This information is current as
of June 17, 2017.
Thymocyte Maturation: Selection for
In-Frame TCR α-Chain Rearrangement Is
Followed by Selection for Shorter TCR β
-Chain Complementarity-Determining
Region 3
Maryam Yassai and Jack Gorski
J Immunol 2000; 165:3706-3712; ;
doi: 10.4049/jimmunol.165.7.3706
http://www.jimmunol.org/content/165/7/3706
Subscription
Permissions
Email Alerts
This article cites 33 articles, 16 of which you can access for free at:
http://www.jimmunol.org/content/165/7/3706.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2000 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
References
Thymocyte Maturation: Selection for In-Frame TCR ␣-Chain
Rearrangement Is Followed by Selection for Shorter TCR
␤-Chain Complementarity-Determining Region 31
Maryam Yassai and Jack Gorski2
roduction of a functional peripheral T cell repertoire requires a number of maturation steps to take place in the
thymus. These are in part a result of the rearrangement
process, which provides receptor flexibility at a cost of generating
cells with nonproductive rearrangements. To facilitate the process,
selection takes place for each TCR chain separately. In the transition from double-negative (DN)3 to double-positive (DP) thymocytes, the quality of the ␤-chain rearrangement product is controlled by its ability to pair with the pre-T ␣-chain (1). This occurs
in the CD44⫺ compartment of DN mouse thymocytes (2, 3) and is
referred to as ␤ selection (4). The subset of DN cells in which the
␤-chain is selected in man is not known, although recent data have
implicated the CD4⫹CD3⫺CD8⫺ immature single-positive (ISP)
cells (5). Later stages, which would include pairing of the TCR
␤-chain with a productively rearranged TCR ␣-chain and recognition of the peptide-MHC ligand by TCR␣␤, are less well
understood.
Thymic selection has been divided into two conceptual frameworks, referred to as positive selection and negative selection
(6, 7). Negative selection is easily understood as elimination of T
cells whose receptor/coreceptor affinity for self-peptide-MHC is
too high (8). Positive selection can be defined quite broadly, ranging from the rescue of thymocytes from programmed cell death to
P
the specific stimulation of a thymocyte by a peptide mimic of the
future Ag. The most accepted definition of positive selection implicates only those events in which the thymocyte is interacting
with self-MHC-peptide (see Ref. 9 for review). Historically, this
form of selection has been closely linked to lineage selection,
which was used as the readout. There have been recent studies
indicating a division between positive selection of thymocytes and
lineage selection (10, 11). There has been a large effort in determining the roles of peptides in the selection process (12–14).
Selection has also been assayed independent of the lineage
markers using TCR ␤-chain Tg mice. It has been reported that in
some cases the ␤-chain pairs preferentially with ␣-chains similar
to those with which it was paired in the hybridoma of origin. By
assaying the stage at which this ␣-chain selection is observed, this
form of positive selection has been mapped to the CD69⫹ subset
of DP cells (15).
We have been investigating the rearrangement status of the TCR
␣- and ␤-chain loci during human thymocyte maturation. Evidence
for ␤- and ␣-selection was obtained. As part of these studies we
have identified an additional stage in the maturation process that
involves the accumulation of SP cells that contain TCR ␤-chains
with shorter lengths of complementarity-determining region 3
(CDR3). These results are discussed in terms of our current understanding of thymic selection.
The Blood Research Institute, The Blood Center of Southeastern Wisconsin, Milwaukee WI 53201
Materials and Methods
Received for publication March 13, 2000. Accepted for publication July 17, 2000.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by the Blood Center Research Foundation.
2
Address correspondence and reprint requests to Dr. Jack Gorski, The Blood Research Institute, The Blood Center of Southeastern Wisconsin, POB 2178, Milwaukee, WI 53201-2178. E-mail address: [email protected]
3
Abbreviations used in this paper: DN, double negative; DP, double positive; ISP,
immature single positive; CDR, complementarity-determining region; RF, relative
frequency; FAM, 5⬘-carboxyfluorescein.
Copyright © 2000 by The American Association of Immunologists
Cells
Thymi were obtained as surgical tissue discards from The Children’s Hospital of Wisconsin. PBMC were obtained as discards after removal of indwelling catheters. All materials were obtained under an institution review
board-approved protocol.
Fluorescent staining and sorting
Thymi were disaggregated by passing them through a wire mesh. Cells
were suspended in RPMI medium (Life Technologies, Gaithersburg, MD),
0.1% sodium azide, and 2% FCS. To determine whether the thymi were
normal, 0.5 ⫻ 106 cells were stained using mouse mAbs to human cell
0022-1767/00/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
Thymocyte maturation consists of a number of stages, the goal of which is the production of functioning T cells that respond to
foreign antigenic peptides using their clonotypic receptors. Selection of a productively rearranged TCR ␤-chain is the first stage
in the process and occurs at the double-negative to double-positive (DP) transition. Later maturation stages are based on changes
in markers such as CD5, CD69, or IL-7R. A stage in which ␣-chains are selected has also been identified using ␤-chain transgenic
mice. Here we identify two additional selection stages in human thymocytes based on characteristics of the TCR. ␣ selection is
measured directly by identification of in-frame rearrangements and is associated with the appearance of CD3 on the DP thymocyte
surface. The next stage has not yet been described and involves selection of thymocytes that express shorter TCR ␤-chain
complementarity-determining region 3 (CDR3). This stage is associated with the acquisition of high levels of CDR3 by DP cells and
the transition to SP thymocytes. The extent of CDR3 length selection observed is a function of the TCR V and J genes. We propose
that CDR3 length selection is based on recognition of the MHC. Thus, there exist limitations on the allowable length of that portion of
the TCR most intimately in contact with MHC and peptide. This may be a physical representation of positive selection. The Journal
of Immunology, 2000, 165: 3706 –3712.
The Journal of Immunology
surface markers; CD3-FITC conjugate, TCR␣␤-FITC conjugate, CD4-Tricolor conjugate, and CD8-R-PE conjugate (Caltag, San Francisco, CA).
The stained cells were analyzed using FACScan (Becton Dickinson, San
Jose, CA). Thymi that had normal CD3, CD4, and CD8 profiles were then
stained for sorting. Three color sorts were performed using FACStar (Becton Dickinson), and different populations were collected. Primary gating
was on the CD3 marker, which resolved the thymocytes into three populations, referred to as CD3neg, CD3low, and CD3high (Fig. 1A). These three
populations were then further divided on the basis of CD4 and CD8 expression (Fig. 1A). Cells were collected into 0.5 ml of FCS so that the final
concentration in the tube was 10% (5-ml final volume).
Preparation of nucleic acids from sorted cells
For DNA, sorted cells were spun down and resuspended in nucleic lysis
buffer, pH 8.2 (10 mM Tris, 0.4 M NaCl, and 2 mM EDTA), in the presence of SDS and proteinase K, then the cells were incubated overnight at
45 C to ensure the complete lysis. After the incubation, proteins were
precipitated by adding 5.3 M NaCl, and DNA was isolated from the supernatant by ethanol precipitation (16). RNA was made using TRIzol reagent (Life Technologies, Gaithersburg, MD).
Rearrangement analysis
1B. Volumes of the DNA preparations were chosen so that the signals
would be identical. The titration procedure was described in greater
detail previously (19).
RT-PCR
Levels of pre-T␣ and ␣-chain mRNA were measured by RT-PCR using
primers specific for each cDNA. One microgram of total RNA was converted to cDNA using Moloney murine leukemia virus reverse transcriptase. The cDNA from a different population of thymocytes was titrated to
determine the amount needed to obtain an equivalent actin ␤-chain mRNA
signal. Serial dilutions of cDNA were amplified for 24 cycles using two
primers, one in exon 2 and the other in the exon 3 of the actin ␤-chain
locus. Based on the ␤-actin titration, levels of pre-T␣ and ␣-chain mRNA
were measured using primers specific for each cDNA. Three concentrations of cDNA were used for the PCR to insure a linear response of fluorescent signal to input. The sequences of the primers used are as follows:
␤-actin direct, 5⬘-CGTGTGGCTCCCGAGGAGCACC-3⬘; ␤-actin antiFam labeled, 5⬘-CCCTGTACGCCTCTGGCCGTACCAC-3⬘; pre-T␣ direct, 5⬘-GGCACACCCTTTCCTTCTCTG-3⬘; pre-T␣ anti-Fam labeled,
5⬘-GCTTCTACAGCCAGGACCTGC-3⬘; C␣ direct, 5⬘-GATATCCAGA
ACCCTGACCC-3⬘; and C␣ anti-Fam labeled, 5⬘-ATGACGCTGCGGCT
GTGGTCCAG-3⬘.
Results
Recombination analysis of thymocyte Ag receptors
We have used the variations in intensity of CDR3 length of the
TCR ␤-chain to assay changes in the T cell repertoire. It was of
some interest to extend these studies to thymocytes, as mature SP
thymocytes represent the immediate precursor of naive circulating
T cells. The recombination assay consists of generating a PCR
product that amplifies the CDR3 using V family-specific and J
region-specific primers. The length of the CDR3 thus amplified is
resolved on denaturing acrylamide gels. We have published an
analysis of the thymic rearrangement profiles of normal V genes
compared with pseudogenes (18) and have also used this method
for analysis of the relationship of ␥␦ thymocytes to ␣␤ thymocytes
(17). The technique and approach are similar to those described by
Hayday and colleagues (4, 20).
Analysis of TCR V ␤-chain genes during thymocyte maturation
FIGURE 1. Titration of DNA from human thymocyte subsets. A, Human thymocyte subsets fractionated on the basis of three surface markers.
The first panel shows the CD3neg (R1), CD3low (R2), and CD3high (R3)
populations. The next three panels show the CD4 and CD8 profiles of each
of the three CD3 populations. Arrows identify regions used for each thymic
subpopulation. B, Titration of DNA isolated from five different subsets of
the same thymus. PCR is performed at a number of different DNA concentrations to insure that the amplified product is a function of input DNA.
The subsets are identified in the inset.
The thymic maturation series in man, as assayed by the surface expression of CD3, CD4, and CD8, consists of the following stages,
CD3negCD4negCD8neg cells (triple negative) followed by two DP
compartments, CD3negCD4⫹CD8⫹ and CDlowCD4⫹CD8⫹. This division of DP thymocytes into almost equivalent numbers of CD3neg
and CD3low is specific to man and is not found in nontransgenic
laboratory mice. In man, triple-negative cells proceed to the DP
CD3neg stage through an ISP CD3⫺CD4⫹CD8⫺ compartment (21).
The most mature cells are SP, showing the following markers:
CD3highCD4⫹CD8⫺ and CD3highCD4⫺CD8⫹ (reviewed in Ref. 22).
The flow cytometric profile of a typical human thymus is shown in
Fig. 1A. The three levels of CD3 expression are shown in the first
panel, and the CD4 and CD8 profiles of the three CD3 gates are
shown in the following panels.
An example of a typical recombination analysis is shown in Fig.
2. The amount of DNA used for each amplification has been normalized by titration using a common set of primers amplifying
exon 1 of the ␤ constant region gene. The banding pattern shows
a 3-bp spacing indicative of in-frame selection. The intensity of
rearranged genes is the same throughout the maturation series,
indicating no further ␤-chain rearrangements. We have analyzed
six human thymi, representing all age groups in which a reasonable
amount of thymic tissue is still found, and have observed the same
patterns of BV rearrangement in these five subsets.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
Rearrangement analysis was performed by PCR amplification of the CDR3
using V and J region-specific primers. A description of the methods has
been published (17–19). The J region primer was labeled with 5⬘-carboxyfluorescein (FAM), the PCR products were analyzed on denaturing polyacrylamide gels, and the fluorescent PCR products were quantitated using
a FluorImager (Molecular Dynamics, Sunnyvale, CA). Data were collected
as a 16-bit Tiff file. Band intensities could be further analyzed using ImageQuant and spreadsheet software. For calculation of CDR3 length
changes, band intensities originally measured as relative fluorescence units
by the FluorImager were converted to the relative frequency (RF) of each
band over the total band intensity. The relative band intensities correct for
minor fluctuations in the data. The use of RF to calculate shortening is
shown in Fig. 5, and a general description is given in Ref. 19.
Rearrangement analysis was performed on DNA samples that had
been titrated to insure equal efficiency of amplification of the ␤-chain
DNA constant region. An example of such a titration is shown in Fig.
3707
3708
THYMOCYTE SELECTION AFFECTS TCR ␤-CHAIN CDR3 LENGTH
FIGURE 2. Analysis of BV2-BJ2.7 rearrangements in thymocyte subsets.
Subsets are identified above each lane and have been defined in the text.
Analysis of TCR V ␣-chain genes during thymocyte maturation
FIGURE 3. Recombination analysis of AV TCR in thymocyte subsets.
A, Analysis of AV8-AJ49 recombinations in thymocyte subsets. The DNA
used in these analysis is the same as that used for the BV2-BJ2.7 analysis
in Fig. 2. B, Analysis of AV1S2-AJ4 rearrangements in the two DP thymocyte subsets from thymus T112. The subsets are identified above each
lane. The lane labeled RNA shows PCR products from peripheral T cell
cDNA and identifies the in-frame CDR3 sizes. C, Analysis of AV12S1AJ4 rearrangements in the two DP subsets from thymus T111.
FIGURE 4. RT-PCR analysis of pre-T␣ and VA mRNA levels in thymocyte subsets. A, Pre-T␣ cDNA. B, VA constant cDNA. The subsets
analyzed are identified in the inset. cDNA concentrations for the three
compartments were normalized on the basis of a titration with ␤-actin
mRNA primers. PCRs for pre-T␣ and AV were performed using the normalized amount of cDNA as well as half and twice the amount, respectively, to insure a linear response relative to the input.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
Recent experiments using ␤-chain TCR transgenic mice have identified a stage occurring in DP cells in which the ␣-chain is selected
(10). These took advantage of a propensity of the ␤-chain to pair
with ␣-chains similar to that with which it was paired in the hybridoma of origin (15, 23). This identifies a stage that can be referred to as ␣ selection. A candidate for such a stage in man is the
CD3low DP subset. Expression of detectable CD3 on the surface
implies that both TCR chains are expressed on the surface, i.e., that
pairing has taken place.
We tested this supposition by using rearrangement analysis to
determine whether the TCR ␣-chain was selected in the CD3low
DP cells. Fig. 3A shows analysis of the same five thymocyte subsets as were analyzed for BV rearrangement in Fig. 2. The data are
for the AV8 family. It is obvious that there are very few rearranged
␣-chain genes in the CD4 ISP and CD3neg DP compartments.
There is a large jump in the overall intensity of rearranged AV
genes in the CD3low DP population, which would indicate that
rearrangement is taking place at this stage. The increases in inten-
sity in the SP populations indicate further rearrangement. However, the rearrangement profiles show a single base pair spacing
throughout the different compartments. This is probably due to the
amplification of rearranged genes that are on excised circles generated as part of the continuing rearrangement process (24, 25).
The accumulation of circles with rearranged ␣-chain genes, most
of which are out-of-frame, thwarts the ability of the analysis to
define at which point thymocytes with in-frame rearrangements are
selected.
To overcome the difficulty posed by recombination circles we
used available AV (26) and AJ (27) gene maps to confine the
recombination analysis to those genes that cannot be further excised because of their distal (AV) or proximal (AJ) positions.
Analysis was performed using the AV1S1 locus, which is the first
AV locus and thus must be retained on the chromosome. This
analysis shows that the selection for in-frame AV genes is first
observed in the CD3low DP compartment (Fig. 3B), whereas the
CD3neg DP compartment has accumulated thymocytes with outof-frame rearrangements for the same AV gene. The analysis was
also performed with AJ4 primers. AJ4 is the first AJ locus that is
used to any great extent in AVJ joining and thus must be maintained on the chromosome. In the analysis with AJ4 and AV12S2
primers (Fig. 3C), the same in-frame ␣ selection was observed
between the two DP stages. The results in Fig. 3, B and C, are from
two different thymi, and similar results were obtained with three
other thymi.
Further data supporting the observation the ␣ selection takes
place between the CD3neg and CD3low DP compartments come
from analysis of pre-T␣ and TCR ␣-chain mRNA levels. Before
pairing with the ␣-chain, the TCR ␤-chain is paired with the preT␣. It would be expected that at the point at which ␣ selection has
occurred, the levels of pre-T␣ mRNA would decrease, whereas the
levels of TCR ␣-chain mRNA would increase. We performed such
an analysis on the RNA from three thymocyte subsets, and the
results are shown in Fig. 4. Quantitative RT-PCR was performed
The Journal of Immunology
3709
at a number of dilutions of the cDNA for either pre-T␣ (Fig. 4A)
or TCR ␣-chain (Fig. 4B). Before the analysis, the cDNA was
titrated to determine the amount needed to obtain an equivalent
␤-chain mRNA signal. The data are presented as a dilution series
to ensure that the response of the PCR is a linear function of the
input cDNA. It can be seen that there is a very significant decrease
in pre-T␣ mRNA between the two DP stages. Likewise, the most
significant increase in TCR ␣-chain mRNA was observed between
the two DP stages. This fits with ␣ selection taking place at the
CD3neg to CD3low DP boundary.
Maturation of SP thymocytes includes selection of cells with
shorter CDR3
Shortening is observed between the CD3low DP and SP stages
To determine whether the selection of thymocytes with shorter
CDR3 is a continuing process or whether there is a particular stage
at which this occurs, we analyzed the CDR3 length distributions in
the two DP compartments and the SP compartments of the thymus.
An example of such an analysis is shown for the BV7 and BV5.1
families (Fig. 6). The shortening for these families between the
CD3low DP and CD4 SP is readily discernible by visual inspection
of the gel data (Fig. 6A). The ⌬RF analysis clearly shows that there
is shortening that occurs when the CD3low DP is compared with
the CD4 SP (Fig. 6B). The same is not seen when the CD3neg DP
is compared with the CD3low DP (Fig. 6C). Thus, a characteristic
of SP thymocytes is that they tend to have shorter TCR BV CDR3.
To show the general nature of these observations, thymocyte
populations prepared from two different thymi are shown in Fig. 6.
As a control for the ⌬RF analysis, the two CD3neg DP populations
from two thymi were compared. Formally these profiles should be
similar and represent the CDR3 size distribution of the initial rearrangement process if no selection has taken place. Comparison
of the CD3neg DP from the two thymi gave very similar profiles
(Fig. 6D), as evidenced by the low ⌬RF. In all, we have clearly
observed the phenomenon in five different thymi and for 10 different BV genes.
Thus far, the data have shown the selection for CD4 SP thymocytes. The same can be observed for CD8 SP thymocytes. A representative gel analysis of a recombination analysis (Fig. 7A)
shows that the average CDR3 length of CD4 SP and CD8 SP are
very similar. The ⌬RF data for a number of V gene rearrangements
from two different thymi (Fig. 7, B and C) show that this is a
FIGURE 5. Measurement of CDR3 shortening. Data from the CD3low
DP and CD4 SP lanes of Fig. 2 are analyzed. A, The gel data are converted
to fluorescence intensity, expressed as relative fluorescence units. The original gel lanes are shown above and below the graph. Bands corresponding
to shorter CDR3 are on the left. B, Conversion of raw peak height data into
the RF for each band. C, Generating the ⌬RF by subtracting the RF of the
CD4 SP bands from that of the CD3low DP bands. A positive ⌬RF indicates
that the DP band is more intense. A negative ⌬RF indicates that the SP
band is more intense.
general phenomenon. Thus, there does not appear to be a difference between the two major SP thymocyte lineages with respect to
this phenotype.
The possible role of BJ genes was also explored. Thus far, all
the data we have presented used BJ2.7. This J gene was chosen
because it is the most frequently used J and thus provides increased
signal strength. The shortening observed is not just a function of
this J gene, as analysis of rearrangements using the BJ2.1 showed
the same phenomenon (not shown). The ⌬RF data show evidence
of shortening, although the overall pattern of shortening differed
by the J gene used.
The average CDR3 length of CD3high DP thymocytes is shorter
than that of the other DP subsets
While a majority of DP thymocytes show either no or low expression of CD3 on the surface, there is a small, but distinct, population
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
One of the more interesting data to come from the rearrangement
analyses is the observation that the average length of the CDR3
shortens between the DP CD3low and SP stages. This can be observed to some extent in the data presented in Fig. 2, although it is
not overtly apparent. For BV2-J2.7 rearrangements (Fig. 2), visual
inspection of the data is difficult, so we will use these data to
introduce a more quantitative approach for analysis of the band
fluorescence intensities. The steps involved in this are shown in
Fig. 5. The FluorImager data are converted to relative fluorescence
units using ImageQuant software (Fig. 5A). This is shown for the
CD3low DP and the CD4 SP lanes. The intensity of each band is
converted into the RF of the band with respect to the total band
intensity (Fig. 5B). The difference of the RF of equivalent bands in
any two samples compared yields the ⌬RF. If two band distributions are similar, the ⌬RF will hover around zero. If there is shortening, then the ⌬RF will be positive for the higher m.w. bands
(right) and negative for the lower mw bands (left). The data will
show a well-defined shift from negative to positive values. This
form of analysis shows that for BV2-BJ2.7 rearrangements there is
indeed shortening between the CD3low DP and CD4 SP populations (Fig. 5C).
3710
THYMOCYTE SELECTION AFFECTS TCR ␤-CHAIN CDR3 LENGTH
of DP thymocytes that express higher levels of CDR3. It is postulated that CD3high DP cells may the precursors of CD3high SP
cells (28, 29). The generation of SP thymocytes from transferred
CD3low DP thymocytes has been reported (30), but a direct precursor relationship of CD4high DP and SP thymocytes has not yet
been shown. The rearrangement status of the CD3high DP population was investigated to determine whether the selection was already taking place at this stage. The CD3high thymocytes (Fig. 8A)
were fractionated by their CD4 and CD8 expression (Fig. 8B), and
the DP and the CD4 SP cells were collected. CD3neg DP and
CD3low DP were collected from the respective populations as described previously (see Fig. 1). The four thymocyte populations
were analyzed, and an example is shown in Fig. 8C. The results
show that the selection for shorter CDR3 begins to be observed in
the CD3high DP population. The difference between the CD3low
DP and CD3high DP is not obvious by visual inspection (Fig. 8C),
but ⌬RF analysis shows that there is a selection step between the
two stages (Fig. 8E) that continues between the CD3high DP and
CD4 SP stages (Fig. 8F). As shown before, there was no evidence
for selection between the CD3neg DP and CD3low stages (Fig. 8D).
These data are compatible with the three stages, CD3low DP,
CD3high DP, and CD4 SP, constituting a sequential maturation
pathway characterized by increasing selection for short CDR3.
The relationship of shortening to thymus transit
There is a simple explanation for the accumulation of thymocytes
with shorter CDR3 in the SP subset. This is that the SP thymocytes
with longer CDR3 rapidly exit the thymus, and thymocytes with
shorter CDR3 are retained. While it is has been impossible for us
to obtain both thymus tissue discards and peripheral blood cells
from the same individual, the general nature of this phenomenon
should insure that comparison of thymus and peripheral T cells
between different individuals is sufficient. We compared two pairs
of age-matched samples, one from thymus and one from PBMC.
Rearrangement analysis of total thymocytes (predominantly DP
cells) and peripheral T cells showed easily visualized evidence of
CDR3 shortening between the two compartments (Fig. 9). The
mean length in the periphery was similar to that observed in SP
thymocytes. Thus, it is thymocytes with short CDR3 that are exported and accumulate in the periphery.
Discussion
In addition to the relatively well understood stage of ␤-selection,
our data have identified two additional stages of thymocyte selection. The second stage identified involves ␣-chain gene rearrangement and pairing of the ␣-chain with the ␤-chain. This takes place
within the DP thymocyte population and results in DP cells that are
CD3low, as they now express TCR␣␤ on the surface. Commensurate with this division of DP thymocytes is the decrease in levels
of pre-T␣ mRNA and the increase in AV mRNA in the CD3low DP
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 6. Measurement of CDR3 shortening for two different BV
families in two different thymi. A, Rearrangement analysis gel data. The
BV family and thymus analyzed are identified above each gel. Subsets are
identified above each lane. B, ⌬RF calculation for CD3low DP and CD4 SP
subsets. C, Measurement of ⌬RF between CD3neg DP and CD3low DP
subsets. D, Measurement of ⌬RF between the CD3neg DP subsets of thymus T114 and T111.
FIGURE 7. Analysis of CD4 SP and CD8 SP subsets. A, Gel data for
the BV5.1 rearrangement for thymus 114. ⌬RF data for results from T108
(B) and T114 (C) are shown. The V families analyzed are identified in the
inset. All rearrangements are for BJ2.7.
The Journal of Immunology
3711
FIGURE 9. Rearrangement analysis comparing thymocytes and peripheral T cells. An example of the gel data (BV7) is shown at the left. The
tissue source, thymus (T) or PBMC (P), is identified above the lane. The
⌬RF for three different BV families are shown in the panel on the right.
The data for the BV5.3 family are from one thymus-PBMC pair. The data
for the BV6.1 and BV7 families are from another thymus-PBMC pair.
population. A similar observation was reported previously, although the DP populations were not resolved (31). Direct evidence
for ␣ selection is obtained from rearrangement analyses of AV-AJ
rearrangements that must be maintained on the chromosome. For
such rearrangements, the first thymocyte population in which inframe rearrangements are observed is the CD3low DP population.
The third stage of maturation is characterized by accumulation
of SP thymocytes that have shorter CDR3. This is an unexpected
characteristic of thymocyte maturation. The shorter CDR3 observed in the CD4 SP and CD8 SP subsets is not a result of rapid
exit of thymocytes with longer CDR3, as peripheral T cells also
have short CDR3. The short length of peripheral TCR BV CDR3
had been noted previously (32), but it was not clear whether this
was a function of the rearrangement mechanism itself or of a shortening postrearrangement. The selection for thymocytes with
shorter CDR3 can be observed at the CD3high DP stage, strongly
supporting the precursor relationship between CD3low and CD3high
DP cells (10, 15, 28, 29).
The shortening phenomenon is a general one, having been observed for a number of individuals. However, it is not clear
whether it will be observed for the same V-J combination in all
individuals. Within any thymus, the phenomenon is characterized
by a dependence on the BV-BJ combination used for the analysis.
For example, it is less evident for the BV2 family used to obtain
the data in Fig. 2, whereas it is much more evident in the data for
the other families shown. The same sensitivity to the recombined
J gene has also been observed. While our studies have not been
exhaustive, the data have always shown some evidence of shortening, no matter which V or J gene was studied. In contrast, comparison of the CD3neg DP subset from different thymi does not
show any evidence for shortening. This would be expected if this
less mature subset had not undergone any selection and the com-
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 8. Analysis of the CD3high DP subset. The FACS profile for
CD3 (A) and CD4 vs CD8 (B) are shown to define the sorted population.
C, An example of the gel data. D–F, The ⌬RF of the analyses for four
different BV families with BJ2.7. The ⌬RF are calculated for CD3neg DPCD3low DP (D), CD3low DP-CD3high DP (E), CD3high DP-CD4 SP (F). BV
families are identified in the inset (D).
ponents of the rearrangement machinery responsible for determining CDR3 length were not polymorphic.
We propose that the observed accumulation of thymocytes with
shorter CDR3 in the transitions between the two TCR-expressing
DP populations, CD3low and CD3high, as well as that leading to SP
cells is a direct result of selection on the TCR ligand, i.e., peptideMHC molecules. This is the most reasonable interpretation of the
observed dependence of the selection on the V-J combination being analyzed. Direct evidence for the role of peptide-MHC in the
shortening process will have to come from work in the mouse,
where inbred strains and mutants are available. We have observed
selection of thymocytes with shorter CDR3 in CD4 SP thymocytes
in the mouse. Results using inbred mouse strains show that there
is an effect on the extent of shortening observed for a particular V-J
combination if MHC disparate or recombinant strains are examined (our manuscript in preparation). The extent of shortening is
much higher in 129 and B10 (H2b) than in B10.PL and PL (H2u)
mice, indicating that the MHC plays a role in the process. These
mouse data also speak to the generality of the CDR3 length
selection.
TCR and Ig employ the same machinery for generating recombinational diversity, whereas the recognition events for these two
classes of immune receptors are different. Therefore, it is not surprising that the recognition of peptide-MHC may require differences in the length of the contact specificity portion of the molecule. Shorter CDR3 could more easily form the flatter recognition
surface characteristic of those observed in TCR-MHC crystal
structures (33–35).
In addition to the role of the interaction of TCR with MHC:
peptide, another molecule that may dictate a need for shorter
CDR3 is the coreceptor, CD4 or CD8. The coreceptors are present
at the time of selection and may also impose structural limitations
on the preferred length of the CDR3.
Because the stage at which TCR ␣-chains are selected is different from that at which the shortening takes place, we do not think
that the shortening is related to pairing of the two chains. Our data
imply that ␣ selection is a distinct phenomenon from the selection
for thymocytes with shorter ␤-chain CDR3. If CDR3 shortening
represents the first selection on peptide-MHC molecules, then the
observation that ␣ selection precedes CDR3 shortening would imply that ␣ selection could be solely based on pairing and not on
peptide-MHC recognition.
While the data presented here provide a novel measure of thymocyte maturation, there remain a number of interesting issues
that will require further investigation. For example, it would be of
great interest to determine how the short CDR3 phenotype fits with
current models of positive selection at the DP to SP boundary (10,
3712
THYMOCYTE SELECTION AFFECTS TCR ␤-CHAIN CDR3 LENGTH
15, 29). If there is coreceptor involvement, the possible role of
CDR3 length selection in lineage commitment could be explored.
In the context of our current understanding of thymic maturation,
two opposing explanations for the selection process could be envisaged, falling under the rubric of either positive or negative selection. If only short CDR3 are compatible with the interaction of
the TCR with peptide-MHC (coreceptor) complexes needed to
maintain viability, this could be considered positive selection, with
elimination of unselected thymocytes by the “neglect” mechanism.
While perhaps less likely, it is possible that a longer CDR3 demonstrates a high affinity interaction with the peptide-MHC ligand
due to the generation of deeper, more complex, Ig-like contacts.
This would result in elimination of these thymocytes by negative
selection mechanisms. It will be interesting to determine which
mechanism is at work. The data presented here provide a physical
basis for considering the issues involved in thymocyte selection
and open up further avenues for studying the process.
Acknowledgments
References
1. Fehling, H. J., A. Krotkova, C. Saint-Ruf, and H. von Boehmer. 1995. Crucial
role of the pre-T-cell receptor ␣ gene in development of ␣␤ but not ␥␦ T cells.
[Published erratum appears in 1995 Nature 378:419.] Nature 375:795.
2. Hoffman, E. S., L. Passoni, T. Crompton, T. M. Leu, D. G. Schatz, A. Koff,
M. J. Owen, and A. C. Hayday. 1996. Productive T-cell receptor ␤-chain gene
rearrangement: coincident regulation of cell cycle and clonality during development in vivo. Genes Dev. 10:948.
3. Godfrey, D. I., J. Kennedy, T. Suda, and A. Zlotnik. 1993. A developmental
pathway involving four phenotypically and functionally distinct subsets of
CD3⫺CD4⫺CD8⫺ triple-negative adult mouse thymocytes defined by CD44 and
CD25 expression. J. Immunol. 150:4244.
4. Mallick, C. A., E. C. Dudley, J. L. Viney, M. J. Owen, and A. C. Hayday. 1993.
Rearrangement and diversity of T cell receptor ␤ chain genes in thymocytes: a
critical role for the ␤ chain in development. Cell 73:513.
5. Blom, B., M. C. Verschuren, M. H. Heemskerk, A. Q. Bakker, E. J. Gastel-Mol,
I. L. Wolvers-Tettero, J. J. van Dongen, and H. Spits. 1999. TCR gene rearrangements and expression of the pre-T cell receptor complex during human T-cell
differentiation. Blood 93:3033.
6. von Boehmer, H. 1994. Positive selection of lymphocytes. Cell 76:219.
7. Nossal, G. J. 1994. Negative selection of lymphocytes. Cell 76:229.
8. Kappler, J. W., N. Roehm, and P. Marrack. 1987. T cell tolerance by clonal
elimination in the thymus. Cell 49:273.
9. Sebzda, E., S. Mariathasan, T. Ohteki, R. Jones, M. F. Bachmann, and
P. S. Ohashi. 1999. Selection of the T cell repertoire. Annu. Rev. Immunol. 17:
829.
10. Bhandoola, A., R. Cibotti, J. A. Punt, L. Granger, A. J. Adams, S. O. Sharrow,
and A. Singer. 1999. Positive selection as a developmental progression initiated
by ␣␤ TCR signals that fix TCR specificity prior to lineage commitment. Immunity 10:301.
11. Keefe, R., V. Dave, D. Allman, D. Wiest, and D. J. Kappes. 1999. Regulation of
lineage commitment distinct from positive selection. Science 286:1149.
12. Janeway, C. A., Jr. 1999. T-cell development: a role for self-peptides in positive
selection. Curr. Biol. 9:R342.
13. Barton, G. M., and A. Y. Rudensky. 1999. Requirement for diverse, low-abundance peptides in positive selection of T cells. Science 283:67.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
We thank Drs. Burt Litwin and James Tweddell, Children’s Hospital of
Wisconsin, for the thymic tissue. The many discussions with our colleagues and especially members of the Gorski laboratory is gratefully
acknowledged.
14. Bevan, M. J. 1997. In thymic selection, peptide diversity gives and takes away.
Immunity 7:175.
15. Sant’Angelo, D. B., B. Lucas, P. G. Waterbury, B. Cohen, T. Brabb,
J. Goverman, R. N. Germain, and C. A. Janeway, Jr. 1998. A molecular map of
T cell development. Immunity 9:179.
16. Miller, S. A., D. D. Dykes, and H. F. Polesky. 1988. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16:
1215.
17. Margolis, D., M. Yassai, A. Hletko, L. McOlash, and J. Gorski. 1997. Concurrent
or sequential ␦ and ␤ TCR gene rearrangement during thymocyte development:
individual thymi follow distinct pathways. J. Immunol. 159:529.
18. Currier, J. R., M. Yassai, M. A. Robinson, and J. Gorski. 1996. Molecular defects
in TCRBV genes preclude thymic selection and limit the expressed TCR repertoire. J. Immunol. 157:170.
19. Yassai, M., E. Naumova, and J. Gorski. 1997. Generation of TCR spectratypes by
multiplex PCR for T cell repertoire analysis. In The Antigen T cell Receptor:
Selected Protocols and Applications. J. R. Oksenberg, ed. Landes Bioscience,
Austin, pp. 326 –372.
20. Passoni, L., E. S. Hoffman, S. Kim, T. Crompton, W. Pao, M. Q. Dong,
M. J. Owen, and A. C. Hayday. 1997. Intrathymic ␦ selection events in ␥␦ cell
development. Immunity 7:83.
21. Kraft, D. L., I. L. Weissman, and E. K. Waller. 1993. Differentiation of
CD3⫺4⫺8⫺ human fetal thymocytes in vivo: characterization of a CD3⫺4⫹8⫺
intermediate. J. Exp. Med. 178:265.
22. Weissman, I. L. 1994. Developmental switches in the immune system. Cell 76:
207.
23. Burns, R. P., Jr., K. Natarajan, N. J. LoCascio, D. P. O’Brien, J. A. Kobori,
N. Shastri, and R. K. Barth. 1998. Molecular analysis of skewed Tcra-V gene use
in T-cell receptor ␤-chain transgenic mice. Immunogenetics 47:107.
24. Takeshita, S., M. Toda, and H. Yamagishi. 1989. Excision products of the T cell
receptor gene support a progressive rearrangement model of the ␣/␦ locus. EMBO
J. 8:3261.
25. McCormack, W. T., M. Liu, C. Postema, C. B. Thompson, and L. A. Turka. 1993.
Excision products of TCR V␣ recombination contain in-frame rearrangements:
evidence for continued V(D)J recombination in TCR⫹ thymocytes. Int. Immunol.
5:801.
26. Boysen, C., L. Inyoul, T. M. Smith, A. Smit, K. Wang, L. Rowen, and L. Hood.
1997. T-cell receptor ␣␦ locus complete nucleotide sequence. GenBank accession
no. AE000521.
27. Koop, B. F., L. Rowen, K. Wang, C. L. Kuo, D. Seto, J. A. Lenstra, S. Howard,
W. Shan, P. Deshpande, and L. Hood. 1994. The human T-cell receptor TCRAC/
TCRDC (C␣/C␦) region: organization, sequence, and evolution of 97.6 kb of
DNA. Genomics 19:478.
28. Punt, J. A., H. Suzuki, L. G. Granger, S. O. Sharrow, and A. Singer. 1996.
Lineage commitment in the thymus: only the most differentiated (TCRhibcl-2hi)
subset of CD4⫹CD8⫹ thymocytes has selectively terminated CD4 or CD8 synthesis. J. Exp. Med. 184:2091.
29. Akashi, K., M. Kondo, and I. L. Weissman. 1998. Two distinct pathways of
positive selection for thymocytes. Proc. Natl. Acad. Sci. USA 95:2486.
30. Swat, W., H. von Boehmer, and P. Kisielow. 1994. Small CD4⫹8⫹TCRlow thymocytes contain precursors of mature T cells. Eur. J. Immunol. 24:1010.
31. Ramiro, A. R., C. Trigueros, C. Marquez, J. L. San Millan, and M. L. Toribio.
1996. Regulation of pre-T cell receptor (pT␣-TCR␤) gene expression during
human thymic development. J. Exp. Med. 184:519.
32. Rock, E. P., P. R. Sibbald, M. M. Davis, and Y. H. Chien. 1994. CDR3 length in
antigen-specific immune receptors. J. Exp. Med. 179:323.
33. Ding, Y. H., K. J. Smith, D. N. Garboczi, U. Utz, W. E. Biddison, and
D. C. Wiley. 1998. Two human T cell receptors bind in a similar diagonal mode
to the HLA-A2/Tax peptide complex using different TCR amino acids. Immunity
8:403.
34. Garboczi, D. N., U. Utz, P. Ghosh, A. Seth, J. Kim, E. A. VanTienhoven,
W. E. Biddison, and D. C. Wiley. 1996. Assembly, specific binding, and crystallization of a human TCR-␣/␤ with an antigenic Tax peptide from human T
lymphotropic virus type 1 and the class I MHC molecule HLA-A2. J. Immunol.
157:5403.
35. Garcia, K. C., M. Degano, R. L. Stanfield, A. Brunmark, M. R. Jackson,
P. A. Peterson, L. Teyton, and I. A. Wilson. 1996. An ␣␤ T cell receptor structure
at 2.5 A and its orientation in the TCR-MHC complex. Science 274:209.