| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Laboratoire d’Immunologie, Centre National de la Recherche Scientifique, Faculté de Médecine, Limoges, France; and
Centre de Développement des Techniques Avancées pour l’Expérimentation Animale, Centre National de la Recherche Scientifique, Orleans, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Within the Eµ enhancer, the core element was shown to be necessary for complete V(D)J rearrangement and expression of the locus, while its flanking 5' and 3' matrix-associated region (MAR)3 were dispensable (7). By contrast, partial DJ rearrangements persisted even upon deletion of the whole enhancer (8, 9).
Several targeted insertions of a neomycin (neo) resistance cassette in the JH-Cµ intron resulted in complex phenotypes, which accommodate the idea that insertions downstream to Eµ do not significantly affect IgH recombination and expression (9, 10). By contrast, a neo insertion within Eµ, in between the 5' MAR and the core enhancer, disrupted the architecture of the intronic enhancer and altered both accessibility to V(D)J recombination and germline JH transcription (9). These alterations were reported in heterozygous mutant B cells from chimeric mice, but not from homozygous mutant animals (9). In addition, it was not clear whether this phenotype mainly resulted from the neo insertion or from disruption of the architecture of the enhancer element.
The analysis of neo insertions may greatly aid in understanding how multigene loci are regulated and how a limited number of enhancer elements manage to cooperate with multiple promoters. We therefore introduced a mutation by which the Eµ element and its flanking MARs remain intact, whereas a thymidine kinase (tk) promoter-driven neo gene is inserted in between the JH4 segment and Eµ. We then checked whether the inserted gene and its promoter were able to modify accessibility of the JH region to recombination and/or IgH locus transcription.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plasmid pUC18-JH, containing a 6.1-kb
EcoRI genomic fragment of the mouse IgH locus, was digested
with NaeI, and a neomycin resistance gene
(hsvTK-neor), flanked by loxP
sites, was inserted (see map on Fig. 1
A). The construct was
linearized before transfection using the unique pUC-derived
NdeI site. Cells of the embryonic stem (ES) cell line CK35
(provided by Dr. C. Kress, Institut Pasteur, Paris, France) were
transfected by electroporation and selected using G418 (400 µg/ml).
BamHI restriction and Southern blot analysis with an
external 3' probe (CH1 µ probe, a 0.4-kb
BglII-BamHI genomic fragment) identified
recombinant clones showing homologous recombination (Fig. 1B). Two ES cell clones were injected into C57BL/6
blastocysts, and the resulting male chimeras were mated with C57BL/6
females. Germline transmission was assessed by coat color, and the
presence of the tk-neor knockin IgH allele
was checked by Southern blot and PCR (Fig. 1C). In parallel,
mutant homozygous mice were mated with EIIa-Cre transgenic
mice (a gift from Dr. H. Westphal, used under a noncommercial research
license agreement from DuPont, Wilmington, DE) (11). The
progeny were checked by PCR (NaeI PCR and neo PCR
(32 cycles at 55°C) as shown in Fig. 1C) for the
occurrence of a Cre-mediated deletion, and this was
confirmed by Southern blot, yielding a 10.5-kb BamHI band
(positive by the neo PCR) for the neo knockin and
a 9-kb BamHI band (negative by the neo PCR and
showing a knockin loxP element by the NaeI PCR)
for the Cre-deleted allele (Fig. 1, B and
C). In the NaeI PCR, the amplified product goes
from 152 (wt) to 367 bp (
/) due to the
presence of the loxP site and flanking polylinker regions
from the targeting plasmid.
|
Single-cell suspensions from peripheral blood or lymphoid tissues were washed in PBS/5% FCS and stained (5 x 105 cells/assay) with various Abs: anti-B220 conjugated with SpectralRed, anti-c-Kit, and anti-CD43 (BD Biosciences, Heidelberg, Germany) and anti-CD25 (Beckman Coulter, Brea, CA) conjugated with PE; anti-IgM conjugated with FITC (Jackson ImmunoResearch, West Grove, PA); and anti-IgMa conjugated with FITC (a gift from Dr. C. Berek, University of Berlin, Berlin, Germany). Cells were analyzed on a Coulter XL apparatus (Beckman Coulter, Fullerton, CA).
Spleen cell cultures and ELISA assays
Single-cell suspensions of spleen cells were cultured 4–5 days at 6 x 105 cells/ml in RPMI medium supplemented with 10% FCS and 20 µg/ml LPS.
Sera from heterozygous mutant mice, homozygous mutant mice, and normal littermates were analyzed for the presence of the various Ig classes and subclasses by ELISA. All Ig evaluations were performed in duplicate. ELISAs were performed in polycarbonate 96-multiwell plates (Maxisorb; Nunc, Roskilde, Denmark), coated overnight at 4°C (100 µl/well) with suitable capture Abs diluted in 0.05 M sodium bicarbonate buffer (2 µg/ml for IgM, IgG1, IgG2a, and IgG2b; 3 µg/ml for IgG3; and 4 µg/ml for IgE and IgA). After blocking and washing steps, 50 µl of sera (first diluted to 1/50) or isotypic standard Ig were diluted into successive wells in 1% BSA/PBS buffer and incubated for 2 h at 37°C. The mouse standard panel included antisera specific for IgM, IgG1, IgG2a, IgG2b, IgG3, and IgA (Southern Biotechnologies, Birmingham, AL) and for IgE (BD PharMingen, San Diego, CA). After washing, 100 µl/well appropriate conjugated Abs diluted in PBS were added and adsorbed during 1.5 h at 37°C. Alkaline phosphatase-conjugated goat antisera specific for mouse Ig classes (Southern Biotechnologies) were used in 0.1% Tween 20/PBS buffer, either at 1 µg/ml for IgM, IgG1, IgG2a, IgG2b, and IgG3 or at 2 µg/ml for IgE and IgA. After washing, phosphatase alkaline activity was assayed on 1 mg/ml alkaline phosphatase substrate (Sigma-Aldrich, St. Louis, MO) and blocked with addition of 3 M NaOH; optic density was measured at 400 nm in a Spectracount photometer (Packard, Meriden, CT).
Generation of hybridomas
Splenic B lymphoblasts from heterozygous mice and myeloma cells
were mixed (ratio 3/1) by adding the appropriate volume of myeloma cell
suspension to the lymphoblasts as previously described
(12). The cell pellet was loosened. The following steps
were performed at 37°C. One milliliter of PEG-1500 solution (Roche,
Mannheim, Germany) was added dropwise. Then serum-free medium was
slowly added during 5 min; the suspension was incubated for a further 5
min at 37°C and was centrifuged. The supernatant was discarded, and
the cells were resuspended in prewarmed selection medium: RPMI 1640,
20% FCS, 100 µM hypoxanthine, 1 µg/ml azaserine, 150 µg/ml
oxalacetate, 50 µg/ml pyruvate, and 8.2 µg/ml insulin
(Sigma-Aldrich). The fusion was distributed onto 96-well plates
prepared with peritoneal macrophages that had been isolated 3 days
previously from an unimmunized mouse and seeded at 
6 x
104 cells/well in selection medium. Selection
took place over at least 10 days.
V(D)J recombination analysis
The Southern blot strategy described by Chen et al.
(9) was used to score
DH-JH recombination.
Hybridoma DNA was digested by SstI and transferred on nylon
sheets (Amersham, Little Chalfont, U.K.). Blots were hybridized with
two probes, as shown in Fig. 1
A (neo probe, a
1.0-kb EcoRI fragment encompassing the neo-coding
region; DJ-IVS probe, a 380-bp
ApaI/SstI genomic fragment located within the
DQ52/JH1 intervening sequence).
For analysis of VDJ recombination, bone marrow DNA from homozygous and heterozygous mutant mice was analyzed by PCR (35 cycles at 60°C) with consensus forward primers specific for the VHJ558 and VH7183 gene families. Backward primer N3 was complementary to the neo gene and thus specifically amplified rearrangements on the neo knockin allele. Amplified products (expected sizes ranking from 450-1900 bp for VDJH4 to VDJH1) were cloned into the TopoI vector (Invitrogen, Groningen, The Netherlands) and sequenced. Alternatively, a J4 primer (complementary to sequence 44–66 bp downstream JH4) simultaneously amplified rearrangements of various JH segments on both wild-type and mutated alleles.
Transcription analysis
For Northern blots, total bone marrow and spleen RNA was prepared using TriPure (Roche, Mannheim, Germany). Northern blotting was conducted by migrating 10 µg of total RNA on a 1% agarose denaturing gel, followed by transfer on nylon sheets (Amersham). The probes used for hybridization were the following: for Cµ transcripts, a 1.2-kb XbaI-HindIII genomic fragment containing the murine Cµ1 to Cµ3 region; for actin transcripts, a 0.45-kb genomic fragment containing exons 4 and 5; and for neo transcripts, the 1.0-kb fragment encompassing the neo-coding sequence.
RT-PCR was conducted on DNase I (Invitrogen)-treated RNA and was checked to be negative in the absence of RT to rule out contamination by genomic DNA. RT was conducted for 1 h with Superscript II (Invitrogen), starting with 1 µg of total RNA.
For germline transcripts, amplification was performed for 32 cycles, hybridization was at 55°C for Iµ transcripts (Iµ/Cµ primers; expected size, 245 bp for amplified products), for germline DQ52 transcripts (GL DQ52/Cµ primers; expected size, 784 bp; hybridizing with the DJ-IVS probe), for VHJ558 germline transcripts (VHJ558/VHRT primers; expected size, 286 bp; hybridizing with a VHJ558 probe), for VH7183 germline transcripts (VH7183a/VH7183 RT primers; expected size, 185 bp; hybridizing with a VH7183 probe), and for VHS107 germline transcripts (VHS107/VHS107 RT primers; expected size, 250 bp; hybridizing with a VHS107 probe).
PCR was performed under similar conditions (hybridization at 55°C)
for the detection of all rearranged DJ transcripts: mature DQ52-Cµ
transcripts (5'DQ52/Cµ primers; expected size, 105 bp), primary
DQ52-JH3 transcripts (5'DQ52/J3 primers; expected
size, 350 bp), mature D-Cµ transcripts other than DQ52 (DHL/Cµ
primers, with the DHL primer defined as a consensus recognizing
multiple D elements except DQ52 (13); expected size,
125 bp), primary D-JH3 transcripts other than
DQ52 (5'DHL/J3 primers; expected size, 370 bp)
PCR was performed with hybridization at 57°C for rearranged mature
VHJ558DJ-Cµ transcripts
(VHJ558/Cµ primers; expected size, 365 bp)
and for rearranged mature VH7183DJ-Cµ
transcripts (VH7183b/Cµ primers; expected size,
175 bp), at 56°C for rearranged primary
VHJ558-JH2 transcripts
(VHJ558-J2 primers; expected size, 340 bp) and
for rearranged primary
VH7183-JH2 transcripts
(VH7183b/J2 primers; expected size, 150 bp),
and at 59°C for rearranged primary
VHJ558-JH3 transcripts
(VHJ558/J3 primers; expected size, 400 bp;
hybridizing with a 850-bp BamHI-NaeI probe
encompassing JH3 and JH4)
and for rearranged primary
VH7183-JH3 transcripts
(VH7183b/J3 primers; expected size, 210 bp,
hybridizing with the same probe encompassing JH3
and JH4).
PCR was performed with hybridization at 55°C for neo transcripts
(N1/N2 primers; expected size, 335 bp) and at 50°C for 
-actin
(Acti4/Acti5 primers; expected size, 260 bp corresponding to spliced
transcripts from exons 3 and 4 of the -actin gene) used as an
internal control.
DNA methylation analysis
Five individual hybridomas were first identified by Southern blot (see above) as carrying a germline JH/neo knockin allele. Genomic DNA was restricted to completion with HindIII and then submitted to a second digestion with either MspI or HpaII. After Southern blotting, nylon filters were hybridized with the DJ-IVS probe. A methylated pattern was characterized by the presence on an uncut 2240-bp HindIII band after HpaII restriction.
Estimation of VDJ rearrangement frequency of both IgH alleles in
wt/ mice
The abundance of VDJ rearrangements on both alleles among
wt/ splenocytes was estimated by comparing
VDJH4 segments with two different sizes amplified
from both alleles. The VH7183b-3'NaeI
PCR (32 cycles at 59°C) strategy used for this assay was designed by
choosing a backward primer located downstream from the NaeI
site normally located 200 bp downstream from JH4.
The presence of a knockin loxP element resulted in a 515-bp
amplified fragment instead of 300 bp for rearrangements of the
wild-type allele.
PCR primers
The following primers were used: 5' NaeI, 5'-TCAGGTAAGAATGGCCTCTCC-3'; 3' NaeI, 5'-TCCTAAAGGCTCTGAGATCCC-3'; Iµ, 5'-CTCTGGCCCTGCTTATGTTTG-3'; Cµ, 5'-GAAGACATTTGGGAAGGACTGACT-3'; VHJ558, 5'-CTSAGCTKGTKARSCCTGGGGCTT-3'; VHRT, 5'-TGWGGTWSYMWCACTGTG-3'; VH7183a, 5'-CAACCATTAGTAGTGGTGGTAG-3'; VH7183b, 5'-CGGTACCAAGAASAMCCTGTWCCTGCAAATGASC-3'; VH7183RT, 5'-TTTCCTCATTGTGTCTCTTGC-3'; VHS107, 5'-CTTCTGGGTTCACCTTCAG-3'; VHS107RT, 5'-GTCCTCTCACTGTGTGCAT-3'; DHL, 5'-GGAATTCGMTTTTTGTSAAGGGATCTACTACTGTG-3'; 5'DQ52, 5'-CGGAGCACCACAGTGCAACTG-3'; GL DQ52, 5'-CAACTGGGACCACGGTG-3'; JH2, 5'-CTGTGAGAGTGGTGCCTTGG-3'; J3, 5'-GTCTAGATTCTCACAAGAGTCCGATAGACC-3'; J4, 5'-TCTGCAATGCTCAGAAAACTCC-3'; N1, 5'-GGATGATCTGGACGAAGAGCAT-3'; N2, 5'-TCCCCTCAGAAGAACTCGTCAA-3'; N3, 5'-TCTTGACGAGTTCTTCTGAGGGGATCGGCA-3'; Acti-4, 5'-TACCTCATGAAGATCCTCA-3'; and Acti-5, 5'-TTCGTGGATGCCACAGGAC-3'.
Degenerate nucleotide positions are coded as follows: R = A or G, Y = C or T, S = C or G, M = A or C, and W = A or T. Primers for VHJ558, VH7183b, VHRT, DHL, and JH3 were adapted from Ref. 13 ; Iµ and Cµ primers were adapted from Ref. 14 .
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A gene-targeting vector was constructed to insert a neo
resistance cassette at the NaeI site located 200 bp
downstream from JH4 (see map in Fig. 1A). The neo gene was driven by a tk
promoter and was flanked by two loxP sites. This construct
allowed the growth of geneticin-resistant ES cell clones. Four of 480
ES clones were selected because they gave hybridizing fragments of the
size expected for the
JH-neo-Eµ (N)
mutation (i.e., insertion of the neo gene) by Southern
blotting and hybridization with a probe located outside of the
construct (Fig. 1B and data not shown).
Heterozygous N mutant ES clones were injected into C57L/B6
blastocysts, which were implanted into foster mothers to derive somatic
chimeras. Chimeras were bred with either wild-type (wt)
animals to obtain heterozygous (wt/N)
F1 mice carrying the heterozygous
JH-neo-Eµ mutation, or with
EIIa-Cre transgenic 129sv mice (11) to yield
heterozygous JH-neo-Eµ
animals (wt/) after Cre deletion of the
neo gene. Further breeding coupled with PCR and Southern
blot testing gave two homozygous lines of JH-neo-Eµ
(N/N) and JH-neo-Eµ (/)
mutant animals (Fig. 1C and data not shown).
B cell development in mutant animals
Pathological studies of tissues from homozygous animals and their
littermates were conducted. Spleens of reduced size devoid of any
germinal centers and the absence of Peyer’s patches were noticed in
N/N, but not / mutant, animals. By flow
cytometry, / animals could not be distinguished from
wt with regard to B cell compartments (not shown). By
contrast, when bone marrow from mutant N/N animals was
checked by flow cytometry, double staining with anti-B220 and
anti-µ antisera showed the lack of surface
IgM+ B cells (Fig. 2A). B cells were also absent
from spleen and peripheral blood (Fig. 2B and not shown).
N/N bone marrow featured an expanded
B220+, c-Kit+,
CD43+, CD25-,
sIgM- pro-B compartment, while the
B220+, CD43-,
CD25+ pre-B cell compartment appeared to be empty
(Fig. 2A and data not shown).
|
/, animals (Fig. 3). IgG serum levels were conserved in
F2 homozygous animals born from heterozygous
females due to the persistence of maternal Abs over 6 mo (data not
shown); however, all serum Ig classes fell to undetectable levels in
F3 animals born to homozygous
F2 N/N females (Fig. 3).
Neo insertion thus resulted in B cell development arrest at
the pro-B cell stage.
|
Whether such a defect in surface IgM expression reflects the
absence of productive V(D)J rearrangements or blocked transcription of
the Cµ gene was first explored by RT-PCR and Northern blot
analysis (Fig. 4). Germline transcription
is usually considered an initial landmark of accessibility of the IgH
locus at the pro-B cell stage, with germline DQ52 transcripts, the
so-called µO transcripts, being markers of transcriptional competence
in the germline DQ52/JH region (15).
Since mutant animals revealed an accumulation of cells blocked at the
pro-B cell stage, the presence of IgH germline transcripts was first
assayed in bone marrow from N/N animals. Iµ transcripts
and germline VH (VHJ558,
VH7183, and VHS107)
transcripts were readily detected in homozygous N/N mutant
bone marrow pro-B cells. By contrast, germline DQ52 transcripts were
undetectable in all 13 homozygous mice assayed (Fig. 4B and
data not shown). Dµ transcripts corresponding to transcription of
partial DJ rearrangements were also undetectable as both mature and
primary transcripts whether they included DQ52 or D segments other than
DQ52 (Fig. 4B).
|
, B and C). Neo transcripts in homozygous and heterozygous animals
Neo transcription was detectable at a low level only by
RT-PCR in bone marrow from all homozygous N/N mice and in
lymphoid tissues from both homozygous and heterozygous animals (Fig. 4
and data not shown). It appeared to be regulated in parallel to VDJ
transcription in normal B cells, since it was induced by LPS
stimulation in cultured heterozygous wt/N splenocytes,
reaching a threshold of detection by Northern blot (Fig. 4C). It was also up-regulated in about half the
neo gene-positive wt/N hybridomas obtained by
fusion with plasmacytoma cells (Fig. 4). Neo transcription
thus contrasted with the silencing of JH and VDJ
segments. Incidentally, it also indicated that, even in heterozygous B
cells expressing a wt IgH locus, the allelically excluded
and targeted IgH allele had not undergone irreversible long-range
inactivation and could be transcribed in activated B cells.
Untranscribed VDJ rearrangements of the IgH knockin allele reveal altered D and JH usage
Since both germline transcripts from the JH
region (DQ52 germline transcripts) and mature V(D)J transcripts were
undetectable in N/N animals, it was questionable whether the
JH region was able to undergo V(D)J
rearrangements in those animals. Bone marrow DNA was thus assayed to
detect rearrangements of the VH7183 family gene
segments on JH segments by using a PCR either
amplifying any VDJ rearrangement (3'JH4 backward
primer) or specific for the neo knockin allele (N3 backward
primer). Complete rearrangements of the mutated allele were readily
detected in both homozygous N/N and heterozygous
wt/N animals at sizes corresponding only to rearrangements
on JH1, JH2, and
JH3; these amplified VDJ junctions were cloned
and sequenced (Fig. 5A). To
have controls with a primer outside neo, we also assayed VDJ
recombination by PCR with a backward primer located immediately
downstream from JH4 that readily detected
rearrangements in wt and / animals (Fig. 5B). This assay was strongly biased toward the amplification
of JH4 rearranged genes due to their much shorter
size. Nevertheless in homozygous N/N animals, there was no
rearrangement detectable on the JH4 segment,
while amplified fragments of larger size corresponding to the upstream
JH segments could be detected (Fig. 5B). A bias also appeared in the use of D elements, and
while at least five different D were used, including the most
5'-located segment, DFL16.1, there was a strong over-representation of
DQ52 (20 cases of 33 cloned VDJ sequences; Fig. 5A). The
structure of VDJ junctions appeared to be grossly normal, although more
than half of them were out of phase, probably due to the lack of any
selection for expression of a functional protein. The presence of both
P and N insertions in heterozygous animals
indicated that both DJ and VD junctions happened at a cell
differentiation stage where terminal deoxynucleotidyl transferase was
still expressed, and altogether there was no indication of any delay in
the gene rearrangement process.
|
C). Although neither blocked nor delayed, the occurrence of
a DJ rearrangement, which is usually biallelic in normal B cells, thus
appeared to be less constant than normal on the knockin allele. Methylation status of the germline JH-targeted allele in heterozygous hybridomas
We then checked the methylation status of the germline
JH allele in the five Ig-producing heterozygous
hybridomas that had kept the targeted JH allele
in its germline configuration (Fig. 6).
By comparison to nonlymphoid cells, in which the
JH region was almost completely methylated, all
five hybridomas displayed a clear pattern of demethylation. Assuming
that demethylation can be considered a molecular marker of DNA
accessibility, the JH region thus appeared to be
accessible to DNA rearrangement even in such
hybridomas.
|
In homozygous / animals, the concentration of all
Ig isotypes was normal. B cells were present at normal levels in
peripheral lymphoid organs.
Since modifications of gene expression simply resulting from insertion
of a loxP element have been reported in some instances
(16), we checked the expression and rearrangement of the
loxP knockin locus in competition with a wt
allele in heterozygous IgH /wt animals. To that goal,
heterozygous mice were selected, and alleles were compared with regard
to the expression of their protein product by flow cytometry using an a
allotype-specific anti-µ Ab. Splenocytes from
a/bwt mice
were compared with those from
awt/bwt
control mice and were found to express similar levels of the IgMa
allotype BCR (not shown). Alleles were also directly compared with
regard to the frequency of rearrangements using a
VH7183 consensus primer
(VH7183b primer) and a PCR strategy
differentiating by size rearranged fragments from the wt or
the loxP allele. In all mice studied, equal amounts of
rearranged fragments from both alleles were obtained (Fig. 7).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1)
(17). By contrast, a number of neo insertions within the JH-Cµ intron had virtually no effect on either V(D)J rearrangement or IgH expression. This was especially true for neo insertions in either orientation, replacing the Cµ CH1 exon (10), as well as for a neo insertion replacing the 3' MAR of the intronic enhancer (9). In a single case, which was also the only neo insertion reported upstream from the core Eµ, a defect affecting the IgH locus was observed (9). The mutation was studied in heterozygous mutant B cells and was shown to severely impair DJ recombination and JH region demethylation together with germline transcription.
In the current study we obtained a mutation that did not alter either the Eµ enhancer by itself or its flanking MARs. However, we showed that it resulted in a complete B cell defect and blockade of Ig production in homozygous animals. That the B cell defect was only due to the neo insertion and not to any unexpected additional mutation of the IgH locus was checked by Cre deletion of neo, which restored normal B cell development and Ig production.
The molecular mechanisms responsible for the blockade of B lineage
differentiation at the pro-B cell stage were explored and showed that
complete VDJ rearrangements of the mutant allele were readily
detectable in both homozygous and heterozygousmutant bone marrow
cells. These recombination events involved various V, D, and J
segments, but revealed a strongly biased D and JH gene
usage. There was a complete lack of rearrangement on the distal
JH4 segment and use of the DQ52 element in
60% of cases instead of the 5–10% previously estimated for
wild-type VDJ rearrangements (18). The observed
JH bias is strikingly similar to the phenotype
reported upon deletion of the DQ52 promoter/enhancer element
(18), further suggesting that accessibility of the distal
JH4 segment to recombination may rely on
synergistic interactions between both Eµ and DQ52 regulatory
elements. Such interactions would be disrupted in a similar manner by
the DQ52 deletion or interposition of a knockin neo gene.
Beside this bias, VDJ rearrangements had normal features, including
N insertions, but they did not yield any mature transcripts.
As a result, there was no possible selection for a functional BCR, and
about two-thirds of the rearranged VDJ segments were out-of-frame.
Although the mutant mice had an expanded bone marrow pro-B cell compartment, neither germline transcripts from the JH region (germline DQ52 transcripts readily detectable in normal bone marrow) nor Dµ transcripts originating from partial DJ rearrangements could be detected in homozygous mutant mice.
Contrasting with the lack of any transcripts from the JH region, i.e., originating upstream from the inserted neo gene, Iµ transcripts were detectable in all mutant animals together with neo transcripts. The tk promoter usually yields constitutive low level expression in mouse cells; in heterozygous wt/N splenocytes, although by force located on the allelically excluded IgH locus, the tk-neo gene captured stimulating effects from the IgH locus and appeared to be strongly induced by LPS stimulation. Upon fusion with terminally differentiated plasma cells, some hybridoma clones also up-regulated neo transcription. It thus appears that in the present case, the so-called neo effect merely consists of promoter competition and acts in a directional way only on transcripts initiated upstream from the insertion as if interposition of the tk promoter interrupted the interactions between Eµ and adjacent upstream promoters.
It is known that activation of most VH promoters requires approximation to the Eµ enhancer through VDJ rearrangement. A notable exception occurs at the pro-B cell stage, where certain VH genes, especially those belonging to the VHJ558 family, are transcribed in their germline environment (19, 20). In our homozygous mutant mice, although there was no detectable transcription of rearranged VH genes, germline transcription was maintained in bone marrow cells for unrearranged members of various VH families (including VHJ558, VH7183, and VHS107). This observation confirms that VH germline transcription probably relies on specific cis regulatory elements other than Eµ; such putative elements lying in the VH gene cluster would be silenced or deleted when B cells progress beyond the pro-B cell stage, after which germline VH transcription ceases. In the present case these elements do not appear to be affected by the neo insertion and therefore are probably independent from Eµ, a finding in agreement with their negative regulation upon B cell stimulation (19).
V segments usually rearrange to transcribed DJ segments, and both transgenic studies and targeted mutations have shown that transcriptional regulatory elements are strongly involved in the lineage-specific and stage-specific regulation of V(D)J recombination (1, 2, 7, 8, 9, 18, 21, 22). However, there is no unequivocal evidence that transcription by itself is needed for VDJ recombination. Models for the regulation of accessibility to V(D)J recombination have proposed that cis regulatory elements may either directly recruit putative accessibility factors or act via transcription-related mechanisms (22). Both pathways may also be interdependent, since a number of transcription factors are known to contain or to recruit chromatin remodeling activities (23). The herein reported neo insertion thus appears to uncouple two different effects of Eµ on the JH region: induction of accessibility to recombination and induction of transcription. Although not reported in the endogenous locus, such an uncoupling has already been observed for Ig transgenes, which were either transcribed and not rearranged (21) or were able to undergo rearrangement but not transcription (24, 25, 26). While the wild-type JH region usually undergoes germline transcription before recombination, endogenous germline VH segments provide other examples of rearrangements involving untranscribed DNA regions (20). It is thus clear that germline transcription of VH segments is not a prerequisite for their accessibility to recombination. From our study the same appears to hold true for the DQ52 and JH segments. Indeed, the neo knockin IgH locus features a complete lack of transcription in the DQ52/JH region that contrasts with a high level of rearrangements involving this region in both homozygous and heterozygous mutant animals. In contrast to constant gene germline transcripts that probably play a mechanistic role in class switch recombination (reviewed in Ref. 27), VH, D, and JH germline transcripts may rather be nonmandatory by-products of accessible genes. Meanwhile, chromatin accessibility would constitute by itself a prerequisite to VDJ recombination.
The high frequency of complete VDJ rearrangement maintained in our N/N mice together with demethylation of JH in the B cell lineage constitute remarkable differences with a previously studied neo insertion upstream from the core enhancer (9). In the latter case, disruption of interactions between the 5' MAR and core Eµ probably resulted in a more drastic effect on demethylation and chromatin remodeling. Indeed, MAR regions have been shown in transgenes to extend the remodeling effect of the core enhancer (28, 29).
In the present case the neo gene may behave as a decoy site for factors normally controlling IgH transcription, as demonstrated by the inducibility of the knockin tk promoter. That the neo gene effect varies depending upon its position in the locus also probably indicates that promoter competition mostly happens when neo is inserted in between an enhancer and its target endogenous promoters. In the latter configuration documented in this study, promoter competition insulates Eµ and precludes its activity on immediately adjacent upstream D or rearranged VH promoters so efficiently that it blocks B cell differentiation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to: Dr. Michel Cogné, Centre National de la Recherche Scientifique, UMR 6101, Laboratoire Immunologie, 2 rue du Dr. Marcland, 87025 Limoges, France. E-mail address: cogne{at}unilim.fr
3 Abbreviations used in this paper: MAR, matrix-associated region; BCR, B cell receptor; ES, embryonic stem; LCR, locus control region; neo, neomycin; tk, thymidine kinase; wt, wild type.
Received for publication May 29, 2002. Accepted for publication October 8, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 dramatically enhances VJ1 rearrangement. J. Exp. Med. 193:699.
enhancers indicates that different elements within the enhancer may mediate activation of transcription and recombination. J. Immunol. 164:795.This article has been cited by other articles:
|
|
 |
|
  I. I. Kuzin, L. Bagaeva, F. M. Young, and A. Bottaro Requirement for Enhancer Specificity in Immunoglobulin Heavy Chain Locus Regulation J. Immunol., June 1, 2008; 180(11): 7443 - 7450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Samara, Z. Oruc, H.-L. Dougier, T. Essawi, M. Cogne, and A. A. Khamlichi Germ line transcription in mice bearing neor gene downstream of I{gamma}3 exon in the Ig heavy chain locus Int. Immunol., April 1, 2006; 18(4): 581 - 589. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Heltemes-Harris, X. Liu, and T. Manser An antibody VH gene that promotes marginal zone B cell development and heavy chain allelic inclusion Int. Immunol., November 1, 2005; 17(11): 1447 - 1461. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Delpy, C. Sirac, C. Le Morvan, and M. Cogne Transcription-Dependent Somatic Hypermutation Occurs at Similar Levels on Functional and Nonfunctional Rearranged IgH Alleles J. Immunol., August 1, 2004; 173(3): 1842 - 1848. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Delpy, C. Sirac, E. Magnoux, S. Duchez, and M. Cogne RNA surveillance down-regulates expression of nonfunctional {kappa} alleles and detects premature termination within the last {kappa} exon PNAS, May 11, 2004; 101(19): 7375 - 7380. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. D. Trinklein, S. F. Aldred, S. J. Hartman, D. I. Schroeder, R. P. Otillar, and R. M. Myers An Abundance of Bidirectional Promoters in the Human Genome Genome Res., January 1, 2004; 14(1): 62 - 66. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
