|
|
||||||||


*
Laboratory of Molecular Immunology, and
Howard Hughes Medical Institute, The Rockefeller University, New York, NY 10021
| Abstract |
|---|
|
|
|---|
and to the small, underused
VH5 family. To determine whether Abs commonly expressed by
V-preB+L+ B cells show similar features, we
analyzed Ig H chains from three highly expressed VH
families, VH1, VH3, and VH4, and
Ig-
. We find that VH1 and VH3 Abs expressed
by V-preB+L+ B cells resemble VH5
in that they display increased JH6 use, long CDR3s, and an
increased frequency of D-D fusions. Abs in all three of these
VH families also show skewed D reading frame use resulting
in predominance of hydrophobic amino acids, which are counterselected
in conventional B cells. Like Ig-
genes, the Ig-
genes in
V-preB+L+ B cells show long CDR3s, but they
differ from Ig-
genes in that they display no evidence of receptor
editing. We conclude that a large number of H and L chain Abs expressed
by V-preB+L+ B cells display features
associated with self-reactive Abs. | Introduction |
|---|
|
|
|---|
In mice, receptor editing makes an important contribution to the establishment of the normal Ab repertoire (16), but little is known about the relative contributions of deletion, anergy, and editing to tolerance in humans.
V-preB+L+ B cells are a
recently described subpopulation of normal human B cells found in the
circulation and tonsils, and they accumulate in the joints of some
patients with rheumatoid arthritis (17, 18). These cells
differ from other peripheral B cells in that in addition to V-preB they
also express low levels of recombinase-activating gene (RAG) mRNA.
Although the initial analysis was limited to Ig-
(L) chain and the
VH5 gene family, which makes a small contribution
to the normal repertoire (19, 20), it showed an unusual Ab
repertoire consistent with self-reactivity and Ig-
receptor editing
(18). Based on these observations, it was proposed that
V-preB+L+ B cells may
represent B cells that have been tolerized (18). To
further examine how tolerance might be established in developing B
cells in the human, we extended the characterization of the Ab
repertoire in V-preB+L+ B
cells to VH1, VH3,
VH4, and Ig-
2.
| Materials and Methods |
|---|
|
|
|---|
V-preB+L+B cells
were obtained from three nonrelated healthy donors by a combination of
MACS and cell sorting with anti-V-preB mAbs (donors 13; Ref.
18). The anti-V-preB Ab was produced against a region
of V-preB that shows no Ig homology, and extensive physical
measurements failed to detect such cross-reactivity with Ig L or H
chains (21). In brief,
V-preB+L+ B cells were
enriched from PBMCs by negative selection against non-B cells followed
by positive selection with anti-VpreB-labeled with PE
(22) and MACS anti-PE microbeads (Miltenyi Biotec,
Aurburn, CA). Conventional
V-preB-L+ B cells were B
cell-enriched PBMC that did not bind to the MACS column. After the
initial enrichment by negative and positive selection, the cells were
stained with FITC human anti-
and anti-
, and
allophycocyanin anti-CD19 (BD PharMingen, San Diego, CA) and
V-preB+L+CD19+
B cells and conventional
V-preB-L+CD19+
B cells were purified by sorting on a FACSVantage (BD Biosciences,
Mountain View, CA; Ref. 18).
RNA and RT-PCR
Total RNA was extracted from
104105 purified cells
with TRIzol reagent (Life Technologies, Rockville, MD). After DNase I
(Boehringer Mannheim, Indianapolis, IN) treatment, RNA was
reverse-transcribed in 10 µl with SuperScript II (Life
Technologies). For RT-PCR, 1 ml of cDNA was amplified for 35 cycles of
30 s at 94°C, 30 s at 58°C
(VH1-Cµ,
VH3-Cµ or
VH4-Cµ) or at 55°C
(V
-C
), and 30 s
at 72°C with a final 10-min extension at 72°C with HotStarTaq
DNA polymerase (Qiagen, Chatsworth, CA) and the following primers:
V
18 family consensus sense,
5'-GGG(G/A)TC(T/C)CTGA(C/T/G)CG(A/C/G)TTCTCTGG(C/G)TCC-3';
V
9 sense, 5'-ATCCCTGATCGCTTCTCAGTCTTG-3';
V
10 sense, 5'-GATCTCAGAGAGATTATCTGCATCC-3';
C
antisense,
5'-CACAC(T/C)AGTGTGGCCTTGTTGGCTTG-3'. Sense FR1
VH1, VH3, and
VH4 and antisense Cµ primers were as described
previously (17, 23). RT-PCR products were analyzed on 2%
agarose gels.
Cloning and sequencing
PCR products were gel purified (QIAquick; Qiagen) and cloned
into TA vectors (Invitrogen, San Diego, CA). The dsDNA sequences were
obtained with antisense Cµ or
C
primers and dye terminator cycle sequencing
(PE Applied Biosystems, Foster City, CA). Sequences were analyzed by
comparison with Ig BLAST. When two or more identical sequences were
found, they were counted as a single clone. IgH CDR3 length was
determined by counting amino acid residues between positions 94 and 102
(conserved tryptophan in all JH segments), and D
segments were identified following the criteria of Corbett et al.
(24). Ig-
CDR3 length included amino acids between
conserved cysteine 88 and the phenylalanine residue embedded in all
J
s (25). Differences in gene
distribution between
V-preB+L+ B cells and
conventional B cells were analyzed with
2
tests, and they were considered significant when p
0.05. Students t test was used for CDR3-length
analysis.
| Results |
|---|
|
|
|---|
To extend the analysis of the Ig H chain repertoire of V-preB+L+ B cells, we cloned and sequenced VH1, VH3, and VH4 genes from V-preB+L+ and conventional B cells obtained from peripheral blood of three healthy donors. The distribution of the VH sequences from each of the three donors was similar, and they are presented together for simplicity. We chose the VH1, VH3, and VH4 families because these genes are well characterized and account for most of the peripheral B cell repertoire (19, 20, 26, 27, 28, 29).
The VH1, VH3, and
VH4 repertoire from conventional B cells was
similar to that reported by others (Fig. 1
; Refs. 20, 26, 27, 28, 29). The
VH repertoire of VH1 and
VH 4 Abs expressed by
V-preB+L+ B cells differed
from conventional B cells. Half of the VH1 genes
were differentially expressed in
V-preB+L+ and conventional
B cells. VH118 and -46 gene use was increased
from 11% and 5%, respectively, in conventional B cells to 29% and
16% in V-preB+L+ B cells
(Fig. 1
, top), whereas VH12 and -69
gene use was decreased in
V-preB+L+ B cells (Fig. 1
, top). Among the VH4 genes,
VH434, was counterselected in
V-preB+L+ B cells to 4.5%
(Fig. 1
, bottom) compared with 15% in conventional B cells
(n = 117; Ref. 30). In contrast to the
VH1 and VH4 genes, we found
no statistically significant differences in the
VH3 repertoire between
V-preB+L+ and conventional
B cells. However, the VH3 family is a very large
family, and therefore, it is difficult to analyze the contribution of
individual family members to the repertoire (31). We
conclude that the VH1 and
VH4 repertoire of
V-preB+L+ B cells differs
from conventional B cells.
|
VH5 genes expressed in
V-preB+L+ B cells show long
CDR3s with increased JH6 use and frequent
diversity gene segment (D) fusions. VH1 and
VH3 genes expressed in
V-preB+L+ B cells resembled
VH5 in that they displayed increased
JH6 use when compared with conventional B cells
(Fig. 2
A). Increased
JH6 use was particularly favored by
VH169 genes (62.5% in
V-preB+L+ B cells vs 27.9%
in conventional B cells; p < 0.001).
JH6 is the longest of the human
JHs, and consistent with its overuse,
V-preB+L+ B cells have
longer VH1 IgH CDR3s than conventional B cells
(Fig. 2
B). In addition, D-D fusions were found in 4% of
VH1 and VH3 Abs expressed
by V-preB+L+ B cells,
whereas only 1 of 351 such fusions were identified in the
VH1 and VH3 Abs cloned from
conventional B cells (4 of 109 VH1 and 4 of 117
VH3; p = 0.006). In contrast to
VH1 and VH3,
VH4 Abs from
V-preB+L+ B cells showed no
bias to JH6, normal CDR3s, and no D-D fusions
(Fig. 2
, A and B, and data not shown). We
conclude that VH1, VH3 and
VH5 genes expressed by
V-preB+L+ B cells show
increased JH6 use, long CDR3s, and D-D
rearrangements, whereas VH4 genes exhibit none of
these features.
|
Human Abs show preferential use of certain D segments
(24). Therefore, we analyzed D region use in
V-preB+L+ B cells. We found
that D use in VH1 and VH3
Abs in V-preB+L+ B cells
differs from D use in conventional B cells (Fig. 3
A). VH1
and VH3 genes were pooled for this analysis
because they displayed similar features. Several of these differences
reached statistical significance, including D22 and D613, which
were overrepresented in
V-preB+L+ B cells, and
D423 use, which was underrepresented (Fig. 3
A). In
contrast, D segment gene use was not significantly altered in most
VH4 Abs in
V-preB+L+ B cells (Fig. 3
B).
|
|
expression in V-preB+L+ B cells
Ig-
genes expressed by
V-preB+L+ B cells show long
CDR3s and are skewed toward upstream V
s and
downstream J
s consistent with increased
secondary V(D)J recombination (18). By analogy to the
Ig-
locus, the human Ig-
locus should also allow deletional
replacement of VJ
s by receptor editing (32) and would
bias the repertoire to more downstream J
s and
upstream V
s.
To determine whether
V-preB+L+ B cells show
evidence of Ig-
receptor editing, we cloned and sequenced the Ig-
genes expressed by these cells.
V-preB+L+ B cells showed no
increase in downstream J
use when compared
with conventional B cells; in fact, there was a small increase in
upstream J
1 segment use (Fig. 5
A; Refs. 33 and
34). In addition, there was no bias for distal
V
segment use (35, 36 ; Fig. 5
B). We conclude that Ig-
repertoire in
V-preB+L+ B cells shows no
skewing to either upstream V
s or downstream
J
s and therefore no evidence of
VJ
receptor editing.
|
gene family analysis in
V-preB+L+ B cells revealed
a bias to V
1 gene use, which increased from
24.4% in conventional B cells to 40.3% in
V-preB+L+ B cells (Fig. 5
136, 44, and 51 in
V-preB+L+ B cells in all
three normal donors, whereas V
140 and 47
gene use remained unchanged (V
136/44;
p = 0.04, data not shown).
Ig-
genes expressed in
V-preB+L+ B cells resemble
the Ig-
genes in that they show an increase in CDR3 length (Fig. 6
). The 11-aa-long Ig-
CDR3s increased
from 35% in conventional B cells to 50% in
V-preB+L+ B cells, whereas
short Ig-
CDR3s (9 aa) decreased from 24% in conventional to 9% in
V-preB+L+ B cells (Fig. 6
).
We conclude that the Ig-
repertoire in
V-preB+L+ B cells is skewed
to long CDR3s and selected subset of V
1
genes.
|
| Discussion |
|---|
|
|
|---|
genes with long CDR3s reminiscent
of anti-self Abs (37, 38, 39, 40).
In this report, we have extended the analysis of the Ig repertoire of
V-preB+L+ B cells to the
VH1, VH3 and
VH4 gene families and Ig-
.
VH1, VH3, and
VH4 genes account for most normal human Abs, and
Ig-
is expressed by 3040% of human B cells (19, 20, 26, 27, 28, 41). The VH1 and VH3
repertoire expressed by
V-preB+L+ B cells is
similar to VH5 in that it shows an increase in
JH6 use, long CDR3s, and D-D fusions. These Abs
also show biased D use and overrepresentation of Ds with hydrophobic
amino acids (24, 29). In addition, Ig-
genes with long
CDR3s are enriched in
V-preB+L+ B cells. Long
IgH, Ig-
, and Ig-
CDR3s and IgH D-D fusions are normally
counterselected in conventional B cells in humans and mice (24, 29, 37). However, these features are prevalent in autoantibodies
(37, 40, 42, 43, 44). We conclude that the Ab repertoire in
V-preB+L+ B cells
frequently shows features consistent with autoreactivity.
Chronic lymphocytic leukemia B cells (B-CLL) frequently express autoantibodies (42, 45, 46). In particular, VH169 genes are highly overrepresented in the Abs expressed in B-CLL cells (20% of all cases; Refs. 47, 48). Although VH169 gene use was decreased in V-preB+L+ B cells relative to controls, VH169 remains one of the two most-used VH1 genes in these cells, representing over 20% of all VH1 genes. The VH169 Abs in V-preB+L+ B cells resemble those found in B-CLL cells because B-CLL VH169 genes display long CDR3s, biased JH6 use, and a preference for D22 and D33 use. Therefore, these data suggest that V-preB+L+ B cells may be precursors of B-CLL (23, 49).
Why is the VH4 repertoire in V-preB+L+ B cells different from VH1, VH3, and VH5? It has been suggested that VH4 genes may be particularly prone to producing self-reactive Abs (50, 51, 52). In particular, Abs with VH434 variable regions show intrinsic self-reactivity and recognize i/I carbohydrate self-determinants displayed on red blood cells and other cell types (51, 53, 54). It has been shown that germline-encoded VH434 gene has intrinsic self-reactivity and that the CDR3 in VH434 Abs modulates the affinity for self (55, 56). We would propose that this intrinsic self-reactivity combined with long hydrophobic CDR3s (which themselves have low levels of self-reactivity; Refs. 37, 38, 39) might form Abs with high affinity for self that are deleted centrally. Consistent with this idea, VH434 was specifically depleted from the VH4 repertoire of V-preB+L+ B cells.
In addition to their unusual Ab repertoire,
V-preB+L+ B cells also
showed evidence of Ig-
receptor editing (18), but we
found no evidence for Ig-
receptor editing in
V-preB+L+ B cells.
Secondary rearrangements on the Ig-
locus are theoretically
possible, and have been reported in patients with systemic lupus
erythematosus (57) and in some transformed cell lines
(58, 59). However, other cell lines that edit Ig-
fail
to rearrange Ig-
(60), and there is no evidence that
secondary Ig-
rearrangements occur under physiologic circumstances.
Ig-
editing requires persistent high levels of RAG expression during
a 2-h arrest at the pre-B cell stage of development, whereas B cells
that proceed beyond the late pre-B cell and early immature B cell
compartment appear to be unable to edit (16). Because
Ig-
gene rearrangement precedes Ig-
recombination, the window for
Ig-
gene editing in B cell development may be very narrow, and
secondary Ig-
recombination may be an exception rather than the
rule.
Long CDR3s are selected against because their expression interferes
with normal B cell development either by hindering IgH and L pairing or
because Abs with these features are self-reactive. Why, then, are these
features found in V-preB+L+
B cells? It has been suggested that
V-preB+L+ B cells differ
from other B cells in that the Abs they express prevent them from
completing B cell development. In this model, B cells expressing Abs
that are partially self-reactive would not fully extinguish expression
of developmentally regulated genes such as RAG and V-preB and would be
selected to become
V-preB+L+ B cells. An
alternative explanation is that improper interaction between H and L
chains would result in Abs that simply fail to produce the adequate
signals for immature B cell development. Improper BCR assembly might
allow V-preB+L+ B cell
survival but not be sufficient for RAG and V-preB gene down-regulation.
Our analysis of the VH1,
VH3, VH4,
VH5, Ig-
, and Ig-
repertoire of
V-preB+L+ B cells is
consistent with either the selection model or altered development and
suggests that V-preB+L+ B
cells might be more prone to becoming B-CLL cells.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Michel C. Nussenzweig, The Rockefeller University, Howard Hughes Medical Institute, 1230 York Avenue, New York, NY 10021. E-mail address: nussen{at}mail.rockefeller.edu ![]()
3 Abbreviations used in this paper: BCR, B cell receptor; RAG, recombination activating gene; RF, reading frame; B-CLL, B-type chronic lymphocytic leukemia. ![]()
Received for publication February 23, 2001. Accepted for publication June 14, 2001.
| References |
|---|
|
|
|---|
L+µ-) proB complex: cell surface expression and biochemical structure of a putative transducing receptor. Blood 93:4336.
genes. J. Exp. Med. 172:609.
repertoire. J Mol Biol 268:69.[Medline]
locus is similarly well expressed in mice and humans. J. Exp. Med. 189:1611.
light-chain locus on chromosome 22q11.2. Hum. Mol. Genet. 4:983.
repertoire. J. Mol. Biol. 264:220.[Medline]
light chains expressed in patients with rheumatoid arthritis and in normal individuals. J. Clin. Invest. 96:831.
light chains with unusually long CDR3 regions in patients with rheumatoid arthritis. J. Rheumatol. 25:1067.[Medline]
light chains expressed in rheumatoid arthritis synovia and PBL, and the influence of V
gene segment utilization on CDR3 length. Mol. Med. 4:525.[Medline]
and heavy chain gene usage in early untreated systemic lupus erythematosus suggests intensive B cell stimulation. J. Immunol. 163:1027.
rearrangements in cell culture. Eur. J. Immunol. 23:1501.[Medline]
This article has been cited by other articles:
![]() |
A. Lim, B. Lemercier, X. Wertz, S. L. Pottier, F. Huetz, and P. Kourilsky Many human peripheral VH5-expressing IgM+ B cells display a unique heavy-chain rearrangement Int. Immunol., January 1, 2008; 20(1): 105 - 116. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. C. Watson, C. S. Moffatt-Blue, R. Z. McDonald, E. Kompfner, D. Ait-Azzouzene, D. Nemazee, A. N. Theofilopoulos, D. H. Kono, and A. J. Feeney Paucity of V-D-D-J Rearrangements and VH Replacement Events in Lupus Prone and Nonautoimmune TdT-/- and TdT+/+ Mice J. Immunol., July 15, 2006; 177(2): 1120 - 1128. [Abstract] [Full Text] [PDF] |
||||
![]() |
H Morbach, S K Singh, C Faber, P E Lipsky, and H J Girschick Analysis of RAG expression by peripheral blood CD5+ and CD5- B cells of patients with childhood systemic lupus erythematosus Ann Rheum Dis, April 1, 2006; 65(4): 482 - 487. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Stamatopoulos, C. Belessi, A. Hadzidimitriou, T. Smilevska, E. Kalagiakou, K. Hatzi, N. Stavroyianni, A. Athanasiadou, A. Tsompanakou, T. Papadaki, et al. Immunoglobulin light chain repertoire in chronic lymphocytic leukemia Blood, November 15, 2005; 106(10): 3575 - 3583. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Meffre, A. Schaefer, H. Wardemann, P. Wilson, E. Davis, and M. C. Nussenzweig Surrogate Light Chain Expressing Human Peripheral B Cells Produce Self-reactive Antibodies J. Exp. Med., January 5, 2004; 199(1): 145 - 150. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Sanz, M. Alvarez-Mon, C. Martinez-A, and A. de la Hera Human cord blood CD34+Pax-5+ B-cell progenitors: single-cell analyses of their gene expression profiles Blood, May 1, 2003; 101(9): 3424 - 3430. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |