Rapid Demethylation of the IFN-γ Gene
Occurs in Memory but Not Naive CD8 T
Cells
This information is current as
of June 18, 2017.
Ellen N. Kersh, David R. Fitzpatrick, Kaja Murali-Krishna,
John Shires, Samuel H. Speck, Jeremy M. Boss and Rafi
Ahmed
J Immunol 2006; 176:4083-4093; ;
doi: 10.4049/jimmunol.176.7.4083
http://www.jimmunol.org/content/176/7/4083
Subscription
Permissions
Email Alerts
This article cites 52 articles, 21 of which you can access for free at:
http://www.jimmunol.org/content/176/7/4083.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 © 2006 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 June 18, 2017
References
The Journal of Immunology
Rapid Demethylation of the IFN-␥ Gene Occurs in Memory
but Not Naive CD8 T Cells1
Ellen N. Kersh,* David R. Fitzpatrick,† Kaja Murali-Krishna,‡ John Shires,* Samuel H. Speck,*
Jeremy M. Boss,* and Rafi Ahmed2*
M
emory CD8 T cells protect from re-exposure to the
same Ag by mounting a faster and more vigorous response than naive cells. This is not only due to the
increased number of Ag-specific memory T cells, but also is due to
important functional differences between naive and memory cells
on a per cell basis (1– 4). Memory CD8 T cells have an increased
ability to produce cytokines and to start proliferation, are more
cytotoxic, express distinct surface markers, and are located at different anatomical sites, all allowing for a functionally more potent
immune response by memory CD8 T cells. The IFN-␥ cytokine,
which is critical for the function of CD8 T cells (5, 6), shows
strong differential expression in naive and memory CD8 T cells.
Naive CD8 T cells require several rounds of proliferation before
they can produce large amounts of IFN-␥ (7–11). Memory CD8 T
cells, however, quickly and effectively produce IFN-␥ upon restimulation (12). This functional change is stable and passed on to
daughter memory cells during homeostatic proliferation, and is
maintained in the absence of Ag (13).
These observations raise the question of what intracellular
mechanisms enable the enhanced functional properties of memory
CD8 T cells such as increased IFN-␥ expression. More efficient
*Emory Vaccine Center and Department of Microbiology and Immunology, Emory
University School of Medicine, Atlanta, GA 30322; †Amgen, Seattle, WA 98119; and
‡
Department of Immunology, University of Washington, Seattle, WA 98195
Received for publication October 14, 2005. Accepted for publication January
19, 2006.
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 National Institutes of Health Grant AI30048 (to R.A.),
Primary Caretaker Technical Assistance Supplement 3 U19 AI057266-02S1 (to R.A.
and E.N.K.), and National Cancer Institute Grant RO1 CA096810 (to J.M.B.).
2
Address correspondence and reprint requests to Dr. Rafi Ahmed, Emory Vaccine
Center, Rollins Research Building, Room G211, 1510 Clifton Road, Atlanta, GA
30322. E-mail address: [email protected]
Copyright © 2006 by The American Association of Immunologists, Inc.
TCR signal transduction (14), increased usage of costimulatory
molecules (15), and differential gene expression (16 –19) are all
mechanisms that contribute to memory CD8 T cell regulation. Differential gene expression can be achieved through the action of
selective transcription factors, and through epigenetic regulatory
mechanisms like histone modifications and gene demethylation. In
this study, we have examined the differential regulation of gene
methylation of the IFN-␥ promoter in naive vs memory CD8 T
cells.
DNA methylation negatively affects gene expression (20). In
mammals, DNA methylation occurs at cytosines within CpG dinucleotide sequences (20), and often coincides with repressed or inactive chromatin. Cytosine methylation is an important regulatory
event as evidenced by the fact that targeted disruption of all known
DNA methyltransferases (Dnmts),3 enzymes that add methyl
groups to CpGs, causes developmental defects and death (21–23).
Specific abrogation of the maintenance methyl-transferase Dnmt1
in T cells leads to ectopic cytokine expression as well as to impaired proliferative capacity (24, 25). Cell differentiation is often
associated with the demethylation of genes (7, 8, 26, 27). It is
generally assumed that DNA demethylation is due to a failure of
DNA methyltransferases to methylate the newly replicated DNA
strand during cell division. However, it has also been postulated
that a currently unknown enzymatic factor actively directs the demethylation process independent of cell division (28).
IFN-␥ gene expression is regulated by DNA methylation, although other mechanisms such as specific transcription factors and
modulation of chromatin structure also contribute to its transcriptional regulation (9, 29 –34). In naive CD8 T cells, the IFN-␥ gene
is methylated at several of the CpG sites in the promoter, and this
prevents gene expression (35). Upon activation and proliferation
3
Abbreviations used in this paper: Dnmt, DNA methyltransferase; LCMV, lymphocytic choriomeningitis virus; MFI, mean fluorescence intensity.
0022-1767/06/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
DNA methylation is an epigenetic mechanism of gene regulation. We have determined that specific modifications in DNA methylation at the IFN-␥ locus occur during memory CD8 T cell differentiation in vivo. Expression of the antiviral cytokine IFN-␥ in
CD8 T cells is highly developmental stage specific. Most naive cells must divide before they express IFN-␥, while memory cells
vigorously express IFN-␥ before cell division. Ag-specific CD8 T cells were obtained during viral infection of mice and examined
directly ex vivo. Naive cells had an IFN-␥ locus with extensive methylation at three specific CpG sites. An inhibitor of methylation
increased the amount of IFN-␥ in naive cells, indicating that methylation contributes to the slow and meager production of IFN-␥.
Effectors were unmethylated and produced large amounts of IFN-␥. Interestingly, while memory cells were also able to produce
large amounts of IFN-␥, the gene was partially methylated at the three CpG sites. Within 5 h of antigenic stimulation, however,
the gene was rapidly demethylated in memory cells. This was independent of DNA synthesis and cell division, suggesting a yet
unidentified demethylase. Rapid demethylation of the IFN-␥ promoter by an enzymatic factor only in memory cells would be a
novel mechanism of differential gene regulation. This differentiation stage-specific mechanism reflects a basic immunologic principle: naive cells need to expand before becoming an effective defense factor, whereas memory cells with already increased
precursor frequency can rapidly mount effector functions to eliminate reinfecting pathogens in a strictly Ag-dependent
fashion. The Journal of Immunology, 2006, 176: 4083– 4093.
4084
RAPID IFN-␥ GENE DEMETHYLATION IN MEMORY CD8 T CELLS
Materials and Methods
IFN-␥ mRNA determination
IFN-␥ mRNA levels were determined by real-time RT-PCR using an iCycler iQ (Bio-Rad). RNA was prepared using the RNeasy kit (Qiagen). After
DNaseI treatment, RNA was reverse-transcribed with oligo dT using the
Superscript First-Strand Synthesis System (Invitrogen Life Technologies).
For real-time RT-PCR of IFN-␥, we amplified a 200-bp region spanning
exons 2 and 3 using primers CTT CTT GGA TAT CTG GAG GAA CTG
GCA AAA and CTC AAA CTT GGC AAT ACT CAT GAA TGC ATC
(37). To further ensure specific amplification of mRNA only, we analyzed
samples without added reverse transcriptase in every experiment. Realtime RT-PCR was done using the SYBR Green kit with Hot Start Taq
(Bio-Rad) to further eliminate nonspecific products. mRNA levels were
standardized by determining the expression of the HPRT gene for every
sample with primers GAT TCA ACT TGC GCT CAT CTT AGG C and
GTT GGA TAC AGG CCA GAC TTT GTT G.
Bisulfite modification, PCR, and sequencing
The methylation status of CpG sequences was determined by “bisulfite
sequencing”. Bisulfite treatment deaminates all cytosines to uracil unless
protected by methylation. Uracils are read as thymidines during PCR, and
sequencing then allows identification of methylated cytosines. Genomic
DNA was prepared as described (35). DNA was denatured and modified
with sodium metabisulfite, purified, desulfonated, and amplified in seminested PCR using primers: IFN-␥-1 (GGT GTG AAG TAA AAG TGT
TTT TAG AGA ATT TTA T) and IFN-␥-4 (CAA TAA CAA CCA AAA
ACA ACC ATA AAA AAA AAC T), then IFN-␥-1 and IFN-␥-3 (CCA
TAA AAA AAA ACT ACA AAA CCA AAA TAC AAT A). For the
noncoding IFN-␥ strand, primers were IFN-␥-7 (GTT AGA AAT AGT
TAT GAG GAA GAG TTG TAA AGT T) and IFN-␥-10 (ACA AAA ACT
CCC TAT ACT ATA CTC TAT AAA TAA A), then IFN-␥-7 and IFN-␥-9
(ACA ATT TCC AAC CCC CAC CCC AAA TAA TAT AAA A). For
H19, we did nested PCR with primers H19-1 (GAT TAG ATA GTA TTG
AGT TTG TTT GGA GT) and H19-4 (CCT AAA ATA CTA AAC TTA
AAT AAC CCA CAA), then H19-2 (GAG AAA ATA GTT ATT GTT
TAT AGT TTT), and H19-3 (ACC ATT TAT AAA TTC CAA TAC CAA
AAA TAA). For IL-2, sites ⫺380 to ⫹64 were examined as described in
Ref. 28. Touchdown PCR with decreasing temperatures from 59 to 52°C
and a final annealing temperature of 50°C for 30 cycles was used. To
prevent contamination, all utensils were UV-irradiated and control reactions showed no amplification of nonspecific reactions. PCR products were
separated on agarose gels, excised, and cloned into the pGemT Vector
System I (Promega). This allowed the selection of bacterial colonies with
a disrupted lacZ reading frame. DNA from individual bacterial colonies
was then amplified by PCR using primers T7 and SP6, and sequenced at
Agencourt Bioscience by automated sequencing with primers SP6 and T7.
Mice and viral infection
P14 TCR transgenic mice were obtained from The Jackson Laboratory and
backcrossed to C57BL/6 (B6) for at least 10 generations. Female B6 mice
were from National Cancer Institute. We generated effector and memory
P14 cells in normal B6 mice that had adoptively received 2.5 ⫻ 105 P14,
Thy1.1⫹ (CD90.1⫹) splenocytes, followed by infection with 2 ⫻ 105 PFU
LCMV Armstrong i.p. This was done to reduce the precursor frequency of
transgenic P14 cells and to prevent premature elimination of LCMV before
all P14 cells were efficiently activated. The technique also allows identification of P14 cells with anti-Thy1.1 Abs, as the recipient B6 mouse is
Thy1.2⫹. All mice were maintained according to Emory University’s Institutional Review Board.
FACS and FACS sorting
For cell purification by cell sorting, T cells from the spleen or lymph node
were stained with anti-CD8␣ and anti-Thy1.1 (clones 53-6.7 and His 51,
respectively; BD Pharmingen) to avoid direct stimulation of the TCR, and
to anti-CD44 for activation status, and sorted on a FACSVantage to a
minimal purity of 96%. For enrichment of memory P14 cells, we depleted
splenocytes with anti-Thy1.2-coated columns (Miltenyi Biotec) according
to the instructions of the manufacturer.
Cell culture, intracellular cytokine staining
Intracellular staining for IFN-␥ was previously described (38), as was the
use of 5-azacytidine (35). For thymidine incorporation, 106 splenocytes
were cultured with 1 ␮M peptide gp33 (KAVYNFATM) and [3H]thymidine was added at 1 ␮Ci/well. Inhibitors were cycloheximide, mizoribine,
mitomycin c, actinomycin d, cyclosporin a, and aminopterin (all from Sigma-Aldrich), all used at 50 ␮g/ml.
Results
Different IFN-␥ mRNA levels in naive and memory P14 CD8 T
cells
Several genes are clearly differentially expressed in naive and
memory cells (16 –18). The cytokine IFN-␥ is a prime example.
Naive peripheral CD8 T cells are slow and inefficient at producing
IFN-␥ in response to stimulation. Memory CD8 T cells, however,
rapidly produce large amounts of IFN-␥ upon stimulation. This
finding is recapitulated in Fig. 1A. To analyze IFN-␥ expression
control, the current system used CD8 T cells from P14 TCR-transgenic mice, which are specific for epitope gp33– 41 of the glycoprotein of lymphocytic choriomeningitis virus (LCMV) (39), and
contained the allelic surface marker Thy1.1 (CD90.1). P14 mice
have previously been used to document changes in gene expression during memory CD8 T cell differentiation (16). In naive mice
not exposed to LCMV, only 5% of P14 cells can produce intracellular IFN-␥ after peptide stimulation in vitro. When effector
cells were harvested on day 7 post-LCMV infection, 95% of P14
cells produced IFN-␥ after stimulation in vitro (Fig. 1A). Importantly, more IFN-␥ was produced, as the mean fluorescence intensity (MFI) had increased ⬃10-fold compared with naive cells.
Memory P14 cells were harvested at least 30 days postinfection,
when cellular phenotype was CD127high, CD44high, bcl-2high, and
functional, rapid recall responses were observed. Nearly all memory P14 cells (91%) produced large amounts of IFN-␥. This distinct IFN-␥ production profile exists over a large range of Ag concentration (Fig. 1B).
The differences in IFN-␥ protein production were reflected in
different IFN-␥ mRNA levels (Fig. 2). IFN-␥ mRNA levels
were determined from P14 cells directly ex vivo. Cells were
purified by FACS sorting (Fig. 2A), and mRNA levels were
measured by comparative real-time RT-PCR (Fig. 2B). Effector
cells contained 65 times more IFN-␥ mRNA than naive cells,
while memory P14 cells contained 10 –20 times more. mRNA
levels following activation of naive and memory T cells were
assessed in short-term cell cultures for various lengths of time.
Production of IFN-␥ mRNA increased rapidly in memory cells,
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
into effector cells, the IFN-␥ gene is demethylated (33, 35, 36), but
it is not clear what happens during differentiation into Ag-specific
memory cells in vivo.
In this study, we address whether regulated gene methylation is
an intracellular mechanism contributing to the differential expression of critical genes in naive and memory CD8 T cells. We assess
IFN-␥ gene methylation in vivo during differentiation of naive
cells into memory cells, and after re-exposure to Ag. In these studies, freshly isolated, virus-specific memory CD8 T cells with precisely known activation history were examined. IFN-␥ promoter
methylation changed from predominantly methylated at three CpG
sites to unmethylated, and to partial methylation of the three sites
during the transition from naive to effector and then to memory
CD8 T cells, respectively. IFN-␥ gene methylation changes were
site specific, as the IL-2 promoter was also demethylated in effector cells, but remained demethylated in memory cells. The IFN-␥
promoter was rapidly demethylated upon restimulation of memory
cells, while it remained unchanged in naive cells. Thus, IFN-␥
gene methylation was surprisingly dynamic and precisely regulated during all stages of CD8 T cell differentiation and during T
cell stimulation. These results suggest a novel mechanism of memory CD8 T cell gene control involving the rapid loss of methylation at a key immune response gene.
The Journal of Immunology
4085
reaching levels of 1,000- to 2,000-fold compared with unstimulated memory cells. This occurred rapidly within the first 2 h
(Fig. 2C). Naive cells also increased the amount of IFN-␥
mRNA in this time frame, but more modestly. At 2 h of stimulation, the increase was 130-fold for naive cells compared with
resting naive cells, while it was 1,600-fold for memory cells
compared with memory cells before stimulation (Fig. 2C).
When naive and memory P14 cells were directly compared, and
resting naive cells were taken as a point of reference, the difference was 20,000-fold (Fig. 2D). Thus, the IFN-␥ gene is an
ideal candidate for the study of differential transcriptional regulation because there is a remarkable difference in IFN-␥
mRNA levels between naive and memory CD8 T cells.
IFN-␥ gene methylation is differentially regulated in naive,
effector, and memory cells
There are nine CpG sites within 300 bp of the IFN-␥ transcription start site (Fig. 3A). Methylation of these sites negatively
affects transcription (35, 36). For CpG methylation analysis, we
used the PCR-based “bisulfite sequencing” method (40). This
requires much less biologic material than Southern blotting
techniques, a critical factor when studying memory CD8 T cells
of limited quantities. We obtained genomic DNA from P14
cells highly purified by FACS sorting to at least 96% purity.
Sorts for naive cells were performed to obtain CD8⫹Thy1.1⫹
CD44low cells to exclude previously activated cells, while sorted effector and memory cells were CD8⫹Thy1.1⫹CD44high. DNA was
treated with sodium bisulfite, resulting in the deamination of cytosines
to uracil, unless protected by methylation. The modified DNA was
amplified by PCR, and uracils were replicated as thymidines. PCR
products were then subcloned, sequenced, and compared with the
wild-type IFN-␥ gene. All cytosines not contained in CpG sequences
were mutated to thymidines, indicating that the bisulfite treatment was
efficient (data not shown). CpG methylation at each site is displayed
FIGURE 2. Different IFN-␥ mRNA levels in naive, effector, and memory P14 CD8 T cells. A, FACS purification of P14 cells. Naive (N), effector
(E), and memory (M) P14 CD8 T cells from the spleen were stained with
CD90.1 and CD8 and sorted on a FACS Vantage. The numbers refer to P14
cells as percentage of total. B, IFN-␥ mRNA was quantified in purified P14
cells using real-time RT-PCR and normalized to hypoxanthine phosphoribosyltransferase (HPRT). The amount of IFN-␥ mRNA in naive P14 cells
was set to 1. Data are representative of the average of duplicate PCRs; three
independent PCRs were performed. C and D, mRNA levels after stimulation of naive and memory P14 CD8 T cells. Memory P14 splenocytes
(Thy1.1⫹) were enriched by depletion of unrelated T cells and stimulated
with 1 ␮M gp33 peptide for the indicated times. IFN-␥ mRNA was quantified by real-time RT-PCR. IFN-␥ mRNA was normalized to HPRT, and
set to 1 for resting naive cells (C, left panel), or set to 1 for resting memory
cells (C, right panel). In D, both data sets are directly compared with naive
resting cells to illustrate the range of mRNA levels. The data are the average of duplicate PCRs and representative of six independent experiments.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 1. Unlike naive P14 cells, effector and memory P14 cells rapidly produce large amounts of IFN-␥. A and B, Intracellular production of IFN-␥.
Naive (N) splenocytes were from P14 Thy1.1⫹ mice. We adoptively transferred P14 Thy1.1⫹ cells into B6 recipients (Thy1.2⫹) and infected them with
LCMV 7 days (E, Effector) or ⬎30 days (M, Memory) before the experiment. Splenocytes were cultured for 5 h with 1 ␮M (A) or increasing doses (B)
of gp33 peptide. Cells were then stained with anti-IFN-␥ and anti-CD90.1 (Thy1.1), and analyzed by FACS. Numbers indicate the percentage of P14 cells
producing IFN-␥.
4086
RAPID IFN-␥ GENE DEMETHYLATION IN MEMORY CD8 T CELLS
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 3. CpG methylation of the IFN-␥ gene promoter in naive, effector and memory P14 CD8 T cells. A, The IFN-␥ gene contains nine CpG sites
at the indicated positions between ⫺205 to ⫹120 bp relative to the transcription start site. B, Methylated CpG (mCpG) sites in naive, effector, and memory
P14 CD8 T cells. Each line represents one DNA strand; E, unmethylated CpGs; F, methylated CpGs. P14 CD8 T cells from the spleen were purified by
FACS sorting to 96% purity or greater; naive cells were CD44low, effector and memory cells were CD44high. CpG methylation was analyzed by “bisulfite
sequencing”. We performed multiple independent PCRs separately for the coding and noncoding strand, and sequenced multiple PCR-derived subclones.
Duplicates were eliminated because they could potentially have arisen from the same PCR subclone. For the coding strands, we examined 5 (N, naive),
The Journal of Immunology
CpG methylation of H19, an unrelated, imprinted gene
Is the demethylation of the IFN-␥ gene in effector P14 cells locus
specific? Because effector cells might undergo global loss of methylation, a process that might contribute to their vulnerability to
apoptosis (41), a control for this system was required. The H19
gene is a paternally imprinted gene whose methylation is modulated during oogenesis and spermatogenesis (42), but is not expected to specifically change during immune responses. Thus, we
performed bisulfite sequencing using FACS-sorted P14 cells from
female mice, examining a previously studied region upstream of
the H19 gene with 11 CpG sites (Fig. 4A) (35, 36). All P14 cells
showed significant methylation of this genetic locus (Fig. 4B).
Quantification of the data showed that the average methylation of
all sites was 62, 59, and 64% for naive, effector, and memory cells,
respectively (Fig. 4C). Thus, there was no extensive global methylation change of this imprinted gene. Therefore, changes in DNA
methylation were not observed at all genetic loci during CD8 T
cell differentiation.
CpG methylation of the IL-2 promoter
For comparison, the CpG methylation of another cytokine gene,
IL-2, which is regulated by gene methylation in CD4 T cells
(28), was also examined during memory CD8 T cell differentiation. The IL-2 promoter is methylated at several sites in naive
CD4 T cells, and this specifically silences transcription (28).
CpG methylation of the five CpG sites closest to the transcription start were examined during CD8 T cell differentiation (Fig.
5A). In CD4 T cells, these sites are methylated, and Ag stimulation leads to the rapid demethylation of sites ⫺380, ⫺262,
and ⫺217, but not sites ⫺69 and ⫹64 (28). The methylation of
the sites in CD8 T cells is unknown, as is their methylation in
effector and memory T cells.
In naive CD8 T cells, the promoter was methylated at sites
⫺217, ⫺69, and ⫹64, with site ⫺69 the most extensively methylated (Fig. 5B). Fewer sites than in CD4 T cells were methylated (28), likely due to different IL-2 promoter regulation in
CD4 and CD8 T cells, or due to mouse strain differences. Methylation was severely reduced at all sites in effector cells from
the peak of virus-induced T cell expansion, although a few
methylated DNA strands remained (Fig. 5B). Thus, both the
IFN-␥ and IL-2 promoters are demethylated in effector CD8 T
cells. In contrast to the IFN-␥ promoter, however, the IL-2 promoter was also demethylated in memory CD8 T cells, and the
low level of methylation observed in effector cells was even
further reduced. Thus, IL-2 and IFN-␥ gene methylation are
both specifically but differently regulated by DNA methylation
during memory CD8 T cell differentiation, and the state of DNA
methylation of both genes is highly locus specific at all of the
examined developmental stages.
IFN-␥ gene methylation changes rapidly during restimulation of
memory P14 CD8 T cells
The CpG methylation of cytokine loci can rapidly change after
T cell stimulation in a 4 –7 h time frame (28). To examine the
methylation status during the rapid induction of IFN-␥ expression during restimulation of memory CD8 cells, IFN-␥ gene
methylation was determined in naive and memory P14 cells
following 5 h of stimulation with Ag (Fig. 6A). During this
short period of Ag stimulation, memory CD8 T cells express
4 (E, effector), and 5 (M, memory) independent cell samples. We performed 7 (N), 14 (E), and 14 (M) independent PCRs, and sequenced 53 (N), 57 (E),
91 (M) DNA strands. For the noncoding strand, the numbers were: 3 (N), 3 (E), and 4 (M) independent cell samples; 6 (N), 6 (E), and 7 (M) independent
PCRs; 44 (N), 10 (E), 19 (M) sequenced DNA strands. C, The percentage of DNA strands with methylation at the indicated CpG site was determined from
data shown in B of coding and noncoding strands combined. The differences between naive and memory cells and between effector and both other cell types
in the extent of methylation at sites ⫹ 17, ⫹97, and ⫹120 were statistically different (p ⬍ 0.05 using Fisher’s exact test). D, The number of methylated
sites per individual DNA strand was determined from data shown in panel B of coding and noncoding strands combined. The average number of mCpGs
per DNA strand was 2.8, 0.3, and 2.7 for naive, effector, or memory P14 cells, respectively.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
such that each line represents the methylation status of one DNA
strand.
Methylation in naive P14 cells was concentrated at the three
CpGs after the transcription start (positions ⫹17, ⫹97, and
⫹120) (Fig. 3B). Several examined DNA strands had additional
methylated sites. The data shown are from five independent
FACS sorts, from which seven independent PCRs were performed, and 53 PCR subclones were sequenced. To avoid a bias
associated with duplicate results arising from the same initial
PCR amplicons, duplicates from the same PCR were discarded
and thus, only 32 DNA strands are depicted. Elimination of
duplicate PCR amplicons in this study did not influence the
overall results or patterns observed. The noncoding strand was
also examined. Specific sites were methylated at similar frequencies in the coding and noncoding strands, suggesting that
methylation was symmetrical at most sites. The three CpG sites
after the transcription start were methylated in ⬎50% of all
examined DNA strands from naive cells (Fig. 3C). Methylation
at site ⫺53 was less than previously reported (35), and this was
consistently observed in naive T cells from either the spleen or
lymph nodes (data not shown). Most individual DNA strands
from naive cells had two or more CpG sites methylated (Fig.
3D), and the average number of methylated sites was 2.8. Importantly, only 2% of the DNA strands from naive cells had no
methylated CpG site (1 of 53 analyzed strands). Thus, in summary, freshly isolated naive P14 T cells are methylated at multiple sites around the IFN-␥ transcription start site.
The IFN-␥ locus was almost completely unmethylated in effector cells at days 7 and 8 postinfection with LCMV (Fig. 3B). As
expected, this correlated with high mRNA levels (Fig. 2). Thus,
IFN-␥ gene methylation is distinctly different between naive and
effector cells.
In contrast to effectors, memory P14 cells harvested 6 –12 wk
postinfection were partially methylated at CpG sites ⫹17, ⫹97,
and ⫹ 120, but to a lesser degree than naive cells (Fig. 3B). The
coding and noncoding strand showed similar methylation patterns.
In comparison to naive cells, methylation was significantly less
focused on sites ⫹17, ⫹97, and ⫹117 after the transcription start
site in memory cells, but appeared more evenly distributed (Fig.
3C). The average number of methylated CpGs per DNA strand was
2.7, very similar to 2.8 in naive cells. However, a larger number of
memory cell DNA strands had no methylation, as 23% remained
unmethylated, in stark contrast to 2% in naive cells (Fig. 3D).
Thus, IFN-␥ gene methylation in memory CD8 T cells is more
heterogeneous than in naive cells in two ways: it is distributed
more heterogeneously within the 300-bp region, and individual
DNA strands differ more widely. The relative methylation patterns
of naive, effector, and memory P14 cells were also observed in
nontransgenic LCMV-specific CD8 T splenocytes from normal B6
mice (data not shown).
4087
4088
RAPID IFN-␥ GENE DEMETHYLATION IN MEMORY CD8 T CELLS
5-Azacytidine induces IFN-␥ expression in naive but not
memory CD8 T cells
5-Azacytidine is a nucleoside analog of cytidine that specifically
inhibits DNA methylation by trapping Dnmts during replication
(43). This effectively leads to demethylation and the expression of
genes silenced by DNA methylation. The effect of 5-azacytidine
on IFN-␥ production in naive and memory P14 cells was determined by intracellular FACS analysis (Fig. 7). P14 cells from the
spleen were cultured with peptide gp33 and 5-azacytidine for 72 h.
This led to a dose-dependent increase in the number of naive P14
cells able to produce IFN-␥ (Fig. 7). The percentage of IFN-␥producing naive P14 cells rose from 12 to 56% with the highest
dose of 5-azacytidine. The amount of intracellular IFN-␥ also increased, as the MFI of IFN-␥ cells shifted (Fig. 7) (35). The effect
was first noticeable after 48 h of cell culture (data not shown),
consistent with a need for cell replication, during which 5-azacytidine blocks Dnmt action. Labeling with CFSE revealed that most
naive cells only produced IFN-␥ after cell division (data not
FIGURE 4. CpG methylation of the imprinted H19 gene is unaltered
during T cell differentiation. A, CpG sites of the examined H19 locus. We
amplified a 300-bp stretch of genomic DNA ⬃4 kb upstream of the H19
transcription start site. The numbers represent the position of 11 CpG sites
within the 300-bp region. B, CpG methylation of the H19 gene as assessed
by bisulfite sequencing of DNA from P14 CD8 T cells from the spleens of
female mice. F, Methylated; E, unmethylated cytosines. Data were collected from at least two independent cell preparations per sample. For
naive, effector, and memory cells, five, five, and three independent PCRs
were performed, respectively. C, The extent of CpG methylation for all
sites combined was calculated from the data shown in B.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
exceedingly high levels of IFN-␥ mRNA whereas naive CD8 T
cells contain significantly less IFN-␥ mRNA (20,000-fold less)
(Fig. 2, C and D). CpG methylation at the IFN-␥ transcription
start site did not change significantly in stimulated naive P14
cells (Fig. 6B). These stimulated naive P14 cells were similarly
methylated for both extent and distribution of CpG methylation
of the IFN-␥ gene when the data were quantified, and no statistically significant methylation differences were found (Fig.
6D). In addition, the coding and noncoding DNA strands remained symmetrically methylated.
In stark contrast to naive cells, memory P14 cells lost IFN-␥
gene methylation during 5 h of stimulation (Fig. 6, C and D).
For resting and stimulated memory P14 cells, a total of four
independent experiments were performed with P14 cells from
the same LCMV immune mice. Demethylation was nearly complete on both the coding and noncoding strands. Thus, the
IFN-␥ gene loses CpG methylation at an extremely rapid rate,
suggesting that this loss of methylation is a transcriptional control mechanism that is differentially regulated in naive and
memory CD8 T cells.
In contrast to the IFN-␥ gene, the IL-2 gene underwent some
methylation changes during the stimulation of naive cells, as
previously reported for CD4 T cells (28) (data not shown).
Methylation at site ⫺217 was completely lost in naive P14 cells
during short-term stimulation for 5 h (data not shown), while
methylation of sites ⫺69 and ⫹64 remained constant as it does
in CD4 T cells (28). At this time point, naive P14 CD8 T cells
increased IL-2 mRNA production 2500-fold over the resting
state (not shown), a much greater increase than the 200-fold
increase observed for IFN-␥ mRNA in naive CD8 T cells at this
time point (Fig. 2C). Stimulated memory CD8 T cells increased
IL-2 mRNA levels 5000-fold and remained demethylated at the
IL-2 gene during the 5-h incubation time (data not shown). In
summary, rapid gene demethylation can be observed at both the
IFN-␥ and the IL-2 locus during short-term stimulation, but at
different developmental states, demonstrating the specific regulation of this process.
The Journal of Immunology
4089
shown). Thus, DNA methylation directly silences the expression
of IFN-␥ in naive CD8 T cells. As expected, 5-azacytidine did not
significantly increase the IFN-␥ production of memory cells, as
these cells were already rapidly undergoing IFN-␥ gene demethylation and already efficiently produced IFN-␥ (Fig. 7). This suggests that IFN-␥ expression is silenced by DNA methylation in
naive CD8 T cells, but not in memory cells, and further illustrates
the differential regulation by gene methylation in the two cell
types.
Replication inhibitors do not block IFN-␥ mRNA induction and
IFN-␥ gene demethylation
Because a DNA demethylase enzyme has not been identified (20,
44), it is generally assumed that methylated CpG sites become
demethylated due to lack of maintenance methylation during replication, when existing cytosine methylation is copied from the
parent DNA strand onto the newly synthesized daughter strand by
the maintenance methyltransferase, Dnmt1. If Dnmt1 activity is
absent or cannot keep pace with newly synthesized DNA, CpG
methylation should disappear by 50% with each cell division. During the first cell division, it should furthermore remain intact on
one DNA strand but not the other. We did not observe this expected behavior for the IFN-␥ gene in stimulated memory cells, as
IFN-␥ CpG methylation disappeared by ⬎50% on both DNA
strands within the first 5 h of stimulation (Fig. 6C). Memory CD8
T cells start to divide 18 –24 h post restimulation, as measured with
CFSE (data not shown). However, DNA replication might begin
before cell division is completed at 18 h post restimulation of
memory CD8 T cells. To address the level of DNA replication
within 5 h of memory T cell restimulation, [3H]thymidine incorporation in P14 spleen cultures from LCMV immune mice
was measured. P14 cells stimulated with or without 1 ␮M gp33
peptide in vitro in 12 duplicate wells showed no significant
[3H]thymidine incorporation within the first 5 h of activation
(Fig. 8A). Similar results were seen using BrdU incorporation to
measure DNA replication (data not shown). Therefore, during
5 h of restimulation, P14 memory cells did not proliferate or
synthesize new DNA to an extent that could be measured with
these techniques. A small, but reproducible increase in [3H]thymidine incorporation was observed at 7 h after activation, indicating the beginning of measurable DNA synthesis. There
was massive replication at 24 h of culture. As expected, replication was blocked by DNA synthesis inhibitors mitomycin C
and aminopterin (Fig. 8A).
We also directly addressed the role of DNA synthesis in IFN-␥
protein and mRNA production. DNA synthesis was blocked with
a variety of inhibitors that act through different mechanisms. If
inhibiting DNA synthesis in turn inhibits DNA demethylation, we
expected a lack of new IFN-␥ transcription and also protein production. Memory P14 splenocytes were stimulated in culture in the
presence of the DNA synthesis inhibitors mitomycin C (cross-links
DNA), aminopterin (inhibits thymidine synthesis), and mizoribine (inhibits nucleotide synthesis), and then examined for intracellular IFN-␥ production by FACS analysis (Fig. 8B) and
mRNA (Fig. 8C). All three drugs were unable to block IFN-␥
protein or mRNA production in memory P14 cells. Preincubation of the above cultures with the drugs for 3 h before the
experiment to allow for drug uptake and depletion of endogenous nucleotides did also not block IFN-␥ transcription (data
not shown). Additionally, fluorouracil (inhibits pyrimidine synthesis), cytosine-b-D-arabinofluranoside (nucleotide analog),
amethopterin (blocks thymidine synthesis), and hydroxyurea
(blocks nucleotide synthesis) were used as DNA synthesis inhibitors, and they too did not block IFN-␥ production (data not
shown). Thus, blocking DNA synthesis had no effect on IFN-␥
transcription. In contrast, IFN-␥ transcription and protein production were severely affected with drugs inhibiting T cell activation (cyclosporin A), transcription (actinomycin D), or protein synthesis (cycloheximide).
To test directly whether DNA synthesis was necessary for
IFN-␥ gene demethylation in memory P14 T cells, memory P14 T
cells were stimulated with peptide in the presence or absence of
mitomycin C, a DNA synthesis inhibitor. Stimulated memory cells
were demethylated regardless of the presence of mitomycin C (Fig.
8D). Thus, mitomycin C did not block demethylation. Therefore,
IFN-␥ gene demethylation in memory CD8 T cells appeared not to
be dependent on DNA replication and cell division.
Discussion
A striking feature of memory cells is their ability to respond to Ag
re-encounter with the rapid and extensive production of the antiviral cytokine IFN-␥ while naive cells require proliferation before
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 5. Changes in CpG methylation of the IL-2 gene promoter in naive, effector, and memory P14 CD8 T cells. A, Five CpG sites of the IL-2
promoter were examined at the indicated positions relative to the transcription start site. B, CpG methylation of the IL-2 promoter was determined in the
naive, effector, and memory P14 cell samples described in Fig. 3. The percentage of DNA strands with methylation at the indicated CpG sites of the coding
strand is shown. Thirty-four, 39, and 53 cloned PCR products were sequenced for naive, effector, and memory cells, respectively. The IL-2 promoter, like
the IFN-␥ promoter, was demethylated in effector cells, but remained demethylated in memory P14 cells.
4090
RAPID IFN-␥ GENE DEMETHYLATION IN MEMORY CD8 T CELLS
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 6. Rapid demethylation of IFN-␥ CpGs upon Ag stimulation of memory P14 CD8 T cells. A, Experimental design: naive and memory P14 CD8
T cells from the spleen were activated with gp33 peptide in vitro for 5 h. Cells were then FACS sorted, and IFN-␥ and IL-2 CpG gene methylation was
analyzed. B, IFN-␥ gene methylation is unaltered after stimulation of naive P14 CD8 T cells. IFN-␥ CpG methylation was determined by bisulfite
sequencing for the coding and noncoding strand. The data are from two independent cell sorts and a total of 12 independent PCRs. C, Rapid IFN-␥ CpG
demethylation in resting and restimulated memory P14 cells. In coding and noncoding strands, the IFN-␥ site was demethylated after 5 h of activation (right
panels). We performed 4 independent FACS sorts, and 15 or 13 independent PCRs for resting and activated cells, respectively. Resting and activated cells
came from the same cell samples. D, The data from B and C were quantified, and the percentage of DNA strands with methylation at the indicated location
is given. There were no statistically significant differences in the methylation of individual sites in the unstimulated and stimulated naive samples; sites ⫹97
and ⫹120 were statistically different in unstimulated and stimulated memory samples (p ⬍ 0.05 using Fisher’s exact test).
they can do so. Our current findings identify changes in DNA
methylation as a novel mechanism that differentially controls the
expression of this critical gene in naive vs memory CD8 T cells.
We found specific changes in IFN-␥ gene methylation during the
differentiation of naive into effector cells, and of effector cells into
memory cells. Importantly, although naive CD8 T cells are largely
methylated at three CpG sites within the IFN-␥ promoter and
memory CD8 T cells are also partially methylated at these sites,
exposure to Ag induces rapid demethylation of these sites only in
memory cells. We contrast this to the IL-2 gene, which is rapidly
The Journal of Immunology
4091
Naive
IFNγ
- peptide
+ peptide
+ peptide
+1 µM 5-Aza.
+ peptide
+ 2.5 µM 5-Aza.
0
12.0
22.1
56.2
0.3
87.5
92.3
79.1
Memory
CD90.1
demethylated in naive cells upon stimulation, and then remains
demethylated during differentiation. Thus, there is both genespecific and differentiation state-specific regulation of DNA
methylation in CD8 T cells.
Our study has revealed a novel mechanism by which the timing
and magnitude of IFN-␥ gene expression can be controlled across
different CD8 T cell developmental stages. Methylation in memory
cells followed by rapid demethylation after antigenic stimulation
ensures that memory CD8 T cells have a strict requirement to see
Ag if they are to produce IFN-␥, a potent proinflammatory cytokine. In addition, rapid demethylation, specifically in memory, not
in naive cells, reflects a basic immunologic principle: naive cells
need to expand in numbers before they can become an effective
defense factor, whereas the primary objective of memory cells
with already increased precursor frequency is the rapid elimination
of reinfecting pathogens.
Our most surprising result was the rapid IFN-␥ gene demethylation in memory cells upon restimulation. The mechanism
appears to not require cell division, as demethylation and IFN-␥
mRNA production cannot be blocked by DNA synthesis inhibitors. This suggests that demethylation is probably not due to
lack of maintenance methylation during DNA replication, but
rather could be achieved by an enzymatic factor. The existence
of such an enzyme has been previously postulated, when the
rapid demethylation of the IL-2 gene was reported (28). There
is currently no enzyme known with confirmed DNA demethylase activity. The recent report of a long-sought histone demethylase (45), whose existence had been doubted (46), makes the
discovery of a CpG-demethylating enzyme more likely. However, other possibilities remain: cell restimulation could induce
selective apoptosis of cells with a methylated IFN-␥ locus or
lead to the selective loss of these cells during cell sorting. In
addition, DNA replication could originate close to the IFN-␥
gene in memory CD8 T cells, and replication of the IFN-␥ DNA
could occur soon after restimulation in a time frame before
DNA synthesis inhibitors act.
Our results suggest that the molecular machinery that controls
gene methylation and demethylation operates differently in naive
and memory CD8 T cells. This machinery is currently incom-
pletely understood. In addition, different signals might be generated at the cell surface of naive and memory CD8 T cells. Signaling through the IL-7R has been shown to regulate gene
methylation in thymocytes (47). Therefore, differences in cytokine
receptor signaling could result in differential regulation of gene
methylation in naive vs memory CD8 T cells.
The passing of genetic information other than through DNA
sequence information has been called epigenetic regulation. It ensures long-term inheritance of gene expression patterns in differentiated cells. Therefore, it was surprising to find such rapid
changes in IFN-␥ gene methylation. Memory CD8 T cells are a
unique cell type, however, because they are highly but not terminally differentiated (reviewed in Ref. 48). Restimulation of memory cells marks the beginning of a rapid differentiation process into
secondary effector cells. Thus, rapid CpG demethylation reflects
the ability of memory cells to rapidly differentiate into effector
cells.
NK cells like memory CD8 T cells are an early source of
IFN-␥. Both cells do not need to proliferate before making
IFN-␥ (12, 49). Despite these similarities, IFN-␥ gene methylation is regulated differently. Freshly isolated NK cells do not
have a methylated IFN-␥ gene locus (49). This is perhaps why
freshly isolated NK cells have a higher level of constitutive
IFN-␥ mRNA than CD8 memory cells. NK cells express 200fold more IFN-␥ mRNA than naive T cells (50) while we found
only 20-fold more in memory T cells. However, irrespective of
the IFN-␥ gene methylation status, we now demonstrate that
memory CD8 T cells are not limited by their basal IFN-␥ gene
methylation because they can quickly lose it upon stimulation
and start transcribing IFN-␥ very rapidly.
Changes in DNA methylation have been implicated in cell
lineage decisions (reviewed in Ref. 8). The IFN-␥ promoter is
demethylated in effector CD8 T cells, but partially methylated
again in memory cells. It is not clear whether demethylation and
remethylation are sequential events in the linear differentiation
of cells from naive to effector and memory cells. Alternatively,
a subset of effector cells with higher methylation or faster remethylation might selectively survive to become memory cells. It
will next be important to determine the IFN-␥ gene methylation
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 7. 5-Azacytidine increases IFN-␥ production in naive but not memory P14 CD8 T cells. Intracellular production of IFN-␥ in the presence of
5-Aza (5-azacytidine), a drug that causes DNA demethylation upon proliferation. Naive (N) and memory (M) splenocytes from P14 Thy1.1⫹ mice were
cultured for 72 h with 0.5 ␮g/ml gp33 peptide and 5-azacytidine where indicated. Cells were then restimulated with peptide for 2 h where indicated and
treated with brefeldin A to block secretion of IFN-␥. Cells were stained with anti-IFN-␥, anti-CD90.1 (Thy1.1), anti-CD8, and analyzed by FACS. Data
are presented after gating on CD8⫹ cells, numbers indicate the percentage of P14 cells producing IFN-␥. Three independent experiments were performed.
5-Aza increases IFN-␥ production in naive but not memory P14 CD8 T cells, indicating differential regulation by DNA methylation.
4092
RAPID IFN-␥ GENE DEMETHYLATION IN MEMORY CD8 T CELLS
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 8. IFN-␥ transcription and IFN-␥ gene demethylation appear independent of replication of memory P14 cells. A, Memory P14 cells do not
incorporate [3H]thymidine within 5 h of culture with 1 ␮M peptide gp33. Splenocytes from LCMV immune P14 mice were cultured for 5, 7, or 24 h. DNA
synthesis inhibitors mitomycin C (MitC) and aminopterin (Amin.) were added at 50 ␮g/ml where indicated. The error bars represent SEM of 12 samples.
The experiment was independently performed four times. B, Intracellular IFN-␥ production in memory P14 CD8 T cells is not blocked by DNA synthesis
inhibitors. Cells were cultured for 5 h with 1 ␮M gp33 peptide, stained with anti-IFN-␥-FITC and anti-CD90.1 (Thy1.1)-PE, and analyzed by FACS.
Numbers indicate the percentage of P14 cells producing IFN-␥. These inhibitors were added at a concentration of 50 ␮g/ml: mitomycin C (MitC),
aminopterin (Amin.), mizoribine (Miz.), cyclosporin A (CsA), actinomycin D (ActD), or cycloheximide (Chx). Two independent experiments were
performed. C, Induction of IFN-␥ mRNA is not blocked by DNA synthesis inhibitors in memory P14 splenocytes. IFN-␥ mRNA was quantitated by
real-time RT-PCR. We cultured P14 CD8 T cells from the spleen as in B. mRNA was normalized to HPRT, and the amount of IFN-␥ mRNA of stimulated
cells was set to 100%. The average of two duplicate PCRs is shown; two independent experiments were performed. D, IFN-␥ gene demethylation after
stimulation of memory P14 CD8 T cells in the presence or absence of the DNA synthesis inhibitor mitomycin C (50 ␮g/ml). Memory P14 CD8 T cells
from the spleen were stimulated for 5 h as described in Fig. 6. IFN-␥ CpG methylation was determined by bisulfite sequencing for the coding strand.
profile of the recently identified subset of effector cells that
selectively develops into memory cells (51, 52). These memory
precursor cells express higher levels of the IL-7R␣ than other
effector cells during the height of the CD8 T cell response to
LCMV infection (51, 52). Further exploring of intracellular
transcriptional control mechanisms such as DNA methylation
will improve the molecular understanding of memory CD8 T
cell differentiation processes. This might ultimately allow the
selective induction of more potent memory responses to pathogens and vaccines.
The Journal of Immunology
Acknowledgments
We thank Dr. Gil Kersh for comments on the manuscript and assistance
with cell proliferation assays, Tao Zhou for technical assistance, and members of the Ahmed laboratory for discussions.
Disclosures
The authors have no financial conflict of interest.
References
28. Bruniquel, D., and R. H. Schwartz. 2003. Selective, stable demethylation of the
interleukin-2 gene enhances transcription by an active process. Nat. Immunol. 4:
235–240.
29. Lee, D. U., O. Avni, L. Chen, and A. Rao. 2004. A distal enhancer in the interferon-␥ (IFN-␥) locus revealed by genome sequence comparison. J. Biol. Chem.
279: 4802– 4810.
30. Yano, S., P. Ghosh, H. Kusaba, M. Buchholz, and D. L. Longo. 2003. Effect of
promoter methylation on the regulation of IFN-␥ gene during in vitro differentiation of human peripheral blood T cells into a Th2 population. J. Immunol. 171:
2510 –2516.
31. White, G. P., P. M. Watt, B. J. Holt, and P. G. Holt. 2002. Differential patterns
of methylation of the IFN-␥ promoter at CpG and non-CpG sites underlie differences in IFN-␥ gene expression between human neonatal and adult CD45RO⫺
T cells. J. Immunol. 168: 2820 –2827.
32. Penix, L. A., M. T. Sweetser, W. M. Weaver, J. P. Hoeffler, T. K. Kerppola, and
C. B. Wilson. 1996. The proximal regulatory element of the interferon-␥ promoter mediates selective expression in T cells. J. Biol. Chem. 271: 31964 –31972.
33. Zhou, W., S. Chang, and T. M. Aune. 2004. Long-range histone acetylation of the
Ifn␥ gene is an essential feature of T cell differentiation. Proc. Natl. Acad. Sci.
USA 101: 2440 –2445.
34. Shnyreva, M., W. M. Weaver, M. Blanchette, S. L. Taylor, M. Tompa,
D. R. Fitzpatrick, and C. B. Wilson. 2004. Evolutionarily conserved sequence
elements that positively regulate IFN-␥ expression in T cells. Proc. Natl. Acad.
Sci. USA 101: 12622–12627.
35. Fitzpatrick, D. R., K. M. Shirley, L. E. McDonald, H. Bielefeldt-Ohmann,
G. F. Kay, and A. Kelso. 1998. Distinct methylation of the interferon ␥ (IFN-␥)
and interleukin 3 (IL-3) genes in newly activated primary CD8⫹ T lymphocytes:
regional IFN-␥ promoter demethylation and mRNA expression are heritable in
CD44highCD8⫹ T cells. J. Exp. Med. 188: 103–117.
36. Fitzpatrick, D. R., K. M. Shirley, and A. Kelso. 1999. Cutting edge: stable epigenetic inheritance of regional IFN-␥ promoter demethylation in CD44highCD8⫹
T lymphocytes. J. Immunol. 162: 5053–5057.
37. Carding, S. R., D. Lu, and K. Bottomly. 1992. A polymerase chain reaction assay
for the detection and quantitation of cytokine gene expression in small numbers
of cells. J. Immunol. Methods 151: 277–287.
38. 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:
177–187.
39. Zimmerman, C., K. Brduscha-Riem, C. Blaser, R. M. Zinkernagel, and
H. Pircher. 1996. Visualization, characterization, and turnover of CD8⫹ memory
T cells in virus-infected hosts. J. Exp. Med. 183: 1367–1375.
40. Frommer, M., L. E. McDonald, D. S. Millar, C. M. Collis, F. Watt, G. W. Grigg,
P. L. Molloy, and C. L. Paul. 1992. A genomic sequencing protocol that yields
a positive display of 5-methylcytosine residues in individual DNA strands. Proc.
Natl. Acad. Sci. USA 89: 1827–1831.
41. Farrar, W. L., F. W. Ruscetti, and H. A. Young. 1985. 5-Azacytidine treatment
of a murine cytotoxic T cell line alters interferon-␥ gene induction by interleukin
2. J. Immunol. 135: 1551–1554.
42. Lucifero, D., C. Mertineit, H. J. Clarke, T. H. Bestor, and J. M. Trasler. 2002.
Methylation dynamics of imprinted genes in mouse germ cells. Genomics 79:
530 –538.
43. Issa, J. P., H. M. Kantarjian, and P. Kirkpatrick. 2005. Azacitidine. Nat. Rev.
Drug Discov. 4: 275–276.
44. Jaenisch, R., and A. Bird. 2003. Epigenetic regulation of gene expression: how
the genome integrates intrinsic and environmental signals. Nat. Genet.
33(Suppl.): 245–254.
45. Shi, Y., F. Lan, C. Matson, P. Mulligan, J. R. Whetstine, P. A. Cole, and
R. A. Casero. 2004. Histone demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell 119: 941–953.
46. Couzin, J. 2004. Molecular biology. Long-sought enzyme found, revealing new
gene switch on histones. Science 306: 2171.
47. Durum, S. K., S. Candeias, H. Nakajima, W. J. Leonard, A. M. Baird, L. J. Berg,
and K. Muegge. 1998. Interleukin 7 receptor control of T cell receptor ␥ gene
rearrangement: role of receptor-associated chains and locus accessibility. J. Exp.
Med. 188: 2233–2241.
48. Kaech, S. M., and R. Ahmed. 2001. Memory CD8⫹ T cell differentiation: initial
antigen encounter triggers a developmental program in naive cells. Nat. Immunol.
2: 415– 422.
49. Tato, C. M., G. A. Martins, F. A. High, C. B. DiCioccio, S. L. Reiner, and
C. A. Hunter. 2004. Cutting edge: innate production of IFN-␥ by NK cells is
independent of epigenetic modification of the IFN-␥ promoter. J. Immunol. 173:
1514 –1517.
50. Stetson, D. B., M. Mohrs, R. L. Reinhardt, J. L. Baron, Z. E. Wang, L. Gapin,
M. Kronenberg, and R. M. Locksley. 2003. Constitutive cytokine mRNAs mark
natural killer (NK) and NK T cells poised for rapid effector function. J. Exp. Med.
198: 1069 –1076.
51. Kaech, S. M., J. T. Tan, E. J. Wherry, B. T. Konieczny, C. D. Surh, and
R. Ahmed. 2003. Selective expression of the interleukin 7 receptor identifies
effector CD8 T cells that give rise to long-lived memory cells. Nat. Immunol. 4:
1191–1198.
52. Madakamutil, L. T., U. Christen, C. J. Lena, Y. Wang-Zhu, A. Attinger,
M. Sundarrajan, W. Ellmeier, M. G. von Herrath, P. Jensen, D. R. Littman, and
H. Cheroutre. 2004. CD8␣␣-mediated survival and differentiation of CD8 memory T cell precursors. Science 304: 590 –593.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
1. Marrack, P., and J. Kappler. 2004. Control of T cell viability. Annu. Rev. Immunol. 22: 765–787.
2. Sallusto, F., J. Geginat, and A. Lanzavecchia. 2004. Central memory and effector
memory T cell subsets: function, generation, and maintenance. Annu. Rev. Immunol. 22: 745–763.
3. Welsh, R. M., L. K. Selin, and E. Szomolanyi-Tsuda. 2004. Immunological memory to viral infections. Annu. Rev. Immunol. 22: 711–743.
4. Gourley, T. S., E. J. Wherry, D. Masopust, and R. Ahmed. 2004. Generation and
maintenance of immunological memory. Semin. Immunol. 16: 323–333.
5. Dalton, D. K., S. Pitts-Meek, S. Keshav, I. S. Figari, A. Bradley, and
T. A. Stewart. 1993. Multiple defects of immune cell function in mice with
disrupted interferon-␥ genes. Science 259: 1739 –1742.
6. Bartholdy, C., J. P. Christensen, D. Wodarz, and A. R. Thomsen. 2000. Persistent
virus infection despite chronic cytotoxic T-lymphocyte activation in ␥ interferondeficient mice infected with lymphocytic choriomeningitis virus. J. Virol. 74:
10304 –10311.
7. Ansel, K. M., D. U. Lee, and A. Rao. 2003. An epigenetic view of helper T cell
differentiation. Nat. Immunol. 4: 616 – 623.
8. Murphy, K. M., and S. L. Reiner. 2002. The lineage decisions of helper T cells.
Nat. Rev. Immunol. 2: 933–944.
9. Agarwal, S., and A. Rao. 1998. Modulation of chromatin structure regulates
cytokine gene expression during T cell differentiation. Immunity 9: 765–775.
10. Avni, O., D. Lee, F. Macian, S. J. Szabo, L. H. Glimcher, and A. Rao. 2002. Th
cell differentiation is accompanied by dynamic changes in histone acetylation of
cytokine genes. Nat. Immunol. 3: 643– 651.
11. Fields, P. E., S. T. Kim, and R. A. Flavell. 2002. Cutting edge: changes in histone
acetylation at the IL-4 and IFN-␥ loci accompany Th1/Th2 differentiation. J. Immunol. 169: 647– 650.
12. Kambayashi, T., E. Assarsson, A. E. Lukacher, H. G. Ljunggren, and
P. E. Jensen. 2003. Memory CD8⫹ T cells provide an early source of IFN-␥.
J. Immunol. 170: 2399 –2408.
13. Murali-Krishna, K., L. L. Lau, S. Sambhara, F. Lemonnier, J. Altman, and
R. Ahmed. 1999. Persistence of memory CD8 T cells in MHC class I-deficient
mice. Science 286: 1377–1381.
14. Kersh, E. N., S. M. Kaech, T. M. Onami, M. Moran, E. J. Wherry, M. C. Miceli,
and R. Ahmed. 2003. TCR signal transduction in antigen-specific memory CD8
T cells. J. Immunol. 170: 5455–5463.
15. Whitmire, J. K., and R. Ahmed. 2000. Costimulation in antiviral immunity: differential requirements for CD4⫹ and CD8⫹ T cell responses. Curr. Opin. Immunol. 12: 448 – 455.
16. Kaech, S. M., S. Hemby, E. Kersh, and R. Ahmed. 2002. Molecular and functional profiling of memory CD8 T cell differentiation. Cell 111: 837– 851.
17. Teague, T. K., D. Hildeman, R. M. Kedl, T. Mitchell, W. Rees, B. C. Schaefer,
J. Bender, J. Kappler, and P. Marrack. 1999. Activation changes the spectrum but
not the diversity of genes expressed by T cells. Proc. Natl. Acad. Sci. USA 96:
12691–12696.
18. Liu, N., T. Phillips, M. Zhang, Y. Wang, J. T. Opferman, R. Shah, and
P. G. Ashton-Rickardt. 2004. Serine protease inhibitor 2A is a protective factor
for memory T cell development. Nat. Immunol. 5: 919 –926.
19. Glimcher, L. H., M. J. Townsend, B. M. Sullivan, and G. M. Lord. 2004. Recent
developments in the transcriptional regulation of cytolytic effector cells. Nat. Rev.
Immunol. 4: 900 –911.
20. Bird, A. P., and A. P. Wolffe. 1999. Methylation-induced repression-belts, braces,
and chromatin. Cell 99: 451– 454.
21. Bestor, T. H. 2000. The DNA methyltransferases of mammals. Hum. Mol. Genet.
9: 2395–2402.
22. Robertson, K. D., and A. P. Wolffe. 2000. DNA methylation in health and disease. Nat. Rev. Genet. 1: 11–19.
23. Hendrich, B., and A. Bird. 2000. Mammalian methyltransferases and methylCpG-binding domains: proteins involved in DNA methylation. Curr. Top. Microbiol. Immunol. 249: 55–74.
24. Lee, P. P., D. R. Fitzpatrick, C. Beard, H. K. Jessup, S. Lehar, K. W. Makar,
M. Perez-Melgosa, M. T. Sweetser, M. S. Schlissel, S. Nguyen, et al. 2001. A
critical role for Dnmt1 and DNA methylation in T cell development, function,
and survival. Immunity 15: 763–774.
25. Makar, K. W., M. Perez-Melgosa, M. Shnyreva, W. M. Weaver,
D. R. Fitzpatrick, and C. B. Wilson. 2003. Active recruitment of DNA methyltransferases regulates interleukin 4 in thymocytes and T cells. Nat. Immunol. 4:
1183–1190.
26. Smale, S. T., and A. G. Fisher. 2002. Chromatin structure and gene regulation in
the immune system. Annu. Rev. Immunol. 20: 427– 462.
27. Wilson, C. B., K. W. Makar, and M. Perez-Melgosa. 2002. Epigenetic regulation
of T cell fate and function. J. Infect. Dis. 185(Suppl. 1): S37–S45.
4093