptosis as compared with effectors generated from
naive CD8⫹ T cells. Similarly, Harty and colleagues (11) have
shown that the contraction phase of secondary effector CD8⫹ T
cells is slower and reduced in magnitude than that of primary effectors. Thus, the slowed contraction of the ␤-gal⫹CD8⫹ T cells in
the memory phase could be explained, as follows: LCMV infection
leads to expansion of LCMV-specific ␤-gal⫹ primary effector
CD8⫹ T cells. The infection presumably also induces expansion of
the pre-existing (10%) memory-phenotype ␤-gal⫹CD44⫹CD8⫹ T
cells through cytokine-mediated bystander activation (52, 53). As
the virus is cleared, the LCMV-specific ␤-gal⫹CD8⫹ T cells undergo apoptosis and contract, but the secondary effectors generated
from the pre-existing ␤-gal⫹CD44⫹ memory CD8⫹ T cells contract at a substantially slower rate. Current studies aim to directly
test this hypothesis by elucidating the cellular source and specificity of these apparent bystander CD8⫹ T cells by introducing
LCMV gp33-specific TCR transgene into the ROSA26R ⫻
GBC mice.
The current work describes a novel transgenic mouse that permanently marks activated CD8⫹ T cells. We have shown that
␤-gal expression can be induced in double-transgenic CD8⫹ T
cells by TCR-dependent and TCR-independent stimuli. Following
LCMV infection, large numbers of CD8⫹ T cells express ␤-gal.
The overall kinetics of the ␤-gal⫹CD8⫹ T cells show remarkable
similarity to those seen using MHC tetramers. The contraction
phase is delayed and reduced, and a greater number of ␤-gal⫹CD8⫹
T cells are observed in the memory phase when compared with
tetramer studies (Fig. 4) (6). Using a variety of activation markers and
functional assays (cytotoxic killing, expansion upon rechallenge,
limiting viral replication), we have shown that the marked fraction of
CD8⫹ T cells contains all of the antiviral effector and memory CD8⫹
T cells.
A limitation of using ␤-gal as the reporter is that cell divisions
in CD8⫹ T cells cannot be tracked using CFSE because the ␤-gal
substrate, FDG, and CFSE fluoresce in the same wavelength. We
TRACKING ACTIVATED CD8 T CELLS
The Journal of Immunology
34. Miller, J. D., M. Peters, A. E. Oran, G. W. Beresford, L. Harrington, J. M. Boss,
and J. D. Altman. 2002. CD94/NKG2 expression does not inhibit cytotoxic function of lymphocytic choriomeningitis virus-specific CD8⫹ T cells. J. Immunol.
169:693.
35. Ahmed, R., A. Salmi, L. D. Butler, J. M. Chiller, and M. B. Oldstone. 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.
36. Hanson, R. D., G. M. Sclar, O. Kanagawa, and T. J. Ley. 1991. The 5⬘-flanking
region of the human CGL-1/granzyme B gene targets expression of a reporter
gene to activated T-lymphocytes in transgenic mice. J. Biol. Chem. 266:24433.
37. Tough, D. F., and J. Sprent. 1995. Life span of naive and memory T cells. Stem
Cells 13:242.
38. Tough, D. F., S. Sun, X. Zhang, and J. Sprent. 1999. Stimulation of naive and
memory T cells by cytokines. Immunol. Rev. 170:39.
39. Tough, D. F., and J. Sprent. 1994. Turnover of naive- and memory-phenotype T
cells. J. Exp. Med. 179:1127.
40. Murali-Krishna, K., and R. Ahmed. 2000. Cutting edge: naive T cells masquerading as memory cells. J. Immunol. 165:1733.
41. Alexopoulou, L., A. C. Holt, R. Medzhitov, and R. A. Flavell. 2001. Recognition
of double-stranded RNA and activation of NF-␬B by Toll-like receptor 3. Nature
413:732.
42. Salazar-Mather, T. P., R. Ishikawa, and C. A. Biron. 1996. NK cell trafficking and
cytokine expression in splenic compartments after IFN induction and viral infection. J. Immunol. 157:3054.
43. Oldstone, M. B., R. Ahmed, J. Byrne, M. J. Buchmeier, Y. Riviere, and
P. Southern. 1985. Virus and immune responses: lymphocytic choriomeningitis
virus as a prototype model of viral pathogenesis. Br. Med. Bull. 41:70.
44. Suresh, M., J. K. Whitmire, L. E. Harrington, C. P. Larsen, T. C. Pearson,
J. D. Altman, and R. Ahmed. 2001. Role of CD28 –B7 interactions in generation
and maintenance of CD8 T cell memory. J. Immunol. 167:5565.
45. Moser, J. M., A. M. Byers, and A. E. Lukacher. 2002. NK cell receptors in
antiviral immunity. Curr. Opin. Immunol. 14:509.
46. Moser, J. M., J. Gibbs, P. E. Jensen, and A. E. Lukacher. 2002. CD94-NKG2A
receptors regulate antiviral CD8⫹ T cell responses. Nat. Immun. 3:189.
47. Bruno, L., H. von Boehmer, and J. Kirberg. 1996. Cell division in the compartment of naive and memory T lymphocytes. Eur. J. Immunol. 26:3179.
48. Sprent, J. 1993. Life spans of naive, memory and effector lymphocytes. Curr.
Opin. Immunol. 5:433.
49. Surh, C. D., and J. Sprent. 2002. Regulation of naive and memory T-cell homeostasis. Microbes Infect. 4:51.
50. Tough, D. F., and J. Sprent. 1994. Turnover of naive- and memory-phenotype T
cells. J. Exp. Med. 179:1127.
51. Grayson, J. M., L. E. Harrington, J. G. Lanier, E. J. Wherry, and R. Ahmed. 2002.
Differential sensitivity of naive and memory CD8⫹ T cells to apoptosis in vivo.
J. Immunol. 169:3760.
52. Tough, D. F., and J. Sprent. 1996. Viruses and T cell turnover: evidence for
bystander proliferation. Immunol. Rev. 150:129.
53. Sprent, J., X. Zhang, S. Sun, and D. Tough. 1999. T-cell turnover in vivo and the
role of cytokines. Immunol. Lett. 65:21.
54. Wurtz, O., M. Pophillat, A. M. Schmitt-Verhulst, and S. Guerder. 2002. A novel
reporter strain to follow Cre-mediated recombination in T and NK cells. Genesis
32:287.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
17. Yee, C., P. A. Savage, P. P. Lee, M. M. Davis, and P. D. Greenberg. 1999.
Isolation of high avidity melanoma-reactive CTL from heterogeneous populations using peptide-MHC tetramers. J. Immunol. 162:2227.
18. He, X. S., B. Rehermann, J. Boisvert, J. Mumm, H. T. Maecker, M. Roederer,
T. L. Wright, V. C. Maino, M. M. Davis, and H. B. Greenberg. 2001. Direct
functional analysis of epitope-specific CD8⫹ T cells in peripheral blood. Viral
Immunol. 14:59.
19. Yee, C., S. R. Riddell, and P. D. Greenberg. 2001. In vivo tracking of tumorspecific T cells. Curr. Opin. Immunol. 13:141.
20. Appay, V., and S. L. Rowland-Jones. 2002. The assessment of antigen-specific
CD8⫹ T cells through the combination of MHC class I tetramer and intracellular
staining. J. Immunol. Methods 268:9.
21. Appay, V., P. R. Dunbar, M. Callan, P. Klenerman, G. M. Gillespie, L. Papagno,
G. S. Ogg, A. King, F. Lechner, C. A. Spina, et al. 2002. Memory CD8⫹ T cells
vary in differentiation phenotype in different persistent virus infections. Nat. Med.
8:379.
22. Blohm, U., E. Roth, K. Brommer, T. Dumrese, F. M. Rosenthal, and H. Pircher.
2002. Lack of effector cell function and altered tetramer binding of tumor-infiltrating lymphocytes. J. Immunol. 169:5522.
23. Spencer, J. V., and T. J. Braciale. 2000. Incomplete CD8⫹ T lymphocyte differentiation as a mechanism for subdominant cytotoxic T lymphocyte responses to
a viral antigen. J. Exp. Med. 191:1687.
24. Fahmy, T. M., J. G. Bieler, M. Edidin, and J. P. Schneck. 2001. Increased TCR
avidity after T cell activation: a mechanism for sensing low-density antigen.
Immunity 14:135.
25. Zajac, A. J., J. N. Blattman, K. Murali-Krishna, D. J. Sourdive, M. Suresh,
J. D. Altman, and R. Ahmed. 1998. Viral immune evasion due to persistence of
activated T cells without effector function. J. Exp. Med. 188:2205.
26. Appay, V., D. F. Nixon, S. M. Donahoe, G. M. Gillespie, T. Dong, A. King,
G. S. Ogg, H. M. Spiegel, C. Conlon, C. A. Spina, et al. 2000. HIV-specific
CD8⫹ T cells produce antiviral cytokines but are impaired in cytolytic function.
J. Exp. Med. 192:63.
27. Garba, M. L., C. D. Pilcher, A. L. Bingham, J. Eron, and J. A. Frelinger. 2002.
HIV antigens can induce TGF-␤1-producing immunoregulatory CD8⫹ T cells.
J. Immunol. 168:2247.
28. Kostense, S., K. Vandenberghe, J. Joling, D. Van Baarle, N. Nanlohy,
E. Manting, and F. Miedema. 2002. Persistent numbers of tetramer⫹ CD8⫹ T
cells, but loss of interferon-␥⫹ HIV-specific T cells during progression to AIDS.
Blood 99:2505.
29. Jacob, J., and D. Baltimore. 1999. Modelling T-cell memory by genetic marking
of memory T cells in vivo. Nature 399:593.
30. Soriano, P. 1999. Generalized lacZ expression with the ROSA26 Cre reporter
strain. Nat. Genet. 21:70.
31. Zambrowicz, B. P., A. Imamoto, S. Fiering, L. A. Herzenberg, W. G. Kerr, and
P. Soriano. 1997. Disruption of overlapping transcripts in the ROSA ␤ geo 26
gene trap strain leads to widespread expression of ␤-galactosidase in mouse embryos and hematopoietic cells. Proc. Natl. Acad. Sci. USA 94:3789.
32. Mao, X., Y. Fujiwara, and S. H. Orkin. 1999. Improved reporter strain for monitoring Cre recombinase-mediated DNA excisions in mice. Proc. Natl. Acad. Sci.
USA 96:5037.
33. Ahmed, R., A. Salmi, L. D. Butler, J. M. Chiller, and M. B. Oldstone. 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.
2401