| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Segments Before Gene Rearrangement1
*
Division of Biology, California Institute of Technology, Pasadena, CA 91125; and
Institute of Systems Biology, Seattle, WA 98105
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
segments are
dissimilar, but any functional differences between them are masked
after productive gene rearrangement by the dominance of the TCR 3'
enhancer. However, thymocytes of recombination-activating gene-2
(Rag2)-deficient mice allow the transcriptional activity of V
promoters to be studied before rearrangement. Here we report that many
V segments are detectably transcribed in Rag2-/-
thymocytes and that there are significant differences in expression
among different V segments. Primer extension and characterization of
cDNA clones from SCID thymocytes suggest that these germline V
transcripts generally use the same start sites as those previously
determined in mature T cells. The strength of expression before
rearrangement does not correlate with proximity to the known enhancer,
because members of the most distal V cluster (V2.1, V1.1,
V4.1) are relatively strongly expressed and more proximal V
segments (V14.1, V3.1, V7.1, V6.1) are only weakly
expressed. Different V segments also show different developmental
programs of activation in different thymocyte subsets, with the
V5.1(L)-8.2(V) spliced transcript expressed earliest as well as most
strongly overall. Comparison with Rag+ MHC class
I-/- and class II-/- thymocytes confirms
that many of these expression differences are leveled by rearrangement
and/or by selection, before MHC-dependent selection. However, the
expression pattern of V2.1 is highly distinctive and includes cell
types apparently outside the T lineage, suggesting potential
acquisition of specialized roles. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain is controlled by a 3' C enhancer and the
relevant V promoter. In general, the C enhancer is a stronger
regulator than the V promoters, because the choice of V segment
used for rearrangement does not detectably affect the levels of TCR
mRNA in a mature T cell (1, 2). Although the promoter
sequences of different V segments are quite divergent, there has
been little evidence to date that they may also contribute
differentially to the activity of these genes. Several regulatory
elements have been identified in TCR V promoter regions, but their
functional importance is not well understood. This is because the
relative strengths of V promoters are usually masked both by TCR
gene rearrangement at the single-cell level and by clonal selection at
the T cell population level. However, the strengths of individual V
promoters can be revealed by analysis of TCR V transcription before
rearrangement.
Therefore, we have used thymocytes from recombinase activating gene-2
(Rag2)3-deficient
(Rag2-/-) mice to study the regulatory
properties of different TCR V promoters. Because of the lack of
recombinase, the TCR gene remains in germline configuration in these
cells, with the C enhancer hundreds of kilobases away. Although the
Rag2-/- thymocytes are arrested as immature
cells, they do reach a stage of development when normal thymocytes have
become committed to the T cell lineage and are fully capable of
transcribing the TCR-chain gene (3, 4, 5, 6). In this
context, the basal levels of "germline" transcription from
unrearranged V segments can be measured sensitively. A few previous
reports have indicated that transcripts from certain V segments
could be detected in such cells (7, 8). Here we show that
in the absence of rearrangement, different V segments are
transcribed at distinctive levels and with distinctive fine-scale
developmental regulation in the stages of lymphocyte differentiation
encompassing T lineage specification.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The mice used in this study were C57BL/6 (B6), (B6, 129)
Rag2-/-,
B6.129P2-Tcrbtm1Mom
Tcrdtm1Mom
(TCR--/-
-/-),
B6-SCID, and
B6-I-Ab-/-
2-microglobulin-/- (B6
MHC-/-). Except as indicated, all of the
animals were bred in our own facility, and immunodeficient mice were
maintained under sterile conditions in isolators, with autoclaved food,
bedding, and acidified water. In some cases, C.B-17-SCID mice were
purchased from Taconic Laboratories (Germantown, NY). These SCID mice
have a higher incidence of leakiness, as measured by the generation of
CD4+CD8+ thymocytes, than
the SCID animals bred at Caltech (Pasadena, CA).
Cell preparation and fractionation
Thymocytes and bone marrow cells were collected from 
6- to
8-wk-old mice. For bone marrow samples, RBC were lysed in 1 ml RBC
lysis buffer (155 mM NH4Cl, 10 mM
KHCO3, 0.1 mM EDTA) by pipetting up and down for
3 min at room temperature. The lysis was stopped by adding 3 ml
CBSS/BSA/azide (1.25x HBSS without phenol red, plus 2.5 mg/ml BSA and
0.03% sodium azide), and cells were washed one more time in the same
buffer. For cell sorting, cells were incubated with supernatant from
the 2.4G2 hybridoma for 20 min on ice to block Fc receptors before
staining with fluorescent-conjugated Abs. FACS sorting was performed
using a Coulter Epics Elite cell sorter (Coulter, Hialeah, FL). Stained
cells were sorted into chilled tubes containing CBSS/BSA/azide with 5%
FBS. A portion of each sorted sample was reanalyzed for purity. The
rest of the cells were used for RNA extraction. All of the Abs used in
this study were from BD PharMingen (San Diego, CA) or BD Biosciences
(San Jose, CA).
To analyze fine-scale developmental control of V expression,
Rag2-/- thymocytes were stained with
Sca-1(Ly-6A/E)-PE, heat stable Ag (HSA; CD24)-FITC, and CD44-CyChrome,
and fractionated by FACS sorting as previously described
(9, 10, 11, 12). The subpopulations collected were: 1)
Sca-1+ HSA-, which are
stem-like cells that probably retain potential for T, B, NK, and DC
lineages; 2) Sca-1- HSA-,
which are NK-like cells; 3) HSA+
CD44+, which are "pro-T cells" that are
specified but not committed to the T lineage yet; and 4)
HSA+ CD44-, which are
cells committed to the T lineage and undergoing TCR rearrangement.
To obtain B cell precursor and other immature hemopoietic populations,
Rag2-/- bone marrow cells were stained with
B220-FITC, CD19-CyChrome, and CD43-PE. The subpopulations collected
were: 1) B220+ CD43+
CD19+, which are pro-B cells; and 2) and
B220- CD43+, which are a
large, granular population likely to be enriched for myeloid lineage
cells. Fetal thymocytes were collected from B6 mice at 14.5 days
postcoitum (E14.5). Fetal liver cells were collected from B6 mice at
E15.5.
RT-PCR analysis
Total cellular RNA was prepared using RNAzol B (Leedo Medical,
Houston, TX), following manufacturer’s instructions. For sorted
samples, 25 µg tRNA was used as carrier and added to the samples at
the beginning of extraction. RNA samples were first treated with
RNase-free DNase I for 30 min at 37°C to remove the remaining genomic
DNA. Portions of each sample were removed and used as RNA quality
controls for PCR amplification. First-strand cDNA was synthesized from
total RNA using Superscript II (Life Technologies, Gaithersburg, MD)
and oligo(dT)12–18. PCR amplification was done
for 45 s at 94°C, 45 s at 55°C, and 1 min at 72°C.
Thirty to 38 cycles were normally used. PCR products were run on 2%
agarose gels and visualized by ethidium bromide staining. For
quantification, 0.5 µCi [32P]dCTP was added
in the PCR mix. Radioactive signals were quantified by PhosphorImager
(Molecular Dynamics, Sunnyvale, CA) after electrophoresis and transfer
of the DNA to Hybond-N+ membranes (Amersham,
Little Chalfont, U.K.). TCR V primers, which were designed for each
V segment to span the intron between the L and V exons, are listed
in Table I
. Genomic DNA of Rag2-/- thymocytes,
which has equal numbers of copies of each V segment, was used as a
control to normalize PCR efficiency of different primer sets. All cDNA
samples were normalized according to their levels of GAPDH RNA by
dilution in 1x Tris/EDTA (TE) buffer. To keep all samples in the
linear range for the V-specific PCR, B6 and
MHC-/- cDNAs were then diluted to 1/50 of the
normalized Rag2-/- and SCID cDNA level, except
as otherwise indicated.
|
The C.B-17-SCID thymocyte random-primed cDNA library was constructed in the pSPORT1 vector by Michele Anderson (12). The library was arrayed and spotted at high density onto Hybond-N+ membrane using the Q-Bot robot (Genetix, Christchurch, U.K.). DNA probes were labeled with [32P]dATP. Prehybridization, hybridization, and wash conditions were as described (12). The positive clones were selected and sequenced using the ABI Prism dye terminator cycle sequencing ready reaction kit (Perkin-Elmer, Foster City, CA) according to the manufacturer’s procedure.
5'-rapid amplification of cDNA ends (RACE) PCR
Poly(A)+ RNA samples from SCID and normal
B6 thymocytes were used for 5'-RACE analysis using a Marathon cDNA
amplification kit (Clontech, Palo Alto, CA). First- and second-strand
cDNA synthesis and adaptor ligation were performed following the
manufacturer’s instructions. The 5'-Marathon RACE reactions were
conducted using adaptor primer 1 (5'-CCATCCTAATACGACTCACTATAGGGC) and
the appropriate 3'-gene specific primer as shown in Table I
(antisense
primers for V4.1, V8.2, and V14.1). cDNA amplification
conditions were: 94°C for 1 min, followed by 30 cycles at 94°C for
30 s, 60°C for 30 s, and 68°C for 3 min. The size of the
PCR products was determined by running on a 2% agarose/ethidium
bromide gel along with appropriate DNA size markers. To increase the
sensitivity, PCR products were transferred to
Hybond-N+ membrane after electrophoresis and
probed with gene-specific radioactive probes
([
-32P]ATP end-labeled V4 and V14
oligonucleotides, see Table I).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
segments are transcribed without V(D)J rearrangement in
thymocytes
Thymocytes in Rag2-/- and SCID mutant mice
are a well-characterized composite of immature T cell subpopulations.
They undergo developmental arrest and death at the
CD25+CD44- stage due to an
inability to carry out TCR gene rearrangement. To progress beyond this
" selection checkpoint," T cell precursors need to express the
translation product of any successfully rearranged -chain gene,
which the mutant thymocytes cannot do. But because the stages of
development up to the selection checkpoint are well represented in
these mice (9), they include the stages when TCR gene
rearrangement would normally be initiated. Thus they are an excellent
source of cells that activate the same set of transcription factors
that may normally coincide with TCR gene rearrangement, without being
able to move the V genes from the germline configuration. We have
used thymocytes from Rag2-/- and SCID mice in
this work to investigate the extent to which V genes are transcribed
in T cell precursors in the absence of rearrangement.
To confirm that all TCR V gene expression in
Rag2-/- and SCID originates from DNA in a
germline configuration, RT-PCR analysis was conducted to determine
whether there is any transcription of rearranged TCR genes in
Rag2-/- and SCID cDNA samples. PCR primer sets
V8-C and V4-C were designed with the 5' primers in the
respective V regions and the 3' primer in the C region. Without
rearrangement, V segments are several hundred kilobases away from
the C region, and no band will be detected in the PCR product. As
expected, V4-C and V8-C bands are only detected in the B6
(wild-type) thymocytes but not in the Rag2-/-
thymocyte sample and only at a very low level in the SCID thymocyte
sample (Fig. 1). This data is consistent
with the number of thymocytes that have undergone V(D)J rearrangement
in these mice. Because V8.2 is the gene segment most commonly used
for TCR rearrangement, this negative result also renders unlikely
any possible contamination of Rag2-/- and SCID
samples with products of other rearranged V genes.
|
segments are transcribed in
Rag2-/- and SCID thymocytes. Fig. 1
demonstrates this by RT-PCR using primer set L5.1 and V8.2 (Table I) to detect the major transcript of the
V8.2 segment (see below). As shown in Fig. 1, with these primers a
V8.2 band was readily detected in all three samples. Because no
products of TCR rearrangement can be detected in
Rag2-/- thymocytes and little if any in SCID
thymocytes, all V8.2 detected in those samples represents germline
transcripts. Our results confirm that although there is no TCR
rearrangement in Rag2-/- and SCID thymocytes,
there is germline transcription of at least one segment in the V
region without rearrangement.
Differential expression of TCR V segments in the absence of
rearrangement
To compare the transcription of different individual V segments
before rearrangement, semiquantitative RT-PCR analyses were conducted
using primers designed for each V to span the intron between the L
(leader) and V exons. Twelve L-V primer sets were made to cover
representatives of the entire locus, including V2.1, which is
the most distant from the enhancer, and V14.1, which is the closest
to the enhancer (see below). Serial dilutions of
Rag2-/- thymocyte genomic DNA, which has the
same number of copies of each TCR V segment, were used as controls
to normalize the PCR efficiencies of different primer sets (Fig. 2A, lanes 1–4).
Each primer set was tested to compare its ability to amplify aliquots
of cDNA from Rag2-/- thymocytes (R
lanes, Fig. 2) with its ability to amplify cDNA from
unfractionated thymocytes of normal B6 mice (B lanes, Fig. 2), in which all V genes should be rearranged and expressed.
Aliquots of the same cDNA samples were used with all primer pairs, and
all these primer pairs amplified fragments of the expected size from
the control samples of genomic DNA (Fig. 2A, lanes
1–4) and from the normal thymocyte cDNA, in parallel reactions.
Thus both the amplification conditions and the quality of the samples
were fully validated in every experiment. The results indicate that PCR
products could successfully be amplified, in differing yields, from
Rag2-/- immature thymocytes using 10 of the 12
primer sets tested. Similar results were seen with cDNA from SCID
thymocytes (S lanes, Fig. 2, B and C).
Thus the transcription of V genes in the absence of rearrangement
extends to multiple V segments, not just V8.2.
|
segments in samples of
Rag2-/- and SCID immature thymocytes was much
lower than that in samples of wild-type B6 thymocytes (Fig. 2A, compare lane 5 to lane 7; and Fig. 2B, compare lanes 1 and 2 to
lanes 3 and 4). In most cases, V germline
transcription in the immature cells was at least 50 times weaker than
the transcription of rearranged TCR genes in the normal thymocytes, as
indicated by the 50-fold dilution of the B6 samples needed to bring
them into the same range as the immunodeficient samples (Fig. 2A, lanes 5 vs 7; Fig. 2B,
lanes 1 vs 3 and 4). This reflects the
dominant positive regulation of TCR- expression by the enhancer once
rearrangement has occurred (1). More interestingly, the
relative expression levels of germline transcripts of different V
segments are also changed from the pattern seen when the genes can
rearrange. In the B6 sample, most Vs are expressed at similar levels
except for V8.3 and V8.2. In fact, there is even more similarity
in expression because the detection of V8.2 is artificially
depressed here: in mature T cells most V8.2 transcripts use L5.1 for
the promoter and leader regions, whereas the only V8.2 detected here
is the fraction that uses L8.2. However, in
Rag2-/- and SCID samples (Fig. 2, A
and B, lanes R and S) the expression
levels of different Vs were significantly different. These
differences are evident both as normalized to quantitation standards in
parallel control reactions (Fig. 2A, lanes 1–4)
and as compared with expression in parallel samples from normal B6
thymocytes (Fig. 2A, lane 5, and Fig. 2B, lanes 1 and 2, lane
B).
Based on their germline expression levels, the V segments can be
classified into four groups. First, V8.2 (L8.2-V8.2) and
V5.1 (L5.1-V5.1) are the most strongly expressed, at 
1/50
the level in B6 thymocytes. The product of the L5.1 promoter spliced
to the body of the V8.2 is expressed at even higher levels, (1/10
the level of B6 thymocytes; see below) indicating an even stronger
intrinsic activity of the V5.1 promoter. V8.3, though
accumulating to a lower absolute level, is also expressed at about 1/50
its level in B6 thymocytes. Second, V2.1, V4.1, V1.1, V5.2,
and V8.1 are expressed at lower levels, from 1/100 to 1/1000 the
level seen in B6 thymocytes. Third, the expression of V7.1 and
V14.1 is lower yet (<1/1000 the level of B6 thymocytes). Finally,
in the same mutant thymocyte samples, expression of V6.1 and V3.1
is generally undetectable. These expression levels are summarized in
Table II. While RNA stability effects
could contribute to the relative levels of these transcripts, the
results strongly imply that different V promoters are differentially
active before rearrangement.
|
5.2, V8.3, V8.2, and
V8.1 (note that in Fig. 2, V8.2 transcripts were amplified using
a 5' primer from the L8.2 promoter segment and not the more distant
L5.1). However, this does not mean that they necessarily resulted
from contamination of the RNA samples with genomic DNA, because these
bands were not detected in the RNA control samples (RNA without reverse
transcription, Fig. 2B, lane N; compare signals
in corresponding sample with reverse transcription, lane R).
Therefore, the sizes of these bands may be due to incomplete splicing
of the germline transcripts. Germline-sized amplification products were
also generated occasionally from SCID and
Rag2-/- cDNA with the V14.1 and V7.1
primers, but the yields were irreproducible (e.g., Fig. 2A,
lane 7 vs Fig. 2B, lanes 3 and
4). Because these amplifications give little product
overall, the germline-sized bands in these cases could result from
small amounts of contaminating DNA in some preparations (data not
shown), and the true levels of V14.1 and V7.1 transcription
before rearrangement could be even lower.
Several distinct mechanisms could contribute to the difference between
the relative efficiencies of germline transcription in these immature,
developmentally arrested cells and transcription of rearranged genes in
the wild-type thymocyte population: V(D)J rearrangement, selection,
and positive selection. To test whether this difference is due to
positive selection, we examined the V expression in
MHC-/- thymocytes. Because MHC class I and II
products are required for positive selection, most thymocytes from
these doubly deficient animals undergo V(D)J rearrangement and
selection normally but arrest in the
CD4+CD8+ stage because they
cannot undergo positive selection. As shown in Fig. 2A,
lane 6 (lane M), virtually
indistinguishable patterns of V expression were observed with RNA
from MHC-/- and B6 thymocytes, suggesting that
the difference between the V transcription profiles in
Rag2-/- and B6 thymocytes is not due to
positive selection.
Expression before rearrangement is not determined by location in
the V complex
The distinctive levels of expression of different V in the
absence of rearrangement do not correlate with the proximity of the
V segment to the enhancer (Fig. 3A). Instead, not only is the
strongest expression seen for segments in the middle of the complex
(V5.1 and V8.2), but also transcription of the most distal V
segments (V2.1, V1.1, and V 4.1) is stronger than that of the
most proximal Vs analyzed in Fig. 2 (V14, V7.1, V3.1, and
V6.1). Most dramatic is the poor expression of V14.1, despite its
location within 5 kb of the enhancer. This implies that proximity to
the C enhancer is unlikely to be the major element controlling
germline transcription of V segments at this stage. In agreement
with this interpretation, a recent report (13) shows that
germline V transcription is not inhibited by deletion of the TCR
enhancer. One possibility is that the individual V promoters may be
the dominant regulators for V germline transcription. Alternatively,
if the clustering of Vs with similar expression profiles is
significant, a few local positive or negative regulatory elements
scattered throughout the 700-kb complex might affect the activity of
V promoters over neighborhoods of 50–100 kb.
|
s are all relatively closely linked in the 3'
portions of the V complex, raising the possibility that there could
be a local inhibitor of gene expression in this region. Alternatively,
it is possible that these Vs share some common feature of their
promoters. The promoters of different V segments are quite
divergent, with few if any blocks of conserved sequence. Computer
searches for various transcription factor consensus sites (TFSearch and
MOTIFSearch) reveal potential binding sites for T cell factor-1
(TCF-1), E26 transformation-specific gene (Ets), GATA, AP-1, basic
helix-loop-helix (bHLH; "E box"), myeloblastosis transforming gene,
and acute myelogenous leukemia gene-1 (AML1; Runx1) family
transcription factors in highly dissimilar clusterings in the regions
from -600 to +100 around the different V promoters, as diagrammed
in Fig. 3B. It has been reported that transcription from TCR
V promoters depends on a conserved decamer motif similar to the cAMP
response element (CRE), often embedded in a conserved 14-mer
(CAGTGAYRTCACTG) (2, 14) or "16-mer"
sequence (GenBank accession numbers AE000663, AE000664, and AE000665;
see annotation). However, the promoters of V6.1 and V3.1 lack the
"16-mer" and even the core TGANNTCA motif (see Table II). To test
whether the weak expression of these Vs is due to the presence of a
general silencing element in the 3' portion of the TCR complex, we
checked the expression of another V segment in the same cluster,
V15.1, located between V6.1 and V3.1, which does have a
divergent but identifiable 16-mer motif in its promoter region. As
shown in Fig. 2C, V15.1 is clearly expressed before
rearrangement. Thus, there is not a general silencing of genes
clustered with V6.1 and V3.1. This implies that the differential
expression is likely to be controlled, at least partially, by
individual V segment promoters.
Germline TCR V transcription starts at the same site as
rearranged TCR V transcription
Structural analyses of the V transcripts were conducted to
determine the promoters used and to verify whether they were indeed
derived from genes in germline configuration. To obtain representative
V cDNA clones, a high-density arrayed library of SCID thymocyte cDNA
was screened using [32P]dATP labeled probes for
particular V segments. In agreement with the preferential expression
of V8.2 (with an L5.1 leader) and V5.1 segments, over 50
V8.2 clones and one V5.1 clone were detected in a screening of
70,000 clones, as illustrated for a part of the library in Fig. 4A. No other V-positive
clone could be detected in this set using either V4.1 or V14.1
probes (data not shown), consistent with the lower abundance for other
transcripts indicated by RT-PCR. However, sequence analysis of the
clones obtained showed clearly the retention of intact heptamer and
nonamer recombination signal sequences at the 3' ends of the V
segments proper (Fig. 4B, and data not shown), confirming
that they represent transcripts of unrearranged genes.
|
segments in the immature cells are different from the profiles in
thymocytes with rearranged genes, they might use different promoters
from those identified in mature T cells. It has been reported that
germline transcription of the TCR D region depends on a
developmental stage-specific promoter in precursor T cells
(15). Distal V promoters have also been found in the
V5.1 and V8.1 regions (16). To test whether any
unconventional promoters are used for V germline transcription,
5'-RACE PCR analysis was performed comparing
poly(A)+ RNA prepared from SCID and normal B6
thymocytes. Three V segments, V4s1 (V4.1), V8s2 (V8.2),
and V14s1 (V14.1), were chosen as representatives of the three
expression classes, based on their relative locations in the complex
and their relative germline transcription levels.
As shown in Fig. 5B, for all
three V segments, the major 5'-RACE products in both B6
(lane B) and SCID (lane S) samples
were the same sizes. We first tested whether any distal V promoters
of V5.1 may be preferentially used for germline transcription of
V8.2. A major band at 400 bp was detected in both B6 and SCID
samples, corresponding to a transcription start site of -40
(relative to the initiation ATG) for the proximal promoter.
Hybridization with L5.1- and L8.2-specific probes showed that
rearranged (B6) and unrearranged (SCID) gene transcripts both
preferentially used the L5.1 promoter, and that even the
distributions of minor 5'-RACE products, representing upstream start
sites, were similar in both (data not shown). In the case of V4.1,
the major band was 380 bp in both B6 and SCID, corresponding to a start
site at around -143, and for V14.1, the only product seen was 350
bp, indicating a start site at -55. While some differences were seen
in the profiles of minor start sites for the V8.2 and V4.1
segments (species from 500 to 1000 bp), these were not reproducible.
Thus, the majority of germline V transcripts share the same
transcriptional starting site as the transcripts of rearranged
genes.
|
8.2-positive clones from the SCID thymus
library (Fig. 5A) provided an independent assay to check the
transcription start sites. All of these cDNAs were of the form
L5.1-V8.2. In agreement with the 5'-RACE results, four of the
five clones sequenced start at -15 (e.g., Fig. 4B, clone
14L12); only one clone has a longer 5' end with a start at -380
(Fig. 4B, clone 15 M16). These results show that the
distinctive transcriptional activities of the germline V segments
are likely to be controlled by sequence elements associated with the
known promoters as defined in mature T cells, and not with novel
promoters mapping at unknown sites far upstream.
V germline transcription alterations during lymphocyte
development before TCR gene rearrangement
Because individual Vs are expressed at different levels, they
could also be regulated differentially during the earliest stages of T
cell development. To test this hypothesis, we examined expression of a
representative range of V segments in thymocytes at different
developmental stages. Early T cell precursor subsets were separated
from Rag2-/- or
TCR--/--/-
thymocytes by FACS sorting, using criteria established previously to
isolate cells in developmentally distinct states (10, 11, 12).
Four populations were collected: 1)
Sca-1+HSA- cells, which
are thought to be pluripotent precursor-like cells; 2)
Sca-1- HSA- cells, which
are NK-like cells; 3)
HSA+CD44+ cells, which are
rapidly proliferating pro-T cells; and 4)
HSA+CD44- cells, which are
mostly resting, committed T cell progenitors that in normal mice would
be undergoing V-DJ rearrangement (reviewed in Ref. 5).
For this analysis, only V transcripts that are predominantly spliced
in total Rag2-/- cells were monitored, to avoid
any question about low-level DNA contamination in the sorted
populations. Data representative of at least three independent
experiments are shown in Fig. 6A, with a tabulation of
results of all trials in Fig. 6B.
|
, individual V not only were expressed to
different levels, but also showed evidence of differential
developmental regulation. At every stage, the "V 5.1" (L5.1)
promoter was most active, yielding high levels of L5.1-V5.1
spliced transcripts and even higher levels of L5.1-V8.2
transcripts. Activity of this promoter appeared to increase at the
pro-T stage (Fig. 6A, compare lanes 1 and
2 vs lanes 3 and 4 and lanes
5 and 6 vs lanes 7 and 8) but was
readily detectable earlier. The expression of V1.1 was much lower in
the precursor-like and NK-like cells, but appeared to be strongly
up-regulated at the pro-T stage and maintained at this level
thereafter. Activity of the V4.1 promoter appeared to undergo
up-regulation at about the same stage as the V1.1 promoter (Fig. 6A, lanes 7 and 8 are most
representative). V14.1 showed a different pattern: while barely
detectable, in two experiments the only expression seen was in the
precursor-like cells (not shown). Most noteworthy is the case of
V2.1. V2.1 RNA was detectable at a higher level in the
precursor-like fraction (Fig. 6A, lanes 1 and
5) than in the definitive T lineage populations (Fig. 6A, lanes 3, 4, 7, and 8), and at the
highest level of all in the NK-like fraction (Fig. 6A,
lanes 2 and 6). Expression of V2.1 continued
to be readily detectable in both fresh and IL-2-stimulated peripheral
NK cells, along with expression of V5.1–8.2 (data not shown). The
pattern of expression of V2.1 is strikingly different from all of
the others.
The individual expression patterns of these different V segments in
thymocyte subsets led to the question whether some of these promoters
might be active prethymically or in any non-T cells. Therefore, we
monitored V expression in bone marrow, fetal liver, and fetal
thymus. Two subpopulations were separated from
Rag2-/- bone marrow cells: pro-B cells, which
are
B220+CD43+CD19+;
and a population enriched for myeloid lineage cells, which are
B220- CD43+. To detect
what was anticipated to be very low-level expression, RT-PCR assays
were conducted in the presence of radioactive nucleotides, and the
yield of products of the right size was determined by phosphorimager
analysis, using the expression of each V in fetal thymus as a
quantitation standard. As shown in Fig. 6C, the expression
of V1.1, V5.1 (L5.1-V5.1), and V8.2 (L5.1-V8.2)
were much lower in pro-B cells, fetal liver, and the myeloid-enriched
population than in fetal thymus. However, the expression of V2.1 was
much stronger in the
B220-CD43+
myeloid-enriched population. The expression pattern of V4.1 showed
highest levels in the fetal thymus but distinctly elevated expression
in the B220-CD43+
population as well. The identity of the V-expressing cells in the
B220-CD43+ bone marrow
population remains to be determined. There are no mature T or B cells
in Rag2-/- bone marrow, but it is possible that
there are some T cell or NK cell precursors in this extrathymic
population.
These results confirm that the different V promoters respond
differentially to developmental transcriptional regulators in early
lymphocyte development. They are all expressed at the highest levels
during and after T lineage specification (Fig. 6A,
lanes 3 and 4), but specific V promoters such
as V5.1, V2.1, and V4.1 also have distinctive patterns of
expression in more immature cells and in prethymic or extrathymic cells
that may or may not ever enter the T cell lineage.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
segments differ
functionally so that they are transcribed at widely differing levels
before rearrangement. We have also shown that V expression levels
are actively regulated, independently of rearrangement, during T cell
development. Most interestingly, different V segments display
distinctive, individual patterns of expression during these early
stages. This provides strong support for the view that the promoter
regions of different V segments do not act as simple basal
promoters, but may have a complex structure that mediates
tissue-specific and stage-specific regulation of TCR expression
(17).
The data presented here provide an in vivo context in which specific
motifs that are differentially distributed among V promoters may be
tested for functional significance. Until now, lacking evidence for
differential function of various TCR V promoters, the emphasis has
been on shared elements, and only a few have been identified. As shown
in Table II, most of the V promoters that are expressed well, before
rearrangement, share the presence of a CRE-like TGANNTCA motif in their
5'-flanking regions. The V segments that are expressed
poorly before rearrangement, even though they contribute strongly to
the mature T cell pool, generally lack this motif: i.e., V6.1,
V3.1, and V14.1. This motif forms the core of the 14-mer,
CAGTGAYRTCACNG (2), that has
also been suggested as a conserved regulatory element. However, the
TGANNTCA motif cannot be essential for germline expression, for it is
not found in a consensus form in the promoters of V2.1 or V4.1
(scanning from -600 to +100 with respect to the translational
initiation site), which our results show are certainly expressed before
rearrangement, whereas it is clearly present in the promoter of
V7.1, which is not expressed nearly as well. So this motif is
neither strictly necessary nor rate-limiting for expression before
rearrangement. The additional cis-regulatory sites that are
particularly important and the factors that bind to them have yet to be
determined.
The highly divergent V promoters afford more than enough potential
regulatory diversity to account for the noncoordinate expression of
these gene segments (see Fig. 3B). However, other kinds of
sequence elements may also contribute. Another feature that may affect
the levels of expression reported here could be the genomic context
itself, particularly the location of repeat sequences with silencing
activity. The proximity and type of repeat sequences abutting the V
promoters vary greatly, from examples with repeats restricted to over 1
kb away (V4.1, the genes in the V5–8 cluster, and V14) to
examples with repeats encroaching within 350 bp of the presumed
transcriptional start site (V2.1, V20.1, and V3.1) (see
GenBank accession numbers AE000663, AE000664, and AE000665). Finally,
the sensitivity with which particular V transcripts can be detected
could be greatly affected by cis-acting sequences that
affect RNA stability. To resolve the roles of these different kinds of
regulatory elements, transfection and transgenesis experiments will
have to be done. The distinctive patterns of expression of particular
V segments before rearrangement can now provide the assay system to
score results of such experiments.
Our data establish not only that the various V promoters are
regulated independently in vivo; they also establish that certain of
these promoters may be among the first T lineage genes activated in
lymphoid precursors. This confirms and extends earlier work that showed
some prethymic expression of the V5.1–8.2 mRNA (7).
Our studies of this fine-scale regulation are still in an early phase,
but two features deserve comment. The first is the coordinate
up-regulation of several TCR V gene segments at the transition when
precursor cells undergo specification to the T cell lineage. The second
is the observation that certain V segments are also expressed well
in cell populations that are unlikely to be committed to the T cell
lineage.
The V1.1, V4.1, V5.1, and V5.1–8.2 transcripts undergo
up-regulation during T lineage specification and commitment (Fig. 6A, populations 3 and 4). This is probably caused by the
global changes in the expression of transcriptional regulators that
take place at this stage (12, 18) and may reflect a
response to T lineage restricted factors. One candidate for such a
factor acting on V promoters could be TCF-1 (19), which
has clusters of five predicted binding sites within 200–350 bp of the
mRNA cap sites in each of these promoters (see Fig. 2B).
These V segments join a group of genes that are similarly
up-regulated or newly induced at this stage, such as preT
, Rag1, and
TdT (10, 11). The V1.1 and V4.1 transcripts may be
particularly valuable as new lineage-specific indicators of this
critical developmental transition, because the Rag1 and TdT genes are
expressed in B lineage as well as T lineage precursors.
At least two V segments studied here show other striking regulatory
features. The V5.1–8.2 and V2.1 transcripts are distinguished by
their expression in the most immature subsets of thymocytes, which are
still thought to be pluripotent (3, 5). The expression of
V5.1–8.2 is detectable even in a further purified subset of
extremely immature thymocytes with a multilineage gene expression
pattern (H. Wang and F. Chen, unpublished data). V2.1, in contrast,
provides an example of regulation that is qualitatively as well as
quantitatively unique. Unlike all other V segments tested, V2.1
is expressed more strongly in an NK-like subset in the thymus than in
pro-T cell subsets, and expressed in pro-B cells in bone marrow. Its
transcripts continue to be readily detectable in peripheral NK cells.
It is also expressed in an enigmatic
CD43+CD45R- population of
cells, possibly myeloid, in Rag2-/- bone
marrow, at even higher levels than in fetal thymus. V5.1–8.2 and
V4.1 are expressed in this bone marrow population too, but at lower
levels. It may be significant that the promoter region of V2.1
(between about -100 and -200 in Fig. 3B (20))
is relatively deficient in predicted TCF-1 sites, but instead harbors a
unique cluster of Ets sites complexed with a functional AP-1 site
(21), unlike any other V analyzed here. The
V5.1–8.2 and V2.1 segments thus join IL-2 and perforin in a
group of T cell associated genes that are expressed in immature cells
precociously or outside the T lineage (11).
The highly divergent regulation of V2.1 may be a direct or indirect
result of its isolation from the rest of the cluster (see Fig. 3A). It is not only far upstream of other V segments, but
even upstream of the trypsinogen gene cluster at the 5' end of the
whole complex. There is no TCR V segment or pseudogene in a similar
position in the human TCR complex (2) and no human
ortholog of TCR V2 identifiable on the basis of sequence
(22), suggesting that V2.1 may have been deleted
altogether in the lineage leading to humans at some point since the
last common ancestor of humans and mice. It is interesting to speculate
that an early translocation may have moved the V2 ancestral gene to
its unusual site, where it might then have been particularly vulnerable
to deletion without disruption of the locus as a whole. Such a
translocation could juxtapose the V2.1 coding sequences with novel
regulatory elements directly. Because the amino acid sequence of murine
V2.1 is highly divergent from other murine and human V segments,
though, it is also possible that its unusual regulatory pattern may
have been established through genetic drift. These results raise the
possibility that V2.1 in mice may have acquired new biological
roles, in addition to its use in the TCR of conventional T
cells.
Even once activated in pro-T cells, expression of these V germline
transcripts remains very low. Precise figures are not available, but a
rough estimate can be made on the basis of the relative RT-PCR product
yields for different V segments and the frequency of V cDNA
clones in our SCID thymocyte cDNA library. Cells that have undergone T
lineage specification (HSA+) constitute about
90% of the population (23) and presumably dominate the
library. A typical thymic lymphocyte with 1 pg of total RNA (est. 20 fg
of poly(A)+ RNA) is likely to contain about
20,000 molecules of mRNA per cell (taking an average mRNA to contain
1500 nt at 350 g/mol nucleotide, to give about 9 x
10-19 g per mRNA). Thus the 70,000 clone library
that was screened represents approximately three cell equivalents. The
V5.1–8.2 sequence was recovered in about 50 separate clones in the
library, but it is expressed more strongly than most other V
segments by several orders of magnitude. This implies that most of the
V germline transcripts are present at much less than one copy per
average cell even in the
HSA+CD25+ population. This
raises the interesting possibility that most germline V segments are
transcribed only in a minority of pro-T cells. Preliminary studies
indicate that these transcripts need not be allelically excluded. In
thymocytes expressing a TCR transgene, although rearrangement of the
V8.2 segment is inhibited, expression of the V5.1–8.2 germline
transcript continues (data not shown). However, further work will be
needed to determine whether in normal mice the same rare precursor
cells express more than one V segment at a time.
Could such germline transcription of V segments foreshadow or even
direct segment choice for V(D)J rearrangement? This question has been
extensively debated, but it is posed in an immediate form by the highly
biased transcription we see of some segments at the expense of others.
Comparison of the V germline expression levels detected in this
study with the first wave of TCR rearrangement frequencies reported
in day 14 wild-type fetal thymus (24) shows some parallels
(Table II), but our transcription data and the reported rearrangement
frequencies are not well correlated overall. One interpretation may be
that germline transcription of certain TCR variable regions could be a
consequence of chromatin-level events in a larger subregion of the
TCR locus, rather than a mechanism for targeting these events.
Significant differences among V segments in recombination signal
sequences could independently skew rearrangement frequencies, perhaps
compensating, to some extent, for quantitative biases introduced by
differential germline transcription. However, our data leave open the
possibility that differential germline transcription plays a more
important role in targeting rearrangements in certain compartments of
the TCR repertoire. It is striking that the
NK1.1+ subset in the
Rag2-/- thymus may express only two V
segments, V2 and V5.1(L)-8.2(V). Provocatively, these are also
two of the most prevalent V segments used to form the TCR of
NK1.1+ T cells in the normal T cell population
(25).
In summary, the regulatory elements that control expression of
individual TCR V segments have patterns of activity in the absence
of rearrangement that differ quantitatively, temporally, and in terms
of developmental lineage specificity. The unexpected magnitude of
expression differences between the individual germline V segments
raise a range of novel questions about their regulation and biological
functions, but also provide an opportunity to define the mechanisms
responsible.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Ellen V. Rothenberg, Division of Biology 156–29, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA 91125.
3 Abbreviations used in this paper: Rag2, recombinase-activating gene-2; AML1, acute myelogenous leukemia gene-1 (Runt family transcription factor also known as Runx1, core-binding factor 2, etc.); bHLH, basic helix-loop helix (transcription factor class); CRE, cAMP response element; E14.5, embryonic day of gestation 14.5; Ets, E26 transformation-specific gene (a family of winged-helix transcription factors); RACE, rapid amplification of cDNA ends; TCF-1, T cell factor-1 (high mobility group box transcription factor); TE, Tris-EDTA buffer; B6, C57BL/6; HSA, heat stable Ag.
Received for publication October 10, 2000. Accepted for publication November 15, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cell receptor locus. Science 272:1755.[Abstract]
, , and rearrangements during adult thymic development: T cell receptor rearrangements are present in CD44+ CD25+ pro-T thymocytes. Proc. Natl. Acad. Sci. USA 95:12522.
and TCR proteins during adult thymus development. Int. Immunol. 11:1641.8.2 gene germ-line transcription: an early event of lymphocyte differentiation. Eur. J. Immunol. 24:3073.[Medline]
gene enhancer blocks T-cell development. Proc. Natl. Acad. Sci. USA 93:7877. (B29) in thymocytes: positive and negative gene regulation to establish T-cell identity. Proc. Natl. Acad. Sci. USA 95:6831. gene enhancer during early T cell development. Implications for the control of TCR- locus recombination. J. Exp. Med. 192:625. promoter depends on a conserved decamer motif similar to the cyclic AMP response element. Mol. Cell. Biol. 9:4835.J gene segments in precursor T lymphocytes. J. Immunol. 161:1399. promoters transcribe novel T-cell receptor- transcripts in early development. Immunology 93:213.[Medline]
-chain core promoter. Mol. Cell. Biol. 17:4220.[Abstract]
-chain gene is controlled by multiple regulatory elements. J. Immunol. 144:1111.[Abstract]
2 promoter. J. Immunol. 149:1980.[Abstract]
repertoire development in the mouse embryo. J. Immunol. 159:4227.[Abstract]
requirement for the development of NK1.1+ T cell receptor-/+ cells in mouse liver. J. Exp. Med. 183:1277. transcript levels in thymus and peripheral lymphoid tissues from various mouse strains: inverse correlation of I-E and Mls expression with relative abundance of several V transcripts in peripheral lymphoid tissues. J. Exp. Med. 169:1703.This article has been cited by other articles:
|
|
 |
|
  A. M. Tatum, L. M. Mylin, S. J. Bender, M. A. Fischer, B. A. Vigliotti, M. J. Tevethia, S. S. Tevethia, and T. D. Schell CD8+ T Cells Targeting a Single Immunodominant Epitope are Sufficient for Elimination of Established SV40 T Antigen-Induced Brain Tumors J. Immunol., September 15, 2008; 181(6): 4406 - 4417. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Melenhorst, M. D. H. Lay, D. A. Price, S. D. Adams, J. Zeilah, E. Sosa, N. F. Hensel, D. Follmann, D. C. Douek, M. P. Davenport, et al. Contribution of TCR-{beta} Locus and HLA to the Shape of the Mature Human V{beta} Repertoire J. Immunol., May 15, 2008; 180(10): 6484 - 6489. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Tydell, E.-S. David-Fung, J. E. Moore, L. Rowen, T. Taghon, and E. V. Rothenberg Molecular Dissection of Prethymic Progenitor Entry into the T Lymphocyte Developmental Pathway J. Immunol., July 1, 2007; 179(1): 421 - 438. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E. Busse, A. Krotkova, and K. Eichmann The TCR{beta} Enhancer Is Dispensable for the Expression of Rearranged TCR{beta} Genes in Thymic DN2/DN3 Populations but Not at Later Stages J. Immunol., September 1, 2005; 175(5): 3067 - 3074. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Otahal, S. C. Hutchinson, L. M. Mylin, M. J. Tevethia, S. S. Tevethia, and T. D. Schell Inefficient Cross-Presentation Limits the CD8+ T Cell Response to a Subdominant Tumor Antigen Epitope J. Immunol., July 15, 2005; 175(2): 700 - 712. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Xin, R. E. Mains, and B. A. Eipper Monooxygenase X, a Member of the Copper-dependent Monooxygenase Family Localized to the Endoplasmic Reticulum J. Biol. Chem., November 12, 2004; 279(46): 48159 - 48167. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Staveley-O'Carroll, T. D. Schell, M. Jimenez, L. M. Mylin, M. J. Tevethia, S. P. Schoenberger, and S. S. Tevethia In Vivo Ligation of CD40 Enhances Priming Against the Endogenous Tumor Antigen and Promotes CD8+ T Cell Effector Function in SV40 T Antigen Transgenic Mice J. Immunol., July 15, 2003; 171(2): 697 - 707. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Senoo, L. Wang, D. Suzuki, N. Takeda, Y. Shinkai, and S. Habu Increase of TCR V{beta} Accessibility within E{beta} Regulatory Region Influences its Recombination Frequency But Not Allelic Exclusion J. Immunol., July 15, 2003; 171(2): 829 - 835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Klein, N. Feldhahn, S. Lee, H. Wang, F. Ciuffi, M. von Elstermann, M. L. Toribio, H. Sauer, M. Wartenberg, V. S. Barath, et al. T lymphoid differentiation in human bone marrow PNAS, May 27, 2003; 100(11): 6747 - 6752. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Mathieu, S. Spicuglia, S. Gorbatch, O. Cabaud, C. Fernex, C. Verthuy, W. M. Hempel, A.-O. Hueber, and P. Ferrier Assessing the Role of the T Cell Receptor beta Gene Enhancer in Regulating Coding Joint Formation during V(D)J Recombination J. Biol. Chem., May 9, 2003; 278(20): 18101 - 18109. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Tripathi, A. Jackson, and M. S. Krangel A Change in the Structure of V{beta} Chromatin Associated with TCR {beta} Allelic Exclusion J. Immunol., March 1, 2002; 168(5): 2316 - 2324. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
B sites. Placement of symbols above or below the lines
on the figure is only to aid visualization of closely linked or
overlapping motifs.
