| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
10C Gene in the B Cell Repertoire Is Due to the Loss of Productive VJ Rearrangements During B Cell Development
*
Division of Monoclonal Antibodies and
Division of Cellular and Gene Therapies, Food and Drug Administration, Center for Biologics Evaluation and Research, Rockville, MD 20852
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
10 family of murine light chain Ig genes is composed of
three members, two of which (V10A and V10B) are well used.
V10C, the third member of this family, is not detected in any
expressed Abs. Our previous work showed that V10C is structurally
functional and can recombine, but mRNA levels in spleen were extremely
low relative to those of V10A and V10B. Furthermore, while the
V10C promoter was efficient in B cells, it was shown to work
inefficiently in pre-B cell lines. Here, we extend our analysis of the
V10 family and examine V10 gene accessibility, their
representation in V cDNA phage libraries, and the frequency and
nature of rearrangements during different stages of B cell development.
We demonstrate that V10C is under-represented in V cDNA
libraries, but that the frequency of its sterile transcripts in pre-B
cells surpasses both V10A and V10B, indicating that the gene is
as accessible as V10A and V10B to the recombination machinery. We
also demonstrate that V10C recombines at a frequency equal to that
of V10A in pre-B cells and has a normal nonproductive to productive
recombination ratio. As B cells develop, however, both the frequency of
V10C rearrangements and the presence of productive rearrangements
decline, indicating that these cells are in some fashion being
eliminated. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
light chain locus was
mapped and shown to contain 141 V genes spanning >3 megabases of
DNA. Of these V genes, 74 were functional, while 47 were considered
to be pseudogenes (1). In addition to the 74 structurally
functional genes, 20 V genes were potentially functional (they
contain no defects), but cDNAs have not yet been detected for them. It
is unknown why these genes would not be used.
It has been proposed that the Ig loci must be accessible for
recombination to occur. Although the mechanisms that promote
accessibility are not completely understood, they are thought to
include changes in chromatin configuration, methylation, and
transcriptional activation of the loci (2). Germline
transcripts of VH genes are taken as evidence of
the accessibility of this locus before recombination. Transcripts from
VH families spanning the locus have been detected
in pro- and pre-B cells and germline transcripts from the more
JH proximal families, but not
VHJ558, have been detected in immature and mature
B cells (3). V germline transcripts have also been
detected in pre-B and mature B cell lines (4, 5).
Other factors that can influence the utilization of individual V genes include the promoter and the recombination signal sequences (RSS)2 (1). Slight alterations in VH and VL promoter sequences have been shown to influence transcription efficiencies (6, 7). Similarly, differences in the sequence of the RSS, at the heptamer or nonamer and even the spacer region, have been shown to influence the recombination frequency of individual genes (8, 9, 10).
Orientation of V genes may also influence V gene usage. At least 40%
of functional V genes lie in the opposite transcriptional
orientation to the J region and rearrange by inversion
(11). The mapping of the locus shows that these
inverted genes lie in the center two-thirds of the locus
(1). Because inversional recombination preserves those
V genes lying between the rearranged V and J genes, this
repertoire remains available for receptor editing of rearranged V
genes. These secondary recombinations at the locus can alter V
gene usage, as detected in peripheral B cells.
To more fully understand the mechanisms involved in shaping the
contribution of a V gene to the diversity of the Ig repertoire, we
have analyzed the usage of individual members of the three-gene V10
family. Two genes, V10A and V10B, are known to rearrange by
inversion (11) and are used in Ab responses to a variety
of Ags. The third member of this family, V10C, is structurally
functional but has not been found in an expressed Ab. It was found to
be rearranged to J1 in a reciprocal recombination product
(11). The recent map of the locus shows the V10C
gene to lie between V10A and V10B (5'-V10A-V10C-V10B-3')
in a deletional orientation (12).
Our previous studies of the V10 family have shown that all V10
genes have the ability to recombine in both spleen and bone marrow.
While the V10A and V10B genes were readily expressed in mice,
V10C gene expression was only detected by nested RT-PCR reactions.
Furthermore, we showed that the V10C promoter was less efficient
than that of V10A in pre-B cell lines (13), the point
in development when the locus begins to rearrange. Successful
expression of a mature B cell receptor indicates the transition to the
immature B cell developmental stage. Although it is clear that the
V10C gene can rearrange, we did not analyze the frequency of
recombination or the nature of the recombinations (i.e., productive or
nonproductive) in spleen or bone marrow. In the present study we
analyze recombination frequency of the V10 family and the nature of
the rearrangements in both adult spleen and bone marrow cells, and we
further explore V10C expression in the spleen. We demonstrate that
all three V10 genes rearrange with approximately equal frequency in
pre-B cells before expression of surface IgM, but subsequent to IgM
expression the V10C rearrangement frequency decreases dramatically
in bone marrow and spleen. Furthermore, the ratio of
nonproductive/productive rearrangements is increased in
IgM+ bone marrow and spleen relative to that in
precursor B cells, suggesting that a productive V10C light chain is
either unable to make a mature Ig or undergoes negative selection.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cDNA phage library screening
Two V cDNA phage libraries with 240,000 (library 1) and
700,000 (library 2) phage were provided by Dr. Peter Brodeur and have
been described previously (14, 15). cDNA libraries were
made from poly(A)+ RNA isolated from 3-day LPS
blasts derived from 11- to 15-wk-old BALB/cByJ mice. Each library was
screened with oligonucleotides V10A3 (5'-TTTTGCCAACAGGGTAATAC-3'),
V10B (5'-AGTTGCAGTGCAAGTCAGGG-3'), and V10C
(5'-AGGGCAAGTGAGGACATTAGCAC-3') end-labeled with
[
-32P-dATP (Amersham Pharmacia Biotech, Piscataway,
NJ). The V10C primer differs from the V10A and V10B templates
by 2 and 4 bases, respectively; the V10B primer differs from the
V10A and V10C templates by 2 and 3 bases, respectively; and the
V10A3 primer differs from the V10B and V10C templates by 7
bases. Specific hybridization temperatures for each probe were
determined by hybridizing at successively higher temperatures to nylon
filters containing DNA from phage with copies of germline V10A and
V10C genes. Total C was determined in each experiment by
hybridization with a random-primed pEC probe specific for C. All
oligonucleotides and primers were produced in the Core Facility for
Biotechnology Resources at the Center for Biologics Evaluation and
Research, Food and Drug Administration.
Bone marrow cell sorting
Bone marrow was collected from femurs and tibiae of 12- to 16-wk-old male and female BALB/cAnNCr mice (National Cancer Institute, Division of Cancer Treatment, Frederick, MD). Red cells were lysed with ACK lysis (Biofluids, Rockville, MD) buffer for 3 min on ice, and the remaining cells were washed in PBS and resuspended at 2 x 107 cells/ml in FACS buffer (Biofluids). Cells were stained with FITC-anti-B220 (PharMingen, San Diego, CA) and biotin-anti mouse IgM (Zymed, South San Francisco, CA) for 30 min on ice. After two washes, cells were stained with streptavidin-PE (Molecular Probes, Eugene, OR) for 30 min on ice. Cells were resuspended at 1.5 x 107 cells/ml and sorted on a Becton Dickinson FACStarPlus for IgM-/B220+ cells and IgM+/B220+ cells. All sorts were performed on individual mice, except for one pool of five mice that was used to isolate RNA.
V10 recombination PCRs
Genomic DNAs were isolated from 1.2 x
105 to 5 x 106 sorted
bone marrow cells or total spleen using the Genomic Prep Cells and
Tissue DNA isolation kit (Amersham Pharmacia Biotech)or the Trizol
method (Life Technologies, Gaithersburg, MD) according to the
manufacturer’s instructions. Primers used for amplification were the
Gen 9 primer, which resides in framework 1 and is identical in all
three V10 genes, and the J 5–3 primer site, which is 3' of J5
and will amplify rearrangements that are in the context of C, but
not those that are reciprocal products of VJ recombinations. Genomic
DNA (100 ng) was amplified with the Gen 9 (5'-TCCAGATGACACAGACTAC-3')
and J 5-3 (5'-CTTTTTGCCCCTAATCTCACTA-3') primer pair to amplify
V10 rearrangements to J1, J2, J4, and J5. PCRs were
performed in a DNA thermal cycler 480 (PE Biosystems, Foster City CA)
in triplicate 100-µl reactions with 1.5 mM
MgCl2, 0.05 mM dNTPs, and 50 pmol of each primer
under the following conditions: 95°C for 4 min, 95°C for 1
min/62°C for 1 min/72°C for 2 min (30 cycles), 72°C for 10 min,
and a 4°C hold. Reactions were pooled, precipitated, washed, and
resuspended in a final volume of 35 µl.
Thirty microliters of each sample were electrophoresed through 1.5%
agarose gels for 3–4 h at 150 V, stained for 30 min (1x TAE/0.5
µg/ml ethidium bromide), and photographed. Four microliters of the
PCR products were ligated into the pCR 2.1 TOPO vector (Invitrogen,
Carlsbad CA) according to the manufacturer’s instructions and used to
transform DH5
Escherichia coli. Colonies from
transformations derived from spleens of two mice (identified as mice 6
and 7) were picked for Southern blot analysis, which was performed as
previously described (13).
Recombination colony lifts
Recombination PCRs and cloning were performed with spleen,
IgM+ bone marrow, and IgM-
bone marrow genomic DNA (100 ng) from four mice as described above.
Colony lifts were performed as previously described (16).
Wash temperatures were determined experimentally for each
oligonucleotide probe. V10A3, V10B, and V10C probes were
specific for their targets after washing four times in 20 ml 2x
SSC/0.1% SDS at 51, 54, and 59°C, respectively. V10C colonies
from IgM- and IgM+ colony
lifts were picked for sequencing.
Sterile transcripts
RNA was isolated from IgM- bone marrow cells using the Trizol method (Life Technologies). Residual DNA was removed by digestion with DNase I (Life Technologies) according to the manufacturer’s instructions. cDNA was synthesized using 115 pmol of ST2 primer (5'-GGAGGTTTATGTTATGAC-3'), which binds within the RSS spacer and nonamer, incubated with 1 µg of total RNA at 70°C for 10 min, and placed on ice. Twelve microliters of 5x first-strand buffer, 6 µl of 10 mM dNTPs (1 mM final concentration), 6 µl of 0.1 M DTT (0.01 M final concentration), and 2 µl of Superscript II reverse transcriptase (200 U; Life Technologies) were added to the tube and incubated for 10 min at room temperature, followed by 50 min at 42°C. Reactions were terminated by incubation at 90°C for 5 min, and RNA was removed by digestion with RNase H for 20 min at 37°C. Samples that did not receive reverse transcriptase were included as controls for the presence of genomic DNA.
Sterile transcript PCRs were performed in 100-µl reactions with 1.5 mM MgCl2; 0.05 mM dNTPs; 50 pmol of ST3 primer (5'-GACTTATATCATTGTGGGAGG-3'), which binds within the heptamer/spacer and coding region; 50 pmol of Gen 9 primer; and 2 µl of cDNA under the following conditions: 95°C for 4 min, 95°C for 1 min/53°C for 1 min/72°C for 1 min (30 cycles), 72°C for 10 min, and a 4°C hold. Products were ligated into pCR TOPO 2.1, and transformants were colony lifted as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
10 expression in spleen
Two cDNA phage libraries generated from BALB/cByJ spleen LPS
blasts (14, 15) were screened with V10A, -B, and
-C-specific oligonucleotides and with a probe to detect C-bearing
phage. Table I
shows the results of both
screens. Library 1 contained about 30% C phage, while library 2
contained 50% C phage. V10A and V10B represent 2.9 and 1.7%
of the total C+ phage, respectively. This is
in agreement with our previous data from a semiquantitative PCR, which
showed that these two genes represent from 0.3 to 3.4% of the total
mRNA (13). V10C represents 0.15% of total
C+ phage and thus, is present at levels
10–20 times lower than V10A and -B. Because the cDNA libraries
are amplified, we cannot completely rule out that the frequency of
individual V clones is partly due to unequal amplification. These
results are consistent, however, with our previous semiquantitative
PCR. Therefore, we conclude that in either experiment, V10C is a
minor component of total V10 mRNA expression in adult spleen.
|
10 recombination frequency in splenic B cells
To determine the relative frequency of recombination for each
V10 gene, we performed PCR reactions with primers that would amplify
all V10 recombinations in preparations of spleen genomic DNA from
two BALB/c mice and on spleen and bone marrow DNA from three and four
additional mice, respectively. The 3' primer, J5–3, binds 3' of
J5, and PCR products are the result of amplification of
rearrangements in the context of C and not those of reciprocal
products to VJ joins. Fig. 1 shows the
percentages of recombinants for each gene for the individual mice.
Fifty-seven and 77 V10 recombination clones from spleen of mice 6
and 7, respectively, were analyzed by Southern hybridization to
oligonucleotide probes specific for each of the V10 genes (Table II.). V10A recombination products were
dominant (64% of the total V10 clones) followed by V10B
recombinations (25% of the total V10 clones). Seven V10C clones
were detected in each mouse, representing 10% of the total V10
clones examined. Rearrangements to all four J elements (Table II)
were observed for V10A and V10B from both mice. V10C
rearrangements to J1 were not detected. This was unexpected, as we
previously analyzed V10CJ1 cDNA sequences. It is likely that
analysis of more clones would show such recombinations. This analysis
shows that V10C recombinations are detected in the spleen, but at
lower levels than V10A and V10B. This lower frequency, however,
is not low enough to explain the extremely low levels of mRNA expressed
in the spleen, as measured here by the screening of amplified cDNA
libraries with a V10C-specific oligonucleotide or previously by
semiquantitative PCR (13).
|
|
Clones for each V10 gene were sequenced and assessed for the
nature of the rearrangement, e.g., productive or nonproductive.
Rearrangements to each J were sequenced to ensure an analysis of
unique recombinations. Rearrangements to the same J genes were
included in the analysis only if the sequence at the VJ junction
was different from other recombinations using the same V10J pair.
The results shown in Table III are
presented as the ratio of nonproductive rearrangements to productive
rearrangements (NP/P). The NP/P ratios are 0.08 and 0.62, respectively,
for V10A and V10B. This indicates that productive recombinations
are more numerous than nonproductive recombinations. In contrast, the
NP/P ratio for V10C is 4.5. Of 11 analyzed V10C sequences, 9 were
nonproductive. Thus, the majority of V10C rearrangements detected in
the spleen are nonfunctional. This together with the lower levels of
V10C rearrangements explain the low levels of mRNA detected in the
spleen.
|
10 sterile transcripts in bone marrow
Because V10A and V10B recombinations are more frequent than
V10C in the spleen, we wanted to assess the ability of the genes to
recombine in precursor B cells. Accessibility of a locus is required
for recombination to occur, and sterile germline transcripts from V
genes are markers of accessibility. We examined the frequency of V10
sterile transcription products in an RT-PCR assay that used RNA from
sorted IgM-/B220+ bone
marrow cells. This population contains all early precursor B cells,
including the small resting pre-B cell where rearrangement begins.
A generic 5'-V10 PCR primer and a 3'-primer that binds within the
RSS of the V10 genes were used to amplify V10 sterile
transcripts. Cloning of PCR products and colony lift analysis of 1027
V10 sterile transcript transformants with V10A, V10B, and
V10C probes revealed that sterile transcripts of all three V10
family members were present (Table IV).
V10C had the highest representation (45.4%), followed by V10A
(35.2%) and V10B (19.5%). Therefore, the V10C gene is readily
accessible for recombination.
|
10 recombination in sorted bone marrow cells and total spleen
Before a developing B cell expresses a fully functional B cell
receptor on its surface, selective pressure based on expression of the
light chain is absent. Light chain genes with equally accessible loci
would be expected to recombine at similar frequency in such an
environment. Because the V10 genes appeared to be equally
accessible, as evidenced by the presence of sterile germline
transcripts in pre-B cells, we wanted to determine whether the
frequency of recombination of each gene in IgM-
pre-B cells was similar. We examined the frequency of V10
recombination in bone marrow cells from four individual mice sorted
into the pro/pre-B population
(IgM-/B220+) and the
immature B cell population
(IgM+/B220+) from four
BALB/c mice. Splenic B cells from three of these mice were also
included in the analysis. Genomic DNA was prepared from these
populations, and V10 PCRs were performed as previously described. As
before, the PCR was designed to amplify rearrangements to all J
genes in the context of C (Fig. 2A). PCR products were cloned,
and colonies were hybridized under conditions specific for each V10
gene. Rearrangements of all four J elements were present in each DNA
preparation. As Fig. 2B and Table V show, the percentages of V10C and
V10A recombinants in
IgM-/B220+ pre-B cells
were identical at 40% of the 2133 clones analyzed, while V10B
recombinants were detected at half this frequency. Interestingly, these
percentages are very similar to the percentage of V10 sterile
transcripts shown above. In PCR products from immature
IgM+/B220+ B cells, the
frequency of V10C recombinants decreased from 40.9 to 18.8%, while
that of V10A increased from 39.5 to 57.4%. The percentage of
colonies carrying V10B recombinations remained approximately the
same (19.5 vs 23.8%). A similar pattern was evident in the spleen,
with V10A, V10B, and V10C recombinations present in 61.5,
27.1, and 11.4%, respectively, of the 5156 colonies examined. These
data suggest that V10C can recombine as efficiently as V10A or
V10B in pre-B cells that do not express surface IgM, but that such
recombinations are lost as the B cell matures.
|
|
10C rearrangements in the bone marrow
We showed that most of the V10C rearrangements in the spleen
were nonproductive. Because V10C is fully competent to recombine in
pre-B cells, and these recombinations are lost upon B cell development,
it was of interest to determine the nature of the recombinations in the
IgM-/B220+ pre-B cell and
IgM+/B220+ immature B cell
populations. As before, V10C rearrangements to each J were
sequenced to ensure an analysis of unique recombination events, and
rearrangements to the same J gene were included in the analysis only
if the sequences at the V10CJ junction were different. Only
V10C recombinations were analyzed from the bone marrow populations.
In the pre-B cell population, it was expected that only one of three
rearrangements would be productive, and two of three would be
nonproductive. This was seen for V10C in the
IgM-/B220+ population, as
the NP/P ratio was 2:1 (Table III
). Remarkably, this ratio increased to
3.25:1 in the IgM+/B220+
population and to 4.5:1 in the spleen. These data combined with the
sterile transcript and frequency of V10C recombination data suggest
that there are no recombination defects in the V10C gene. The
possible reasons for its underutilization in the B cell repertoire are
discussed below.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 x
107 IgM+ immature B cells
produced in the bone marrow each day do not migrate to the periphery
(29, 30).
In this study we have extended our analysis of the underutilization of
the V10C gene. We previously reported that this gene is structurally
functional and is capable of recombination, but is expressed at levels
at least 1000-fold lower than those of V10A and V10B in adult
BALB/c spleen (13). Indeed, this light chain has never
been detected in a functional Ab. Additionally, the V10C promoter
was shown to function significantly less efficiently in pre-B cells
than the V10A promoter, but the major promoter element, the octamer,
is identical in both genes. Here we show that the V10C gene is
accessible for recombination and recombines as efficiently as other
family members, but productive rearrangements are lost as B cell
development progresses.
It is important here to discuss our findings in light of the recent
analysis of the V locus by Zachau and his colleagues (1, 12, 31), which includes in the definition of pseudogenes those
sequences with deviations from the canonical octamer promoter element,
the canonical heptamer of the RSS, or drastic deviations from the RSS
nonamer. The V10A sequence (ce9) (12) is considered
functional, V10B (cp9) (12) is considered to have RSS
defects but cDNAs are detected, and V10C (by9) (12) is
considered to be nonfunctional due to a defect in the RSS. Our previous
analysis of the V10 family as well as that of Zachau’s group show
that all three genes have identical RSSs including a noncanonical
heptamer. As shown in Fig. 3, all V10
germline sequences, except AJ1, from a variety of inbred strains have
identical RSSs (3). We believe that the difference in the
last base of the AJ1 nonamer is probably due to sequencing error. As
V10A and V10B are clearly functional genes, the noncanonical
heptamer sequence in V10C should not define it as a pseudogene. In
addition, the V10 family, which contains three strongly hybridizing
bands upon Southern analysis, has now been combined with the V9
family into a larger heterogeneous group. Each member of this combined
family has at least 80% sequence similarity with some family members
while sharing <80% similarity with others. The three V10 genes are
94–97% homologous with each other (1, 13), so we
continue to consider this a separate three-member family.
|
10C gene in naive B
cells, we analyzed the frequency of V10C recombinations relative to
V10A and V10B in bone marrow populations and the spleen, and we
examined the accessibility of each of these genes to the recombinase
machinery. The presence of Ig gene germline sterile transcripts is
often correlated with accessibility for recombination and frequency of
rearrangement (2, 7). We used the frequency of V10A,
V10B, and V10C sterile transcripts as markers of accessibility.
Interestingly, V10C sterile transcripts were present at the highest
frequency (45.4%) compared with V10A (35.2%) and V10B (19.5%),
indicating that V10C is as accessible in pre-B cells, if not more
so, than the expressed members of this family.
To determine the status of V10 genes during B cell development, we
examined the frequency and nature of V10 gene recombinations in
total spleen, bone marrow
IgM+/B220+ immature B
cells, and IgM-/B220+
pre-B cells. The PCR assay was designed to detect rearrangements only
in the context of C and not reciprocal products to VJ joins. In
cells not subjected to selection, i.e.,
IgM-/B220+ pre-B cells,
one would expect that recombination frequencies would be similar among
individual members of a family of genes that are equally accessible and
have identical RSS. Indeed, the
IgM-/B220+ pre-B cells
showed an overall V10C recombination frequency identical with that
of V10A, but 2-fold higher than that of V10B. These numbers
closely approximated the frequency of detection of V10A, V10B,
and V10C sterile transcripts found in pre-B cells and suggest that
the recombination frequency of these genes in the absence of selection
correlates with sterile transcription. Furthermore, both V10A and
V10C are more accessible and recombine more frequently than
V10B.
In the IgM+ B cell population from the bone
marrow, the V10C recombination frequency dropped to around 18% of
total Vk10 recombinations with a resulting increase in the frequency of
both V10A and V10B recombinations. In total spleen, there was a
further diminution of the V10C recombination frequency to 11.4% of
the total V10 recombinations.
Examination of the nature of V10C rearrangements from spleen,
IgM+/B220+, and
IgM-/B220+ bone marrow
cells revealed that the NP/P increased as developmental stage of the B
cells progressed from pre-B cells to immature B cells to splenic B
cells. Based on reading frames, one would expect an in-frame
recombination to occur one of three times. This 33% frequency of
productive light chain rearrangements has been observed for many other
genes in both bone marrow and fetal liver (15, 32, 33). In IgM-/B220+
pre-B cells the ratio of nonproductive to productive V10C
rearrangements was 2:1, indicating that, as expected, one-third of
the recombinations were in-frame. This ratio climbed to 3.25:1 in
immature B cells and to 4.5:1 in spleen. These data suggest that
productive V10C rearrangements are selectively lost during B cell
development. Together, the overall decrease in frequency of V10C
rearrangements and, more importantly, the specific loss of productive
V10C rearrangements in the spleen explain the lack of expression of
this gene in the Ab repertoire.
There are three explanations for the loss of productive V10C
rearrangements during the pre-B to immature B cell transition. First,
we previously showed in vitro that the V10C promoter is less
efficient than that of V10A in pre-B cells. Therefore, if not enough
L chain protein can be synthesized, mIgM may not be expressed. The
second possibility is that V10C L chains cannot make effective
VH-VL pairs. In either
case, if no mIgM is expressed, secondary recombinations would occur
that delete V10C rearrangements or remove them from the context of
C. The third possibility is that V10C L chains paired with H
chains encode mIgM, which undergoes negative selection. Receptor
editing would then eliminate the V10C rearrangement.
(Mechanistically, receptor editing is the same as secondary
recombination. For the purposes of this discussion, secondary
recombination refers to events that occur before successful expression
of mIgM, while receptor editing refers to events that occur
thereafter.) All three possibilities would have the same outcome of
loss of the productive V10CJ rearrangement in the
IgM+ cell population. Evidence suggesting that
only one allele at a time undergoes rearrangement (34)
supports the idea that the V10CJ rearranged allele would undergo
secondary recombination before initiating recombination on the other
allele. We would predict that productive V10C rearrangements in the
IgM+ bone marrow population or in the spleen
could be detected in 
+ B cells with RS
rearrangements or in rearrangements in the context of reciprocal
joints. Alternatively, they could reflect a low level of productive
V10C-expressing Ab.
Presently there is no evidence to favor any of the three possibilities,
but all are testable. The first two possibilities, however, are
selection neutral and would not be viewed differently by the B cell
than if the initial recombination event was nonproductive. That is, if
mIgM was not expressed due either to a nonproductive rearrangement or
to a productive rearrangement that did not result in a mature IgM,
selection could not occur, and recombination would continue. Secondary
recombination on the same allele would result in the loss of the
productive V10C rearrangement. This accounts for the reduced
frequency of all V10C rearrangements in immature B cells but does
not explain the selective loss of productive rearrangements. Negative
selection of mIgM containing V10C light chain could explain the
selective loss of these productive rearrangements.
Additional arguments can be made against each possibility. 1) While our
in vitro data suggest lower levels of V10C transcription in pre-B
cells (13), the current analysis of sterile transcripts
suggests that the V10C promoter is equally efficient as V10A. It
is not known, however, how promoter efficiency would be influenced
subsequent to recombination after which the promoter is brought into
the vicinity of the IgK enhancers. 2) Several reports have suggested
that heavy and light chains have an intrinsic ability to pair without
bias (35, 36, 37). In contrast, it has recently been shown
that a particular heavy chain is incapable of pairing with many light
chains (38). It is possible that V10C L chains cannot
form VH-VL pairs with many
H chains. We do not think this is likely, as V10C protein has only
three amino acid residues that are different from both V10A and
V10B. Two of these differences are in complementarity-determining
region 1, and the third is a Thr to Ala change in framework 3
(13). Although this change is unusual, it has been
detected in at least one V10A-expressing anti-DNA Ab
(39). We also previously observed that many productive
V10CJ1 cDNA junctions lacked the invariant
Pro95 residue. All junctions analyzed in the
current study contained this proline. Because these junctions are
derived from DNA, we do not know whether they are transcribed. It
remains a possibility that productive V10C recombinations lacking
Pro95 that are expressed cannot form proper
VH-VL pairs. 3) Although
negative selection of mIgM-expressing V10C L chains is an attractive
possibility, we would have to argue that the V10C protein structure
alone contributes to negative selection, such that there is no
VH protein with which it can be expressed in the
periphery. Additionally, V10A and V10B are often components of
autoantibodies from autoimmune mice (39, 40, 41, 42, 43), but V10C
has not been detected in these Abs either. It would be expected that on
genetic backgrounds incapable of delivering negative signals, L chains
used in autoantibodies would be expressed.
The V10C gene is not the only apparently functional gene not used in
the Ab repertoire. cDNAs for 20 potentially functional genes have not
been detected (1), and there may be other V genes
defined as pseudogenes that are, in fact, similar to V10C. It
remains to be determined why 14% of the V repertoire is only
potentially functional and not expressed.
|
Acknowledgments |
|---|
cDNA libraries, and Drs.
Steven Bauer, Kurt Brorson, and Rose Mage for critical review of this
manuscript. |
Footnotes |
|---|
2 Abbreviations used in this paper: RSS, recombination signal sequence; RS, recombining sequence; NP/P, ratio of nonproductive rearrangements to productive rearrangements; mIgM, membrane IgM.
3 In our previous paper (13) we compared our V10C sequence to AJ1 and AJ2 and noted the 1-base difference between the AJ1 nonamer and those of V10C and AJ2. We also mistakenly printed this difference in the wrong position, at the base following the nonamer sequence. Nonetheless, differences in V10 gene usage cannot be due to RSS differences.
Received for publication January 21, 2000. Accepted for publication April 25, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
locus. Eur. J. Immunol. 29:2072.[Medline]
gene rearrangement correlates with induction of germline gene transcription. Cell 58:1001.[Medline]
genes in the BLIN-1 human pre-B cell line correlates with germline J-C and V transcription. J. Exp. Med. 173:639. gene segment is associated with reduced germ-line transcription and a low frequency of rearrangement. J. Immunol. 154:1748.[Abstract]
usage in vivo. J. Exp. Med. 187:1495. genes rearrange. J. Immunol. 139:3834.[Abstract]
locus. Eur. J. Immunol. 29:2065.[Medline]
exons: differences in the utilization of V10 exons. J. Immunol. 161:2290. families and V exons: implications for the available B cell repertoire. J. Immunol. 147:3194.[Abstract]
repertoire of the mouse: an analysis of the usage of two V gene segments. J. Immunol. 160:4904. genes, pseudogenes, relics and orphons in the mouse genome. Eur. J. Immunol. 29:2082.[Medline]
L chain-deficient B lineage cells. Immunity 11:317.[Medline]
light chain rearrangement in mouse fetal liver. J. Immunol. 153:1150.[Abstract]
light-chain variable gene families occurs in polyclonally activated B cells. Proc. Natl. Acad. Sci. USA 87:4932. genes encode predominating antibody variable regions elicited in strain A mice by immunization with p-azophenylarsonate. J. Exp. Med. 166:1.10 genes encoding L chains of antibodies bearing the Ars-A and A48 cross-reactive idiotypes. Immunogenetics 34:231.[Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
