CD4-CD8 Lineage Commitment Is Regulated
by a Silencer Element at the ThPOK
Xi He,1Kyewon Park,1,2Haitao Wang,3Xiao He,1Yi Zhang,1Xiang Hua,1Yi Li,1and Dietmar J. Kappes1,*
1Fox Chase Cancer Center, 7701 Burholme Ave., Philadelphia, PA 19111, USA
2Drexel University, School of Biomedical Engineering, Science & Health Systems, 3141 Chestnut St., Philadelphia, PA 19104, USA
3Department of Medicine, University of Pennsylvania, School of Medicine, Philadelphia, PA 19104, USA
The transcription factor ThPOK is necessary and
sufficient to trigger adoption of the CD4 lymphocyte
fate. Here we investigate the regulation of ThPOK
expression and its subsequent control of CD4+T
with anti-TCR (T cell receptor) showed that TCR sig-
nals were important in ThPOK induction and that the
CD4+8lostage was the likely target of the inductive
TCR signal. We identified at the ThPOK locus a key
distal regulatory element (DRE) that mediated its
differential expression in class I- versus II-restricted
CD4+8lothymocytes. The DRE was both necessary
for suppression of ThPOK expression in class I-
restricted thymocytes and sufficient for its induction
in class II-restricted thymocytes. Mutagenesis analy-
sis defined an essential 80bp core DRE sequence
and its potential regulatory motifs. We propose a si-
lencer-dependent model of lineage choice, whereby
inactivation of the DRE silencer by a strong TCR
signal leads to CD4 commitment, whereas continued
silencer activity leads to CD8 commitment.
Development of ab T cells in the thymus proceeds through three
molecules CD4 and CD8, i.e., CD4?CD8?(double negative or
DN), CD4+CD8+(double positive or DP), and CD4+CD8?or
most thymocytes pass through an intermediate CD4+8lostage,
irrespective of whether they are destined to become SP CD4
or CD8 cells (Guidos et al., 1990; Kydd et al., 1995; Lucas and
Germain, 1996; Lundberg et al., 1995; Suzuki et al., 1995; Barth-
pressed at the DP stage, allowing its engagement by intrathymic
peptide-MHC ligands. TCR signaling can induce two different
processes in developing thymocytes, i.e., negative selection
leading to death by apoptosis, or positive selection leading to
thymocyte activation and differentiation into SP T cells.
Coincident with positive selection, thymocytes undergo alter-
nate commitment to either the cytotoxic or the helper T cell
lineages, generating class I-restricted SP CD8+and class II-re-
relation between MHC restriction and functional phenotype is
achieved has remained controversial. Initially, it was suggested
distinct signals initiated upon TCR and coreceptor engagement
by class I- or II MHC ligands (instructive model) (Robey et al.,
1991), or that it occurred randomly and was followed by a selec-
tion step that eliminates thymocytes that express the inappropri-
ate coreceptor (stochastic-selective model) (Chan et al., 1993;
Davis et al., 1993). However, experiments to test these hypothe-
ses gave inconsistent outcomes, suggesting that the models
needed to be refined (Robey et al., 1991; Borgulya et al., 1991;
Chan et al., 1994; Matechak et al., 1996; Robey et al., 1994).
A quantitative-instructive model instead proposes that stronger
and weaker TCR signals, respectively, promote CD4 and CD8
commitment (Matechak et al., 1996). This model is based on
the fact that the cytoplasmic tail of CD4 binds the critical signal-
ing factor Lck with substantially higher affinity than that of CD8a
(Ravichandran and Burakoff, 1994; Veillette et al., 1988). Studies
in which TCR signaling is modulated quantitatively have lent
strong support to this model (Hernandez-Hoyos et al., 2000;
Legname et al., 2000; Liu and Bosselut, 2004). It has been sug-
gested that lineage commitment is not completed until the inter-
mediate CD4+8lostage, because many of these cells are not yet
irreversibly committed to a particular lineage (Bosselut et al.,
2003; Brugnera et al., 2000). Indeed, the distinct kinetic-signal-
ing model postulates that lineage commitment is determined at
the CD4+8lostage, on the basis of the relative ability of corecep-
tors to contribute to TCR signaling (Singer and Bosselut, 2004).
Because CD8 is selectively downmodulated in CD4+8lothymo-
cytes, TCR signaling will be specifically impaired for class I-
restricted thymocytes at this stage. The intracellular pathway
downstream of Lck and Zap70 that regulates alternate lineage
commitment remains unknown. Although a preferential role
for the Ras-Mek-Erk pathway in CD4 development has been
suggested (Bommhardt et al., 1999; Sharp et al., 1997), genetic
approaches do not support this (Alberola-Ila and Hernandez-
Hoyos, 2003; Fischer et al., 2005).
Some insight into the intracellular control of lineage commit-
ment was provided by the discovery of HD mice, which carry a
346 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.
to mature into SP CD8 rather than CD4 T cells (Dave ´ et al., 1998;
Keefe et al., 1999). In light of the quantitative-instructive model,
However, all biochemical assays of TCR signaling proved nor-
mal, and the efficiency of positive selection, as assessed by the
proportions of SP thymocytes, and negative selection, both pro-
cesses that are highly sensitive to alterations in TCR signaling,
are unaffected (Keefe et al., 1999; He et al., 2005). These data
argue strongly that the HD mutation does not affect the initiation
of the CD4 commitment signal, but rather the downstream inter-
further implies that lineage commitment and positive selection
areatsome level mechanistically distinct, because the HD muta-
tion affects only the former process. Given that both processes
and specialization in signaling pathways downstream of the TCR
scription factor ThPOK, also known as Zbtb7b, Zfp67, or cKrox
(He et al., 2005; Kappes et al., 2006). The expression pattern of
ThPOK during thymic development is tightly regulated in a man-
ner that indicates a specific role in CD4 commitment. Thus,
ThPOK RNA is first detected at the CD4+8lostage, and ThPOK
mRNA expression is substantially higher in class II- than
class I-restricted cells at this stage, consistent with preferential
sion of ThPOK during thymic development causes redirection of
class I-restricted thymocytes to the CD4 lineage. Thus, ThPOK
expression is both necessary and sufficient for CD4 commit-
ment. Importantly, neither overexpression nor inactivation of
ThPOK is sufficient to drive development to the SP stage in the
absence of a positive selection signal, consistent with the view
that ThPOK does not mimic or modulate TCR signaling.
Several additional transcription factors play selective roles
in CD4 or CD8 development and might therefore be involved in
lineage commitment, notably Gata-3 and Runx3. Constitutive
expression of Gata-3 specifically blocks CD8 development,
whereas its conditional deletion blocks CD4 development (Her-
nandez-Hoyos et al., 2003; Nawijn et al., 2001; Pai et al.,
2003). In neither case are affected thymocytes redirected to
the opposite lineage. Hence, Gata-3 is necessary but not suffi-
cient for CD4 development. Runx factors clearly play an impor-
tant role in regulating the CD8 development program, including
the CD8 lineage-specific suppression of CD4 (Sawada et al.,
1994; Siu et al., 1994; Taniuchi et al., 2002; Sato et al., 2005;
Ehlers et al., 2003; Egawa et al., 2007). However, a direct role
of Runx factors in initiating lineage commitment appears unlikely
on the basis of results from constitutive Runx3 transgenic mice.
Thus, although development of SP CD4 thymocytes is impaired
in Runx3 transgenic mice, this is corrected by constitutive CD4
expression (Grueter et al., 2005). CD4 deficiency is known to
cause redirection of class II-restricted thymocytes to the CD8
lineage, presumably by weakening or interrupting TCR signaling
(Matechak et al., 1996; Tyznik et al., 2004). Hence the primary
effect of Runx factors on lineage commitment appears to lie in
controlling CD4 expression.
nate CD4 or CD8 lineage choice remain poorly understood. The
finding that lineage commitment is controlled by differential ex-
these pathways. In particular, through the determination of how
differential expression of ThPOK is controlled, it should ulti-
mately prove possible to identify the upstream pathways that
control lineage commitment. Accordingly, in the present study
weexaminedtwoaspects ofThPOKregulation, i.e.,its upstream
control by TCR signaling and its immediate control by cis-acting
regulatory elements. This study led to two important insights.
First, we provide direct evidence that TCR signaling is an impor-
tant contributor to ThPOK induction. Second, we demonstrate
that differential ThPOK expression is controlled by a transcrip-
tional silencer element. We propose a model whereby CD4 com-
mitment requires inactivation of the silencer, whereas continued
activity of the silencer leads to CD8 commitment.
Regulation of ThPOK by TCR Signaling
ThPOK mRNA amounts are markedly higher in class II- than
class I-restricted CD4+8locells, suggesting that ThPOK induc-
tion is regulated instructively by relative TCR signal strength
(He et al., 2005; Sun et al., 2005). However, in vitro anti-TCR
treatment of uncommitted DP thymocytes failed to induce
ThPOK mRNA (data not shown). Reasoning that additional
in vivo stimuli might be required, we carried out anti-TCR cross-
linking in vivo. For this purpose, we administered anti-TCRb
antibody to MHC class II-deficient (IAb?/?) mice whose thymo-
cytes can only undergo positive selection by class I ligands
and normally express little ThPOK. Antibody treatment led to
strong induction of ThPOK transcription at the CD4+8lostage
by 48 hr (Figure 1A). This was accompanied by the appearance
of SP CD4 thymocytes, as previously reported (Figure 1B)
(Nasreen et al., 2003). Interestingly, antibody treatment did not
induce ThPOK in DP thymocytes, even though they had clearly
received a TCR stimulus, as evidenced by increased expression
are either not susceptible to TCR-mediated ThPOK induction, or
the response is delayed until transition to the CD4+8lostage.
If ThPOK expression in CD4+8lothymocytes requires ongoing
TCR stimulation, then interrupting this signal should lead to ter-
mination of ThPOK expression. This was tested with CD4+8lo
thymocytes from B2m?/?Zbtb7bHD/HDmice, which are under-
going positive selection by class II ligands and exhibit strong
ThPOK induction (Figure 1D, top panel) but cannot—because
of the lack of functional ThPOK—undergo commitment to the
CD4 lineage; as a result, ThPOK is only transiently induced.
When TCR signaling was interrupted by placing sorted CD4+8lo
cells from these mice in suspension culture overnight, ThPOK
transcripts disappeared, consistent with the notion that ThPOK
expression required continued TCR signaling (Figure 1D). Cul-
tured CD4+8locells remained viable and eventually developed
into SP CD8 cells (Figure 1E). If cessation of ThPOK expression
in culture was caused by interruption of TCR signaling, then sup-
plying a TCR stimulus in vitro might restore ThPOK expression.
Indeed, stimulation of cultured thymocytes with anti-CD3 partly
restored ThPOK expression (Figure 1D). Incomplete restoration
could indicate that some cells are already too advanced toward
CD8 commitment to respond to TCR stimulation, or that full
induction requires additional signals not available in vitro. Sorted
Regulation of CD4-CD8 Lineage Commitment
Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc. 347
CD4+8locells from B2m?/?Zbtb7b+/HDmice maintained ThPOK
expression after overnight culture, indicating that most of these
cells are committed to the CD4 lineage, leading to permanent
activation of the ThPOK locus. A possible alternative explanation
for the specific loss of ThPOK transcripts in cultured B2m?/?
Zbtb7bHD/HDthymocytes is that sustained ThPOK induction
involves a ThPOK-mediated positive-feedback loop that is
blocked in the absence of functional ThPOK. However, the fact
that freshly isolated CD4+8locells from B2m?/?Zbtb7bHD/HD
and B2m?/?Zbtb7b+/HDmice expressed similar amounts of
ThPOK transcripts argues against such a feedback loop (data
To test whether TCR signaling was also required for mainte-
nance of ThPOK expression after CD4 commitment, we isolated
SP CD4 thymocytes from antibody stimulated class II-deficient
mice 12 days after anti-TCR treatment, when circulating anti-
body is expected to have largely disappeared. Nevertheless,
ThPOK expression remained high in SP CD4 thymocytes (Fig-
ure S1A available online). Similarly, when class I-restricted thy-
mocytes were redirected to the CD4 lineage by constitutive
expression of ThPOK (He et al., 2005), the resulting SP CD4
thymocytes expressed endogenous ThPOK transcripts, even
though their TCRs are class I specific (data not shown). Con-
versely, when class II-restricted thymocytes were redirected to
the CD8 lineage in B2m?/?Zbtb7bHD/HDmice, the resulting SP
CD8 thymocytes failed to express ThPOK, even in the presence
suggest that ThPOK expression in SP cells is regulated by
Defining the Minimal Genomic Region Required
for Correct Lineage-Specific ThPOK Expression
ThPOK expression during thymic development is regulated at
the level of mRNA, suggesting a transcriptional control mecha-
nism. To identify the responsible cis elements, we first sought
to define the minimal genomic region required for normal ThPOK
regulation. For this purpose, BAC transgenes encompassing the
ThPOK locus were crossed to the Zbtb7bHD/HDbackground, to
determine whether CD4 development was restored, and to the
Figure 1. Strong TCR Stimulus Induces ThPOK mRNA Expression in CD4+8loThymocytes
MHC class II-deficient mice were injected with TCRb antibody, and analyzed at indicated time points. Shown are (A) real-time RT-PCR analysis of ThPOK mRNA
expression for indicated sorted thymocyte subsets. The graphs represent mean ± standard deviation (SD) (n = 2), (B) CD4 and CD8 expression pattern of total
thymocytes at indicated time points, and (C) a comparison of CD69 expression in gated DP and CD4+8losubsets before or 4 days after antibody treatment. (D)
Sorted CD4+8lothymocytes from B2m?/?Zbtb7bHD/HDmice were incubated overnight in the presence or absence of indicated antibodies, and real-time RT-PCR
analysis of ThPOK mRNA was performed on indicated populations. The graphs represent mean ± SD (n = 2).(E) CD4 and CD8 expression pattern of total B2m?/?
compare CD4 and CD8 expression profiles of cultured CD4+8locells after indicated times in culture (in absence of anti-CD3 stimulation). Numbers in the FACS
plot represent percentages of the associated gate. Each experiment was repeated at least three times with similar results.
Regulation of CD4-CD8 Lineage Commitment
348 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.
IAb?/?or OT-1 TCR transgenic backgrounds, to determine
whether MHC class I-restricted cells developed exclusively to
the CD8 lineage. Both a full-length 200 kb BAC transgene,
126.34, and a 26 kb BAC fragment, SPE7.6, extending from
fulfilled both of these requirements, indicating that they each
contained all cis elements required for proper ThPOK regulation
(Figures 2A and 2B; Figure S2). Further reducing the ThPOK ge-
nomic transgene to 17 kb, construct REC3.33, still rescued CD4
development on the Zbtb7bHD/HDbackground, but also severely
impaired CD8 development, indicating that ThPOK expression
was not appropriately suppressed in class I-restricted cells
(Figure S2B). This suggested that a key regulatory element
required for CD4 lineage-specific expression of ThPOK was dis-
rupted in REC3.33 transgenic mice. Comparison of the SPE7.6
and REC3.33 transgenes indicates that the latter lacks a con-
served 500 bp region at the 50end of the ThPOK locus, which
may therefore be necessary for lineage-specific regulation of
ThPOK transcription (Figure S2A, indicated by arrow).
Analysis of public CAGE and EST databases revealed two
alternate ThPOK transcripts, initiating at different genomic posi-
tions(Carninci etal.,2006;CAGE trackofmouse [mm5]genomic
elements database hosted by RIKEN, Japan) (Figure 2C; data
not shown). These transcripts utilize different upstream noncod-
ing exons but the same ThPOK coding exons and are predicted
to encode the same protein. A genome-wide analysis of histone
modifications in human CD4+T cells demonstrated enrichment
for H3-K4 methylation, which is strongly associated with active
promoter regions, upstream of both noncoding exons, consis-
tent with the occurrence of promoters at these positions (Barski
etal., 2007;human genome histone methylation maps byA. Bar-
ski et al., hosted by NHLBI, USA). Real-time RT PCR and 50
RACE analyses demonstrated that both distal and proximal tran-
scripts were produced in murine thymocytes, but with different
Figure 2. Minimal ThPOK Genomic Fragment Necessary for Normal Regulation of Lineage Commitment Comprises 26 kb and Encodes Two
Alternate Promoters with Different Stage-Specificity
(A) CD4 and CD8 expression pattern of gated mature (TCRhiCD69?) thymocytes and PBLs from ThPOK BAC transgenic mice on the Zbtb7bHD/HDbackground,
and similar analysis of thymocytes and PBLs from BAC transgenic mice crossed to the class I-restricted OT-1 TCR transgenic or to MHC class II-deficient mice.
(B)Schematic representation ofBAC transgenic constructs,showing location ofZbtb7b gene. Thenumber offoundersthatshowed transgene expression aswell
as total founders is indicated.
(C) Schematic of Zbtb7b locus, showing positions of exons and splicing patterns of transcripts produced from alternate distal and proximal promoters.
(D)Real-time RTPCRanalysisofsortedthymocytesubsetsfrom B2m?/?mice,showingalternatetranscriptsproduced from distal (darkbars) orproximal Zbtb7b
promoters (white bars). Graphs are mean ± SD (n = 2).
Regulation of CD4-CD8 Lineage Commitment
Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc. 349
developmental kinetics (Figure 2D and data not shown). Thus,
the distal transcript was expressed at highest amounts at the
CD4+8lostage, whereas maximal expression of the proximal
Distal and proximal promoter usage did not correlate strictlywith
uncommitted or CD4-committed status. Thus proximal tran-
mice, which cannot undergo CD4 commitment, whereas distal
transcripts were still detected in mature peripheral CD4+T cells,
i.e., long after CD4 commitment has occurred. The former result
also shows that delayed activity of the proximal promoter does
tal promoter, i.e., does not reflect autoregulation of the proximal
promoter by ThPOK. The fact that both promoters are utilized at
the CD4+8lostage before CD4 commitment suggests common
Figure 3. Multiple DHS Sites at ThPOK
Potential Regulatory Elements
(A)Restriction mapof ThPOK locus,showing loca-
tion of probes used for DHS analysis, and position
of DHS sites or clusters identified in thymocytes
(R = EcoRI, p = PstI, K = KpnI, B = BamHI,
E = EcoRV). (B) Representative DHS analyses us-
ing indicated probes and restriction enzymes.
Analyses were carried out with total thymocytes
from AND or OT-1 TCR transgenic mice, or
Cd3d?/?mice. Red arrows mark DHS sites, B, D,
and F, that are preferentially detected in class
regulation by a common TCR-responsive
element that regulates both promoters.
Identification of Putative
Regulatory Elements at the ThPOK
Locus by DNase I Hypersensitive
DNase I Hypersensitive Site (DHS) analy-
sis was carried out to identify potential
cis-acting elements within the minimal
26 kb ThPOK fragment. Thymocytes
from three kinds of mice were used for
this analysis: (1) MHC class II-restricted
AND TCR transgenic mice, in which
30%–50% of thymocytes belong to the
CD4+8loand SP CD4 subsets and exhibit
high ThPOK expression; (2) MHC class
I-restricted OT-1 TCR transgenic mice,
in which thymocytes develop to the CD8
lineage and exhibit low ThPOK expres-
sion only in the CD4+8losubset; and (3)
Cd3d?/?mice, in which thymocytes are
blocked in positive selection and lack
ThPOK expression (Dave ´ et al., 1997;
data not shown). Five closely spaced
probes (Figure 3A) were used for South-
ern analysis of DNaseI-treated genomic
DNA samples from total thymocytes of
these strains (Figure 3B). Six DHS sites, A–F, were identified
within the region corresponding to construct SPE7.6. Site A co-
incided with the conserved element at the 50end of the ThPOK
locus mentioned above, which isspecifically lacking in construct
REC3.33. Sites A, C, and E were present in all strains of mice ex-
amined, even Cd3d?/?thymocytes, indicating that these sites
are accessible and loaded with transcription factors (TFs) prior
to ThPOK transcription, although specific TFs bound may vary.
These sites did not coincide with putative promoters and may in-
Sites B, D, and F were preferentially detected in AND TCR trans-
genic thymocytes, suggesting that they are associated with TFs
and proximal promoters, whereas site F lies upstream of the first
coding exon. DHS analysis of sorted thymocyte subsets from
Regulation of CD4-CD8 Lineage Commitment
350 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.
AND TCR transgenic mice demonstrated that sites B, D, and
F are lacking in preselection DP CD69?thymocytes from these
mice (data not shown).
Reporter Transgene Recapitulates Normal
Developmental Expression Pattern of ThPOK
A GFP reporter construct, F2F3, was generated that included all
DHS sites identified above, as well as both ThPOK promoters
(Figure 4A). Several independent founders exhibited the same
thymocyte-subset-specific expression pattern, although there
was some founder-specific variation in the amount of GFP
expressed and in the proportion of GFP+cells (data not shown).
Importantly, GFP expression in T cells was restricted to SP CD4
cells in the thymus and periphery, and some CD4+8lothymo-
Figure 4. A 12 kb Genomic Fragment
Regulation of ThPOK Expression
(A) Top panel shows exon and intron organization
of the ThPOK locus, sequence conservation
between mouse and human (% homology is
indicated by color; green = 78–84; yellow = 85–89;
orange = 90–95; red = > 95%), position of DHS
sites, and design of GFP reporter construct F2F3.
(B) Histograms of GFP reporter expression for
gated thymocyte and PBL subsets, as indicated.
Right panel shows real-time RT PCR analysis of
GFP transcripts initiating at alternate distal or
proximal ThPOK promoters in sorted F2F3 thymo-
cyte subsets. Assays utilize specific forward
primers for distal and proximal noncoding exons
in combination with a common reverse GFP
(C) Analysisof GFP expression ingated thymocyte
subsets of F2F3 mice crossed to class I- (OT-1) or
class II- (AND) restricted TCR transgenics. Top
panels show gates used to define indicated sub-
sets. Bottom panels shows GFP expression in in-
cytes, recapitulating the normal expres-
sion of ThPOK (Figure 4B). Relative distal
and proximal promoter usage, as de-
tected by RT-PCR, was also similar to
that of endogenous ThPOK (Figure 4B).
No GFP was detected in other thymocyte
subsets, or in any other peripheral lym-
which normallyalsoexpress endogenous
ThPOK (data not shown). For separate
examination of reporter expression in
MHC class I- and class II-restricted
CD4+8lothymocytes, F2F3 transgenics
were crossed to mice expressing the
OT-1 or AND TCR transgenes, respec-
tively. This revealed substantially higher
GFP expression in class II- than class I-
restricted CD4+8lothymocytes (Figure
4C, bottom panels). This difference in
GFP expression at the CD4+8lostage,
i.e., before lineage commitment, argues
strongly that the F2F3 construct contains an element that
mediates differential responsiveness to MHC class I- versus II-
restricted TCR signals.
Identification of a Dual Silencer-Enhancer Element that
Controls CD4 Lineage-Specific Expression of ThPOK
On the basis of the BAC reconstitution studies discussed above
(Figure S2), we speculated that the 50end of the F2F3 construct
marked by DHS site A might encode a CD8 lineage-specific
silencer. Consistent withthis hypothesis, deletion of this putative
distal regulatory element (DRE), from construct F2F3 (Figure 5A,
construct F2F3DDRE) resulted in promiscuous reporter expres-
sion in both the CD4 and CD8 lineages (Figure 5B, top). A shorter
4 kb construct that includes DHS sites C–F and the proximal
Regulation of CD4-CD8 Lineage Commitment
Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc. 351
Figure 5. Identification of a Distal Regulatory Element Encoding Dual Silencer and Enhancer Functions
(A) The design of GFP reporter constructs is shown. Red, gray, and green boxes indicate positions of DRE element, minimal human CD2 promoter, and GFP
reporter, respectively. The ratio next to each construct name indicates the number of founders that showed GFP expression out of total founders identified
by PCR. This ratio varies between constructs, presumably in accordance with their relative susceptibility to position-dependent silencing; different founders
(in some instances, only a single informative founder for a particular construct is available; such data are only included if they are confirmed by results from
relevant closely related constructs).
(B) Histograms illustrating DRE silencer function show GFP expression for gated thymocyte and PBL subsets from indicated reporter lines.
(C) Histogramsillustrating DRE enhancer function show analysis of GFP expressionin F2c-pCD2 reporter line 261, on a non-TCR transgenic background (top), or
crossed to class I- (OT-1) or class II- (AND) TCR transgenics.
(D)DREmediates specificexpressioninclassII-restricted CD4+8lothymocytesincombination withthedistalThPOKpromoter. HistogramsofGFPexpression for
gated thymocyte and PBL subsets from indicated reporter lines. Data for construct F5.3, which is not expressed at any stage, is not shown.
(E) Real-time RT PCR analysis of GFP and endogenous ThPOK expression in sorted thymocyte subsets from F2 transgenic mice. Note that the GFP?CD4+8lo
subset shows no expression of GFP mRNA, indicating a tight correlation between the onset of reporter transcription and GFP protein expression.
Regulation of CD4-CD8 Lineage Commitment
352 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.
promoter, but excludes the DRE and the distal promoter,
showed a similar promiscuous expression pattern (Figure 5A,
construct F3e). Inserting the DRE in front of construct F3e
restored CD4 specificity (construct F2cF3e), indicating that the
DRE silencer functions independently of its distance from
the promoter or enhancer elements that it regulates. Reversing
the orientation of the DRE also did not affect silencer function
(data not shown). Deletion of the DRE from the REC3.33 ThPOK
class I-restricted thymocytes, consistent with the observed phe-
notype of REC3.33 transgenic mice (Figure S2). We conclude
that the DRE encodes a position- and orientation-independent
in the SP CD8 lineage.
insufficient for transgene expression (Zhumabekov et al., 1995).
Construct F2c-pCD2 mediated GFP expression in SP CD4 thy-
mocytes and peripheral CD4+T cells (Figure 5B), indicating that
mice expressing the OT-1 or AND TCR transgenes, respectively.
GFP expression was 6–10-fold higher in class II- than class I-
restricted CD4+8lothymocytes. Importantly, the entire CD4+8lo
population in AND transgenic mice showed elevated GFP ex-
pression relative to CD4+8locells from OT-1 mice (Figure 5C).
Given that only a fraction of class II-restricted cells is committed
infer that F2c-pCD2 reporter expression in these cells precedes
CD4 lineage commitment. These data indicate that the DRE
element in conjunction with the heterologous CD2 promoter is
sufficient for differential ThPOK expression in CD4 versus CD8-
committed cells, and in uncommitted class II- versus class I-
restricted CD4+8locells. Specificity of the DRE enhancer could
be inherent or might be imposed by the associated silencer.
It was important to establish whether the DRE also functions
as an enhancer in the context of the endogenous ThPOK locus,
in particular with respect to the distal promoter, which is the pre-
dominant promoter at the CD4+8lostage. For testing of this, four
reporter constructs were generated that contained the distal
promoter together with different extents of 50flanking DNA (Fig-
ure 5A, constructs F5, F5.1, F2, and F5.3). Analysis of corre-
sponding transgenic mice revealed three main points: First, all
constructs that included the DRE exhibited CD4 lineage-specific
expression (Figure 5D). CD4 specificity requires the DRE, be-
cause a control construct in which the hCD2 enhancer is placed
infrontof thedistalpromoter insteadof theDRE showedexpres-
sion in both lineages (data not shown). Second, deletion of the
F5.3), indicating that the DRE enhancer is essential for expres-
sion of this construct (although other endogenous enhancers
may contribute to distal promoter activity in vivo). Third, GFP
expression by all three constructs was much lower in mature pe-
ripheral CD4+T cells than in SP CD4 thymocytes, in contrast to
the parental F2F3 construct, which is expressed equally at both
stages. At the mRNA level, reporter expression was even more
restricted, i.e., essentially it was limited to CD4+8lothymocytes
(Figure 5E). The discordance between mRNA and protein ex-
pared to mRNA. It is informative that GFP protein persists only in
SP CD4 but not SP CD8 thymocytes, because it implies that
reporter transcription at the CD4+8lostage is limited to class II-
restricted thymocytes, which are the direct precursors of SP
CD4 cells. Hence, the DRE enhancer in the context of the distal
promoter supports mRNA expression specifically in class II-
restricted CD4+8lothymocytes. Restricted expression to the
CD4+8losubset appears to require the combination of the DRE
with the distal ThPOK promoter, because neither the combina-
tion of the DRE with the hCD2 promoter nor the combination of
the hCD2 enhancer with the distal ThPOK promoter (data not
shown) shows this stage-specificity.
Identification of a Non-Lineage-Specific Enhancer
at the ThPOK Locus
Deletion of the DRE from construct F2F3 did not abolish reporter
expression, indicating the presence of an additional enhancer
within this construct. In contrast, deletion of the DRE from the
smaller F2 construct completely abolished reporter expression
(construct F5.3), indicating that the additional enhancer is lack-
ing from F2 or acts only on the proximal promoter. To map this
second enhancer, a further series of reporter transgenes was
generated that were centered on the proximal promoter (Fig-
ure S3A). Analysis of corresponding transgenic mice revealed
three main points: First, all constructs that included the region
around DHS site C exhibited expression in both CD4 and CD8
lineages (constructs F3, F3d–f). Some of these constructs addi-
tionally exhibited expression at the DP stage, in particular con-
struct F3d (Figure S3). Second, deletion of DHS site C caused
dramatic reduction in reporter expression, particularly in the
CD8 lineage (constructs F3b, F3bF6), indicating that this region
encoded an important positive regulatory element, designated
as the ‘‘general T lymphoid element’’ (GTE). In contrast, deleting
sion (Figure S3B, construct F3d-DPRE). Third, although deletion
of the GTE severely diminished reporter expression, some ex-
pression remained, implying the existence of another, weaker
enhancer. This remaining expression is abolished by simulta-
neous deletion of DHS sites C and E (construct F3b.1), indicating
that another positive regulatory element, designated as the
‘‘proximal regulatory element’’ (PRE), maps near DHS site E.
Interestingly, constructs that depend on the PRE for enhancer
function exhibit preferential expression in the CD4 lineage (con-
structs F3b, F3bF6), although it is unclear whether this prefer-
ence is dependent on the PRE, the proximal promoter, or the
combination of both. Only a subset of SP CD4 thymocytes and
peripheral CD4+T cells exhibited PRE-dependent reporter ex-
pression, which may mark a particular Thelper subset or may re-
flect a high degree of random transgene silencing for these con-
structs. Finally, to test the functional relevance of DHS site F,
located immediately upstream of the ThPOK coding exons, we
modified construct F3 by inserting a reporter cDNA directly
into the proximal noncoding exon, allowing the downstream
splice acceptor-reporter cassette that includes DHS site F to
be omitted (Figure S3B). Omission of DHS site F did not alter
the reporter expression pattern for constructs F3 or F3b, indicat-
ing that it does not markedly contribute to transcription in the
context of these constructs (Figure S3B, and data not shown).
Regulation of CD4-CD8 Lineage Commitment
Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc. 353
Figure 6. Mutational Dissection of DRE Element Defines 80 bp Core Element
(A) Schematic of DRE mutants indicating deletion boundaries, mutations, and positions of conserved TF consensus sites. The gray box indicates the 80 bp core
region that is essential for silencer function.
(B) Histograms of GFP expression for gated SP CD4 and CD8 peripheral T cells from indicated reporter lines. Numbers in the histograms indicate % GFP+cells.
(C) The bar graph summarizesthe proportion of GFP+T cells belonging to the CD8 lineage for multiple founders of each construct (inexample at left % GFP+CD8
Regulation of CD4-CD8 Lineage Commitment
354 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.
Consistent with lack of function for DHS site F, construct F3b.1,
which lacks the GTE and PRE elements but retains DHS site F,
shows no expression.
Functional Dissection of the DRE Element
The above results demonstrate that the DRE is the main element
responsible for specificity of ThPOK transcription during thymic
development. In particular, our data suggest a mechanism
whereby the DRE silencer prevents ThPOK transcription, unless
restricted TCR signal. To elucidate the pathway(s) that control
this switch, it is important to identify the transacting factors
(TFs) that regulate DRE function. Interspecies nucleotide com-
parison of the DRE indicates that a region of 300 bp is strongly
(Figure S4A, colored regions). A number of conserved sites of
potential functional relevance occur within this region, including
one Gata, two Runx, two NF-kB, and three E-box consensus
sites. Three deletion mutants of the DRE element were gener-
ated in order to further narrow the region under consideration:
DF2H, D2, and D3, which each lacked about 170 bp of the con-
tion, the variant elements were inserted upstream of reporter
construct F2F3DDRE, which lacks a DRE element and thus ex-
hibits promiscuous expression in both CD4 and CD8 lineages
(Figure S4B), and the resulting constructs were used to generate
transgenic mice. In the transgenic assay, silencing function of
the variant DRE elements was measured by their ability to sup-
press GFP reporter expression in CD8 cells (Figures 6B and
6C). Constructs D2 and DF2H showed quite efficient suppres-
sion of GFP expression in CD8 cells, demonstrating that the de-
leted regions are dispensable for silencing. Of note, the D2 and
DF2H deletions removed the Gata and Runx consensus motifs,
respectively, indicating that neither is essential for silencer func-
tion. In contrast, construct D3 largely fails to repress GFP ex-
pression in CD8 cells, indicating that a central 80 bp core region
(Figure 6A, the 80 bp core indicated in gray). Comparison of mul-
tiple D3 founders suggests that this construct may retain some
weak partial silencing activity, so that a minor contribution to
silencing by sites outside of this region cannot be excluded
(Figure 6C, note that the average proportion of GFP+CD8 cells
seems somewhat reduced in D3 compared to DDRE mice). By
inserting the DF2H DRE variant in front of the hCD2 promoter
(construct F2cDF2H-pCD2), which lacks its own enhancer, we
further demonstrate that Runx sites were also not required for
enhancer activity of the DRE element (Figure 6B). The 80 bp
core region required for silencer function contains both an
NFkB and E-box motif. E-box motifs could mediate regulation
by E proteins such as E2A and HEB whose activity is known to
be controlled by TCR-mediated signals. However, mutation of
all three E-box motifs within the DRE element does not markedly
impair silencing activity (Figures 6B and 6C). A functional role for
the NFkB site cannot be excluded at this point, although previ-
ous analyses of knockout mice affecting this pathway do not
support a role in CD4-CD8 lineage commitment. Hence, it
appears likely that the DRE silencer is regulated by a factor (or
factors) whose role in lineage commitment is not yet recognized.
ThPOK induction represents the earliest available indicator of
CD4 commitment, thus providing a valuable new access point
from which to elucidate the responsible upstream pathways.
The present study contributes to this goal in two respects. First,
we provide compelling evidence of mechanistic linkage between
TCR signaling and ThPOK induction. Second, we demonstrate
that ThPOK induction is controlled at the level of transcription
and identify a key element, the distal regulatory element, that
mediates selective expression in MHC class II-restricted thymo-
cytes. The DRE encodes dual silencer and enhancer functions,
thus making it an ideal molecular switch. We propose a mecha-
depend on overcoming DRE silencer function.
Our observations that in vivo antibody-mediated TCR treat-
ment leads to ThPOK induction in CD4+8lothymocytes and
that CD4+8lothymocytes require TCR stimulation for continued
ThPOK expression in vitro together provide strong support for
a causal relationship between TCR engagement and ThPOK in-
duction. The simplest and most plausible interpretation of these
experiments is that TCR stimulation of CD4+8lothymocytes
directly triggers ThPOK transcription in those cells. The observa-
tion that ThPOK is not induced until the CD4+8lostage is most
consistent with the kinetic signaling model of lineage commit-
ment, and less so with the instructional model, which postulates
that lineage-determining TCR signals are transmitted at the DP
stage.Nevertheless therearetwoimportant caveatsto thisinter-
pretation. First, ThPOK induction at the CD4+8lostage could
represent a delayed response to TCR signals received at the
DP stage, which might, for instance, initiate changes in chroma-
tin conformation that are essential for later ThPOK induction.
Second, we cannot exclude the possibility that additional stimuli
are required for CD4 commitment. Definitive resolution of these
issues will require elucidation of the upstream pathways that
control ThPOK induction.
We have identified several important cis elements at the
ThPOK locus that control its differential expression in class I-
versus class II-restricted thymocytes. The ThPOK locus is
marked by a dense concentration of conserved noncoding
DNA regions, many of which are preserved even between pla-
cental and marsupial mammals, implying important regulatory
functions. Six distinct DHS sites and clusters are accessible in
thymocytes and map within or near these conserved regions.
Two DHS sites, B and D, correlate with ThPOK transcription
and mark the distal and proximal promoters, respectively,
whereas three constitutive sites, A, C, and E, correspond to dis-
of the latter elements acts as an enhancer but with different line-
age specificity. The PRE in the context of the proximal promoter
shows partial preference for the CD4 lineage, although this
If the proportion of GFP+CD8 cells out of the total GFP+T cells is the same as this value, it indicates that silencing is abolished, whereas if it is significantly lower,
silencing is maintained. Graphs are mean ± SD (n is indicated above each bar).
Regulation of CD4-CD8 Lineage Commitment
Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc. 355
preference may require cooperation with the proximal promoter.
The GTE in the context of the proximal promoter shows promis-
cuous activity in both CD4 and CD8 lineages. Finally, the DRE
in the context of the distal promoter or the heterologous hCD2
promoter is active only in class II-restricted thymocytes.
We propose that the DRE is the primary cis element at the
ThPOK locus responsible for receiving and translating upstream
lineage-specifying signals, for the following reasons: The DRE
mediates exclusive expression in class II-restricted thymocytes
both in combination with the endogenous distal promoter and
the heterologous hCD2 promoter. The latter result indicates
that specificity to class II-restricted thymocytes is encoded
by the DRE rather than the distal promoter. The DRE mediates
mitment may be decided. In fact, in the context of the distal pro-
moter, the DRE mediates transcription exclusively at this stage.
The DRE imposes specificity for class II-restricted cells on other
cis elements that lack this inherent capacity, notably the GTE el-
ement. Finally, our studies have identified no other element that
confers specificity to class II-restricted cells and is also active
at the CD4+8lostage. Although the PRE mediates preferential
expression in the CD4 lineage, it supports little to no expression
at the critical CD4+8lostage and may be more relevant for
A striking feature of the DRE is that it functions as both a
silencer and an enhancer. Silencer function is essential for
enforcing CD4 lineage-specific transcription and for proper reg-
ulation of lineage commitment. Whether or not the enhancer
Figure 7. Model of ThPOK Transcriptional
Schematic representation of the activity of ThPOK
cis elements at different thymocyte stages, as in-
dicated. Approximate positions of promoters,
enhancers, and the DRE element are indicated
by boxes, triangles, and ovals, respectively. Red,
green, and striped shading of elements indicates
that they are repressed by the DRE silencer, or ex-
hibit full and partial activity, respectively. Strong
or weak promoter activity is indicated by thick or
thin arrows. Arrows extending from one cis ele-
ment to another indicate functional interactions,
whereby + and ? symbols denote positive or
function of the DRE is also essential for
lineage commitment is unknown and
will be difficult to assess unless these
functions can be separated. If these
it would indicate that the same regula-
tory site controls both silencer and en-
hancer activities, and that the choice be-
tween them depends on competition
between silencing and activating factors
for binding to this site. A key unresolved
question is the identity of the factors that
control DRE activity. On the basis of our
mutational analysis, we can rule out di-
rect regulation of the DRE element by at least two potential sus-
pects, i.e., Gata-3 and Runx3. However, this does not exclude
indirect regulation of DRE activity by one of these factors. Stud-
ies to identify the factors that directly control DRE activity are
currently underway employing affinity purification and other
We propose the following model for transcriptional control of
ThPOK during thymic development (Figure 7): At the DP stage,
ThPOK transcription is actively repressed by a DRE-dependent
mechanism, consistent with induction of ThPOK when the DRE
is deleted. At the CD4+8lostage, strong TCR signaling mediated
by class II ligands activates or induces a transacting factor that
converts the DRE from silencer to enhancer mode and initiates
ThPOK transcription. Alternatively, TCR-mediated derepression
of the DRE silencer may be initiated at the DP stage but not
completed until the CD4+8lostage. In SP CD4 thymocytes and
peripheral CD4+T cells, ThPOK transcription is maintained
by a TCR-independent mechanism. This seems to require the
GTE and PRE elements, because the DRE loses the ability to
act as an enhancer at the SP CD4 stage, at least in the context
of the distal promoter. Inactivation of the ThPOK locus in SP
CD8 thymocytes and peripheral CD8+T cells requires the DRE,
at least for initiation but perhaps not for permanent maintenance
of silencing. In this model, other regulatory elements at the
ThPOK locus, including enhancers and promoters, are not
targets of TCR signaling but become active by default when
DRE silencer activity is blocked. Although these elements are
not involved in initiating CD4 commitment, they are probably
Regulation of CD4-CD8 Lineage Commitment
356 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.
important in achieving the appropriate level of ThPOK expres-
sion required to complete CD4 commitment. Whether any
elements other than the DRE are individually indispensable for
CD4 commitment remains to be established.
CD4 (Killeen and Littman, 1993) (kindly provided by S. Reiner) and OT-1-TCR
(Hogquistetal.,1994)transgeniclines,aswellasZbtb7b?/?(Dave ´ etal.,1998),
MHC IAb?/?, and Cd3d?/?(Dave ´ et al., 1997) mice, have been described
previously. All other transgenic lines described in this paper were generated
by the FCCC Transgenic Facility. Animal care was in accordance with National
Institutes of Health (NIH) guidelines.
TCR Stimulation Assays
For in vivo a-TCR stimulation, young adult IAb?/?mice were injected intraper-
stimulation, 106sorted thymocytes were cultured in DMEM with 10% FSC on
96-well plates precoated with 20 mg/ml each of a-CD3 (2C11) and a-CD28.
BAC constructs 368.8, SPE7, and REC3.33 were derived form BAC clone
368D24 (C57BL/6 RPCI-23 BAC library, Genome Sequence Centre, BC Can-
cer Agency) by restriction digestion or recombineering, according to estab-
lished protocols (see the recombineering web site hosted by NCI-Frederick).
ThPOK coding exon in the context of a 1.3 kb genomic fragment (this extends
from 1.3 kb upstream of the first ThPOK coding exon to the ATG start codon
and includes the splice acceptor site). Final reporter constructs were gener-
ated by inserting the desired ThPOK genomic fragments into a HindIII site up-
stream of this splice acceptor cassette. For construct F2cpCD2, the EGFP
cassette was inserted downstream of the minimal hCD2 promoter (Zhumabe-
kov et al., 1995). For constructs utilizing the dsRed reporter (pDsRedT.4.str,
noncoding exon (constructs F3-dsRed and F3dDPRE-dsRed). Precise
sequences of all constructs are provided in supplementary materials.
Cells were prepared from thymus and peripheral blood and analyzed by flow
cytometry according to standard procedures. All antibodies were obtained
Real Time RT-PCR
Real-time RT PCR analysis for Th-POK and EGFP was carried out according
to the probe-based method and analyzed by the comparative Ct method
(compared to b-actin). Primer and probe sequences are provided in the
DNase I Hypersensitivity Analysis
DHS analysis was carried out according to standard procedures. In brief,
107–108primary thymocytes were allowed to swell in ice-cold high-salt buffer
(10 mMTris pH 7.4, 100 mM NaCl, 30 mM MgCl2, 0.1% NP-40) for 15 min, and
nuclei were isolated by centrifugation. The nuclear pellet was then resus-
pended in low-salt buffer (10 mM Tris pH 7.4, 10 mM NaCl, 3 mM MgCl2,
0.1% NP-40 buffer), divided into 120 ml aliquots, and briefly treated with vary-
ing concentrations of DNase I (Worthington: DPRF). DNase I digestion was
stopped by addition of an equal volume of cell lysis buffer and DNA was
isolated with Genomic DNA purification kit (Puregene). Southern blotting
was carried out according to standard procedures with different probes, as
Additional Experimental Procedures and four figures are available at http://
flow cytometry, E. Nicolas for real-time PCR, and G. Olson for technical assis-
tance. This research was supported by grants from the NIH to D.J.K. (AI42915)
and Fox Chase Cancer Center (CA06927), and by an appropriation from the
Commonwealth of Pennsylvania.
Received: December 22, 2007
Revised: January 28, 2008
Accepted: February 7, 2008
Published online: March 13, 2008
the control of positive selection. Immunol. Rev. 191, 79–96.
Barski, A., Cuddapah, S., Cui, K., Roh, T.Y., Schones, D.E., Wang, Z., Wei, G.,
Chepelev, I., and Zhao, K. (2007). High-resolution profiling of histone
methylations in the human genome. Cell 129, 823–837.
Barthlott, T., Kohler, H., Pircher, H., and Eichmann, K. (1997). Differentiation of
CD4(high)CD8(low) coreceptor-skewed thymocytes into mature CD8 single-
positive cells independent of MHC class I recognition. Eur. J. Immunol. 27,
Bommhardt, U., Basson, M.A., Krummrei, U., and Zamoyska, R. (1999).
Activation of the extracellular signal-related kinase/mitogen-activated protein
kinase pathway discriminates CD4 versus CD8 lineage commitment in the
thymus. J. Immunol. 163, 715–722.
Borgulya, P., Kishi, H., Muller, U., Kirberg, J., and von Boehmer, H. (1991). De-
velopment of the CD4 and CD8 lineage of T cells: Instruction versus selection.
EMBO J. 10, 913–918.
vealingparadox:Whymajorhistocompatibility complex I-signaledthymocytes
‘‘paradoxically’’ appear as CD4+8lo transitional cells during positive selection
of CD8+ T cells. J. Exp. Med. 197, 1709–1719.
Brugnera, E., Bhandoola, A., Cibotti, R., Yu, Q., Guinter, T.I., Yamashita, Y.,
Sharrow, S.O., and Singer, A. (2000). Coreceptor reversal in the thymus:
Signaled CD4+8+ thymocytes initially terminate CD8 transcription even
when differentiating into CD8+ T cells. Immunity 13, 59–71.
Carninci, P., Sandelin, A., Lenhard, B., Katayama, S., Shimokawa, K.,
Ponjavic, J., Semple, C.A., Taylor, M.S., Engstrom, P.G., Frith, M.C., et al.
(2006). Genome-wide analysis of mammalian promoter architecture and
evolution. Nat. Genet. 38, 626–635.
Chan, S.H., Cosgrove, D., Waltzinger, C., Benoist, C., and Mathis, D. (1993).
Another view of the selective model of thymocyte selection. Cell 73, 225–236.
Chan, S.H., Waltzinger, C., Baron, A., Benoist, C., and Mathis, D. (1994). Role
of coreceptors in positive selection and lineage commitment. EMBO J. 13,
Chan, S., Correia-Neves, M., Dierich, A., Benoist, C., and Mathis, D. (1998).
Visualization of CD4/CD8 T cell commitment. J. Exp. Med. 188, 2321–2333.
Dave ´, V.P., Zhensheng, C., Browne, C., Alarcon, B., Fernandez-Miguel, G.,
Lafaille, J., de la Hera, A., Tonegawa, S., and Kappes, D.J. (1997). CD3d
deficiency arrests development of the ab but not the gd T cell lineage.
EMBO J. 16, 1360–1370.
Dave ´, V.P., Allman, D., Keefe, R., Hardy, R.R., and Kappes, D.J. (1998). HD
mice: A novel mouse mutant with a specific defect in the generation of
CD4(+) T cells. Proc. Natl. Acad. Sci. USA 95, 8187–8192.
Davis, C.B., Killeen, N., Crooks, M.E., Raulet, D., and Littman, D.R. (1993).
Evidence for a stochastic mechanism in the differentiation of mature subsets
of T lymphocytes. Cell 73, 237–247.
Ehlers, M., Laule-Kilian, K., Petter, M., Aldrian, C.J., Grueter, B., Wurch, A.,
Yoshida, N., Watanabe, T., Satake, M., and Steimle, V. (2003). Morpholino
antisense oligonucleotide-mediated gene knockdown during thymocyte
development of CD4-/CD8+ thymocytes. J. Immunol. 171, 3594–3604.
Regulation of CD4-CD8 Lineage Commitment
Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc. 357
Egawa, T., Tillman, R.E., Naoe, Y., Taniuchi, I., and Littman, D.R. (2007). The Download full-text
role of the Runx transcription factors in thymocyte differentiation and in
homeostasis of naive T cells. J. Exp. Med. 204, 1945–1957.
Fischer, A.M., Katayama, C.D., Pages, G., Pouyssegur, J., and Hedrick, S.M.
(2005). The role of erk1 and erk2 in multiple stages of T cell development.
Immunity 23, 431–443.
Grueter, B., Petter, M., Egawa, T., Laule-Kilian, K., Aldrian, C.J., Wuerch, A.,
Ludwig, Y., Fukuyama, H., Wardemann, H., Waldschuetz, R., et al. (2005).
Runx3 regulates integrin alpha E/CD103 and CD4 expression during develop-
ment of CD4-/CD8+ T cells. J. Immunol. 175, 1694–1705.
Guidos, C.J., Danska, J.S., Fathman, C.G., and Weissman, I.L. (1990). T cell
receptor-mediated negative selection of autoreactive T lymphocyte precur-
sors occurs after commitment to the CD4 or CD8 lineages. J. Exp. Med.
He, X., He, X., Dave ´, V.P., Zhang, Y., Hua, X., Nicolas, E., Xu, W., Roe, B.A.,
and Kappes, D.J. (2005). The zinc finger transcription factor Th-POK regulates
CD4 versus CD8 T-cell lineage commitment. Nature 433, 826–833.
Hernandez-Hoyos, G., Sohn, S.J., Rothenberg, E.V., and Alberola-Ila, J.
(2000). Lck activity controls CD4/CD8 T cell lineage commitment. Immunity
Hernandez-Hoyos, G., Anderson, M.K., Wang, C., Rothenberg, E.V., and
Alberola-Ila, J. (2003). GATA-3 expression is controlled by TCR signals and
regulates CD4/CD8 differentiation. Immunity 19, 83–94.
Hogquist, K.A., Jameson, S.C., Heath, W.R., Howard, J.L., Bevan, M.J., and
Carbone, F.R. (1994). T cell receptor antagonist peptides induce positive
selection. Cell 76, 17–27.
Kappes,D.J.,He,X.,and He,X.(2006). Role of thetranscriptionfactorTh-POK
in CD4:CD8 lineage commitment. Immunol. Rev. 209, 237–252.
Keefe, R., Dave ´, V., Allman, D., Wiest, D., and Kappes, D.J. (1999). Regulation
of lineage commitment distinct from positive selection. Science 286, 1149–
Killeen, N., and Littman, D.R. (1993). Helper T cell development in the absence
of CD4-p56lck association. Nature 364, 729–732.
Kydd, R., Lundberg, K., Vremec, D., Harris, A.W., and Shortman, K. (1995).
Intermediate steps in thymic positive selection. Generation of CD4–8+ T cells
in culture from CD4+8+, CD4int8+, and CD4+8intthymocytes with up-regulated
levels of TCR-CD3. J. Immunol. 155, 3806–3814.
Legname, G., Seddon, B., Lovatt, M., Tomlinson, P., Sarner, N., Tolaini, M.,
Williams, K., Norton, T., Kioussis, D., and Zamoyska, R. (2000). Inducible
expression of a p56Lck transgene reveals a central role for Lck in the differen-
tiation of CD4 SP thymocytes. Immunity 12, 537–546.
Liu, X., and Bosselut, R. (2004). Duration of TCR signaling controls CD4–CD8
lineage differentiation in vivo. Nat. Immunol. 5, 280–288.
Lucas, B., and Germain, R.N. (1996). Unexpectedly complex regulation of
CD4/CD8 coreceptor expression supports a revised model for CD4+CD8+
thymocyte differentiation. Immunity 5, 461–477.
Lundberg, K., Heath, W., Kontgen, F., Carbone, F.R., and Shortman, K. (1995).
Intermediate steps in positive selection: Differentiation of CD4+8int TCRint
thymocytes into CD4–8+TCRhi thymocytes. J. Exp. Med. 181, 1643–1651.
Matechak, E.O., Killeen, N., Hedrick, S.M., and Fowlkes, B.J. (1996). MHC
class II-specific T cells can develop in the CD8 lineage when CD4 is absent.
Immunity 4, 337–347.
Nasreen, M., Ueno, T., Saito, F., and Takahama, Y. (2003). In vivo treatment of
class II MHC-deficient mice with anti-TCR antibody restores the generation of
circulating CD4 Tcells and optimalarchitectureofthymicmedulla.J.Immunol.
Nawijn, M.C., Ferreira, R., Dingjan, G.M., Kahre, O., Drabek, D., Karis, A.,
Grosveld, F., and Hendriks, R.W. (2001). Enforced expression of GATA-3
during T cell development inhibits maturation of CD8 single-positive cells
and induces thymic lymphoma in transgenic mice. J. Immunol. 167, 715–723.
Pai, S.Y., Truitt, M.L., Ting, C.N., Leiden, J.M., Glimcher, L.H., and Ho, I.C.
(2003). Critical roles for transcription factor GATA-3 in thymocyte develop-
ment. Immunity 19, 863–875.
Ravichandran, K.S., and Burakoff, S.J. (1994). Evidence for differential intra-
cellular signaling via CD4 and CD8 molecules. J. Exp. Med. 179, 727–732.
Robey, E.A., Fowlkes, B.J., Gordon, J.W., Kioussis, D., von Boehmer, H.,
Ramsdell, F., and Axel, R. (1991). Thymic selection in CD8 transgenic mice
supports an instructive model for commitment to a CD4 or CD8 lineage. Cell
Robey, E., Itano, A., Fanslow, W.C., and Fowlkes, B.J. (1994). Constitutive
CD8 expression allows inefficient maturation of CD4+ helper T cells in class
II major histocompatibility complex mutant mice. J. Exp. Med. 179, 1997–
Sato, T., Ohno, S., Hayashi, T., Sato, C., Kohu, K., Satake, M., and Habu, S.
(2005). Dual functions of Runx proteins for reactivating CD8 and silencing
CD4 at the commitment process into CD8 thymocytes. Immunity 22, 317–328.
Sawada, S., Scarborough, J.D., Killeen, N., and Littman, D.R. (1994). A line-
age-specific transcriptional silencer regulates CD4 gene expression during
T lymphocyte development. Cell 77, 917–929.
Sharp, L.L., Schwarz, D.A., Bott, C.M., Marshall, C.J., and Hedrick, S.M.
(1997). The influence of the MAPK pathway on T cell lineage commitment.
Immunity 7, 609–618.
Singer, A., and Bosselut, R. (2004). CD4/CD8 coreceptors in thymocyte devel-
opment, selection, and lineage commitment: Analysis of the CD4/CD8 lineage
decision. Adv. Immunol. 83, 91–131.
Siu, G., Wurster, A.L., Duncan, D.D., Soliman, T.M., and Hedrick, S.M. (1994).
A transcriptional silencer controls the developmental expression of the CD4
gene. EMBO J. 13, 3570–3579.
Sun, G., Liu, X., Mercado, P., Jenkinson, S.R., Kypriotou, M., Feigenbaum, L.,
Galera, P., and Bosselut, R. (2005). The zinc finger protein cKrox directs CD4
lineage differentiation during intrathymic T cell positive selection. Nat. Immu-
nol. 6, 373–381.
Suzuki, H., Punt, J.A., Granger, L.G., and Singer, A. (1995).Asymmetric signal-
ing requirements for thymocyte commitment to the CD4+ versus CD8+ T cell
lineages:A newperspective on thymiccommitmentandselection. Immunity2,
and Littman, D.R. (2002). Differential requirements for Runx proteins in CD4
repression and epigenetic silencing during T lymphocyte development. Cell
Tyznik, A.J., Sun, J.C., and Bevan, M.J. (2004). The CD8 population in CD4-
deficient mice is heavily contaminated with MHC class II-restricted T cells.
J. Exp. Med. 199, 559–565.
Veillette, A.,Bookman, M.A.,Horak, E.M.,and Bolen,J.B.(1988). TheCD4and
CD8 T cell surface antigens are associated with the internal membrane
tyrosine-protein kinase p56lck. Cell 55, 301–308.
Zhumabekov, T., Corbella, P., Tolaini, M., and Kioussis, D. (1995). Improved
version of a human CD2 minigene based vector for T cell-specific expression
in transgenic mice. J. Immunol. Methods 185, 133–140.
Regulation of CD4-CD8 Lineage Commitment
358 Immunity 28, 346–358, March 2008 ª2008 Elsevier Inc.