| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Immunology, The Scripps Research Institute, La Jolla, CA 92037; and
Torrey Pines Institute for Molecular Studies, San Diego, CA 92121
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and 
, chain
molecules, respectively. However, this combinatorial diversity is
constrained by the fact that V, D, and J segments recombine at very
different frequencies. An example of unusually high frequency of
rearrangement is observed with the 81X gene, which is the
most 3' functional VH gene in the murine
VH locus. Many laboratories have observed that
this VH gene family rearranges very frequently
early in ontogeny and also in the adult bone marrow
(1, 2, 3, 4, 5, 6, 7, 8, 9). The overuse of the 7183 family,
especially 81X, gave rise to the chromosomal proximity
hypothesis of Alt and colleagues, which proposed that the rearrangement
mechanism preferentially rearranges genes that are in closer proximity
in the chromosome (1, 10). Supporting the generality of
this proximity observation, mice that lack the IL-7 receptor show
decreased rearrangement of VH genes, displaying a
gradient of rearrangement throughout the entire
VH locus, ranging from almost normal levels of
rearrangement of the VH7183 family to
essentially no rearrangement of the distal
VHJ558 and VHJ606
families (11). Also, overuse of V-proximal J genes has
been observed in the TCR loci in fetal and newborn thymi
(12), and only later do the more distal J genes rearrange.
Most strikingly, there is a strict chromosomally ordered gene
rearrangement in the V
loci (13). However,
other loci, such as the human VH loci, do not
display this phenomenon of chromosomally ordered rearrangements
(14). It has been shown that many of the
VH3 genes that are over-represented in
rearrangements map in the middle or 5' portion of the locus
(15). However, the most 3' human VH
gene, VH6, is over-represented in rearrangements
early in ontogeny (16). A direct tracking mechanism as an
explanation for this phenomenon has been deemed unlikely, but the
reason for these several examples of positionally graded rearrangement
frequencies is not clear.
An alternative, or additional, explanation is that each individual gene
controls the frequency of its rearrangement to a large extent by virtue
of its relatively unique recombination signal sequence
(RSS).4 Most natural
TCR and Ig RSS vary from the consensus RSS, and this variation could
result in differences in rearrangement frequency (17). In
these cases, the chromosomal location would be irrelevant, and thus
could explain the inconsistent correlation of high frequency of
rearrangement with proximal chromosomal location for only some, but not
other, V and J loci. We and others, particularly Wu and colleagues,
have, in fact, shown that the natural variation in the RSS of V genes
can affect rearrangement frequencies (18, 19, 20, 21, 22, 23). Although
the classic studies of Gellert and colleagues have shown that some
positions in the heptamer and nonamer are critical for rearrangement,
while others are less critical, those studies have not analyzed all
possible changes in the RSS (24). We have shown that an
allelic change involving a single substitution in the heptamer can lead
to a 4.5-fold reduction in rearrangement frequency in a V
gene critical for an effective Ab response to
Haemophilus influenzae type b (21, 25).
Furthermore, Navajos, who have the allele with the nonconsensus
heptamer, have a greatly increased susceptibility to H.
influenzae type b disease. Thus, variations in RSS leading to
relatively modest differences in rearrangement frequencies can alter
the composition of the Ab repertoire and can have severe biological
ramifications (18).
It has been proposed that the reason that 81X
rearranges so often is due to its RSS (23). Using miniloci
recombination substrates transiently transfected into pre-B cells, the
relative frequency of rearrangement of 81X vs
VHJ558 genes was determined, and 81X
rearranged 
15 times more often than VHJ558
(22, 23). However, the choice of a
VHJ558 gene as a competitor was likely to show a
bias in favor of VH7183, because
VHJ588 genes have a poor RSS. It stands to
reason that if 81X rearrangements comprise over
half of the VH7183 rearrangements, the other
18 functional members of the VH7183 family
cannot all rearrange at that same high rate. Hence one would predict
that 81X would have a unique RSS if its high frequency of
rearrangement were due to a particularly potent RSS.
Therefore, in this current study we have sequenced the RSS of 19 of 20 members of the VH7183 family, and they fall into two similar groups. All but two genes share the same nonconsensus nonamer. Strikingly, over half of the VH7183 genes have the same nonconsensus heptamer as 81X, although 81X has a unique spacer sequence. The other VH7183 genes have RSS with a consensus heptamer and a slightly different spacer. Twelve of the 19 RSS are identical with one of the two prototypic RSS, and the other seven have a small number of changes. We determined the relative potential of the various RSS to support recombination in a recombination substrate assay, and, not surprisingly, the RSS with the consensus heptamer rearranged more often than RSS with the nonconsensus heptamers. Even the RSS of 81X, with its unique spacer, supported less rearrangement than the VH7183 RSS with the consensus heptamer.
We determined the chromosomal order of the VH7183 genes from a bacterial artificial chromosome (BAC) contig covering the entire VH7183 region of the IgH locus, and we also determined the relative frequency of rearrangement of each of the VH7183 genes. All but one of the frequently rearranging genes were in the 3' half of the region, and the rearrangement frequency fell off in gradient fashion 5' of 81X, with the exception of one poorly rearranging gene in the 3' portion. Thus, these data clearly show that variation in RSS potency is not the major determinant in the nonrandom VH usage of the VH7183 gene family or in the overuse of 81X in rearrangements, and that the RSS variation is overridden by other factors. It appears that chromosomal location may play a more major role in the over-rearrangement of many of the 3' genes. Finally, despite identity in RSS and similarity in coding regions, similar genes rearrange at quite different frequencies in different parts of the locus, suggesting that the local region surrounding each VH gene is a major determinant of the frequency of rearrangement.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
µMT mice (26) were obtained from The Jackson Laboratory (Bar Harbor, ME) and were bred at The Scripps Research Institute’s animal facility.
PCR and sequencing
The following oligonucleotides were used, with engineered restriction enzyme sites underlined: 5' primers: AF191, 5'-CGGGCTCGAGTTGGTTTTCCTTGTCC-3' (leader); AF215, 5'-GAGGCTCGAGGAAACTCTCCTGTG-3' (starts at bp 43 of the coding region); AF303, 5'-GGGGCTCGAGGAGTCTGGGGGA-3' (starts at bp 6 of the coding region); AM3, 5'-GTCCGAATTCCTCTCCTGTGCAA-3' (starts at bp 48 of the 50.1 coding region); AM6, 5'-AATAGGAATTCTGGTAGCAGGTAT-3' (starts at bp 154 of the 68–5N coding region); RR190, 5'-ATGCTGGTGGAGTCTGGG-3' (starts at bp 7 of the coding region); Ox2 5', 5'-GCTGAAGCTTTCAGGCAATGGCCA-3'; 3' primers: AF192, 5'-CTCCGCGGCCGCTGCTGGTCCT-3' (starts 16 bp 3' of the RSS); AF13N, 5'-AAAAGCGGCCGCTTACCTGAGGAGACGGTCA-3' (JH); AF269, 5'-GGTCGCGGCCGCTCTGCAGGAGGTTT-3' (starts 5 bp from the end of the nonamer of the RSS); AM5, 5'-GCAGAAGCTTTAGTTTGAGCTCA-3' (starts in the middle of the spacer of the VH7183 family RSS and extends through the nonamer); RR192, 5'-GACCTCAGACTGCTCATTTGC-3' (starts at bp 243 of the coding region); Ox2 3', 5'-GCGCGAATTCCTGATCATCTTTCTAGTCA-3'; and E4.psi.seq, 5'-ACTGATGGTAGTTTCA-3' (starts at bp 157 of E4.psi coding region).
All oligonucleotides were supplied by Genosys (The Woodlands, TX) or Life Technologies (Gaithersburg, MD). For the repertoire analysis, newborn liver DNA was prepared using the method described by Schlissel et al. (27). AF303 and AF13N were used to amplify VDJ rearrangements from newborn liver DNA. Thirty-five cycles of 30 s at 94°C, 30 s at 57°C, and 1 min at 68°C, followed by one cycle of 10 min at 68°C, were performed using an Ericomp cycler (Ericomp, San Diego, CA). Elongase (Life Technologies) was used for all PCRs because it contains a proofreading polymerase. The resulting PCR fragments were digested with XhoI and NotI and cloned into pBluescript (Stratagene, La Jolla, CA). Individual colonies were miniprepped and sequenced with a T7 primer using Sequenase version 2.0 (U.S. Biochemical Corp., Cleveland, OH) sequencing kit and [35S]ATP. For the repertoire analysis, once repeat sequences (i.e., the same V, D, and J segments and the same junctional sequence) were identified coming from the same PCR, no further sequences from that PCR were obtained. Identification of the VH genes was routinely based on >150 bp of VH sequence information from the 3' half of the gene.
AF191 and AF192, AF191 and AF269, RR190 and RR192, or AF215 and AM5 were used for amplifying the genomic sequences from individual BAC DNA for the mapping analysis. Twenty-five cycles of 30 s at 94°C, 30 s at 55°C, and 1.5 min at 68°C, followed by one cycle of 10 min at 68°C, were performed using Elongase. The PCR fragments were digested with XhoI and NotI, or EcoRI and HindIII, depending upon the primers used. Clones were sequenced from both sides, and a minimum of two sequences from independent PCRs were determined for each VH gene to ensure the absence of errors.
Additionally, in the assembly of the BAC contig (see below) each VH7183 gene, except 7183.10, 7183.11, 81X, E4.psi, D6.96, and 3:3.39, has recently been sequenced directly from BAC DNA without PCR amplification, confirming the sequences obtained by PCR and cloning. Direct sequencing was performed on agarose gel eluates purified with the Qiaquick kit (Qiagen, Valencia, CA). Cycle sequencing used BigDye terminators (PE Biosystems, Foster City, CA) for one cycle of 95°C for 5 min, followed by 100 cycles of 96°C for 30 s, 50°C for 15 s, and 60°C for 4 min.
Recombination substrate assay
The competition recombination substrate assay has been described
previously (21). The RSS and 100–200 bp of flanking
regions were cloned into a competition recombination substrate plasmid
5' of the transcriptional stop, and an inverted
DSP2 segment was cloned into the 3' side of the
transcriptional stop using the following oligonucleotides (restriction
enzyme sites are underlined): P50a, 5'-GGGTACGCGTGCCTCCAGGGAAG-3';
P50b, 5'-CCTGGCGGCCGCGAGCACCCCTGCA-3'; P58A2,
5'-CCCTGCGGCCGCAAATGAGCAGTC-3'; P61b,
5'-CTCTCGTCGACCCCTGCTGGTCCT-3'; P76b,
5'-TCTGGTCGACCAGAACACCCCTGCA-3'; P82,
5'-TTCAGGCGGCCGCCCTGAGAGCTGAGG-3'; P97a,
5'-TGAAGCGGCCGCCTGGACAGGGACTTG-3'; and P97b,
5'-TTCTGTCGACTCAATCCCAGTGC-3'.
P50a and P50b were used to amplify the VHS107/V1 RSS for cloning into the external site; P82 and P76b were used for cloning the V1 gene into the internal RSS cloning site. P58A2 and P61b were used to clone 81X, 3.3.39 (representative 7183-I gene), and 37.1 (representative 7183-II gene) into the internal RSS cloning site. P97a and P97b were used to clone the VHJ558 RSS into the internal RSS cloning site. The resulting constructs were sequenced to confirm that no mutations had been introduced during the PCR.
The RSS competition assay was performed as previously described
(21). Briefly 18.8 Abelson-murine leukemia
virus-transformed pre-B cells were transfected using a Bio-Rad
GenePulser with 300 V and 960 µF capacitance (Bio-Rad, Hercules, CA).
Transfected cells were cultured for 48 h in the presence of
0.5–1.0 mM caffeine (ICN Pharmaceuticals, Costa Mesa, CA), then the
plasmids were recovered using a standard alkaline lysis protocol.
Recovered plasmid was digested with SpeI, transformed into
Electrocompetent XL1 Blue (Stratagene), and plated on Luria Bertoni
agar containing isopropyl -D-thiogalactoside and
chloramphenicol. The bacterial colonies were screened using a PCR-based
assay as previously described to determine the relative rearrangement
of the DSP RSS to the external vs internal
VH RSS (21).
Assembly of VH7183 BAC contig
The BAC library of the mouse strain 129/Sv (129S3) from the embryonic stem cell line CJ7 distributed by Research Genetics (Huntsville, AL) was screened using PCR assays for VHIII, D12Nds2 (VHS107/V1), VHE4.psi, and BAC ends CT7–19B1-T7 and CT7–057J13-Sp6 (C. Chevillard et al., manuscript in preparation). BAC ends were isolated by vector-hexamer PCR (28) and hybridized to Southern blots of HindIII-digested BAC DNAs to identify overlap patterns. High density BAC library membranes were probed with pooled VH7183 PCR products from several 7183+ BACs, and additional BACs were identified. All BACs were digested with EcoRI and blotted; replicate blots were probed with amplified inserts of 81X and VHQ52 plasmids. The gene content of each hybridizing fragment was identified from at least two representative BACs by excision of the fragment from agarose gel and sequencing its VH7183 gene. At least one sequence in each case was obtained directly from eluted restriction fragment without PCR, with the exception of six genes, as noted above; confirmatory sequences were in some cases obtained from PCR amplification from eluted restriction fragments. Taken together, these data allowed the unambiguous ordering of BACS and VH7183 genes.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The RSS of 81X is nonconsensus, but has been shown to be much more efficient at rearrangement than a VHJ558 RSS (22). We wanted to determine the sequence of the RSS of each VH7183 gene to determine the extent of diversity of RSS within this VH gene family and to determine whether the relative frequency of recombination of the individual family members correlated with the sequence of their RSS. To determine the sequence of the RSS of each VH7183 gene, we initially used a primer in the leader, and a primer 3' of the RSS. The leader sequence is often very conserved within members of a VH family, and the sequence just 3' of the RSS is relatively conserved. Therefore, we used this primer set to amplify as many germline VH7183 genes as possible. PCRs were performed on the BAC DNA from 129/Sv mice, which, like BALB/c, are of the Igha haplotype. We were able to amplify 13 genes in this way, with multiple independent sequences of those genes. Because there were >13 EcoRI bands on Southern blots that hybridized with the VH7183, and because some known genes were missing initially, we also amplified some of the BAC DNAs with framework 1 (FR1) primers and either AF269 (which overlaps the 3' end of the RSS by 5 bp) or AM5 (which contains the nonamer and half the spacer) in an attempt to choose regions more likely to be conserved. E4.psi did not amplify with the primers, so we used a sequencing primer in FR3 of the published sequence of E4.psi to sequence the RSS. In other cases direct sequencing of DNA obtained from EcoRI bands of BAC DNA excised from agarose gels was performed for genes that we were unable to amplify from the BAC.
The sequence of the RSS and the adjacent coding end sequence for all
members of the family are shown in Fig. 1
(full-length sequences are available from GenBank under accession no.
AF290959–AF290972). The RSS of the VH7183 genes
fell into two major groups, which we called 7183-I and
7183-II. 7183-I RSS had the same nonconsensus heptamer as
81X (CACAATG; the one change underlined) and a
nonconsensus nonamer (ACTAAAACC), while the
7183-II RSS had the same nonconsensus nonamer as the
7183-I genes, but had a consensus heptamer. The spacers were
very similar, with the 7183-I and 7183-II
prototypes differing only in three positions. Seven genes had the
identical prototypic 7183-I RSS, and five had the identical
prototypic 7183-II RSS. In addition to these genes, seven
other genes were found to vary from the prototypes at a few positions
in the RSS, predominantly in the spacer, including 81X,
which differed from the 7183-I RSS by two changes in the
spacer. Only two genes have heptamers that vary from these two
prototypes, and they are classified as variants of 7183-I or
-II based on spacer sequence.
|
During our analysis of the BAC contig we found a previously unknown functional member of the VH7183 family by using FR1 and FR3 primers. It is identical with 7183.9 in the coding region, other than two changes at the 5' end and a single base pair change of a G to a T in FR3, so we call this gene 7183.9T. We have extended the sequence through the RSS, and it is also identical with 7183.9 in that region, but its leader/intron region shows four differences between the two genes. We also identified a second new, apparently functional, gene that we have provisionally called 7183.19. All these VH genes will be renamed numerically once all the interspersed pseudogenes and VHQ52 genes are identified. It is only 94% homologous to the most similar VH7183 gene and has a 7183-I variant RSS. A new presumably nonfunctional gene of the VH7183 family, called 7183.20, was also identified. Its RSS was related to the 7183-I RSS; however, it contains a C to G mutation at the first position in the heptamer, an invariant position essential for RSS function (24). Therefore, this RSS is highly likely to be nonfunctional.
Several years ago we reported the sequence of four new VH7183 genes from BALB/c mice and also made a compilation of all the VH7183 genes described at that time in the BALB/c strain (also Igha) (29). The sequences obtained in this current study enable us to refine that list. We have added three new sequences as just described. We also now eliminate MOPC21 as a germline gene; rather, it is almost certainly a 61-1P gene with three somatic mutations. Also, we previously listed a sequence from the literature, V-BK, which only differs from 7183.12 by 3 bp, two of which are a GA dinucleotide instead of an AG dinucleotide in FR3. We never observed V-BK in rearrangements, nor did we find it in the BAC contig, and we assume that it is the same as the germline 7183.12 gene. Also, we never observed the published sequence for 283 (30). The published coding sequence of 283 is almost identical with that of E4.15, but they vary in the intron (1). 283 is the only published VH7183 gene that ends with . . .TA, so we conclude that the two published sequences, 283 and E4.15, really represent the same gene.
Analysis of the role of the RSS in the nonrandom VH gene use
It has been shown that 81X rearranges 7- to 30-fold
more frequently than the VHJ558 gene in a
competition substrate (22, 23). However, the
VHJ558 genes have very poor RSS, and the nonamer
is so different from the consensus that it is not clear whether they
also have a shortened, less functional, 22-bp spacer (see Fig. 1
)
(24). Therefore, the first issue we wanted to address was
whether the RSS of 81X is much better than that of a more
conventional RSS. We chose the VHS107 gene
V1 as a representative gene. This gene has a consensus
heptamer, and the nonamer has one change from consensus. The spacer
length is the optimal 23 bp, as are all of the
VH7183 RSS. We wished to assess the relative
recombination supported by V1 vs 81X, so we made
competition recombination substrates comparing the relative frequency
of recombination of these two RSS to recombine to a DSP2
fragment. RSS fragments are made by PCR and are usually
100–200 bp in size, containing the RSS as well as some of the 5'-
and 3'-flanking regions. The substrates are transiently transfected
into pre-B cells, and the recovered plasmid is plated on a
chloramphenicol plate. Colonies are screened by PCR to determine the
relative frequency of rearrangement of the internal vs the external
VH fragment (Fig. 2).
|
shows that a control competition substrate containing
V1 in both internal and external positions demonstrates the
minimal effect of position (internal vs external) of the two competing
RSS fragments in the substrates. This is consistent with the controls
we have previously published using this competition recombination
system (18, 21). When we cloned a
VHJ558 RSS fragment in the competition
substrate, it was vastly disfavored, with 91% of the rearrangements
occurring to V1 instead of J558 (Fig. 2). Thus,
we conclude that VHJ558 genes do, in fact, have
poor RSS, as predicted by inspection of their RSS. However, when the
J558 fragment is replaced by 81X in the above
substrate, the relative frequencies of recombination of V1
and 81X are similar. This shows that the
81X RSS is only slightly better than that of a
VHS107 gene and suggests that V1
should rearrange almost as often as 81X in vivo if the RSS
is the major determinant of rearrangement frequency.
Within the VH7183 family, all
VH genes but two share an identical nonconsensus
nonamer. Approximately one-third of the VH genes
have a consensus heptamer, while two-thirds have a nonconsensus
heptamer, identical with that reported for 81X. Based on
classical studies with recombination substrates, one would expect that
the 7183-II RSS, with its consensus heptamer, might
rearrange at a higher frequency than the 7183-I RSS, with
one change in the heptamer sequence (24, 31).
81X, which rearranges at a very high frequency, has a
7183-I-type RSS, but with a unique spacer sequence. It is
possible that its variant spacer is the reason for its high frequency
of recombination. Although the heptamer and nonamer are traditionally
believed to be the most important regions of the RSS, we have
previously shown that the spacer can also play a significant role in
recombination frequency of V genes (18). In
that study we also showed the surprising result that a RSS with a
nonconsensus nonamer but an optimal spacer rearranges more frequently
than a consensus RSS with a poorer spacer, thus underscoring the
necessity to actually test the recombination potential of each RSS
experimentally. Therefore, to determine the relative frequency of
recombination of prototypic 7183-I and 7183-II
RSS, we replaced the internal gene fragment in our control
recombination substrate with that of 3.3.39 (a
representative 7183-I gene) or 37.1 (a
representative 7183-II gene). As predicted, the
7183-II gene, with it consensus RSS, is better than
V1 or the 7183-I genes for supporting
rearrangement in the recombination substrates (Fig. 2). Seventy-one
percent of the rearrangements were to the 7183-II gene,
rather than to V1. Importantly, the 7183-I gene
rearranged much less well than the 7183-II gene, and
81X rearranged at only a marginally higher frequency.
Therefore, for 81X, the two changes in the spacer sequence
did not significantly affect its recombination frequency. Thus, if the
RSS plays a major role in controlling the relative rearrangement
frequency of this VH family in vivo, then one
would predict that the 71813-II genes would rearrange most
frequently, and that 81X rearrangement would occur at
approximately the same frequency as that of the other 7183-I
genes.
Relative frequency of rearrangement of individual VH7183 genes
Therefore, we analyzed the initial rearrangement frequency of the VH7183 gene family members, determining only the biases inherent to the VDJ recombination process and not subsequent biases imposed by preferential selection and clonal expansion. The µMT mice (26), which have a targeted deletion of the cytoplasmic tail of µ, have such an unbiased repertoire. Because these mice cannot express surface Ig nor can they signal through the pre-B receptor, there can be no selection of developing pre-B cells, and furthermore, there is a resulting block of B cell development at the pre-B to B cell transition step. Thus, neither selection nor environmental bias is able to shape the observed repertoire in µMT mice. As would be anticipated in such mice, two-thirds of the sequences are out-of-frame. Using these mice, we asked the following questions. Do the 7183-I vs 7183-II RSS competition substrate results reflect the rearrangement frequency of those genes in vivo? Do all prototypic 7183-I (or all 7183-II) genes rearrange at the same frequency?
To determine the VH7183 family repertoire we
used a 5' pan-VH7183 family-specific primer
(AF303) located in FR1 in a region that is identical among all the
VH7183 family members, and a 3'
JH primer (AF13N). Rearranged
VH7183 genes were amplified from newborn liver
DNA and cloned into pBluescript, and individual clones were sequenced
to determine their identities. The data in Fig. 3 represent 126 sequences derived from
nine independent PCRs.
|
) is quite large. Two of the 7183-I
genes showed minimal if any rearrangement (69.1 and
68–5N), while other 7183-I genes were among the
five most frequently rearranging non-81X genes.
Surprisingly, the prototypic 7183-II genes were all
relatively poor at rearranging in vivo. Of the genes with variant RSS
other than 81X, only 61-1P rearranged at a high
frequency. Mapping of the chromosomal order of the VH7183 genes
Because the frequency of recombination of individual genes did not correlate with variation in the RSS, we wished to determine whether there was a gradient of usage from 3' to 5' throughout the locus as predicted by the chromosomal location hypothesis. The ongoing assembly of a BAC contig of the Igh locus of strain 129 (P. Goebel, N. Janney, W. J. Romanov, C. Hurre, and A. J. Feeney, manuscript in preparation) enabled us to map all the VH7183 genes and correlate their relative positions with their frequency of rearrangement.
VH7183 genes were identified and ordered using
two strategies. In one approach, each BAC in a tiling path across the
VH7183 region was analyzed by PCR amplification
using sets of VH7183-specific primers (primarily
AF191 (leader) and AF192 (3' of the RSS)), followed by cloning and
sequencing multiple clones to determine the content of each BAC. This
approach gave rise to most of the full-length sequences. In the second
approach, hybridizing bands on the Southern blots were ordered by
analysis of the overlapping BACs; the corresponding restriction
fragments were isolated from agarose gel, PCR-amplified, and sequenced
without cloning. This approach supplemented the first for genes that
did not amplify with the external primers. Additionally, most of the
genes were confirmed by direct sequencing from the agarose-isolated
restriction fragment. Using all these approaches, we have identified
almost all EcoRI bands that hybridize with 7183
probes, suggesting we have essentially completed this analysis. It is
likely that most of the hybridizing bands that we have not been able to
amplify represent pseudogenes, which may vary significantly from our
primer sequences. A map of the BAC contig and the
VH7183 gene order is shown in Fig. 4.
|
) with our repertoire data (Fig. 3)
which is displayed with the genes in the correct chromosomal order, we
identified an approximate gradient of rearrangement from the 3' to the
5' within the locus. 3.3.39, D6.96, 7183.9, and
50.1, all located near 81X at the 3' end of the
locus, are all over-represented. However, 61-1P, which
rearranges at a relatively high rate is located at the 5' end of the
locus, and 7183.14, on the 3' side, seldom rearranges. Mapping of VHOx2
The large VHQ52 family is interspersed with the VH7183 genes. One VHQ52 gene, Ox2, was shown to be greatly over-represented in rearrangements from fetal liver-derived Abelson-murine leukemia virus pro-B cell lines, and also from hybridomas from LPS-stimulated adult spleen cells (32). To determine whether this gene mapped near the frequently rearranging 81X at the 3' end of the locus, we mapped it using primers from the published sequence of Ox2, followed by cloning and sequencing. We mapped Ox2 to the most 3' BAC, 167C1, which contains 81X, D6.96, and E4.Psi (data not shown). This supports the hypothesis that VH genes at the 3' end of the locus undergo frequent rearrangement.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Together, their data and ours indicate that the
nonrandom frequency of gene usage seen in the more mature B cell
repertoire is significantly influenced by initial rearrangement
frequencies.
We addressed two main hypotheses that have been proposed to
explain the unequal frequency of rearrangement of V genes. One issue
was the role of the natural variation in RSS in determining the
frequency of rearrangement. We sequenced the RSS of 19 members of the
VH7183 family and showed that 12 of these had
one of two different closely related RSS, the prototypic
7183-I and 7183-II RSS. Most of the rest had a
small number of changes from one of these two prevalent RSS.
81X, which rearranges the most of all genes in this family,
had a unique spacer, and variations in the spacer sequence can affect
recombination frequency (18, 33). Thus, the hypothesis
that the unique RSS of 81X was responsible for its high
frequency of rearrangement was a reasonable hypothesis until we
performed recombination substrate experiments (Fig. 2). These
experiments comparing the various RSS in competition recombination
substrates reinforce previous observations that differences in the
individual elements of the RSS can clearly influence the frequency of
recombination. The VHJ558 RSS with its poor
nonamer and shortened spacer was greatly disfavored compared with the
much more conventional VHS107/V1 RSS. The
VH7183-II RSS containing a consensus heptamer
rearranged better than the VH7183-I RSS
containing a nonconsensus heptamer. However, the 81X RSS,
which differed by only two nucleotides in the spacer from the
prototypic 7183-I RSS, rearranged only marginally better
than the consensus 7183-I RSS, showing that in this case,
the spacer sequence variation did not influence recombination
potential.
However, these in vitro results were not mirrored by rearrangement
patterns in vivo. Firstly, 7183-II genes have the better RSS
as assayed in vitro, but rearrange less frequently in vivo. Secondly,
the frequency of use of the various members of 7183-I is
unequal. For example, rearrangements of 7183.9, 3:3.39,
50.1, and D6.96 are found much more frequently than
those of 68–5N and 69.1, yet both possess
identical 7183-I-type RSS. Previously we and others have
found a strong correlation between the relative frequency of
rearrangement of genes in vivo and the ability of their RSS to support
recombination in vitro in recombination substrates. In some cases
different loci (e.g., vs 
) or different genes within a locus
(e.g., the three VIII genes) were analyzed, so the
correlation could possibly be deemed coincidental (18, 19). However, for the VA2 genes we showed that two
alleles differing at one position in the heptamer rearrange at
5-fold different frequencies in vivo and in vitro, and the alleles
presumably are identical in chromosomal location and sequence (only
three changes were found in 700 bp) strongly suggesting that indeed the
RSS was responsible for the differences in rearrangement frequencies in
vivo vs in vitro (21). Similarly, two alleles of
V3 have been shown to differ drastically in their
representation in the peripheral repertoire (8.1 vs 1.2%), and the
only difference in their sequence is 1 bp in the spacer sequence
(34). Thus, we believe that these two latter cases of
allelic differences in which chromosomal position is not a factor
coupled with the extensive in vitro recombination substrate data
demonstrate that RSS differences can indeed significantly affect
rearrangement frequencies (24, 31, 35). However, the data
shown here indicate that factors other than the RSS can also have an
important role in rearrangement frequency, and for the
VH7183 genes these other factors appear to play
a predominant role, over-riding the variation in RSS potency.
The other previously proposed hypothesis that we were testing was that the location of a gene within the V, D, or J locus plays a key role in its accessibility for rearrangement, with the most proximal genes rearranging first or most often. Here we analyzed BAC clones to determine the gene order of the individual members within the VH7183 region. These data clearly show that there is increased frequency of rearrangement in the 3' end of the locus. Also, we showed that the VHQ52 gene, which was previously shown to rearrange very frequently, was located on the same BAC as 81X, again suggesting that this region at the 3' end of the locus contains genes that rearrange more often. However, there are a few exceptions. 61-1P, which rearranges frequently, is at the far 5' end of the VH7183 region. Also, chromosomal location alone cannot account for the very high frequency of rearrangement of 81X, nor can its RSS. Thus, chromosomal location appears to bias the rearrangements in general, although location proximal to the D cluster is clearly not the only factor that affects recombination frequency.
All frequently rearranging genes other than 61-1P were
7183-I genes. These genes also are located in the 3' half of
the locus, so their location may contribute to their overuse. However,
7183.14, a 7183-II gene, is located within this
region of higher recombination, but rearranges at a much lower level,
suggesting that perhaps 7183-I genes may be more
recombinogenic than 7183-II genes. However, any difference
in 7183-I and 7183-II use would clearly not be a
result of the RSS per se, but could perhaps be due to some other
commonality among the 7183-I vs 7183-II genes in
a region outside of the portion that we have sequenced. Perhaps the
7183-I genes have a common flanking DNA element that
influences their recombination frequency. Further analysis of the
sequence of the flanking regions surrounding these genes should yield
insight into genomic control regions potentially located in the
VH7183 locus. Also, because not all
7183-I genes rearrange at high frequency, this again
suggests that individual genes control their rearrangement frequency by
some flanking DNA. In accord with this concept, 
T cells show
strictly ordered rearrangement in fetal and adult thymus, and swapping
of 1 kb of 5'-flanking regions, including the promoters of the V3
and V2 genes, led to alteration of their recombination pattern
(36). Also, we have recently shown that the transcription
factors E2A and EBF show gene-specific preferences in their ability to
induce rearrangements in transfected epithelial cells, suggesting that
another factor influencing rearrangement frequency could be the
presence and relative location of binding sites for transcription
factors that promote accessibility to
rearrangement.5 Perhaps the 7183 genes,
which do not rearrange as often as other genes with the same RSS, such
as 7183.14, might have a deleterious change in the sequence
of the binding sites for factors that control accessibility, and thus
be less accessible for rearrangement.
In conclusion, we determined the relative rearrangement frequency of all members of the VH7183 family and showed that the individual genes rearranged with a wide range of different frequencies. We compared individual VH gene family usage with RSS efficiency and chromosomal location, two factors that have been proposed to be major contributors to rearrangement frequencies of a given gene segment in vivo. We determined that the RSS was clearly not the predominant reason for preferential rearrangement, but, rather, that other factors must override the relative efficacy of the RSS. Because the mapping of all the individual genes showed that most frequently rearranging genes were located in the 3' half of the VH7183 region, this suggests that chromosomal location may have the most major influence on VH7183 gene rearrangement.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Institut National de la Santé et de la Recherche Médical, Unité 399, Immunology and Genetics of Parasitic Diseases, Faculty of Medicine, 27 boulevard J. Moulin, Marseilles 13385, France.
3 Address correspondence and reprint requests to Dr. Ann J. Feeney, Department of Immunology, IMM-22, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037. E-mail address: feeney{at}scripps.edu
4 Abbreviation used in this paper: RSS, recombination signal sequence(s); BAC, bacterial artificial chromosome; FR, framework.
5 P. Goebel, N. Janney, J. R. Valenzuela, W. J. Romanow, C. Murre, and A. J. Feeney. 2001. Localized gene-specific induction of accessibility to V(D)J recombination induced by E2A and EBF in non-lymphoid cells. Submitted for publication.
Received for publication January 29, 2001. Accepted for publication April 16, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
locus. J. Immunol. 145:2347.[Abstract]
locus correlates with transcription of the unrearranged genes. J. Exp. Med. 177:729. usage in vivo. J. Exp. Med. 187:1495. light-chain recombination signal sequences mediate recombination more frequently than do those of light chain. Proc. Natl. Acad. Sci. USA 88:10721. gene A2b, which is associated with increased susceptibility of Navajos to Haemophilus influenzae type b disease. J. Immunol. 161:6068.A2 allele in Navajos which may play a role in increased susceptibility to Haemophilus influenzae type b disease. J. Clin. Invest. 97:2277.[Medline]
3) expression correlates with allelic polymorphism in the spacer region of the recombination signal sequence. J. Exp. Med. 179:1707. genes is controlled by sequences immediately upstream of the V genes. Immunity 9:159.[Medline]
This article has been cited by other articles:
|
|
 |
|
  I. Retter, C. Chevillard, M. Scharfe, A. Conrad, M. Hafner, T.-H. Im, M. Ludewig, G. Nordsiek, S. Severitt, S. Thies, et al. Sequence and Characterization of the Ig Heavy Chain Constant and Partial Variable Region of the Mouse Strain 129S1 J. Immunol., August 15, 2007; 179(4): 2419 - 2427. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. C. Watson, C. S. Moffatt-Blue, R. Z. McDonald, E. Kompfner, D. Ait-Azzouzene, D. Nemazee, A. N. Theofilopoulos, D. H. Kono, and A. J. Feeney Paucity of V-D-D-J Rearrangements and VH Replacement Events in Lupus Prone and Nonautoimmune TdT-/- and TdT+/+ Mice J. Immunol., July 15, 2006; 177(2): 1120 - 1128. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. C. Ippolito, R. L. Schelonka, M. Zemlin, I. I. Ivanov, R. Kobayashi, C. Zemlin, G. L. Gartland, L. Nitschke, J. Pelkonen, K. Fujihashi, et al. Forced usage of positively charged amino acids in immunoglobulin CDR-H3 impairs B cell development and antibody production J. Exp. Med., June 12, 2006; 203(6): 1567 - 1578. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Johnston, A. L. Wood, D. J. Bolland, and A. E. Corcoran Complete Sequence Assembly and Characterization of the C57BL/6 Mouse Ig Heavy Chain V Region J. Immunol., April 1, 2006; 176(7): 4221 - 4234. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Schelonka, I. I. Ivanov, D. H. Jung, G. C. Ippolito, L. Nitschke, Y. Zhuang, G. L. Gartland, J. Pelkonen, F. W. Alt, K. Rajewsky, et al. A Single DH Gene Segment Creates Its Own Unique CDR-H3 Repertoire and Is Sufficient for B cell Development and Immune Function J. Immunol., November 15, 2005; 175(10): 6624 - 6632. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Espinoza and A. J. Feeney The Extent of Histone Acetylation Correlates with the Differential Rearrangement Frequency of Individual VH Genes in Pro-B Cells J. Immunol., November 15, 2005; 175(10): 6668 - 6675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. I. Ivanov, R. L. Schelonka, Y. Zhuang, G. L. Gartland, M. Zemlin, and H. W. Schroeder Jr. Development of the Expressed Ig CDR-H3 Repertoire Is Marked by Focusing of Constraints in Length, Amino Acid Use, and Charge That Are First Established in Early B Cell Progenitors J. Immunol., June 15, 2005; 174(12): 7773 - 7780. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Fuxa, J. Skok, A. Souabni, G. Salvagiotto, E. Roldan, and M. Busslinger Pax5 induces V-to-DJ rearrangements and locus contraction of the immunoglobulin heavy-chain gene Genes & Dev., February 15, 2004; 18(4): 411 - 422. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Montalbano, K. M. Ogwaro, A. Tang, A. G. W. Matthews, M. Larijani, M. A. Oettinger, and A. J. Feeney V(D)J Recombination Frequencies Can Be Profoundly Affected by Changes in the Spacer Sequence J. Immunol., November 15, 2003; 171(10): 5296 - 5304. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Johnson, C. Angelin-Duclos, S. Park, and K. L. Calame Changes in Histone Acetylation Are Associated with Differences in Accessibility of VH Gene Segments to V-DJ Recombination during B-Cell Ontogeny and Development Mol. Cell. Biol., April 1, 2003; 23(7): 2438 - 2450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Goebel, A. Montalbano, N. Ayers, E. Kompfner, L. Dickinson, C. F. Webb, and A. J. Feeney High Frequency of Matrix Attachment Regions and Cut-Like Protein x/CCAAT-Displacement Protein and B Cell Regulator of IgH Transcription Binding Sites Flanking Ig V Region Genes J. Immunol., September 1, 2002; 169(5): 2477 - 2487. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Yu, A. Taghva, and M. R. Lieber The Cleavage Efficiency of the Human Immunoglobulin Heavy Chain VH Elements by the RAG Complex. IMPLICATIONS FOR THE IMMUNE REPERTOIRE J. Biol. Chem., February 8, 2002; 277(7): 5040 - 5046. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Kompfner, P. Oliveira, A. Montalbano, and A. J. Feeney Unusual Germline DSP2 Gene Accounts for All Apparent V-D-D-J Rearrangements in Newborn, But Not Adult, MRL Mice J. Immunol., December 15, 2001; 167(12): 6933 - 6938. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, genes possessing 7183-I RSS
(prototypic and variant); 
, those with 7183-II RSS
(prototypic and variant). *, Genes with prototypic
7183-I RSS; **, those with a prototypic
7183-II RSS.