International Immunology, Vol. 15, No. 3, pp. 393±402
doi:10.1093/intimm/dxg040, available online at www.intimm.oupjournals.org
ã 2003 The Japanese Society for Immunology
Expression of recombination-activating gene
in mature peripheral T cells in Peyer's patch
Eisuke Kondo1, Hiroshi Wakao1, Haruhiko Koseki2, Toshitada Takemori3,
Satoshi Kojo1, Michishige Harada1, Minako Takahashi1, Sakura Sakata1,
Chiori Shimizu1, Toshihiro Ito1, Toshinori Nakayama1 and Masaru Taniguchi1
1Laboratory for Immune Regulation, RIKEN Research Center for Allergy and Immunology, and Department
of Molecular Immunology, and 2Department of Molecular Embryology, Graduate School of Medicine,
Chiba University, Chiba 260-8670, Japan
3Department of Immunology, National Institute of Infectious Diseases Tokyo 162-8640, Japan
Keywords: green ¯uorescent protein, knock-in mice, RAG2, TCR rearrangement
Abstract
Recombination-activating gene (RAG) 1 and 2 are essential for the gene rearrangement of antigen
receptors of both T and B cells. To investigate RAG gene expression in peripheral lymphoid
organs other than the thymus and bone marrow, we established mice in which a green ¯uorescent
protein (GFP) gene is knocked-in the RAG2 gene locus (RAG2-GFP mice). In the thymus and bone
marrow of heterozygous RAG2-GFP mice, as expected, GFP expression was detected in the
appropriate stages of developing T and B cells. Interestingly, only a fraction of Thy-1.2+ cells in the
Peyer's patch were found to be GFP+ amongst the peripheral lymphoid organs. The GFP+ cells
expressed high levels of surface TCRb and CD3, suggesting mature T cells with rearranged TCRab.
However, they showed activated/memory phenotypes, i.e. CD45RBlow, CD69high, CD44high and
CD62Llow, and belonged to a CD4+CD8+ population expressing c-kit, IL-7R and pTa characteristic of
immature developing lymphocytes. Moreover, RAG+ Peyer's patch T cells seem to be of thymic
origin as judged by their expression of CD8ab. These results show that there exists a fraction of
mature T cells expressing RAG genes in the Peyer's patch, implying a potential for a secondary
rearrangement of TCR in extrathymic tissues.
Introduction
The development of T and B cells is controlled by rearrangement of the genes coding for the variable region of the BCR
or TCR, which are composed of variable (V), diversity (D)
and joining (J) gene segments (1). The lymphocyte-speci®c
proteins, recombination-activating gene (RAG) 1 and 2,
initiate V(D)J recombination events. RAG genes have long
been considered to be expressed only in immature
lymphocytes in bone marrow and thymus, and are readily
down-regulated in mature lymphocytes. In developing
thymocytes, RAG1 and 2 are expressed ®rst in CD4±CD8±
double-negative (DN) thymocytes, which can be further
subdivided based on CD44 and CD25 expression. RAG
expression initiates at the CD44+CD25+ stage and continues
to be detected at the next CD44±CD25+ stage in which
TCRb, g and d rearrangement commences (2,3). Functional
V(D)J rearrangements lead to the cell-surface expression of
pre-TCR composed of TCRb and pTa (4), and down-regulate
RAG gene expression at the CD44±CD25± stage.
Subsequently, during differentiation into CD4+CD8+ doublepositive (DP) thymocytes, TCRa gene rearrangement occurs
with high expression of RAG gene products (2,5). RAG
expression is then down-regulated again during the maturation of DP thymocytes into CD4 or CD8 single-positive (SP)
mature thymocytes (6).
Similarly, RAG gene expression in immature B cells is
developmentally controlled. Two periods of B cell development express RAG genes. First, RAG is detected in pro-B cells
that undergo IgH chain rearrangement (7). Functional IgH
chain expression leads to pre-BCR expression and the downregulation of RAG gene expression (8). Subsequently, RAG
gene expression is up-regulated in later pre-B cells, coincident with the onset of IgL chain rearrangement (8). Then
functional light chains lead to the expression of surface BCR
and down-regulate RAG gene expression.
Correspondence to: M. Taniguchi, Department of Molecular Immunology, Graduate School of Medicine, Chiba University, 1-8-1 Inohana,
Chuo-ku, Chiba 260-8670, Japan. E-mail: [email protected]
Transmitting editor: M. Miyasaka
Received 24 September 2002, accepted 10 December 2002
394 RAG-GFP expression in mature Peyer's patch T cells
Fig. 1. Generation of RAG2 knockout, GFP knock-in (RAG2-GFP) mice. (A) Design of the RAG2 knockout, GFP knock-in targeting construct.
The genomic restriction endonuclease map of the 6.0-kb XbaI fragment of the RAG2 locus clone. XhoI (X), SalI (S), PstI (P) and EcoRI (E)
sites are indicated. The vector was constructed as described in Methods. The 0.85-kb PstI fragment of the RAG2 coding region was replaced
by a 0.65-kb EGFP-coding fragment and pMC1 neo SalI±XhoI 1.2-kb cassette within two ¯anked lox P signals. The predicted structure of the
RAG2-EGFP mutant allele before pMC1 neo cassette deletion is shown at the bottom. SalI±XhoI 0.5-kb 5¢-RAG2 external probe (Ex) is
indicated. (B) Southern blot analysis with a 5¢-external probe on XbaI and XbaI/EcoRI digestion isolated from wild-type ES cells (wild-type)
and ES cells targeted with the RAG2 knockout, GFP knock-in construct (RAG-GFP hetero). This yielded a 7.1-kb fragment by XbaI digestion
and a 3.7-kb fragment by XbaI/EcoRI digestion for the targeted allele.
The developmental regulation of RAG gene expression
described above has been con®rmed by the establishment of
RAG-green ¯uorescent protein (GFP) transgenic mice or
RAG2:GFP fusion gene knock-in mice, in which GFP expression is under the control of endogenous RAG promoters in
bone marrow and thymus (9,10). The majority of GFP+ cells
were found to be pro-B cells, pre-B cells and immature B cells
in bone marrow and DP immature T cells in the thymus, while
no mature B and T cells were found to be GFP+ in the RAG
gene-manipulated mice, suggesting the physiological expression of GFP in immature lymphocytes. It has been previously
reported that the expression of RAG genes was demonstrated
by PCR and immuno¯uorescence in germinal center (GC)
B cells that phenotypically resembled immature bone marrow
B cells (11±15). From these results, it has been hypothesized
that the GC, but not other tissues, is the site in which mature
B cells induce RAG and mediate the secondary rearrangement, responsible for an additional means of somatic receptor
diversi®cation. However, analysis of reporter mice in which a
RAG2:GFP or RAG1:GFP fusion gene replaces the endo-
genous RAG locus suggest that RAG+ immature B cells from
bone marrow accumulate in the spleen by effects of adjuvant
and infectious agents, which could be the source of `RAG+ GC
B cells' (10,16,17).
In the case of mature T cells, RAG gene re-expression in
peripheral CD4+ T cell has been described in a murine TCRb
chain transgenic model, in which tolerogen-mediated chronic
peripheral selection against cells expressing the transgenic
TCR led to the down-regulation of TCR expression and the
up-regulation of RAG gene expression, and resulted in the
surface expression of endogenous TCRb chains (18).
Similarly, RAG gene expression has been demonstrated in
human CD4+CD3low peripheral T cells with low TCR expression (19). These T cells showed unusual characteristics of
altered TCR expression. However, RAG gene expression in
peripheral mature T cells under physiological conditions
remains unclear. It has also not been determined which
organ is responsible for the induction of RAG gene expression in the peripheral lymphoid tissues. Here, we established
RAG2-GFP knock-in mice to investigate RAG gene
RAG-GFP expression in mature Peyer's patch T cells
expression in mature T cells in peripheral lymphoid tissues
under physiological conditions.
Methods
Establishment of RAG2 knockout, GFP knock-in mice
A 4.0-kb SalI±XbaI 129/sv genomic DNA fragment containing
the RAG2 gene was modi®ed (20). A 0.85-kb PstI fragment
containing the RAG2 open reading frame was replaced by a
0.65-kb enhanced GFP (EGFP) coding region, generated by
PCR with template EGFP-C1 (Clontech, Palo Alto, CA). The
EGFP fragment was ligated in-frame at the RAG2 start codon.
A 1.2-kb pMC1 neo cassette (Stratagene, La Jolla, CA) cloned
between two ¯anked 34-bp lox P signal sequences
(Invitrogen, Carlsbad, CA) was inserted at the 3¢-end of the
EGFP fragment. The resulting RAG2 mutated fragment was
subcloned into a pPNT TK vector (targeting vector, Fig. 1A).
Electroporation of the targeting vector to R1 embryonic stem
(ES) cells was performed as described (21). Seven homologous mutant clones were identi®ed by Southern blot analysis
with 32P-labeled 0.5-kb 3¢-external probe (XhoI±SalI 0.5-kb 5¢RAG2 intron fragment). ES clones containing a mutated
RAG2-GFP allele were aggregated with BDF1 female mouse
morula in aggregation drops as described (21). After culturing
at 37°C for 24 h in a CO2 incubator, the aggregated blastocysts were transplanted into the uteri of foster mothers.
Southern blot analysis of tail DNA using a speci®c probe for
the mutant allele was performed. The expected 7.1-kb
fragments in the XbaI digest and 3.7-kb fragment in the
XbaI/EcoRI digest were detected in mice carrying the mutant
allele (Fig. 1B). The heterozygous mutant (RAG2-GFP hetero)
mice were subsequently crossed to generate homozygous
mutant (RAG2-GFP homo) mice.
Lymphocyte preparation
Freshly prepared thymocytes, spleen cells, peritoneal exudate
cells (PEC), mesenteric lymph node cells, liver lymphocytes
and Peyer's patch cells were suspended in PBS supplemented with 2% FCS and 0.1% sodium azide. Intestinal
intraepitherial lymphocytes (iIEL) were prepared with Percoll
(Amersham-Pharmacia, Little Chalfont, UK). Brie¯y, the intestine was put inside out after washing, cut into four pieces and
incubated in 1 3 HBSS supplemented with 5% FCS under
vigorous agitation at 37°C. The resulting suspensions were
passed through the column ®lled with glass wool and pelleted
through 30% Percoll solution. Cells were resuspended in 44%
Percoll solution and overlaid onto 70% Percoll solution. After
centrifugation, the interphase containing iIEL was recovered
and subjected to further analysis.
395
IL-7Ra±biotin, CD3e±biotin, CD45RB±biotin, CD62L±biotin
and CD69±biotin. In order to visualize biotin-conjugated
mAb, Texas red±streptavidin, Red613±streptavidin or
CyChrome±streptavidin was used (PharMingen). For direct
staining, cells were ®rst incubated with an anti-FcgRII/III
(2.4G2; PharMingen) to prevent the non-speci®c binding of
mAb via FcR interactions. Dead cells were excluded by
staining with propidium iodide (PI). Stained cells were analyzed on a FACS Vantage (Becton Dickinson, San Jose, CA)
and an Epics XL (Coulter, Palo Alto, CA). Data were analyzed
with CellQuest (Becton Dickinson) and Flow Jo software (Tree
Star, San Jose, CA).
For B cell analysis, single-cell suspensions were prepared
from bone marrows and the red blood cells depleted by
incubation in 0.83% NH4Cl. Cells were blocked with 20 mg/ml
anti-FcgRII/III (2.4G2), and incubated with a mixture of
biotinylated antibodies against IgM, CD23, CD90, CD5,
TER119 and NK1.1 (PharMingen, San Jose, CA), and IgD,
Mac-1 and Gr-1 (Southern Biotechnology Associates,
Birmingham, AL). After washing, the cells were incubated for
30 min with a mixture of antibodies [APC-conjugated antiB220 (anti-B220APC), PE-conjugated anti-CD43 (anti-CD43PE)
and Texas Red (TX)-labeled anti-CD24 (anti-CD24TX)] and
with TriColor (TC)-conjugated streptavidin (streptavidinTC;
Caltag, Burlingame, CA) and propidium iodide. Cells recognized by the biotinylated mAb and dead cells were excluded
by gating, and viable B220+/dull cells were then selected under
a lymphocyte gate on forward with side light scatter. In some
experiments, the cells were incubated for 30 min with a mixture
of biotinylated antibodies against anti-IgM, IgD and CD23 and
anti-B220APC followed by incubation with PE-conjugated
streptavidin and PI.
RNA preparation and RT-PCR
Approximately 45,000 Thy-1.2+(TCRb+) GFP± and 25,000 Thy1.2+(TCRb+) GFP+ cells from the Peyer's patches were
resuspended with TRIzol solution (Invitrogen, Carlsbad, CA).
Between 1000 and 100,000 thymocytes were also resuspended
with TRIzol solution. RNA extraction was performed with
chloroform followed by isopropanol precipitation with glycogen.
Total RNA prepared from sorted cells or from kidney and
thymocytes was subjected to reverse transcription into cDNA
with oligo(dT) primer using SuperScript II reverse transcriptase
(Invitrogen). PCR reaction with b-actin primers was performed
to ensure the equal amounts of cDNA. The primers used for each
PCR reaction was described previously (23,24).
Results
Flow cytometry analysis
GFP expression in developing thymocytes of heterozygous
and homozygous RAG2-GFP mice
In general, 106 cells were incubated on ice for 30 min with FITC-,
phycoerythrin (PE)-, CyChrome-, allophycocyanin (APC)-,
Cy5- and biotin-conjugated mAb as described (22). The
following mAb were purchased from PharMingen (San Jose,
CA): CD4±PE, CD8±Cy5, CD69±PE, NK1.1±PE, CD25±PE,
HAS (CD24)±PE, CD90±PE, Thy-1.2±PE, TCRab±Cy5, CD4±
Cy5, CD8±Cy5, B220±APC, CD44±biotin, IgM±biotin, Mac-1±
biotin, Gr-1±biotin, CD4±biotin, CD8±biotin, c-kit±biotin,
GFP expression was studied on developing thymocytes. The
expression of RAG2 gene is strictly regulated developmentally
(3,6). The cell number, CD4/CD8 and TCRab/CD69 pro®les,
and CD44/CD25 pro®les of CD4±CD8± (DN) thymocytes were
similar between wild-type and heterozygous RAG2-GFP mice
(Fig. 2A, C and E, left and middle panels). The number of
thymocytes in homozygous RAG2-GFP mice was ~100-fold
lower and only DN thymocytes were detected (Fig. 2A). The
396 RAG-GFP expression in mature Peyer's patch T cells
Fig. 2. GFP expression in the developing thymocytes of RAG2-GFP mice. Flow cytometric analysis was performed on thymocytes isolated
from RAG2-GFP heterozygous mice, homozygous mice and wild-type mice after staining for the indicated cell-surface markers. Cell yield is
indicated by the boxed numbers and the percentages of cells are shown in the respective quadrants. A wild-type histogram (no GFP
expression, shaded area) was overlaid with that of RAG2-GFP heterozygous (solid line) and homozygous (bold line) mice. (A and B)
Thymocytes were stained with anti-CD4±PE and anti-CD8±Cy5. CD4±CD8± DN, CD4+CD8+ DP, CD4 SP and CD8 SP subsets were
electronically gated and analyzed for GFP expression. (C and D) Thymocytes were stained with anti-CD4±APC, anti-CD8±APC, anti-CD25±PE,
anti-CD44±biotin and avidin±Red613. APCbright cells were gated out electronically. CD44+CD25±, CD44+CD25+, CD44±CD25+ and CD44±CD25±
subsets of CD4±CD8± DN thymocytes were gated and analyzed for GFP expression. (E and F) Thymocytes were stained with anti-CD69±PE
and anti-H57 (TCRb)±Cy5. CD69±TCR± (gate #1), CD69intTCRint (gate #2), CD69+TCR+ (gate #3) and CD69±TCR+ (gate #4) subsets were
gated and analyzed for GFP expression.
developmental block appears to be at the CD44±CD25+ DN
thymocyte stage, similar to that of RAG2-de®cient mice
(Fig. 2C) (20).
Representative GFP expression pro®les of electronically
gated DN, CD4+CD8+ (DP), CD4+CD8± (CD4 SP) and CD4±
CD8+ (CD8 SP) subsets are shown in Fig. 2B. As expected,
GFP expression was detected in DN and DP subsets, but not
in the CD4 SP or CD8 SP subset. The expression levels of GFP
in DN cells in homozygous RAG2-GFP mice were signi®cantly
higher than in heterozygous RAG2-GFP mice. In the DN
subset, GFP expression was observed in the CD44+CD25+,
CD44±CD25+ and CD44±CD25± stages, suggesting a normal
expression pattern of the RAG gene (Fig. 2D). These results
are consistent with those of a previous report (10).
RAG-GFP expression in mature Peyer's patch T cells
397
Fig. 3. GFP expression in developing B cells in bone marrow of RAG2-GFP mice. (A) Bone marrow cells were obtained from RAG2-GFP (c)
and littermate mice (a and b). The cells were incubated with biotinylated mAb against IgM, IgD, CD23, CD90, CD5, Gr-1, Mac-1, NK1.1 and
TER119, followed by staining with B220APC, HSATX and CD43PE mAb. These cells were stained with avidin±TriColor and PI. Cells recognized
by the biotinylated mAb and PI+ dead cells were excluded from analysis by gating. Viable cells were analyzed under a lymphocyte gate on
forwarded and side light scatters. B220+/dull cells were gated (G1 in a) and separated into HSAhigh (G2), HSAdull (G3) and HSA± (G4) in (b) and
(c). Histograms represent the relative number of GFP+ cells in the HSAhighB220+ (d), HSAdullB220+ (e) and HSA±B220dull (f) populations of
RAG2-GFP mice (solid line) and littermates (dotted line). (B) Bone marrow cells from heterozygous RAG2-GFP mice were incubated with
biotinylated mAb against IgM, IgD and CD23, followed by staining with B220APC and with avidin±PE and PI. Viable and B220+/dull cells were
analyzed under a lymphocyte gate (a), as described in (A). Ig+ cells, including immature B cells and recirculating B cells, and Ig± B cells were
gated in G1 and G2 respectively. Histograms show the expression of GFP in Ig+ (b) and Ig± B cells (c).
Furthermore, GFP expression was down-regulated between
small resting and large cycling cells in the CD44±CD25+
subset (data not shown). GFP expression was also analyzed
in subsets de®ned by CD69 and TCRab expression (Fig. 2E
and F). The level of both CD69 and TCRab appears to
increase concomitantly (gate #1 ® gate #2 ® gate #3) during
the transition between DP and SP thymocytes (23). As shown
in Fig. 2F, GFP expression was shut off in the late DP subset,
CD69highTCRhigh cells (gate #3).
GFP expression in developing B cells in bone marrow
Similar to T cells, the development of B cells in heterozygous
RAG2-GFP mice was normal as determined by cell number,
IgM/B220 pro®les and HSA/B220 pro®les (data not shown).
Again, as we expected, GFP+ cells were detected in HSA+ and
HSAdull B220+ bone marrow subsets containing pre-B and proB cells (Fig. 3A, d and e) and in IgM±B220+cells (Fig. 3B, c).
GFP expression seemed inversely correlated with the surface
IgM level. In homozygous RAG2-GFP mice, absolute numbers
of bone marrow cells decreased by about half and B cell
development was blocked at the B220+IgM± pro-/pre-B cell
stage (data not shown). These results are consistent with
those of RAG2-de®cient and RAG2:GFP fusion gene knock-in
mice (10,20).
GFP expression in a Thy-1.2+ subset of the Peyer's patch
The results obtained thus far suggest that GFP expression
faithfully re¯ects the expression pattern of the endogenous
RAG2 gene products. Therefore, heterozygous RAG2-GFP
mice would be useful animals for identifying T cell populations
with RAG gene expression in peripheral lymphoid organs.
Consequently, we sought GFP-expressing T cells in peripheral
lymphoid organs including spleen, inguinal lymph nodes
(ILN), iIEL, liver, PEC and Peyer's patch in heterozygous
RAG2-GFP mice. Among the samples tested, we found
signi®cant GFP expression in the Peyer's patch, but not in
other secondary lymphoid organs (Fig. 4A). Since it has been
reported that RAG is expressed in a small CD3± population of
iIEL (5,25,26), we further examined GFP expression in iIEL
after staining with anti-CD3 (Fig. 4B, upper panel). However,
398 RAG-GFP expression in mature Peyer's patch T cells
we did not detect GFP+ cells in the CD3± iIEL populations
(Fig. 4B, lower panel).
The GFP+ cells in the Peyer's patch were Thy-1.2+ and
represented ~0.1±3% of Peyer's patch Thy-1.2+ cells (Fig. 5A).
To examine further the characteristics of these cells, the
expression of various cell-surface markers was compared
between Thy-1.2+GFP+ (gate #3) and Thy-1.2+GFP± (gate #2)
cells (Fig. 5B and C). The GFP+ Peyer's patch T cells showed a
mature T cell phenotype with rearranged TCRab receptors on
the cell surface, as they expressed very high levels of CD3e
and TCRb (Fig. 5B). They also expressed molecules characteristic of developing immature lymphocytes, such as IL-7Ra,
c-kit and CD44. An activation marker antigen, CD69, was also
positive. In addition, they did not express CD62L, which is
positive in the majority of CD4+CD8+ thymocytes. The unique
phenotypic features indicate that GFP+ Peyer's patch T cells
are not immature thymocyte emigrants, although the majority
of the GFP+ T cells were revealed to be of a CD4+CD8+ DP
phenotype (Fig. 5E). As shown in Fig. 5C, GFP+ DP Peyer's
patch T cells expressed CD8b, suggesting a subset developed in the thymus (CD8ab-expressing cells), but not in the
gut (27±30). Interestingly, low but signi®cant levels of IL-7Raexpressing cells were also detected in Peyer's patch T cells
from wild-type mice (Fig. 5F), suggesting physiological consequences of these cells.
GFP signal correlates with the endogenous RAG expression
in GFP-RAG2 heterozygous mice
Finally, we have performed RT-PCR to con®rm that the
observed GFP signal indeed corresponded to the endogenous RAG expression (Fig. 6). Both Thy-1.2+GFP+ and Thy1.2+GFP± cells were sorted from Peyer's patches of RAG2GFP heterozygous mice. After RT-PCR, PCR products were
subjected to the Southern blot analysis (Fig. 6). As anticipated,
both RAG1 and RAG2 mRNA were detected in Thy-1.2+GFP+
cells, but not in Thy-1.2+GFP± cells (Fig. 6, lanes 2 and 3). The
expression level of RAG1 and RAG2 in Thy-1.2+GFP+ population was equivalent to that of 1000 thymocytes respectively
(Fig. 6, cf. lane 3 and 6). Further analysis demonstrated that
there was mRNA for the surrogate TCRa chain (pTa) in the
Thy-1.2+GFP+ and Thy-1.2+GFP± cells (Fig. 6). Intriguingly,
pTab mRNA was almost absent in the Thy-1.2+GFP± populations (Fig. 6, lane 2). As for pTaa expression, a higher level of
the mRNA was present in Thy-1.2+GFP± populations than Thy1.2+GFP+ cells (Fig. 6, cf. lane 2 and 3). This pTa expression
pro®le was in contrast to that of thymus (31). There was less
CD3e in Thy-1.2+GFP+ cells than in Thy-1.2+GFP± cells (Fig. 6,
cf. lane 2 and 5). The level of CD3e in Thy-1.2+GFP+ cells was
almost equivalent to that in 1000 thymocytes (Fig.6, cf. lane 3
and 6). These data clearly demonstrated that the mRNA
expression pro®le in Thy-1.2+GFP+ cells was distinct from that
of thymocytes.
Discussion
RAG2-GFP knock-in mice (RAG2:GFP mice) were established
in order to investigate the presence of lymphocytes expressFig. 4. GFP expression in lymphocytes from secondary lymphoid
tissues. (A) Lymphocytes in the thymus, spleen, ILN, iIEL, liver, PEC
and Peyer's patch were prepared, and GFP expression was
determined by ¯ow cytometric analysis. Shaded areas represent
background ¯uorescence in wild-type mice. (B) GFP expression in
CD3± fractions in iIEL. After staining with anti-CD3e, gated CD3e±
fractions were measured for GFP expression (upper panel).
Histograms of cells from RAG2-GFP heterozygous mice (line) from
gate #1 plus #2 or from gate #2 are overlaid with histograms of wildtype cells (shaded).
RAG-GFP expression in mature Peyer's patch T cells
ing RAG2 gene, particularly in the secondary lymphoid
tissues. In homozygous RAG2-GFP mice, T and B cell
development was completely arrested in a manner similar to
that described previously in RAG2-de®cient mice (20). In
heterozygous RAG2-GFP mice, both T and B cell development
399
appeared to be normal, and GFP expression was found in the
expected subsets of developing lymphocytes (Figs 2 and 3).
Three lines of similar gene-manipulated GFP mice have
been reported (9,10,32), and all focus on the development of
B cells and editing of BCR molecules. Monroe et al. estab-
Fig. 5. GFP expression in Peyer's patch T cells. Flow cytometric analysis was performed with 30 3 106 Peyer's patch cells from pooled RAG2GFP heterozygous and wild-type mice. (A) The gates used for multicolor analysis (#1, #2 and #3) are shown. The percentages of cells in gate
#3 are depicted. (B and C) Peyer's patch lymphocytes were stained with anti-Thy-1.2±PE and several biotinylated antibodies. Thy-1.2+GFP+
(gate #3 shown in A) and Thy-1.2+GFP± (gate #2 shown in A) subsets were electronically gated, and the expression of CD3e, TCRab, IL-7Ra,
c-kit, CD69, CD45RB, CD44, CD62L, CD8b and NK1.1 was evaluated. The background staining histogram (shaded) was overlaid on those
showing the Thy-1.2+GFP± (solid line) and Thy-1.2+GFP+ (bold line) populations. In the panels showing CD8b and NK1.1 staining in (C),
staining pro®les of Thy-1.2+GFP± (shaded) and Thy-1.2+GFP+ (open) cells were overlaid. (D and E) Peyer's patch lymphocytes were stained
with anti-CD4±PE and anti-CD8±Cy5. CD4±CD8± DN, CD4+CD8+ DP, CD4 SP and CD8 SP subsets were gated and analyzed for GFP
expression. Shaded areas represent background ¯uorescence in wild-type mice. (F) IL-7Ra expression on Thy-1.2+ Peyer's patch T cells
(using gate #1 shown in A) from wild-type and RAG2-GFP heterozygous mice. The shaded area represents background staining with an
isotype-matched control mAb.
400 RAG-GFP expression in mature Peyer's patch T cells
Fig. 6. RAG1, RAG2, pTa and CD3e mRNA expression in Thy-1.2+
Peyer's patch cells. After sorting, 1.0 3 104 cells from the Thy1.2+GFP± and Thy-1.2+GFP+ fractions from RAG2-GFP heterozygous
mice were used for RT-PCR analysis. For comparison, equal
numbers of cells from kidney and 1±100,000 thymocytes were
subjected to RT-PCR analysis. b-Actin was used as a control to
ensure the equal amount of mRNA recovered from these fractions.
PCR products were electrophoresed on agarose gel and detected
by probes speci®c for each transcript.
lished mice in which a functional GFP:RAG2 fusion gene is
knocked-in the endogenous RAG2 locus. In these mice, GFPexpressing B cells are B220+CD43high pro-B cells and
B220+CD43low pre-B cells in bone marrow, and
B220+pB130-140high immature B cells in spleen (10). Yu et al.
established transgenic mice carrying an arti®cial bacterial
chromosome that encodes GFP instead of RAG2. In these
mice, GFP is expressed in all immature B cells, including
B220+CD43+ pro-B cells, B220lowCD43±IgM± pre-B cells in
bone marrow and B220+493+ transitional B cells in spleen,
while B220+CD43±, B220+HSA+ mature B cells are negative for
GFP expression (9). Igarashi et al. described heterozygous
RAG1-GFP/neo knock-in mice in which the neo gene is
present in the forward orientation upstream of the RAG
gene. GFP+ cells were found to be B220+ immature B cells
in the GC of lymphoid follicles in the spleen, axillary and
mesenteric lymph nodes, and Peyer's patch (33). As for T cell
development in these gene-manipulated animals, GFP or RAG
gene expression was con®ned to the thymus and was
restricted to immature populations, including CD44±CD25+,
CD44+CD25+ and CD44±CD25± DN subsets. No GFP+ cells
were reported in the peripheral T cell populations.
Unlike these reports, we found that small but signi®cant
GFP+ and Thy-1.2+ populations are present in the Peyer's
patch, but not in other secondary lymphoid tissues so far
analyzed (Fig. 4). The GFP+ cells in the Peyer's patch are
clearly categorized as mature T cells because of their high
level Thy-1.2, CD3e and TCRb expression (Fig. 5).
Interestingly, GFP+ Peyer's patch T cells show unique
phenotypes such as being CD4+CD8+ and CD44+, being
reminiscent of developing immature thymocytes (Fig. 5). In
addition, this population also possesses an activation marker,
CD69, making the cells resemble activated peripheral T cells
or thymocytes undergoing selection events (34). The most
intriguing ®nding is that GFP+ Peyer's patch T cells express
characteristic molecules in developing lymphocytes, such as
IL-7R, c-kit and pTa, despite their high expression of TCRb
and CD3e (Fig. 5B and C).
In addition to these unique features, the expression pro®le of
pTaa and pTab in Thy-1.2+ Peyer's patch cells is worthy of note
(Fig. 6). Given that pTaa retains TCRb chain within the cells,
while pTab allows its cell-surface expression (31), abundance
of pTab in GFP+ cells may permit high level of TCRb
expression (Fig. 5B). The dominant expression of pTab over
pTaa in GFP+ cells suggests that these cells are similar to
mature T cells (31,35). Together, these data support the notion
that Thy-1.2+GFP+ Peyer's patch cells are not immigrants from
thymus that represent merely the immature T cells, but rather
indicate that Thy-1.2+ GFP+ cells in the Peyer's patch constitute a novel T cell subset.
In RAG2-transgenic mice, GFP is not synthesized as a
fusion protein, but this strain carries a 165-kb bacterial arti®cial
chromosome, and ¯uorescence levels are much higher with a
dynamic range of 3±4 log (9). However, the half-life of GFP in
this strain is 2±3 days, probably much longer than that of the
endogenous RAG2 protein (16). In contrast, the dynamic
range of GFP in RAG2-GFP knock-in mice is ~1 log over
background (Figs 2 and 3) (10). Although the GFP signal may
not truly re¯ect the expression of RAG2 in RAG2-GFP knock-in
mice, we detected RAG1 and RAG2 mRNA by RT-PCR in the
GFP+Thy-1.2+, but not in the GFP±Thy-1.2+ fraction in the
Peyer's patch of RAG2-GFP knock-in mice (Fig. 6). This
unequivocally demonstrates that GFP expression observed
in our RAG2-GFP mice directly mirrors the expression of
RAG2.
Some particular microenvironments, such as that in Peyer's
patch, are thus likely to induce signals or provide cellular
scaffolding to promote the re-expression of the recombination
machinery in mature T cells. Some T cells with memory or
immature phenotypes are present in normal Peyer's patch
(Fig. 5F) and anti-ovalbumin TCR transgenic mice (36).
As well as our present data, messages for the RAG1 and/or
2 genes have been found to be present in TCRb chain
transgenic mouse (18) and in human CD4+ T cell clones with
altered surface expression of TCR, such as low or defective
expression of CD3e and TCRab (19). Although the level of
RAG expression is low (103- to 104-fold less than in
thymocytes), RAG1/RAG2+ clones undergo secondary TCR
rearrangements (19). Furthermore, mature T cells with altered
TCR and RAG expression are present in increased numbers in
patients with defective responses to DNA damage (37). Thus,
RAG expression in these cells may correlate with the cellular
requirement to rescue them from defective phenotypes.
However, this is not the case in the present study, since
RAG expression in T cells in Peyer's patch was observed
under normal physiological conditions. Because of the high
level of TCRb and CD3e expression, these GFP+ T cells are
distinct from those with altered TCR expression (Fig. 5B).
Thus, RAG2-GFP mice may be interesting models to show the
presence of receptor revision in mature T cells and to address
its physiological relevance.
RAG-GFP expression in mature Peyer's patch T cells
In line with this expectation, we have detected a high
frequency of rearrangement of TCRa and b loci in GFP+, but
not in GFP±, Peyer's patch T cells as judged by the presence of
DNA double-strand breaks and circular DNA associated with
recombination events. In addition, the pro®le of DNA doublestrand breaks in GFP+ Peyer's patch T cells was different from
that in TCRb+ CD4+CD8+ DP cells in thymus, indicating that
these cells were not migrants from thymus harboring a
residual GFP signal without RAG expression and recruited
into the Peyer's patch (H. Wakao et al., submitted for
publication).
Finally, it is worth noting that GFP+ T cells in Peyer's patch
are not related to recently identi®ed RAG+ immature extrathymic T cells, cryptopatches in the small intestine (38).
Immature cryptopatch T cells are CD8aa T cells. In contrast,
GFP+ Peyer's patch mature T cells express CD8ab, but not
CD8aa, indicating that they are thymus-derived and not gutderived T cells. Although previous studies demonstrated that
there exists RAG transcripts in a CD3± fraction of iIEL, we did
not detect GFP+ cells in the CD3± fraction of iIEL of RAG2-GFP
knock-in mice (Fig. 4B) (5,25,26). This is most likely due to the
difference in sensitivity between FACS and other analysis such
as in situ hybridization and PCR.
In conclusion, we have identi®ed a novel subset of T cells in
the Peyer's patch, whose physiological role is yet to be
elucidated.
Acknowledgements
The authors thank Ms Hiroko Tanabe and Kaoru Sugaya for preparation of this manuscript. This work was supported by grants from the
Ministry of Education, Culture, Sports, Science and Technology
(Japan) (Grants-in-Aid for Scienti®c Research, Priority Areas
Research 13218016 and 12051203, Scienti®c Research A
13307011, B 14370107 and C 12670293, and Special Coordination
Funds for Promoting Science and Technology), the Ministry of Health,
Labor and Welfare (Japan) (the Program for Promotion of Fundamental
Studies in Health Sciences of the Organization for Pharmaceutical
Safety and Research), and the Human Frontier Science Program
Research Grant (RG00168/2000-M206).
Abbreviations
APC
B6
DN
DP
EGFP
ES
FCM
GC
GFP
iIEL
ILN
PE
PEC
PI
RAG
SP
allophycocyanin
C57BL/6
double negative
double positive
enhanced green ¯uorescent protein
embryonic stem
¯ow cytometry
germinal center
green ¯uorescent protein
intestinal intraepitherial lymphocytes
inguinal lymph nodes
phycoerythrin
peritoneal exudate cell
propidium iodide
recombination-activating gene
single positive
References
1 Tonegawa, S. 1983. Somatic generation of antibody diversity.
Nature 302:575.
2 Wilson, A., Held, W. and MacDonald, H. R. 1994. Two waves of
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
401
recombinase gene expression in developing thymocytes. J. Exp.
Med. 179:1355.
Ismaili, J., Antica, M. and Wu, L. 1996. CD4 and CD8 expression
and T cell antigen receptor gene rearrangement in early
intrathymic precursor cells. Eur. J. Immunol. 26:731.
Fehling, H. J. and von Boehmer, H. 1997. Early alpha beta T cell
development in the thymus of normal and genetically altered
mice. Curr. Opin. Immunol. 9:263.
Guy-Grand, D., Vanden Broecke, C., Briottet, C., Malassis-Seris,
M., Selz, F. and Vassalli, P. 1992. Different expression of the
recombination activity gene RAG-1 in various populations of
thymocytes, peripheral T cells and gut thymus-independent
intraepithelial lymphocytes suggests two pathways of T cell
receptor rearrangement. Eur. J. Immunol. 22:505.
Turka, L. A., Schatz, D. G., Oettinger, M. A., Chun, J. J., Gorka, C.,
Lee, K., McCormack, W. T. and Thompson, C. B. 1991.
Thymocyte expression of RAG-1 and RAG-2: termination by T
cell receptor cross-linking. Science 253:778.
Willerford, D. M., Swat, W. and Alt, F. W. 1996. Developmental
regulation of V(D)J recombination and lymphocyte differentiation.
Curr. Opin. Genet. Dev. 6:603.
Grawunder, U., Leu, T. M., Schatz, D. G., Werner, A., Rolink, A. G.,
Melchers, F. and Winkler, T. H. 1995. Down-regulation of RAG1
and RAG2 gene expression in preB cells after functional
immunoglobulin heavy chain rearrangement. Immunity 3:601.
Yu, W., Nagaoka, H., Jankovic, M., Misulovin, Z., Suh, H., Rolink,
A., Melchers, F., Meffre, E. and Nussenzweig, M. C. 1999.
Continued RAG expression in late stages of B cell development
and no apparent re-induction after immunization. Nature 400:682.
Monroe, R. J., Seidl, K. J., Gaertner, F., Han, S., Chen, F.,
Sekiguchi, J., Wang, J., Ferrini, R., Davidson, L., Kelsoe, G. and
Alt, F. W. 1999. RAG2:GFP knockin mice reveal novel aspects of
RAG2 expression in primary and peripheral lymphoid tissues.
Immunity 11:201.
Han, S., Zheng, B., Schatz, D. G., Spanopoulou, E. and Kelsoe, G.
1996. Neoteny in lymphocytes: Rag1 and Rag2 expression in
germinal center B cells. Science 274:2094.
Hikida, M., Mori, M., Takai, T., Tomochika, K., Hamatani, K. and
Ohmori, H. 1996. Reexpression of RAG-1 and RAG-2 genes in
activated mature mouse B cells. Science 274:2092.
Hikida, M., Nakayama, Y., Yamashita, Y., Kumazawa, Y.,
Nishikawa, S. I. and Ohmori, H. 1998. Expression of
recombination activating genes in germinal center B cells:
involvement of interleukin 7 (IL-7) and the IL-7 receptor. J. Exp.
Med. 188:365.
Papavasiliou, F., Casellas, R., Suh, H., Qin, X. F., Besmer, E.,
Pelanda, R., Nemazee, D., Rajewsky, K. and Nussenzweig, M. C.
1997. V(D)J recombination in mature B cells: a mechanism for
altering antibody responses. Science 278:298.
Han, S., Dillon, S. R., Zheng, B., Shimoda, M., Schlissel, M. S. and
Kelsoe, G. 1997. V(D)J recombinase activity in a subset of
germinal center B lymphocytes. Science 278:301.
Nagaoka, H., Gonzalez-Aseguinolaza, G., Tsuji, M. and
Nussenzweig, M. C. 2000. Immunization and infection change
the number of recombination activating gene (RAG)-expressing
B cells in the periphery by altering immature lymphocyte
production. J. Exp. Med. 191:2113.
Gartner, F., Alt, F. W., Monroe, R. J. and Seidl, K. J. 2000. Antigenindependent appearance of recombination activating gene
(RAG)-positive bone marrow B cells in the spleens of immunized
mice. J. Exp. Med. 192:1745.
McMahan, C. J. and Fink, P. J. 1998. RAG reexpression and DNA
recombination at T cell receptor loci in peripheral CD4+ T cells.
Immunity 9:637.
Lantelme, E., Palermo, B., Granziero, L., Mantovani, S.,
Campanelli, R., Monafo, V., Lanzavecchia, A. and Giachino, C.
2000. Cutting edge: recombinase-activating gene expression and
V(D)J recombination in CD4+CD3low mature T lymphocytes. J.
Immunol. 164:3455.
Shinkai, Y., Rathbun, G., Lam, K. P., Oltz, E. M., Stewart, V.,
Mendelsohn, M., Charron, J., Datta, M., Young, F., Stall, A. M.,
et al. 1992. RAG-2-de®cient mice lack mature lymphocytes owing
to inability to initiate V(D)J rearrangement. Cell 68:855.
402 RAG-GFP expression in mature Peyer's patch T cells
21 Akasaka, T., Kanno, M., Balling, R., Mieza, M. A., Taniguchi, M.
and Koseki, H. 1996. A role for mel-18, a Polycomb group-related
vertebrate gene, during the anteroposterior speci®cation of the
axial skeleton. Development 122:1513.
22 Yamashita, M., Katsumata, M., Iwashima, M., Kimura, M., Shimizu,
C., Kamata, T., Shin, T., Seki, N., Suzuki, S., Taniguchi, M. and
Nakayama, T. 2000. T cell receptor-induced calcineurin activation
regulates T helper type 2 cell development by modifying the
interleukin 4 receptor signaling complex. J. Exp. Med. 191:1869.
23 Taniguchi, M., Koseki, H., Tokuhisa, T., Masuda, K., Sato, H.,
Kondo, E., Kawano, T., Cui, J., Perkes, A., Koyasu, S. and Makino,
Y. 1996. Essential requirement of an invariant V alpha 14 T cell
antigen receptor expression in the development of natural killer T
cells. Proc. Natl Acad. Sci. USA 93:11025.
24 Cui, J., Shin, T., Kawano, T., Sato, H., Kondo, E., Toura, I.,
Kaneko, Y., Koseki, H., Kanno, M. and Taniguchi, M. 1997.
Requirement for Valpha14 NKT cells in IL-12-mediated rejection of
tumors. Science 278:1623.
25 Guy-Grand, D., Cerf-Bensussan, N., Malissen, B., MalassisSeris, M., Briottet, C. and Vassalli, P. 1991. Two gut
intraepithelial CD8+ lymphocyte populations with different T
cell receptors: a role for the gut epithelium in T cell
differentiation. J. Exp. Med. 173:471.
26 Lin, T., Matsuzaki, G., Yoshida, H., Kobayashi, N., Kenai, H.,
Omoto, K. and Nomoto, K. 1994. CD3±CD8+ intestinal
intraepithelial lymphocytes (IEL) and the extrathymic
development of IEL. Eur. J. Immunol. 24:1080.
27 Lefrancois, L. and Puddington, L. 1995. Extrathymic intestinal
T-cell development: virtual reality? Immunol. Today 16:16.
28 Sim, G. K. 1995. Intraepithelial lymphocytes and the immune
system. Adv. Immunol. 58:297.
29 Klein, J. R. 1996. Whence the intestinal intraepithelial lymphocyte?
J. Exp. Med. 184:1203.
30 Mowat, A. M. and Viney, J. L. 1997. The anatomical basis of
intestinal immunity. Immunol. Rev. 156:145.
31 Barber, D. F., Passoni, L., Wen, L., Geng, L. and Hayday, A. C.
1998. The expression in vivo of a second isoform of pT alpha:
implications for the mechanism of pT alpha action. J. Immunol.
161:11.
32 Kuwata, N., Igarashi, H., Ohmura, T., Aizawa, S. and Sakaguchi,
N. 1999. Cutting edge: absence of expression of RAG1 in
peritoneal B-1 cells detected by knocking into RAG1 locus with
green ¯uorescent protein gene. J. Immunol. 163:6355.
33 Igarashi, H., Kuwata, N., Kiyota, K., Sumita, K., Suda, T., Ono, S.,
Bauer, S. R. and Sakaguchi, N. 2001. Localization of
recombination activating gene 1/green ¯uorescent protein
(RAG1/GFP) expression in secondary lymphoid organs after
immunization with T-dependent antigens in rag1/gfp knockin
mice. Blood 97:2680.
34 Yamashita, I., Nagata, T., Tada, T. and Nakayama, T. 1993. CD69
cell surface expression identi®es developing thymocytes which
audition for T cell antigen receptor-mediated positive selection.
Int. Immunol. 5:1139.
35 Bruno, L., Rocha, B., Rolink, A., von Boehmer, H. and Rodewald,
H. R. 1995. Intra- and extra-thymic expression of the pre-T cell
receptor alpha gene. Eur. J. Immunol. 25:1877.
36 Saparov, A., Kraus, L. A., Cong, Y., Marwill, J., Xu, X. Y., Elson, C.
O. and Weaver, C. T. 1999. Memory/effector T cells in TCR
transgenic mice develop via recognition of enteric antigens by a
second, endogenous TCR. Int. Immunol. 11:1253.
37 Lantelme, E., Mantovani, S., Palermo, B., Campanelli, R.,
Granziero, L., Monafo, V. and Giachino, C. 2000. Increased
frequency of RAG-expressing, CD4+CD3low peripheral T
lymphocytes in patients with defective responses to DNA
damage. Eur. J. Immunol. 30:1520.
38 Saito, H., Kanamori, Y., Takemori, T., Nariuchi, H., Kubota, E.,
Takahashi-Iwanaga, H., Iwanaga, T. and Ishikawa, H. 1998.
Generation of intestinal T cells from progenitors residing in gut
cryptopatches. Science 280:275.