| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Repertoire1
Department of Internal Medicine and Harold C. Simmons Arthritis Research Center, University of Texas Southwestern Medical Center, Dallas, TX 75235
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
light chain repertoire in humans, a single-cell
PCR technique using genomic DNA obtained from individual peripheral B
cells was employed. Of the 30 known functional V genes, 23 were
detected in either the nonproductive or productive repertoires.
Specific V genes, including 2A2, 2B2, 1G, and 4B, were overexpressed
in the nonproductive repertoire, whereas some V genes, such as 3R,
2A2, 2B2, 1C, 1G, and 1B, were overexpressed in the productive
repertoire. Comparison of the nonproductive and productive repertoires
indicated that no V genes were positively selected, whereas a number
of V genes, including 4C, 1G, 5B, and 4B, were negatively regulated.
All four of the functional J segments were found in both
repertoires, with J7 observed most often. Evidence of terminal
deoxynucleotidyltransferase activity was noted in nearly 80% of
nonproductive VJ rearrangements, and exonuclease activity was
apparent in the majority. Despite this, the mean CDR3 length was 30
base pairs in both productive and nonproductive repertoires, suggesting
that it was tightly regulated at the molecular level. These results
have provided new insights into the dimensions of the human V
repertoire and the influences that shape it. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
light chains, whereas about 40% express light chains
4, 5, 6, 7 . A great deal of information is available about the human
-chain repertoire, but much less is known about the distribution of
expressed -chains and the mechanisms involved in the generation of
the human repertoire. Part of the reason for the limited
information about the human repertoire relates to the relatively
delayed development of knowledge regarding the genetic organization of
the locus 8, 9, 10, 11, 12, 13, 14 . Delineation of the dimensions of the human
repertoire is important not only because this light chain is used by
40% of all human B cells, but also because a large number of
autoantibodies appear to employ -chains 15, 16 .
The human V locus is located on chromosome 22q11.2 and is arranged
so that J/C pairs are downstream of the V genes, which are
organized into three clusters 8 . There are a total of 51 V genes
in this locus, 30 of which are thought to be functional 9 . The V
genes themselves are divided into 10 families according to sequence
homology, and the 10 families are divided into three clusters (A, B,
and C). Cluster A, which is J-proximal, contains the V2 and V3
families and the 4B gene from the V4 family; cluster B contains the
V1, V5, V7, and V9 families; and cluster C, which is
J-distal, contains the V6, V8, and V10 families, as well as
genes 4A and 4C from the V4 family. All V genes are organized in
the same transcription orientation, and, therefore, rearrange by
deletion.
There are seven J segments, of which four (J1, J2, J3, and
J7) are considered functional 11, 12 . J4, J5, and J6 are
not considered functional because the GT dinucleotide sequence in the
5' splice donor site of these three J segments has been deleted 11 .
The J segments are arranged so that one of the seven C segments
is located between each J segment 11, 12, 13, 14 . This is in contrast to
the heavy and gene loci, in which all J segments are grouped
together followed by the C segment(s) 2 .
The goal of this study was to assess the usage of V genes by normal
human peripheral B cells. A single-cell PCR technique was employed,
and, because the template was genomic DNA, the methodology permitted
detection of both productive and nonproductive rearrangements. Analysis
of nonproductive (nonexpressed) rearrangements permits an estimate of
the immediate products of the VJ combinatorial machinery without
the overarching influence of selection. The distribution of V genes
in the productive rearrangements is influenced not only by VJ
recombination, but also by subsequent positive and negative selection
and potential receptor editing at various stages of B cell development.
The data summarized here include the analysis of 55 nonproductive and
172 productive VJ rearrangements randomly found in normal human
peripheral B cells. Analysis of the sequences of these genes provides
the first comprehensive view of the dimensions of the normal human V
repertoire and the influences that generate it.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Peripheral blood was obtained from two healthy donors (donor 1 was a 26-year-old Hispanic man, and donor 2 was a 42-year-old Caucasian man). Mononuclear cells were isolated by Ficoll-Hypaque sedimentation as described 17 . B cells were further enriched using a CD19-positive selection column. The cells were then stained with a biotinylated anti-CD5 mAb (followed by secondary staining with RED613-labeled streptavidin), a phycoerythrin-labeled anti-CD19 mAb, and a FITC-labeled anti-human IgM mAb. The cells were sorted using a FACStarPlus flow cytometer Becton Dickinson, San Jose, CA outfitted with an automatic cell deposition unit, and one cell was deposited into each well of a 96-well PCR plate assembled on a microAmp base. The two populations obtained with this technique were CD19+/IgM+/CD5+ or CD5- B cells. Each well contained 5 µl of an alkaline lysing solution (200 mM KOH/50 mM DTT).
Primer extension preamplification
To obtain sufficient amounts of DNA for multiple subsequent PCR reactions, a preamplification step was employed using random 15-oligomers and 60 rounds of amplification with TAQ polymerase.
Amplification of rearranged VJ sequences
The sequences of all primers used to amplify VJ
rearrangements are shown in Table I
. All
of the V genes (including pseudogenes) in the V base 9 should be
amplified with these primers if rearranged, according to PCR
simulations using the Amplify 1.2 software program (provided by B.
Engels, Genetics Department, University of Wisconsin at Madison). Only
those J segments that have been documented to rearrange (J1,
J2, J3, and J7) would be amplified by these primers. It should
be noted that because of the J primer design, J2 and J3 cannot
be distinguished from each other. For the initial (external) PCR
amplification, 25 µl of a lower reaction mix containing 100 µM
concentration of each deoxynucleoside triphosphate, 0.5 µM
concentration of each primer listed as an external primer in Table I,
and 2 mM MgCl2 were added to each well of a 96-well PCR
plate (Robbins Scientific, Sunnyvale, CA). The lower reaction mix was
isolated by sealing it with a wax pellet. A total of 50 µl of an
upper reaction mix containing 10x PCR buffer and 2 units of TAQ DNA
polymerase (Promega, Madison, WI) were added to each well, followed by
5 µl of template generated from the preamplification step. The
amplification protocol was as follows: step 1, 95°C for 10 min; step
2, 50°C for 30 s; step 3, 72°C for 1 min 30 s; step 4,
94°C for 1 min; step 5, 35 cycles from step 2 through step 4; step 6,
72°C for 5 min. The second (nested) PCR was performed in the same
manner, with the exception that only one V family-specific internal
primer (listed as nested primers in Table I) was used in the lower
reaction mix in any one amplification, and the template was 5 µl of
the initial (external) PCR reaction. The annealing temperature in step
2 of the second (nested) PCR program was 62°C. All V primers were
manufactured by Integrated DNA Technologies (Coralville, IA).
|
All PCR products were separated by electrophoresis on a 1.5%
Seakem Agarose gel (FMC BioProducts, Rockland, ME). Positive bands (as
visualized with ethidium bromide) were cut from the agarose gel and
purified using GenElute agarose spin columns (Supelco, Bellefonte, PA).
Purified products were sequenced directly using the dideoxy termination
method 18 with the ABI PRISM Dye Terminator Cycle Sequencing Ready
Reaction Kit (Perkin-Elmer, Foster City, CA). Sequences were primed
with the internal V primers listed in Table I
and analyzed on an
Applied BioSystems 377 Automated DNA Sequencer (Perkin-Elmer). The V
sequences were identified using the Sequencher software (Gene Codes
Corp, Ann Arbor, MI) and the V BASE Sequence Directory (provided by I.
Tomlinson, Medical Research Council Centre for Protein Engineering,
Cambridge U.K.). The V nomenclature (as indicated in the V BASE
Sequence Directory) was employed. J segments were identified
manually. A rearrangement was considered productive if it contained no
stop codons and the VJ junction maintained the reading frame into
the J segment. All 55 nonproductive rearrangements were out of
frame; none contained stop codons. Rearrangements that involved
pseudogenes were always considered nonproductive. All sequences are
available from the EMBL Data Bank (accession numbers
AJ230234–AJ230460).
Accuracy of the sequencing technique
To validate the reliability of the sequence data, a known
V1 rearrangement (L18F) was subjected to the preamplification
procedure and subsequent nested amplifications multiple times. A single
PCR product (238 base pairs
(bp)3 long) was obtained in
each of 96 amplifications and subjected to sequencing as described
above. Clear sequences were found in 81 of the 96 PCR products. A total
of 19,261 bp were analyzed, and two errors were detected in the
resulting copies (error rate: 1.0 x 10-4). This
outcome is similar to a previous report estimating PCR-induced errors
to be seven errors in 42,000 bp (error rate: 1.7 x
10-4) 19 and indicates that the technique used to obtain
and sequence V rearrangements introduces few if any errors.
ß-Actin analysis
To determine the number of wells that received a cell during the
sorting procedure, PCR amplifications using ß-actin primers was
performed. Both external and internal primers were designed. The
external primers were 5'-ATCTGGCACCACACCTTCTACAATGAGCTGCG-3' and
5'-CGTCATACTCCTGCTTGCTGATCCACATCTGC-3'. The internal primers were
5'-GTTTGAGACCCTTCAACACCC and 5'-CCAGGAAGGAAGGCTGGAAG-3'. All
ß-actin primers were manufactured by Integrated DNA Technologies. A
PCR product of 
400 bp from exon 3 of the ß-actin gene 20 is
generated using the same protocol as used to amplify the V
rearrangements. The percentage of wells containing a ß-actin product
was then divided by the total number of wells sampled for
amplification. A ß-actin product was detected in 84% of wells.
Statistical methods
2 tests were used to compare the distribution of
V and J elements found in the nonproductive and productive
repertoires. Values of p 
0.05 were considered
significant. The goodness-of-fit 2 test was used to
assess differences between the observed frequencies and the expected
frequencies of V genes as would be expected by random usage based on
the number of V genes known to be in the genome.
Analysis of V rearrangements
The V data included in Table II
were previously reported in Ref. 21.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
J rearrangements from individual B cells
Of the 736 individual
CD19+/IgM+/CD5+ or
CD5- B cells that were sorted, 185 contained at least one
rearranged light chain gene. A total of 26.6% of the wells from
donor 1 and 24.6% of the wells from donor 2 contained rearranged
light chain genes. Table II summarizes the distribution of and
rearrangements in the B cells analyzed. Of the total number of wells in
which at least one productive VL rearrangement was detected
(n = 432), 293 (67.8%) contained one productive
rearrangement, whereas 139 (32.2%) contained at least one productively
rearranged -chain gene. This ratio is reflective of the distribution
of and light chains in human B cells.
All 10 V families are detected in both the nonproductive and
productive repertoires
Table III compares the
distribution of the 10 V gene families detected in individual B
cells to the expected distribution based on the presence of V family
members in the genome. The V2 and V1 families were found at the
highest frequencies in the nonproductive repertoire (30.9% and 25.4%,
respectively), followed by the V4 family (18.2%). The V3 family
and the less frequently found families (V5–10) accounted for 25.3%
of the nonproductive repertoire. The V1 and V4 families were
found significantly more often than expected in the nonproductive
repertoire, whereas the V3 family was found significantly less often
than expected. The frequencies of the remaining families (V5–10) in
the nonproductive repertoire were not significantly different from
expected.
|
2 and V1 families were also found most frequently in the
productive repertoire (33.1% and 29.1%, respectively), followed by
the V3 family (15.7%). The V4 family and the less frequent
families (V5–10) accounted for 22.2% of the productive repertoire.
The V1 and V2 families were found significantly more often than
expected in the productive repertoire, whereas the V3, V5, and
V9 families were detected less often than expected. The frequencies
of the remaining families (V4, V6, V7, V8, V9, and
V10) in the productive repertoire were not significantly different
from expected.
When the nonproductive and productive repertoires were compared, the
V4 family was found to be significantly underrepresented in the
productive repertoire (p < 0.001), and the
V3 family was found more often in the productive repertoire
(p < 0.018).
Overall V gene cluster representation is comparable to that
expected from random usage
As shown in Table IV, the
frequencies of clusters A and B V genes in the nonproductive
repertoire were comparable to the expected distribution, whereas
cluster C V genes was overrepresented in the nonproductive
repertoire. The distribution of all clusters in the productive
repertoire was comparable to the expected distribution.
|
genes are overrepresented in the normal repertoire
As shown in Fig. 1, not all of the
V genes were detected in this analysis. Of the 30 putative
functional V genes, 15 were detected in the nonproductive
repertoire. A total of five pseudogenes was also detected in the
nonproductively rearranged repertoire (2A1, 3A2, 3I, 7C, and 5A). Of
the functional V genes that were detected, 2A2, 2B2, 1G, and 4B were
found more often than expected in the nonproductive repertoire. All of
the other genes were found at frequencies that did not deviate
significantly from the expected frequency.
|
genes expressed, 3R, 2A2, 3H,
2B2, 1C, 1G, and 1B were found more often than expected in the
productive repertoire, whereas 3M and 9A were found less often than
expected. Moreover, a number of V genes were not found at all in the
productive repertoire, including 4C, 3P, 3A, 2D, 3L, 3E, 1A, 5B, and
4A.
Comparison of the nonproductive and productive distributions indicated
that 4C (p = 0.001), 1G
(p = 0.012), 5B (p =
0.048), and 4B (p = 0.026) were negatively
selected in that they were found significantly less often in the
productive compared with the nonproductive repertoire. No evidence of
positive selection of V genes was detected, although 3H was
modestly, but not significantly (p = 0.060),
overrepresented in the productive repertoire.
All functional J segments are used in the VJ
rearrangements
Table V depicts the distribution of
J gene elements in the nonproductive and productive repertoires.
J1 was the least used J gene segment in the nonproductive and
productive repertoire (5.5% vs 7.0%). With the PCR conditions
employed, J2 and J3 cannot be distinguished, but were used less
frequently than expected in both the nonproductive and productive
repertoires as well (34.5% vs 39%). J7 was used more frequently
than expected in the nonproductive repertoire (60%) and the productive
repertoire (54.1%). Hence, the order of J usage in both the
nonproductive and productive repertoires was the same (J1 used least
often, followed by J2/3, and J7 used most often). Moreover, the
frequency of each J segment was similar in the nonproductive and
productive repertoires.
|
and J segments
It was possible that the bias in J segment utilization in both
the nonproductive and productive repertoires was related to
preferential rearrangement of certain J segments with certain V
genes. To address this, all V rearrangements were analyzed for the
preferential usage of J segments. Table VI shows the subset of V genes
exhibiting biased association with specific J segments. It was
evident from Table V that J1 was used the least in the nonproductive
repertoire and, as indicated in Table VI, associated with only two V
genes (2A1 and 1G). The J2/3 and J7 segments were rearranged with
the majority of the V genes used in the nonproductive repertoire,
whereas the 2B2 and 1G genes rearranged most often with the J7
segment. Although several other genes not shown in Table VI appeared to
rearrange exclusively with either the J2/3 or J7 segments in the
nonproductive repertoire, the numbers of rearrangements were quite
small and thus the relevance of their distribution is uncertain.
|
1 usage in the productive repertoire was not limited to
rearrangements with specific genes, but was nonetheless used less often
than the other J segments. As in the nonproductive repertoire, the
J2/3 and J7 segments rearranged with the majority of the V
genes used in the productive repertoire. 3R and 2C appeared most often
in the productive repertoire rearranged to the J2/3 segments,
whereas 2B2, 7A, and 4B rearrangements most frequently employed the
J7 segment. Differences in the CDR3 lengths of the nonproductive and productive repertoire
As indicated in Fig. 2, 43% of the
175 productive VJ rearrangements had a CDR3 length of 33 bp and
28.5% of them had a CDR3 length of 30 bp. The mean (±SEM) length of
the CDR3 in productively rearranged VJ genes was 30.9 ± 0.2
bp, with a range from 27 to 45 bp. The CDR3 lengths of the
nonproductive rearrangements were somewhat more broadly distributed
with a range from 23 to 53 bp. Despite this, the mean (±SEM) length of
the CDR3 of the nonproductively rearranged VJ genes was 31.9
± 0.5 bp, not significantly different from that of the productive
rearrangements.
|
J rearrangements
Further analysis sought to determine the imprint of TdT activity
in both the nonproductive and productive repertoires. As shown in Table VII, only 21.4% of the nonproductive
rearrangements exhibited no evidence of TdT activity, whereas 78.6% of
the nonproductive rearrangements had one to nine
n-insertions per rearrangement. A total of 71% of the
productive rearrangements had 1–11 n-insertions per
rearrangement. The mean (±SEM) number of n-insertions (for
those rearrangements that had them) was 3.0 ± 0.3 nucleotides for
nonproductive rearrangements and 2.8 ± 0.1 nucleotides for
productive rearrangements. This indicates that TdT was active during
the VJ rearrangement process.
|
J
rearrangements
Evidence of V (5') exonuclease activity was detected in 44.7%
and 42.3% of the nonproductive and productive rearrangements,
respectively. A range of one to nine nucleotides was removed (Table VIII), although the effect of
5'-exonuclease activity more frequently was to remove one to three
nucleotides. J (3') exonuclease activity was detected in 75% and
70.3% of the nonproductive and productive rearrangements,
respectively. A range of one to nine nucleotides was removed as well
(Table IX), although the more frequent
impact of 3'-exonuclease activity was the removal of three nucleotides.
|
Very few differences were found in the CD5+ and
CD5- populations of the two donors. The V3 and V6
family distributions in the CD5+ and CD5-
populations in the productive repertoire of donor 1 were slightly
different. The CD5- population had more V3 and V6
rearrangements than the CD5+ population
(p = 0.03). However, the total numbers for
these two groups were quite small (n = 5 vs
n = 1 for V3, and n = 3 vs
n = 0 for V6). In addition, the V1 family was
more prevalent in the CD5+ B cells compared with the
CD5-B cells in the nonproductive repertoire of donor 2
(n = 10 vs n = 2, p =
0.02). J segment utilization was not different in either donor in
either B cell subset.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
utilization by individual human peripheral B
cells has documented a number of features of the expressed human V
repertoire and has begun to identify influences that shape its
development. Specifically, there is a distinct bias for individual V
genes to be rearranged as judged by their distribution in the
nonproductive repertoire, but no evidence of positive selection of V
rearrangements was apparent. In addition, the CDR3 length of the
rearranged genes is 1–2 amino acids longer than that of rearranged
genes and is determined by both molecular mechanisms operating on
the VJ joins as well as selection that eliminates excessively
long CDR3s. Moreover, TdT activity occurs frequently in VJ
rearrangements, implying that they occur contemporaneously with
VJ rearrangements.
A number of features of this analysis indicated that a nonbiased sample
has been analyzed and that the distribution of sequences detected
reflected the actual distribution of V-expressing B cells in vivo.
An attempt was made to deposit one cell in each of 736 wells, although
four of these wells contained more than one cell. According to the
recovery of ß-actin sequences, at least one cell was deposited
84% of the time, and so the maximum number of wells in which a
productive VL chain could be detected was 618. According to
previous reports 5, 6 and as demonstrated here, the ratio of B cells
expressing and is 2:1 in humans. Thus, if the efficiency of
amplification was 100%, 419 wells should have contained a productive
-chain and 199 wells should have contained a productive -chain.
However, the data indicate that the amplification efficiencies were
somewhat less, but comparable for the - and -chains, with the
efficiency of detecting productive -chains being 69.9% (293/419)
and that of detecting productive -chains being 69.8% (139/199).
The overall distribution of productive V families is comparable to
that reported previously 10 , although there are no previous data on
the use of V genes in the nonproductive repertoire. As the
nonproductive repertoire directly reflects the action of the VJ
recombination machinery without the superimposed influence of positive
and negative selection, the availability of this information provides
important insight regarding the molecular influences that shape the
final expressed repertoire.
Analysis of the VH and V repertoires of this same set of
B cells indicated that families with the most functional genes
(VH3 and V1) constituted the majority of rearrangements
in the nonproductive repertoire 17, 21 . This was not the case with
the V repertoire. The V3 family contains the most functional
genes in the lambda locus, but was the fourth most frequently employed
V family in the nonproductive repertoire, indicating that there is
bias in the use of V gene families for VJ recombination.
Overrepresentation of V families largely resulted from the
preferential usage of a limited number of individual genes within the
family. Overrepresentation of the V1 and V4 families in the
nonproductive repertoire reflected a significantly higher frequency of
the individual genes, 1G (16.4%) and 4B (12.7%). In contrast, it is
likely that underrepresentation of the V3 family was related to the
lack of rearrangements of this entire family based on a feature that is
common to the V3 family and infrequently found in other V genes.
These results indicate that there is preferential use of specific V
genes of the V1 and V4 families and underutilization of the V3
family in the recombination process of the V locus. It should be
emphasized that the preferential usage of specific V gene segments
in nonproductive rearrangements is similar to that noted for V genes
21 , thus implying similarities in the molecular mechanisms governing
recombination of light chain genes.
Factors that can influence the use of specific V genes in the
recombination process include accessibility of the DNA to the
recombination machinery, diversity in recombination signal sequences
(RSS) with varying efficiencies, promoter and enhancer efficiencies,
and proximity of V genes to the J genes. An analysis of the individual
genes overrepresented in the nonproductive repertoire, such as 1G and
4B, and those not found at all in the nonproductive repertoire may
reveal that one or more of these factors underlies the rearrangement
bias. However, a complete understanding of the regulatory influences of
the promoter(s) and enhancer(s) at the locus is only beginning to
emerge 22, 23 , and thus their role in recombination bias cannot be
estimated. However, efficiency of RSS does not appear to contribute
significantly to the biased use of particular V genes as the RSS are
largely conserved among the V genes 14 .
One of the factors that might influence the distribution of V genes
in the nonproductive repertoire is that genes more proximal to the
J-segments might have a greater propensity to rearrange. This has been
suggested by some studies of VH and V rearrangements
24, 25 , but not others 17, 21 . In the current analysis, it is clear
that distance from the J segments per se does not determine V
utilization, as the most frequently used genes are dispersed throughout
the V locus and are not all within cluster A. In fact, the 4B gene
is the most distal V gene and is the second most frequently
rearranged gene in the nonproductive repertoire. This is similar to the
repertoire, in which overrepresented V genes in the
nonproductive repertoire were also dispersed throughout the J
proximal V cassette, which is similar in length to the locus and
also comparable in that all V genes in this cassette rearrange by
deletion. Hence, there is no preference for rearranging V genes that
are proximal to the J segments, nor is the position of V genes an
apparent hindrance to rearrangement at the locus as evidenced by
the observation that the majority of the more distal V genes in the
nonproductive repertoire were found largely at their expected
frequencies.
The V1 and V2 families were found more often than expected in the
productive repertoire, and the V3 family was found less often than
expected. The overrepresentation of the V1 and V2 families in the
productive repertoire can be attributed to the significantly higher
frequency of a few genes (1C, 1G, 1B, 2A2, and 2B2).
Underrepresentation of the V3 family in the productive repertoire
appears to result from the decreased frequency of rearrangements of all
V3 family genes, as documented in the analysis of the nonproductive
repertoire. Of importance, three of the five genes overrepresented in
the productive repertoire (2A2, 2B2, and 1G) were also overrepresented
in the nonproductive repertoire, indicating that these genes were
subjected to recombinational bias. As there was no significant
differences in the frequency of these genes in the productive and
nonproductive repertoires, it is therefore likely that their
overrepresentation in the productive repertoire was not the result of
selective influences dependent on expression of a protein product.
The prevalence of the V1 and V2 families in the productive
repertoire found here is in general agreement with a previous report
describing the V distribution in a sampling of cDNA from B cells
10 . The overrepresentation of the V1 and V2 families in the
previously reported data base was also explained by the
overrepresentation of a few genes within those families, as was found
here. Moreover, the 2A2 gene was found most often in the previously
reported data base (27% of all rearrangements) as well as in the
productive repertoire reported here (17.4%). However, divergence
between the two reports is evident beyond these observations. For
example, three genes constitute nearly 60% of the previously reported
data base (1E, 2A2, and 2C), whereas 60% of the productive repertoire
reported here is represented by six genes (2A2, 3H, 2B2, 1C, 1G, and
1B). The only common frequent gene in both data bases is 2A2.
It is possible that the differences between the data bases are related
to the sources of the sequences. The previous data base was generated
from cDNA, and thus the repertoire could have been biased toward
activated cells, which have greater amounts of mRNA than resting cells
26 . In addition, the cDNA analyzed in the previous report was cloned
into a phage library, which could have introduced further biases. It is
also possible that the differences between the two data bases can be
attributed to different donor sources. In this regard, the V family
distributions of the four donors used to generate the previous data
base were similar, as was the V family distribution of the two
donors described here. Hence, it is most likely that the divergence
between the two data bases is attributable to the source of the
sequences, which was cDNA in the previous report and genomic DNA in
this report.
Comparison of the productive repertoire to the nonproductive repertoire
allows for a determination of positive and negative selection. The
V4 family was not observed as frequently in the productive
repertoire as it was in the nonproductive repertoire, and was thus
likely to be negatively selected. One possible explanation is that
V4 family genes may have a greater propensity to generate
autoantibodies and thus be deleted from the productive repertoire by
either negative selection or receptor editing. The V3 family is
underrepresented in both the nonproductive and productive repertoires,
and yet appears to be positively selected. This appears to be related
to a combined effect of 3H and 3R, neither of which alone is
significantly more frequent in the productive compared with the
nonproductive repertoires, but together they are significantly
overrepresented in the productive repertoire (p
= 0.012) and sufficiently to cause the appearance of significant
selection of the entire V3 family. Explanation for the combined
overrepresentation of 3R and 3H is not clear, but could relate to
Ag-mediated selection.
Although several individual genes are overrepresented in the productive
repertoire, positive selection of individual V genes was not
apparent in this V distribution. This differs from both the
VH and V repertoires, in which individual V genes are
positively selected 17, 21 or overrepresented in the productive
repertoire and thus could be positively selected 27, 28, 29, 30 . This
suggests that the composition of the CDR3s in VJ rearrangements,
as opposed to other aspects of the structure, may be of greater
importance for positive selection of V genes than in the
VH and V repertoires. In contrast, negative selection of
individual V genes was not observed 21 , whereas several V genes
were negatively selected (4C, 1G, 5B, and 4B). This may indicate that
these particular V genes have a propensity for autoreactivity and
subsequent deletion independent of CDR3 composition. Alternatively,
these V genes may not pair well with heavy chains and thus not
appear in the productive repertoire.
All four of the functional J segments were found in the
nonproductive and productive repertoires. Of note, J7 was used most
frequently in both repertoires (60.0% and 54.1%, respectively). The
similarity in J7 frequency in the nonproductive and productive
repertoires suggests that neither selection nor receptor editing is
playing a role in the overrepresentation of J7 segments. The
explanation for the overrepresentation of J7 remains unclear.
Extensive analysis of the C/J region has not indicated any
inherent characteristics of the functional J segments 31, 32, 33, 34 that
would account for the overrepresentation of J7 rearrangements in
this analysis. In addition, both the RSS and spacers that can influence
recombination 35 are highly conserved within the functional J
segments 14, 35 . Of note, J7 appears to be preferentially
rearranged with a number of the more frequently rearranged V genes,
such as 2B2, 1G, and 4B, although the explanation for this pairing is
uncertain.
A previous report 10 indicated that the J1, 2, and 3 segments were
found comparably in the productive repertoire (27%, 38%, and 34%,
respectively), whereas J7 was employed less often (0.6%). However,
the V distribution in this previously published data base was
heavily skewed to the rearrangement of five genes, 2A2, 2C, 1E, 1C, and
3H, none of which demonstrated biased rearrangement to J7 in the
current analysis. Moreover, 2B2, 1G, and 4B, which are frequently
detected V genes in the current report and recombine frequently with
J7, did not dominate the previously published database. Thus, it is
possible that biased recombination with certain V genes and
individual variation in expression or detection of these V genes
contributed to the overrepresentation of J7 in the current analysis.
The range of CDR3 lengths of the nonproductive and productive VJ
gene rearrangements was comparable and congregated around 30–33 bp,
with a mean length of 30 bp. This contrasts with VJ
rearrangements, which exhibit a mean CDR3 length of 27 bp 21 . It is
noteworthy that the CDR3 length of nonproductive VJ
rearrangements is maintained despite extensive modification of the
joints by TdT and exonuclease activities, implying that the final
assembly of the V CDR3 is tightly regulated at a molecular level.
However, CDR3 assembly of the V CDR3 is somewhat less strictly
regulated than that of VJ rearrangements in that a few
nonproductive VJ rearrangements were found with longer CDR3
regions. However, these were largely eliminated from the productive
repertoire. The net result of the molecular mechanisms and selection
was a mean V CDR3 that is 30 nucleotides in length with 75% of the
rearrangements having a CDR3 length one or two amino acids longer than
that of the V rearrangements. Differences in the lengths of the V
and V CDR3 regions imply that these light chains may contribute
differently to Ag binding of Ab molecules 36 . The apparent greater
importance of the CDR3 of V chains to positive selection may be a
reflection of the differential binding contribution of the V CDR3
region.
At least one molecular determinant of the final CDR3 length of V and
V rearrangements may relate to the potential contribution of the
genetic segments of the respective genes to the CDR3. Before
rearrangement, the potential V contribution to the CDR3 is 21
nucleotides, whereas the potential V contribution to the CDR3 is
24–27 nucleotides. Thus, the increase in length of the CDR3s of
VJ rearrangements could reflect the increased contribution of
V segments to the CDR3 region. Support for this possibility comes
from an analysis of members of the V4 family. 4C contributes 36
nucleotides to the CDR3, generating a mean CDR3 length of 44
nucleotides in nonproductive rearrangements and is negatively selected.
In contrast, 4B contributes 21 nucleotides to the CDR3, generating a
mean CDR3 length of 28 nucleotides. This gene is also negatively
selected. Of the rearrangements containing 4B in the productive
rearrangements, the mean CDR3 length is a V-like 27.6 nucleotides.
These results are consistent with the conclusion that the contribution
of light chain genetic elements to the CDR3 region is one determinant
of the final CDR3 length. Furthermore, marked deviations from a mean
CDR3 length of 30 nucleotides appears to contribute to deletion of the
rearrangement from the productive repertoire, perhaps because of
defective pairing with heavy chains or ineffective positive selection.
Another contribution to the final CDR3 length of VJ
rearrangements appears to be limited 5'-exonuclease activity. In this
regard, 57% of the VJ rearrangements in comparison to 18%
of the VJ rearrangements lack evidence of 5'-exonuclease
activity. The explanation for this is uncertain, but does not reflect a
global diminution of exonuclease activity as nearly 75% of all
nonproductive VJ rearrangements manifest 3'-exonuclease activity.
Regardless of the explanation, limited 5'-exonuclease activity could
also contribute to the somewhat longer CDR3 regions characteristic of
VJ rearrangements.
TdT activity is considered an early B cell event, occurring routinely
when the heavy chain locus is rearranging and decreasing with the onset
of light chain rearrangement in mice 2 . As a result,
n-insertions are rare in murine light chain rearrangements
37 . However, recent evidence in humans has shown that more than
two-thirds of human VJ rearrangements have
n-insertions, albeit less extensively than VH
rearrangements 17, 21 , implying that there is residual TdT activity
during light chain rearrangements in humans. Therefore, comparison of
TdT activity in VJ and VJ rearrangements might provide some
insights into the relative timing of light chain recombination in
humans. The data indicate that a comparable fraction of VJ (78%
of nonproductive and 71% of productive) and VJ (69% of
nonproductive and 58% of productive) rearrangements have
n-insertions. These results reflect previous results 38
and imply that VJ and VJ rearrangements might occur
contemporaneously in humans. This possibility is supported by recent
findings in the mouse, suggesting that the and loci are
targeted for rearrangement at the same time 39, 40 .
No compelling evidence of receptor editing of the repertoire was
noted when the distribution of the use of V and J segments was
analyzed. However, B cells with two productive V rearrangements were
noted (7.6% with one productive and one productive , 3.2% with
two productive s, and 0.9% with two productive s). It is
unlikely that this is an artifact of cell sorting in which some wells
received two cells because if that were the case, the most frequent
occurrence would be two productive VJ rearrangements. It is more
likely that this represented a form of receptor editing in which the
edited rearrangement remains in the cell but the final expressed Ig
receptor employs the light chain encoded by the replacement light chain
rearrangements 41 . If this interpretation is correct, this form of
receptor editing appears to be reasonably frequent in the normal
repertoire.
In summary, a variety of molecular and selective mechanisms appear to
play a role in generating the human V repertoire. The result is a
repertoire dominated by a few V genes and a stereotypical but unique
CDR3 region.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Peter E. Lipsky, Department of Internal Medicine, Harold C. Simmons Arthritis Research Center, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-8884. E-mail address:
3 Abbreviations used in this paper: bp, base pairs; TdT, terminal deoxynucleotidyltransferase; RSS, recombination signal sequence.
Received for publication August 14, 1998. Accepted for publication November 4, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ ratio of human serum immunoglobulins in the course of development. Blut 33:281. (in German). [Medline]
chains and genes in inbred mice. Annu. Rev. Immunol. 3:337.[Medline]
light-chain locus on chromosome 22q11.2. Hum. Mol. Genet. 4:983. repertoire. J. Mol. Biol. 264:220.[Medline]
repertoire. J. Mol. Biol. 268:69.[Medline]
genes. J. Exp. Med. 172:609.7 complex. J. Immunol. 157:4474.[Abstract]
gene locus. Gen. Res. 5:125. gene locus. Gen. Res. 7:250. gene repertoire of IgM+ B cells. J. Clin. Invest. 99:1614.[Medline]
L chain Ig gene complex. J. Immunol. 147:2354.[Abstract]
locus is duplicated and functionally independent of NFB. Genes Dev. 4:978. gene expression does not follow the VH paradigm. J. Exp. Med. 169:1859. segments reveals a strong bias in their usage. Eur. J. Immunol. 24:827.[Medline]
genes with no or little somatic mutation throughout life. Eur. J. Immunol. 23:3272.[Medline]
constant region genes in man. Nature 294:536.[Medline]
6 gene encodes the Ke+Oz- chain and C4 and C5 are pseudogenes. Proc. Natl. Acad. Sci. USA 84:9074. light chain locus: organization and DNA sequences of three genomic J regions. Immunogenetics 25:63.[Medline]
and J-C gene segments of the human immunoglobulin light chain locus are separated by 14 kb and rearrange by a deletion mechanism. Eur. J. Immunol. 21:1513.[Medline]
repertoire. J. Immunol. 158:3761.[Abstract]
and rearrangement occur simultaneously in transformed pre-B cells. J. Immunol. 159:6061.[Abstract]
and genes can occur in all B cells and is initiated around the same pre-B-cell developmental stage. Dev. Immunol. 4:13.[Medline]
This article has been cited by other articles:
|
|
 |
|
  N. S. Longo, P. L. Lugar, S. Yavuz, W. Zhang, P. H. L. Krijger, D. E. Russ, D. D. Jima, S. S. Dave, A. C. Grammer, and P. E. Lipsky Analysis of somatic hypermutation in X-linked hyper-IgM syndrome shows specific deficiencies in mutational targeting Blood, April 16, 2009; 113(16): 3706 - 3715. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Lloyd, D. Lowe, B. Edwards, F. Welsh, T. Dilks, C. Hardman, and T. Vaughan Modelling the human immune response: performance of a 1011 human antibody repertoire against a broad panel of therapeutically relevant antigens Protein Eng. Des. Sel., March 1, 2009; 22(3): 159 - 168. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. S. Longo, G. J. Grundy, J. Lee, M. Gellert, and P. E. Lipsky An Activation-Induced Cytidine Deaminase-Independent Mechanism of Secondary VH Gene Rearrangement in Preimmune Human B Cells J. Immunol., December 1, 2008; 181(11): 7825 - 7834. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Capello, M. Martini, A. Gloghini, M. Cerri, S. Rasi, C. Deambrogi, D. Rossi, M. Spina, U. Tirelli, L. M. Larocca, et al. Molecular analysis of immunoglobulin variable genes in human immunodeficiency virus-related non-Hodgkin's lymphoma reveals implications for disease pathogenesis and histogenesis Haematologica, August 1, 2008; 93(8): 1178 - 1185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Su, J. N. Gordon, F. Barone, L. Boursier, W. Turnbull, S. Mendis, D. K. Dunn-Walters, and J. Spencer Lambda Light Chain Revision in the Human Intestinal IgA Response J. Immunol., July 15, 2008; 181(2): 1264 - 1271. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. S. Longo, C. L. Satorius, A. Plebani, A. Durandy, and P. E. Lipsky Characterization of Ig Gene Somatic Hypermutation in the Absence of Activation-Induced Cytidine Deaminase J. Immunol., July 15, 2008; 181(2): 1299 - 1306. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Forconi, E. Sozzi, D. Rossi, S. S. Sahota, T. Amato, D. Raspadori, L. Trentin, L. Leoncini, G. Gaidano, and F. Lauria Selective influences in the expressed immunoglobulin heavy and light chain gene repertoire in hairy cell leukemia Haematologica, May 1, 2008; 93(5): 697 - 705. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. F. Widhopf II, C. J. Goldberg, T. L. Toy, L. Z. Rassenti, W. G. Wierda, J. C. Byrd, M. J. Keating, J. G. Gribben, K. R. Rai, and T. J. Kipps Nonstochastic pairing of immunoglobulin heavy and light chains expressed by chronic lymphocytic leukemia B cells is predicated on the heavy chain CDR3 Blood, March 15, 2008; 111(6): 3137 - 3144. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Stamatopoulos, C. Belessi, A. Hadzidimitriou, T. Smilevska, E. Kalagiakou, K. Hatzi, N. Stavroyianni, A. Athanasiadou, A. Tsompanakou, T. Papadaki, et al. Immunoglobulin light chain repertoire in chronic lymphocytic leukemia Blood, November 15, 2005; 106(10): 3575 - 3583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Smithson, K. Kolibab, A. K. Shriner, N. Srivastava, S. Khuder, and M. A. J. Westerink Immune Response to Pneumococcal Polysaccharides 4 and 14 in Elderly and Young Adults: Analysis of the Variable Light Chain Repertoire Infect. Immun., November 1, 2005; 73(11): 7477 - 7484. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Ghosh, A. C. Steere, B. D. Stollar, and B. T. Huber In Situ Diversification of the Antibody Repertoire in Chronic Lyme Arthritis Synovium J. Immunol., March 1, 2005; 174(5): 2860 - 2869. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. L. Monson, P. D. Cravens, E. M. Frohman, K. Hawker, and M. K. Racke Effect of Rituximab on the Peripheral Blood and Cerebrospinal Fluid B Cells in Patients With Primary Progressive Multiple Sclerosis Arch Neurol, February 1, 2005; 62(2): 258 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Rahman Autoantibodies, lupus and the science of sabotage Rheumatology, November 1, 2004; 43(11): 1326 - 1336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wardemann, J. Hammersen, and M. C. Nussenzweig Human Autoantibody Silencing by Immunoglobulin Light Chains J. Exp. Med., July 19, 2004; 200(2): 191 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Viau, N. S. Longo, P. E. Lipsky, L. Bjorck, and M. Zouali Specific In Vivo Deletion of B-Cell Subpopulations Expressing Human Immunoglobulins by the B-Cell Superantigen Protein L Infect. Immun., June 1, 2004; 72(6): 3515 - 3523. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Su, L. Boursier, A. Padala, J. D. Sanderson, and J. Spencer Biases in Ig {lambda} Light Chain Rearrangements in Human Intestinal Plasma Cells J. Immunol., February 15, 2004; 172(4): 2360 - 2366. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Capello, M. Cerri, G. Muti, E. Berra, P. Oreste, C. Deambrogi, D. Rossi, G. Dotti, A. Conconi, M. Vigano, et al. Molecular histogenesis of posttransplantation lymphoproliferative disorders Blood, November 15, 2003; 102(10): 3775 - 3785. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Abraham, S. M. Geyer, T. L. Price-Troska, C. Allmer, R. A. Kyle, M. A. Gertz, and R. Fonseca Immunoglobulin light chain variable (V) region genes influence clinical presentation and outcome in light chain-associated amyloidosis (AL) Blood, May 15, 2003; 101(10): 3801 - 3807. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-H. Wang, Z. Zhang, P. D. Burrows, H. Kubagawa, S. L. Bridges Jr, H. W. Findley, and M. D. Cooper V(D)J recombinatorial repertoire diversification during intraclonal pro-B to B-cell differentiation Blood, February 1, 2003; 101(3): 1030 - 1037. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Rahman, I Giles, J Haley, and D Isenberg Systematic analysis of sequences of anti-DNA antibodies--relevance to theories of origin and pathogenicity Lupus, December 1, 2002; 11(12): 807 - 823. [Abstract] [PDF]
|
|
|
|
|
|
S. Yavuz, A. S. Yavuz, K. H. Kraemer, and P. E. Lipsky The Role of Polymerase {eta} in Somatic Hypermutation Determined by Analysis of Mutations in a Patient with Xeroderma Pigmentosum Variant J. Immunol., October 1, 2002; 169(7): 3825 - 3830. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Golby, M. Hackett, L. Boursier, D. Dunn-Walters, S. Thiagamoorthy, and J. Spencer B cell development and proliferation of mature B cells in human fetal intestine J. Leukoc. Biol., August 1, 2002; 72(2): 279 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Perfetti, S. Casarini, G. Palladini, M. C. Vignarelli, C. Klersy, M. Diegoli, E. Ascari, and G. Merlini Analysis of Vlambda -Jlambda expression in plasma cells from primary (AL) amyloidosis and normal bone marrow identifies 3r (lambda III) as a new amyloid-associated germline gene segment Blood, July 18, 2002; 100(3): 948 - 953. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. van der Burg, B. H. Barendregt, E. J. van Gastel-Mol, T. Tumkaya, A. W. Langerak, and J. J. M. van Dongen Unraveling of the Polymorphic C{lambda}2-C{lambda}3 Amplification and the Ke+Oz- Polymorphism in the Human Ig{lambda} Locus J. Immunol., July 1, 2002; 169(1): 271 - 276. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-H. Wang, R. P. Stephan, A. Scheffold, D. Kunkel, H. Karasuyama, A. Radbruch, and M. D. Cooper Differential surrogate light chain expression governs B-cell differentiation Blood, April 1, 2002; 99(7): 2459 - 2467. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Goebel, N. Janney, J. R. Valenzuela, W. J. Romanow, C. Murre, and A. J. Feeney Localized Gene-Specific Induction of Accessibility to V(D)j Recombination Induced by E2a and Early B Cell Factor in Nonlymphoid Cells J. Exp. Med., September 3, 2001; 194(5): 645 - 656. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Meffre, N. Catalan, F. Seltz, A. Fischer, M. C. Nussenzweig, and A. Durandy Somatic Hypermutation Shapes the Antibody Repertoire of Memory B Cells in Humans J. Exp. Med., August 6, 2001; 194(3): 375 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Comenzo, Y. Zhang, C. Martinez, K. Osman, and G. A. Herrera The tropism of organ involvement in primary systemic amyloidosis: contributions of Ig VL germ line gene use and clonal plasma cell burden Blood, August 1, 2001; 98(3): 714 - 720. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. L. Monson, T. J. Fenske, S. Wei, A. J. Okragly, J. L.O. de Jong, M. Haak-Frendscho, J. O’Shea, J. Djeu, and P. M. Sondel A p74 common gamma receptor chain isoform facilitates IL-2 and IL-15 responses by the myelomonocytic cell line Tf-1{beta}2 J. Leukoc. Biol., March 1, 2001; 69(3): 419 - 425. [Abstract] [Full Text]
|
|
|
|
|
|
J. Lee, N. L. Monson, and P. E. Lipsky The V{lambda}J{lambda} Repertoire in Human Fetal Spleen: Evidence for Positive Selection and Extensive Receptor Editing J. Immunol., December 1, 2000; 165(11): 6322 - 6333. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Dorner, N. L. Farner, and P. E. Lipsky Ig {lambda} and Heavy Chain Gene Usage in Early Untreated Systemic Lupus Erythematosus Suggests Intensive B Cell Stimulation J. Immunol., July 15, 1999; 163(2): 1027 - 1036. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, significantly lower frequency than
predicted from its presence in the genome; 
, pseudogene; X,
significant differences between the frequency in the productive and
nonproductive repertoires.