| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Microbiology and Immunology, Loyola University of Chicago, Maywood, IL 60153
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
80% of rearranged VDJ genes (2, 5), and much
of the diversity of the Ab repertoire develops both by somatic
hypermutation (6, 7, 8) and by a somatic gene conversion-like
mechanism in which upstream VH gene segments
serve as the donors (3, 5, 8, 9).
Because normal rabbits preferentially use the
VHa allotype-encoding gene,
VH1, in VDJ gene rearrangements, most of their
serum Ig (80%) has the VHa allotypes
(reviewed in Refs. 10 and 11). Three
allelic forms of the VHa allotypes, a1, a2, and
a3, have been identified in laboratory rabbits by serologic reactions
using anti-allotype Abs. The serologic VHa
specificities are associated with specific amino acids within FR1 and
FR3 (1, 12). Up to 20% of serum Ig in normal rabbits do
not react with anti-a1, anti-a2, or anti-a3 allotypic Abs.
These VHa-negative molecules are generated mainly
from two VH gene segments,
VHx, and
VHy, which do not
encode the VHa allotype-associated amino acids
(2, 7).
In contrast to normal rabbits, serum of young homozygous ali/ali mutant Alicia rabbits contains mostly VHa-negative Ig (13). The mutant Alicia rabbits were derived from a VHa2 parental strain in which a 10-kb segment of DNA, including VH1, was deleted (1). As expected, most of the VHa-negative Ig molecules are encoded by the two VHa-negative-encoding gene segments, VHx and VHy (14, 15). This preferential usage of VHx and VHy occurs even though these gene segments are located at least 50 kb 5' of VH1 (1).
Although Alicia rabbits of a few weeks of age primarily express the VHa-negative-encoding gene segments VHx and VHy, the level of a2 serum Ig and the number of a2 B cells increase as the rabbits age (13, 16). The goal of our study is to determine which VH gene segments are used by VDJ genes from a2 B cells in Alicia rabbits and how VH gene segment usage in Alicia rabbits is regulated. Sehgal et al. (17) determined the nucleotide sequences of VDJ genes from FACS-sorted a2 Ig+ B cells and found that the VH regions of many of the sequences were similar to that of VH4. They suggested that many of the VDJ genes in a2 B cells used VH4, the 3'-most functional VH gene segment of Alicia rabbits (1). The problem with identifying the utilized VH gene segment solely on the basis of the nucleotide sequence of the VDJ coding region is that all rabbit VH gene segments are 80% identical and belong to a single VH family (18, 19, 20). Further, once the VDJ genes have undergone somatic diversification, the utilized VH gene segment cannot be identified with certainty. In the present study we determined the VH gene segment used in VDJ gene rearrangements of a2+ B cells in Alicia rabbits by analyzing the nucleotide sequence of the promoter regions of the VH gene segments. Promoter regions can be used to identify VH gene segments, because promoter regions are distinct for different VH gene segments and they undergo little somatic diversification. We cloned the VDJ genes from a2 Ig-secreting hybridomas and from FACS-sorted a2+ B cells of adult Alicia rabbits and found that >80% of the VDJ genes utilized either VH4 or VH7. To investigate the timing of rearrangements of VH4 and VH7, we determined the VH gene segment usage in VDJ gene rearrangements during fetal development. We found that VH4 was used in 60–80% of nonproductive fetal VDJ gene rearrangements, which suggests that VH4 is predominantly rearranged during fetal development. However, 15% or less of fetal VH4-utilizing VDJ genes were productive. We discuss possible mechanisms by which the predominantly rearranged gene segment VH4 is found infrequently in productive VDJ gene rearrangements during fetal development and yet is expressed in a large number of B cells in the periphery as the rabbits age.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Alicia rabbits were progeny of rabbits originally described by Kelus and Weiss (13). Three-year-old Alicia rabbits (20297 and 20304) were immunized with heat-killed Bacteroides spp., and spleen cells were fused with the rabbit hybridoma fusion partner 240E, as described by Spieker-Polet et al. (21). Hybridoma supernatant fluids were tested by ELISA for the presence of a2 Ig molecules by using microtiter plates coated with affinity-purified anti-a2 Ab. After supernatant fluids were added, biotinylated anti-a2 Ab was added and the ELISA was developed with reagents from the Vectastain ABC kit (Vector Laboratories, Burlingame, CA) and 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) substrate. The optical density was read at 405 nm. From a total of 47 IgG-secreting hybridomas, 35 (75%) secreted a2 IgG. The a2-secreting hybridomas (non-Ag specific) were subcloned by limiting dilution analysis and were used to analyze VH gene segment usage.
PCR amplification and cloning of VDJ genes
The VDJ genes and the promoter regions were PCR amplified from hybridoma genomic DNA of the a2 Ig-secreting hybridomas. The upstream primer, 5'-TCGAATTCTCTGAATCATATCACAGCCAT-3' (the EcoRI site is underlined), was derived from the promoter region 216–236 bp 5' of the ATG start codon (4). This oligomer primes the promoter region of most known VH gene segments but not the promoter region of VHy, the VDJ gene of the rabbit fusion partner. In each hybridoma tested, we were able to amplify a non-VHy-utilizing VDJ gene, which indicates that none of these hybridomas used VHy in their VDJ gene rearrangements. The downstream primer, 5'-GTTAAGCTTCACCTGA(G/A)GAGACGGTGACCAGGGT-3' (the HindIII cloning site is underlined), was derived from the conserved 3' end of rabbit JH gene segments (22), which includes the coding sequence for amino acids TLVTSS and four noncoding nucleotides 3' of the JH gene segments. The PCR was performed by the hot-start method for 30 cycles with denaturation at 94°C for 45 s, annealing at 60°C for 45 s, and extension at 72°C for 45 s. The PCR products were cloned into EcoRI/HindIII-restricted M13 mp18/19, and the nucleotide sequences were determined in both orientations by the dideoxynucleotide termination method using the Sequenase kit (U.S. Biochemical Corp., Cleveland, OH).
To determine the VH gene segments used in the VDJ genes of a2+ B cells, appendix cells from a 2.5-yr-old Alicia rabbit were reacted with biotinylated anti-a2 Ab and FITC-avidin and a2+ B cells were isolated by flow cytometry on the FACStarPlus (Becton Dickinson, Mountain View, CA). Genomic DNA of these cells was prepared for PCR amplification of VDJ genes and their promoters. The PCR amplification protocol described above was used except that the upstream primer was 5'-TAACAAGCTTAAAAATTCATATGATCTGAA-3' (the HindIII cloning site is underlined), derived from 232–253 bp 5' of the ATG start codon. This oligomer primes the promoter region of most known VH gene segments, including the promoter of VHy. The PCR-amplified VDJ genes were cloned into pGEMT (Promega, Madison, WI), and the nucleotide sequences were determined in one orientation on the automated Applied Biosystems (Foster City, CA) Prism310 DNA sequencer, according to the manufacturer’s protocol. From a total of 11 sequences analyzed, 1 had none of the 11 a2 allotype-associated amino acids, and these data are not included in our analysis. We suggest that this sequence derived from a VHa-negative B cell that contaminated the flow cytometry-sorted a2+ B cells.
To determine the VH gene segment usage in rearrangement during fetal development, genomic DNA was isolated from liver and bone marrow of gestation day 21 fetuses and from newborn rabbits. The VDJ genes of genomic DNA were PCR amplified using a seminested PCR scheme. The first round of PCR used 5'- AGGAATTCTGCAGCTCTGGCACAGGAGCTC-3' (the EcoRI cloning site is underlined) as the 5' primer, derived from 57–80 bp 5' upstream of the start codon, and 5'-GGACAAGCTTGAGGAGACGGTGACCAGGGT-3' (the HindIII cloning site is underlined) as the 3' primer, derived from a conserved region at the 3' end of JH that encodes amino acids TLVTSS. In the second round of PCR, 1 µl of the PCR product from the first round of PCR was used as the template, and 5'-GGCTGAATTCGCTTCTCCTGGTCGCTGTG-3' (the EcoRI cloning site is underlined) as the 5' primer, derived from the leader region 21–39 bp 3' of the start codon, and the same 3' primer as that used in the first round of PCR. The PCR products were cloned into either M13 mp18/19 or pGEMT-Easy (Promega), and the nucleotide sequences were determined in one orientation as described above. Nonproductively rearranged VDJ genes were identified as those in which either the reading frame of the JH region was altered or a stop codon was introduced during recombination of VH, D, and JH gene segments.
Cloning of VDJ genes from a size-selected library
The VDJ gene of the VHa2 Ig-secreting hybridoma 8E1 was cloned from a size-selected genomic library. The 6-kb BamHI genomic library was constructed in the pGEM-3 plasmid vector and screened with VH and JH probes for double-positive clones. DNA from the double-positive clones was isolated and restriction mapped. The VDJ and promoter regions were cloned subsequently into M13 mp18/19, and the nucleotide sequences were determined using the Sequenase kit.
Expression of VH4-utilizing VDJ genes
Undiversified VH4- and
VHy-utilizing VDJ genes from Alicia rabbits and
a VH1-utilizing VDJ gene from an a2 rabbit that
had been cloned into M13 for nucleotide sequence analysis were PCR
amplified using, as 5' primer,
5'-TCGGATATCCACCATGGAGACTGGGCTGCGCTGGCTTCTCCTGGTCGCTG-3'
(the EcoRV cloning site is underlined) and, as 3' primer,
5'-AGTGCTAGCTGAAGAGACAGTGACCAGGGTGCC-3' (the
NheI cloning site is underlined). The VDJ genes were cloned
into the EcoRV and NheI sites of the expression
vector PAH4604 (kindly provided by Sherie L. Morrison, University of
California at Los Angeles) (23) and expressed as chimeric
IgG molecules, with the constant region encoded by the human C
1
gene. Rabbit-
L-chain-producing SP2/0 cells
(24) were transfected with the H chain-encoding plasmids,
and stable transfectants were selected with histidinol. Supernatant
fluids were first tested for the concentration of chimeric IgG
using ELISA plates coated with rabbit
anti-human IgG, and the assay was developed with biotinylated
rabbit anti-human IgG using the Vectastain ABC kit (Vector
Laboratories) and 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid)
as substrate. Transfectants secreting comparable levels of chimeric IgG
were used to determine reactivity with anti-a2 Ab by ELISA, using
ELISA plates coated with affinity-purified rabbit anti-a2 Ab (1.4
µg/ml). Two hours after the supernatant of transfectants was added,
biotinylated a3 rabbit anti-a2 Ab (0.4 µg/ml) was added and
incubated for 1 h. The assay was developed using the Vectastain
ABC kit as described above. The optical density at 405 nm was
determined after 24 h at 4°C.
The reactivity of VH4-encoded chimeric IgG
(VH4-IgG) (3739) with anti-a2 Ab was
inhibited with purified VH4-IgG and a2 Ig in an
inhibition assay. Briefly, supernatant from the
VH4-IgG-secreting transfectant was purified on a
protein A-conjugated Sepharose column (Pharmacia, Piscataway, NJ). The
Ig was eluted with 0.1 M citric acid pH 3.0, and the concentration of
purified VH4-IgG was estimated by ELISA using
anti-b4 (anti-rabbit light chain allotype) Ab with purified
polyclonal b4 Ig as standards. In the inhibition assay, ELISA plates
were coated with VH4-IgG (1 µg/ml) overnight
at 4°C. Biotinylated anti-a2 Ab (1 µg/ml) was incubated with
serial (1:2) dilution of 10 µg/ml of inhibitor
VH4-IgG, rabbit polyclonal a1, a2, and a3 Ig.
After 2 h at room temperature, the mixture was added to the
VH4-IgG coated ELISA plate. After 1h at room
temperature, the assay was developed with the Vectastain ABC kit as
described above. The rabbit polyclonal a1, a2, and a3 (used as
inhibitor) and b4 Ig (as standard for quantitation) were obtained from
normal rabbit serum by ammonium sulfate precipitation and
DEAE-cellulose chromatography (25).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We cloned VDJ genes from a2-secreting hybridomas and from
a2+ B cells of adult Alicia rabbits and
identified the utilized VH gene segment by nucleotide
sequence analysis of the VH promoter region. We used the
sequence of the promoter region rather than the sequence of the coding
region to identify the VH gene segment, because all rabbit
VH gene segments reportedly belong to the same
VH gene family, and the coding regions of essentially all
VDJ genes are somatically diversified within the first few weeks of the
rabbit’s life (9). In contrast to the coding regions, we
had evidence as shown below that the promoter regions are distinct for
each VH gene segment and they undergo little somatic
diversification. Therefore, we used promoter regions to identify the
utilized VH gene segment. To confirm that the
VH gene segments of the ali allele have distinct
promoter regions, we determined the nucleotide sequences of the
promoter regions for the known functional germline VH gene
segments from the ali haplotype, VH4,
VH7, VHx, and
VHy, as well as VH9 from
the a2 (E) haplotype (Fig. 1
). We found that the sequence of each
promoter region was different and therefore could be used to identify
the utilized VH gene segment.
|
), indicating that both
VH4 and VH7 were utilized
frequently in the VDJ genes of the a2 Ig-secreting hybridomas and of
a2+ B cells from 2.5- and 3-yr-old Alicia
rabbits.
|
). We found that the restriction map of
DNA 5' of the VDJ gene of clone 8E1 was identical with that 5' of
germline VH4. Further, the nucleotide sequence
of the VH gene segment 5' proximal to the 8E1 VDJ gene was
identical with that of germline VH5 (GenBank
accession no. AF177017), the VH gene segment immediately 5'
of germline VH4 (data not shown). These data
confirm that VH4 was used in the VDJ gene of
hybridoma 8E1.
|
).
The promoter region of one of the VDJ genes (hybridoma 1E6) was
identical with that of VH9 of the
a2 allele (E haplotype), indicating
that VH9 was used in this VDJ gene. Further, the
promoter regions of three VDJ gene sequences, 9C4, 5066, and 5086, were
unlike those of any known VH gene segment (Fig. 1),
indicating that these VDJ genes likely used germline VH
gene segments that have not yet been identified. We conclude that at
least six VH gene segments were used in the VDJ genes of
a2+ B cells and that VH4
and VH7 were used in most (84%) of
them. Origin of VHa2 allotype-associated amino acids
We assessed next whether the a2 allotype-associated amino acids of
the a2-encoding VDJ genes were derived from the germline VH
gene segments used in the VDJ gene rearrangements or from somatic
diversification of the VDJ genes. We determined the nucleotide
sequences of the VDJ genes and compared the deduced amino acid
sequences with the known a2 allotype-associated amino acid residues
(Fig. 3) (26). We found that
20 of the 25 VDJ genes each encoded at least 7 of the 11 known a2
allotype-associated amino acids, whereas the other five VDJ genes (1C3,
4C4, 7D3, 9C4, and 5086) each encoded only four to six of the a2
allotype-associated amino acids. When we compared the sequences of the
VH4- and VH7-utilizing
VDJ genes with those of germline genes, we found that some of the
allotype-associated amino acids derived from the germline and others
resulted from somatic diversification. For example, in the
VH4-utilizing gene segment, 1F7, two of five a2
allotype-associated residues in FR1 and five of six a2
allotype-associated residues in FR3 were encoded by germline
VH4; the others must have arisen through somatic
diversification.
|
). Therefore, we
searched for potential donor genes that could have contributed these a2
allotype-associated nucleotides. By comparing the nucleotide sequences
of the leader intron and FR1 of the VH4- and
VH7-utilizing VDJ genes with the sequences of
all known germline VH gene segments, we found that
VH9-a2 has a sequence nearly identical with the
diversified VH region of nine of the
VH4- and seven of the
VH7-utilizing a2-encoding VDJ genes (Fig. 4). We
conclude that these a2-encoding VDJ genes gained a2 allotype-associated
nucleotides through somatic gene conversion using
VH9 or a VH9-like
germline gene as donor.
|
The frequent use of VH4 and
VH7 in a2-encoding VDJ genes from adult Alicia
rabbits contrasts sharply with the rare usage of these gene segments in
young Alicia rabbits (14, 15). Although the increased
usage of these gene segments could result from de novo B lymphopoiesis
in older rabbits with VH4 and
VH7 gene segments being rearranged frequently,
Crane et al. (9) showed that little, if any, B
lymphopoiesis occurs in adult rabbits. Consequently, we thought the
VH4- or VH7-utilizing B
cells developed early in ontogeny but for some reason were not expanded
until later in development. To determine to what extent
VH4 and VH7 were
rearranged early in ontogeny, we PCR amplified and sequenced VDJ genes
from fetal/neonatal liver and bone marrow of Alicia rabbits. We found
that the most frequently rearranged gene segment wasVH4, found in 60–80% of the nonproductive VDJ genes
(Table II). Although
VH4 was the predominantly rearranged gene
segment, 15% or fewer of the VH4-utilizing
rearrangements, both at 21 days of gestation and at birth, were
productive (Table III). The
VHa-negative gene segments,
VHx and VHy, which are
the gene segments predominantly utilized in B cells soon after birth
(16), were used together in <15% of the nonproductive
VDJ gene rearrangements (Table II); however, in this case, the
percentage of productive VHx and
VHy gene segment rearrangements was 38% at 21
days of gestation and increased to 89% at birth (Table III). Further,
the majority (39 of 44) of the productive VDJ genes at birth used
VHx and VHy (Table II).
These data show that early in development,
VH4 rearranged extensively, but that
the resulting VH4-utilizing B-lineage cells did
not expand during fetal development. In contrast,
VHx and VHy represented
<15% of VDJ gene rearrangements during fetal development
(nonproductive rearrangements), and yet, at birth, they represented
>75% of the productive VDJ genes.
VH7-utilizing VDJ genes were found infrequently
during fetal development, both in productive and nonproductive VDJ gene
rearrangements.
|
|
To determine whether germline VH4, which
encodes 7 of the 11 a2 allotype-associated amino acids, encodes a2
allotypic determinants, we cloned and expressed two
VH4-utilizing VDJ-C genes in murine SP2/0
cells that had been transfected with a rabbit L-chain gene. We then
tested, by ELISA, whether the Ig secreted from the transfected cells
reacted with anti-a2 Ab. We used the undiversified
VH4-utilizing VDJ genes, 3739 and 4053, cloned
from fetal and neonatal rabbits, respectively. In both cases, we found
that the Ig secreted from the transfected cells reacted with
anti-a2 Ab (Fig. 5A).
However, reactivity was considerably less than that of the transfectant
5.2-1 (27), which expressed a similar IgG molecule with
the VH region derived from the prototypic a2-encoding
VH gene segment, VH1 (Fig. 5A). The negative control, Ig, with the VH
region derived from the VHa-negative VH gene
segment, y33 (3719), did not react with anti-a2 Ab. The reactivity
of VH4-encoded Ig with anti-a2 Ab was
confirmed in an inhibition assay. We found that rabbit polyclonal a2 Ig
inhibited the reactivity of VH4-encoded Ig
(3739) with anti-a2 Ab in a manner similar to that of
VH4-encoded Ig (3739). As control, neither
rabbit polyclonal a1 Ig nor a3 Ig inhibited the reaction. These data
indicate that germline VH4 encodes some epitopes
of the a2 allotypic specificity. In data not shown, we found that
VH4-encoded Ig (3739) did not inhibit the
reactivity of polyclonal a2 Ig with anti-a2 Ab, indicating that
VH4 does not encode all of the a2 epitopes found
in polyclonal a2 Ig. Germline VH7 encodes a2
allotype-associated amino acid residues similar to those of
VH4, and accordingly, we expect that
VH7 encodes some but not all of the epitopes of
the a2 allotypic specificity.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) (1). Without
VH1, Alicia rabbits use predominantly the
VHa-negative encoding gene segment
VHy in the first few weeks of life (14, 15). By several weeks of age, however, a2 Ig molecules are
present in the serum of Alicia rabbits and eventually they account for
nearly 50% of total serum Ig (13). We report now that
a2-encoding B cells from the periphery of adult Alicia rabbits utilize
predominantly two a2-encoding gene segments, VH4
and VH7. We suggest that the increase of a2 Ig
in the serum of Alicia rabbits as they become older results from
selective expansion of VH4- and
VH7-utilizing B cells in the periphery.
|
15%. These data indicate that
VH4-utilizing B-lineage cells are not expanded
during fetal development. Possible explanations for this observation
are discussed below. VH gene segment usage in VHa2-producing B cells and the genetic origin of VHa2 allotype-associated amino acids
We undertook this study to determine the origin of the VHa2 allotype Ig in serum of adult Alicia rabbits. We hypothesized that the a2 Ig could be the result of somatic diversification of VHa-negative genes by somatic gene conversion using a2-encoding donor VH gene segments, or late-onset expression of a2-encoding VH gene segments, or both. We found that none of the VDJ genes from the a2+ B cells used VHx or VHy, indicating that a2 serum Ig in adult Alicia rabbits was not derived from B cells utilizing VHa-negative gene segments. Instead, all but four of the a2-encoding VDJ genes used either VH4 or VH7, germline gene segments that encode 7 of the 11 a2 allotype-associated amino acids. Although we cannot rule out the possibility that we missed B cells using other VH gene segments that encode for weak reactivity to anti-a2 Ab, our sorted a2+ appendix B cells included cells spanning a wide range of fluorescence intensity. Further, 6 of the 25 VDJ genes we analyzed encoded only 4 to 6 of the 11 a2-allotype associated amino acids. We expect that these 6 VDJ genes encode for VH regions that react weakly with anti-a2 Ab. Therefore, we consider it likely that our findings are representative of most a2+ B cells.
We expressed an undiversified VH4-utilizing
VDJ-C gene in vitro and found that germline
VH4 encoded a2 epitopes as determined by
reaction with anti-a2 Abs. The VDJ genes in most of the
a2+ B cells had acquired
VHa2-encoding nucleotides in addition to those encoded in
the germline, and, as reported by Sehgal et al. (17), we
found that many of the VDJ genes appeared to acquire
VHa2-encoding nucleotides through somatic gene conversion,
using VH9 as the donor. We think it is unlikely
that these changes attributed to gene conversion were due to enzymatic
reactivity during PCR, because the nucleotide sequence of 8E1, one of
the VDJ clones that appeared to have undergone gene conversion, was
determined from a subcloned genomic fragment. We conclude that the
VHa2 Ig in serum of adult Alicia rabbits is derived from
the use of VHa2-encoding gene segments (e.g.,
VH4 and VH7) in VDJ gene
rearrangements as well as from somatic gene conversion of these genes
using still other VHa2-encoding gene segments as
donors.
Preferential rearrangement of the 3'-most VH gene segment, VH4, in Alicia rabbits
The extensive rearrangement of VH4 early in ontogeny of Alicia rabbits surprised us, because this gene segment has rarely been found to be used in normal rabbits (2, 28). The frequency of rearrangement of specific VH gene segments in mice, humans, and chickens is influenced by several factors, including proximity to D or JH gene segment clusters (29, 30), sequence variation of recombination signal sequence (31, 32, 33, 34), selective homology between VH and D coding end sequences (35, 36, 37, 38, 39), and intronic cis-acting regulatory elements (40, 41). Because the nucleotide sequences of the 3' end of the VH region and the heptamer and nonamer of the recombination signal sequence of VH4 are identical with VH1, the VH gene segment preferentially rearranged in normal rabbits, the limited use of VH4 gene segments in normal rabbits cannot be explained by differences in the recombination signal sequence or by selective homology. Because VH1 in normal rabbits and VH4 in Alicia rabbits are the 3'-most functional VH gene segments, VH gene segment rearrangement might be determined to a large extent by proximity. However, we cannot rule out the possibility that the recombination frequency of various VH gene segments might also be determined by intronic regulatory elements associated with one or more VH gene segments.
Changes in VH4 usage during ontogeny: contribution of B cell selection
In primary lymphoid tissues, we found that
VH4 was used in 76% (48 of 63) of total VDJ
genes (productive plus nonproductive) at 21 days of gestation, and by
birth, its usage decreased to nearly 20% (13/63). In contrast,
VH4 usage in the periphery increased from an
undetectable level at 2 wk of age (15) to 50% of the
a2-encoding VDJ genes and 25% of the total VDJ genes in adult rabbits.
We suggest the decrease in VH4 usage during
fetal development is the result of VH4-utilizing
B lineage cells not differentiating into mature B cells. One of the
signals necessary for differentiation of murine and human pre-B cells
to B cells is mediated through the pre-B cell receptor (pBCR), a
complex formed between µ-chain and surrogate light chain (42, 43). Murine B cell precursors rearranging
VH81x rarely differentiate into mature B cells,
because µ-chains encoded by VH81x cannot pair
with surrogate light chain to form pBCR (44, 45).
Likewise, VH4-derived heavy chains might not
pair efficiently with surrogate light chain, and consequently, B cell
precursors utilizing VH4 would not be selected
for differentiation.
We further suggest that the increase of VH4 usage in the periphery results from the selective expansion of a few VH4-utilizing B cells. These VH4-utilizing mature B cells may be generated during fetal development either because their precursors express particular VH4-utilizing µ-chains that can pair with surrogate light chain or because of as-yet-unknown mechanisms. VH4-utilizing B cells in the periphery could also be derived from de novo postnatal lymphopoiesis. Although Crane et al. (9) showed that little, if any B lymphopoiesis occurs in bone marrow of adult rabbits, we know very little about postnatal B cell development, to what extent it contributes to the peripheral B cell pool, and which VH gene segments are used by newly generated B cells. It will be important to determine whether the development of VH4-utilizing B cell precursors is regulated in adult bone marrow in a manner similar to that in fetal lymphoid tissue.
Another potential source of peripheral VH4-utilizing B cells is B cells that initially express VHy or VHx and then undergo a secondary VDJ gene rearrangement with VH4 on the unrearranged IgH allele in germinal centers of peripheral lymphoid tissues. In mouse, RAG1 and RAG2 are expressed in germinal centers (46, 47) and mature B cells can undergo secondary V(D)J gene rearrangement (48, 49). In rabbit, RAG1 and RAG2 expression in appendix was also reported (50), raising the possibility that rabbit B cells undergo secondary rearrangement of Ig genes. We suggest that the majority of germinal center B cells undergoing secondary VDJ gene rearrangement will rearrange VH4, just as we found in the present study that VH4 is preferentially rearranged during fetal VDJ gene rearrangement. Such preferential rearrangement of VH4 during secondary Ig gene rearrangement could help account for the increase in VH4-utilizing B cells in the periphery.
We further suggest that VH4-utilizing mature B cells, regardless of their origin (fetal or adult), selectively expand once they leave the primary lymphoid organs, resulting in the accumulation of peripheral VH4-utilizing B cells with time. These B cells may be selected for expansion because they express a2 epitopes in the VH framework regions. Because we also found a few VH7-utilizing productive VDJ genes at birth, and because VH7 is also likely to express a2 epitopes, we suggest that VH7-expressing B cells accumulate in adult Alicia rabbits through processes similar to those used by VH4-expressing B cells. The hypothesis that a2-expressing B cells are selectively expanded is supported by the finding that in the appendix of Alicia rabbits, the a2+ B cells proliferate more and undergo less apoptosis than a2- B cells (16). Because germline VH4 encodes at least one a2 epitope and still other a2 allotype-associated amino acids can be gained through somatic gene conversion of VH4- and VH7-utilizing VDJ genes, B cells utilizing these genes could be selected in the appendix (51) or other gut-associated lymphoid tissue (GALT), on the basis of a2 epitopes both before and after somatic diversification. The mechanism that could mediate selective expansion of a2+ B cells is not known; however, Pospisil et al. (52) reported that (Fab')2 interacts with B cell-associated CD5 and that a2 (Fab')2 interacts more than does a2- (Fab')2. Selective interaction of CD5 with a2 Ig could contribute to the selective expansion of a2 B cells. It is also possible that interaction through unknown motifs in variable regions that are associated with a2-allotype drives the selective expansion of a2 B cells. Further, if the selective expansion of a2 B cells occurs primarily in GALT, then it may also be that a B cell superantigen is present in the GALT microbial flora and its interaction with a2+ BCR contributes to the selective expansion of a2 B cells (53, 54). Although it is also possible that a2+ B cells are selectively expanded in GALT as a result of the binding of microbial Ags to the Ag binding site, rather than to the a2 FR epitope(s), we think this possibility cannot account for the preferential expansion of a2+ B cells.
In summary, we found that the 3'-most VH gene segment
in Alicia rabbits, VH4, was predominantly
rearranged during fetal development but that by birth, most of the
productive VDJ genes utilized VHx and
VHy, both of which encode VHa
allotype-negative molecules. By adulthood, however,
a2+ B cells that utilized
VH4- and VH7 accumulate
in the periphery. Accordingly, we propose a model in which different
selection mechanisms regulate the B cell repertoire throughout the
rabbit’s life (Fig. 7). The first level
of selection occurs in the fetal liver and bone marrow early in
development, perhaps during B lymphopoiesis when either B cell
precursors utilizing VH4 are selectively
eliminated and/or B cell precursors utilizing
VHy and VHx are
selectively expanded. We suggest that the next level of selection
occurs a few weeks after birth in the secondary lymphoid tissues,
likely in GALT, where B cells expressing the a2 epitope are selectively
expanded. Although we do not know what triggers the selective expansion
of the a2 B cells, we suggest that this expansion is mediated in some
form by the intestinal microflora. This model should be useful for
investigating the molecular and cellular mechanisms that shape the Ab
repertoire in rabbits.
|
|
Footnotes |
|---|
2 X.Z. and A.B. contributed equally to this study.
3 Current address: Department of Microbiology, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand.
4 Address correspondence and reprint requests to Dr. Katherine L. Knight, Department of Microbiology and Immunology, Stritch School of Medicine, Loyola University of Chicago, Maywood, IL 60153.
Received for publication April 6, 1999. Accepted for publication July 8, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
genes as chimeric rabbit-mouse IgA antibodies. Proc. Natl. Acad. Sci. USA 86:7561. light-chain recombination signal sequences mediate recombination more frequently than do those of 
light chain. Proc. Natl. Acad. Sci. USA 88:10722.
chain J cluster. Science 271:1416.[Abstract]
5 protein in B cell development. Cell 69:823.[Medline]
light chain genes in mature B cells in vitro and in vivo: function of reexpressed recombination-activating gene (RAG) products. J. Exp. Med. 187:795.This article has been cited by other articles:
|
|
 |
|
  G. Yang, H. Obiakor, R. K. Sinha, B. A. Newman, B. L. Hood, T. P. Conrads, T. D. Veenstra, and R. G. Mage Activation-induced deaminase cloning, localization, and protein extraction from young VH-mutant rabbit appendix PNAS, November 22, 2005; 102(47): 17083 - 17088. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-J. Rhee, P. J. Jasper, P. Sethupathi, M. Shanmugam, D. Lanning, and K. L. Knight Positive selection of the peripheral B cell repertoire in gut-associated lymphoid tissues J. Exp. Med., January 3, 2005; 201(1): 55 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Esteves, D. Lanning, N. Ferrand, K. L. Knight, S.-K. Zhai, and W. van der Loo Allelic Variation at the VHa Locus in Natural Populations of Rabbit (Oryctolagus cuniculus, L.) J. Immunol., January 15, 2004; 172(2): 1044 - 1053. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Lanning, P. Sethupathi, K.-J. Rhee, S.-K. Zhai, and K. L. Knight Intestinal Microflora and Diversification of the Rabbit Antibody Repertoire J. Immunol., August 15, 2000; 165(4): 2012 - 2019. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. P. Fitzsimmons, K. J. Clark, H. S. Mostowski, and M. A. Shapiro Underutilization of the V{kappa}10C Gene in the B Cell Repertoire Is Due to the Loss of Productive VJ Rearrangements During B Cell Development J. Immunol., July 15, 2000; 165(2): 852 - 859. [Abstract] [Full Text] [PDF]
|
|
| ||||||||
,
EcoRI; 
, HindIII; 
,
HpaI. The distance between
VH4 and VH5 is
VH4-utilizing VDJ gene 4053; 
,
VH1-utilizing VDJ gene
5.2-1 (27); •, VHy-utilizing
(VHa-negative) gene 3719; 
), and rabbit polyclonal a3 Ig (
).
indicates a nonfunctional
VH gene segment because of the presence of a
stop codon in the coding region (VH3,
VH5,
VH6, and
VH8) and a defective promoter region
(VH3).
D, D gene segments. The distance between adjacent
VH gene segments is 