| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
Schools of
* Microbiology and Immunology and
Medicine, University of New South Wales, Kensington, Australia;
Royal North Shore Hospital, St. Leonards, Australia; and
Department of Medical Oncology, St. Vincent’s Hospital, Darlinghurst, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RI), present upon the surface of mast cells and basophils, the
effector cells of allergy (1), and once receptor bound, is
most stable on the cell surface. The binding of allergen by specific
IgE causes cross-linking of cell-bound IgE, and then aggregation of the
FcRI. This aggregation initiates an enzymatic cascade of events that
ultimately leads to cell activation and degranulation (2, 3). Mast cells and basophils contain potent stores of mediators,
such as histamine, that are responsible for the immediate symptoms of
allergic disorders. These cells may also synthesize IL-4 and 13 that
together up-regulate IgE class switching in B cells
(4). Many studies have attempted to characterize the IgE-allergen interaction by focusing on the allergens (5, 6, 7, 8, 9, 10), but the fine specificity of the IgE Abs themselves remains largely unknown. As IgE Abs are only present in trace amounts (nanograms per milliliter) in the sera of allergic patients (1), conventional serological approaches have not been useful in their characterization. B cells producing IgE are also scarce (11, 12), making their isolation and study equally difficult. Therefore, until recently, much of what was known about the structure, specificity, and molecular genetics of IgE Abs had come from the study of IgE myeloma proteins (13, 14, 15, 16). A more recent approach to the investigation of the structure and molecular genetics of IgE Abs has been the isolation and analysis of their rearranged Ig V region genes. The V region genes of Abs are formed by the joining of one V, one D, and one J gene segment for the H chain, and one V and one J gene segment for the L chain. In every pre-B cell, this process of recombinational rearrangement generates a functional rearranged H and L chain V gene (17, 18). Diversity is essentially generated by these rearrangements; however, further diversity is added by processing at each end of the recombining elements through N and P nucleotide addition, and exonuclease activity (18). Point mutations made by the process of somatic hypermutation within the germinal center reaction may also introduce diversity (19).
Each rearranged V gene, when expressed, will encode a polypeptide
composed of a supporting 
-pleated sheet scaffold or framework region
and three hypervariable loops termed complementarity-determining
regions (CDRs)3
(20). In the majority of Ab-Ag interactions, it is the
CDRs that facilitate Ag binding. The CDR3 of the H chain is the most
diverse of all the hypervariable loops, and is encoded solely by the
V-D-J junction. The human CDR3 can assume a range of shapes and sizes,
and can theoretically be 1 of 1014 different
peptides (21).
Characterization of the Ig V genes can generate useful information about the specificity and molecular biology of IgE, as well as providing insights into the nature of the B cells that produced them. Our current knowledge of the molecular forces that shape the IgE repertoire comes from a handful of studies on patients with atopic asthma (22, 23, 24, 25, 26), atopic dermatitis (27, 28), peanut allergy (29), and allergy to grass pollens (30). Cloning and sequencing the rearranged V genes of IgE sequences from PBMCs (22, 27), a bronchial biopsy (24), and splenocytes (23) of allergic patients have shown a bias to the small VH5 family. This family is rarely seen in the expressed adult B cell repertoire of normal individuals (31). All of the above studies, however, have sequenced random IgE transcripts, rather than transcripts from allergen-specific IgE. Whether or not specific IgE taking part in an allergic response also demonstrates restricted V gene usage is unknown.
Steinberger et al. (30) have used phage Ab display (32) to isolate specific IgE Ab fragments to the timothy grass pollen allergen Phl p 5. Phage display allows the isolation and characterization of specific Abs by making large V gene libraries in an Escherichia coli compatible vector (33, 34). Through manipulation of the E. coli-filamentous phage lifecycle, the V gene products are expressed as phage-Ab fusions (35, 36). Specific phage can be isolated by panning against immobilized Ag, and enriched by propagation in E. coli (37). The study of IgE by Steinberger represented the first look at allergen-specific IgE Abs in an allergic response. In this study, we have combined rearranged V gene analysis with phage Ab display, to study IgE Abs from an egg-allergic atopic dermatitis patient.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
After informed consent, 100 ml peripheral blood was collected from a 40-year-old female atopic dermatitis patient. Radioallergosorbent assay demonstrated IgE reactivity to a wide range of allergens, including egg white protein (2+), house dust mite (4+), peanut (2+), and rye grass pollen (4+). Mononuclear cells were harvested from peripheral blood by Ficoll density gradient centrifugation and used as a source of IgE mRNA.
RNA extraction, cDNA synthesis, and PCR amplification
Total RNA was isolated by the method of Chomczynski and Sacchi
(38). Synthesis of cDNA was performed as previously
described (39). Primers used for PCR amplification
involved V region H chain primer sequences reported by Yip et al.
(40) and Persson et al. (33). Sequences
(5'-3') were VH1a, CAG GTG CAG CTC GAG CAG TCT GGG;
VH1f, CAG GTG CAG CTG CTC GAG TCT GG;
VH3a, GAG GTG CAG CTC GAG GAG TCT GGG;
VH3f, GAG GTG CAG CTG CTC GAG TCT GGG;
VH4f, CAG GTG CAG CTG CTC GAG TCG GG;
VH4g, CAG GTG CAG CTA CTC GAG TGG GG;
VH6a, GAG GTA CAG CTC GAG CAG TCT GG;
VH6f, CAG GTA CAG CTG CTC GAG TCA GGT CCA; and
C1 of IgE CH1, GCT GAA ACT AGT GTT GTC GAC CCA
GTC TGT GGA (30). L chain primers used have also been
reported by Persson et al. (33) PCR amplification of IgE
Fd (H chain variable domain and C1) and 
and 
L
chains was performed with 5 µl cDNA template, with 0.6 µM of each
primer, 0.2 mM dNTPs (Promega, Madison, WI), 2.0 mM
MgCl2, 0.4 U Tth polymerase (Fischer
Biotec, Subiaco, WA), and a buffer supplied by the manufacturer
(Fischer Biotec). Cycling was 94°C for 3 min, followed by 94°C for
30 s, 52°C for 1 min, 72°C for 2 min for 40 cycles, and then
72°C for 10 min.
Ab library construction
Ab libraries were constructed essentially as described by Ward
et al. (39). Briefly, IgE Fd and L chains of 
660 bp
were purified from PCR by agarose gel electrophoresis and then cut with
XhoI and SpeI (IgE Fd) or XbaI and
SacI (L chains). Cut PCR products (1 µg) were then ligated
with the vector MCO3 (1.5 µg), which had been digested with similar
sets of restriction enzymes. L chain and then Fd H chain were
sequentially cloned into the vector using DNA ligase, followed by
electroporation into E. coli XL-1 (Stratagene, La Jolla, CA)
using standard procedures. After 1 h recovery at 37°C, with shaking,
an aliquot of the library was titered on 2xYT (yeast extract and
trypton medium) plates with 50 µg/ml carbenicillin, 2% (w/v) glucose
to determine library size. The remaining culture was scaled up to 100
ml with 2xYT 50 µg/ml carbenicillin and 2% (w/v) glucose and grown
for 1 h at 37°C. Insert frequency was determined by the analysis
of randomly picked clones for presence of insert by PCR, and by
titering colonies produced by mock ligations consisting only of cut
vector and no insert, and cut vector with no insert or ligase.
Scaling up of library and selection of specific phage Fabs
After growth for 1 h, the library culture was infected with 1.5 x 1012 VCSM13 helper phage (Stratagene) grown for 2 h at 37°C, then spun at 4000 rpm for 15 min. Cells were resuspended in 2xYT medium with 50 µg/ml carbenicillin, 70 µg/ml kanamycin, and 10 µg/ml tetracycline, selecting for E. coli infected with helper phage and maintaining recombinant plasmids. Phage-Fab fusions were produced from the culture by growing overnight at 30°C. Phage were selected on microtiter plates (Polysorb; Nunc, Roskilde, Denmark) coated with purified ovomucoid (a gift from R. Sleigh, CSIRO, Sydney, Australia) essentially as previously described (39). One exception was that the amount of Ag decreased over the rounds. In rounds 1 and 2, 100 µg/ml was used, decreasing to 10 µg/ml at round 3, 1 µg/ml at round 4, and 0.1 µg/ml at round 5. Input and output phage were titered on log phase E. coli XL-1 blue (Stratagene) to determine the ratio of input-output phage at each round.
Induction of soluble IgE Fabs, colony screening by ELISA, and inhibition ELISA
Phage isolated from each panning round were used to infect a log
phase culture of the nonsuppressor strain E. coli HB2151,
and single colonies were grown in 96-well plates. The cultures were
induced with isopropyl--D-thiogalactoside for
soluble Fab expression (37). Overnight cultures in 96-well
plates were spun at 2750 rpm for 15 min, and 0.1-ml aliquots of culture
supernatant were transferred to microtiter plates (polysorp; Nunc) that
had been coated in 10 µg/ml ovomucoid in carbonate buffer and blocked
in blocking buffer (2% (w/v) skim milk in PBS, pH 7.4) for 2 h at
37°C. Aliquots were also added to a microtiter plate (polysorp; Nunc)
coated in blocking buffer, as described above. A nonspecific Fab (human
anti-p53) served as an additional negative control. Supernatants
were incubated at room temperature for 1 h before washing three
times with PBS, pH 7.4. Purified anti-c-myc Ab (clone
9E10, 0.5 mg/ml) was diluted 1/1000 in diluting buffer (2% (w/v) skim
milk in PBS, pH 7.4) and added to each well. Incubation and washing
were as above. Goat anti-mouse Ab-HRP conjugate (0.5 mg/ml; Jackson
ImmunoResearch Laboratories, West Grove, PA) was diluted 1/1000 in
diluting buffer, then added and incubated as above. Plates were washed
three times with PBS, pH 7.4, HRP substrate (Kirkegaard & Perry
Laboratories, Gaithersburg, MD) was added, and color was allowed to
develop at room temperature for 20 min. The reaction was stopped by
addition of 2 M H2SO4, and
absorbances were read at 450 nm. Clones were considered positive when
the OD was 3-fold greater than that seen on blank wells with no Fab
sample. To confirm the ELISA results, an inhibition ELISA was performed
by a modification of the ELISA in which 50 µl 2% (w/v) skim milk, 10
µg/ml ovomucoid, or 2% (w/v) BSA (Sigma-Aldrich, St. Louis, MO) in
PBS, pH 7.4, was included in the well at the time of addition of the
Fab sample (50 µl). Inhibition was determined by comparing the
absorbances of wells with Fab sample and immobilized Ag, with wells
containing Fab sample, immobilized Ag, and Ag in the liquid phase.
PCR amplification of cloned inserts
Plasmids were extracted by an alkaline lysis procedure from overnight cultures grown in 2xYT, 50 µg/ml carbenicillin, and 2% glucose. Inserts were amplified using PCR with primers pelBFor (CTA CGG CAG CCG CTG GAT TG) and g3pRev (GAT CCT CTT CTG AGA TGA GT), which anneal to the pelB leader sequence and g3p of filamentous phage, respectively, in MCO3. One microliter of plasmid DNA served as template under identical reaction conditions as for cDNA amplification. The cycling conditions used were 94°C for 3 min, then 30 cycles of 94°C for 15 s, 52°C for 50 s, 72°C for 1.5 min, and then one cycle of 72°C for 10 min, on a PerkinElmer (Wellesley, MA) 9600 Gene Amp thermal cycler. PCR products were purified by agarose gel electrophoresis and a gel extraction kit (Qiagen, Chatsworth, CA).
V gene sequencing and V gene sequence alignments
Cloned Fd inserts were sequenced using a Dye Terminator Kit (Applied Biosystems, Foster City, CA), with 90 ng purified PCR product and 7.5 pmol forward (pelBFor) or reverse primer (g3pRev). Sequences were analyzed on an ABI 377 automated sequencer (Applied Biosystems). Cycling conditions were 96°C for 10 s, 50°C for 5 s, and 60°C for 4 min for 25 cycles. Sequences were aligned using the DNAPLOT algorithm available at the V BASE database (41) and are named using the IMGT nomenclature (42). The numbering system for the CDRs and framework regions complies with the system of Chothia et al. (43).
Design of cyclic peptides
The cyclic peptide CP-1 was designed according to the CDR3
sequence of a number of VH6-utilizing
ovomucoid-binding Fabs. A control peptide CP-N was designed from the
CDR3 sequence of a nonbinding VH3-utilizing clone
(clone 46). Both peptides are illustrated in Fig. 1
. The peptides were obtained from Auspep
(Parkville, Australia) as a lyophilized powder, and were reconstituted
with sterile PBS, pH 7.4, before use.
|
Binding of CP-1 and the control peptide CP-N to a range of Ags was assessed by ELISA and dot-blot assay. Microtiter plates (polysorp; Nunc) were coated in 100 µl 20 µg/ml ovomucoid, 20 µg/ml OVA, 10 µg/ml BSA, 28 µg/ml tetanus toxoid, 10 µg/ml recombinant human early pregnancy factor, human bcl-2 (DAKO, Carpenteria, CA) diluted 1/10, the 4-mer peptide arg-gly-asp-ser (Sigma-Aldrich) diluted 1/10, normal rabbit serum diluted 1/100, Intragram (60 µg/ml human polyclonal IgG; CSL, Victoria, Australia) diluted 1/100, 8 µg/ml extracellular domain of human erbB-2, 20 µg/ml purified bovine von Willebrand factor (vWF), 7 µg/ml purified human vWF, and 8.5 µg/ml recombinant human p53. All Ags were diluted in carbonate-coating buffer, and coating occurred overnight at 4°C. After three washes in PBS, pH 7.4, containing 0.05% Tween 20 (PBS/Tween), the wells were blocked in PBS containing 0.1% (w/v) casein hydrolysate for 2 h at 37°C. One hundred microliters of CP-1 or CP-N (diluted to 10 µg/ml in PBS, pH 7.4, containing 0.1% (w/v) casein hydrolysate) were added per well, and binding was allowed to occur at room temperature for 1 h before washing three times with PBS/Tween. Streptavidin-HRP conjugate (DAKO) was diluted 1/1000 in PBS/casein hydrolysate, added to all wells, and incubated for 1 h at room temperature. After three washes of PBS/Tween, tetramethylbenzidine substrate (Kirkegaard & Perry Laboratories) was added. Color was allowed to develop for 20 min, and the reaction was then stopped by the addition of 1 M H2SO4. Absorbances were read at 450 nm.
Dot-blot assays were conducted on polyvinylidene difluoride membrane (Immun-Blot; Bio-Rad, Hercules, CA). Membranes were blocked overnight at 4°C in 0.1% (w/v) casein hydrolysate and washed four times in PBS/Tween. A total of 10 µl 200 µg/ml ovomucoid, 200 µg/ml OVA, 100 µg/ml BSA, 280 µg/ml tetanus toxoid, 100 µg/ml human recombinant early pregnancy factor, human bcl-2, 4-mer peptide arg-gly-asp-ser, normal rabbit serum diluted 1/10, Intragram diluted 1/10, 80 µg/ml extracellular domain of human erbB-2, 200 µg/ml bovine vWF, 70 µg/ml human vWF, and 85 µg/ml recombinant human p53 diluted in PBS, pH 7.4, were blotted onto the membrane. Membranes were incubated in CP-1 or CP-N at 10 µg/ml in PBS, pH 7.4, for 1 h at room temperature. A peptide-free control membrane was also included. After washing four times in PBS/Tween, membranes were incubated in streptavidin-alkaline phosphatase conjugate (Zymed Laboratories, San Francisco, CA) diluted 1/100 in 0.1% (w/v) casein hydrolysate for 1 h at room temperature. Membranes were washed four times in PBS/Tween, and incubated in nitroblue tetrazolium chloride (Research Organics, Cleveland, OH)/5-bromo-4-chloro-3-indolyl-phosphate (Boehringer Mannheim, Indianapolis, IN) substrate (0.2 mg/ml each of nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl-phosphate in carbonate buffer). Color was allowed to develop for 10 min, and stopped by immersing membranes in water.
The ability of CP-1 to inhibit binding by serum IgE was assessed using
the Pharmacia ImmunoCAP system (Pharmacia and Upjohn, Uppsala, Sweden).
Immunocaps specific for ovomucoid were incubated with a 1/4 dilution of
serum in PBS, pH 7.4, containing CP-1 at 20 µg/ml, CP-1 at 10
µg/ml, CP-1 at 5 µg/ml, CP-1 at 2.5 µg/ml, CP-N at 20 µg/ml, or
PBS only for 30 min at room temperature. Immunocaps were washed in an
Assay Washer 96 (Pharmacia and Upjohn), and incubated for 2.5 h at
room temperature in -galactosidase/anti-IgE conjugate (Pharmacia
and Upjohn). After washing, immunocaps were incubated in
4-methylumbelliferyl--D-galactoside substrate for 10 min
at room temperature. Fluorochrome was eluted from immunocaps with
sodium carbonate buffer, and fluorescence of the eluate was measured at
450 nm.
Statistical analysis
Where appropriate, data were analyzed for statistical significance using either ANOVA with Bonferroni’s multiple comparison test, or the unpaired two-tailed t test. A confidence interval of 95% was used for all tests.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DNA encoding the Fd and L chain was amplified from cDNA using PCR. A phage display library containing IgE Fd and L chains was constructed using the vector MCO3. The size of the library was 4.4 x 107 primary clones with an insert frequency of >99%, as determined by titering E. coli XL-1 blue after transformation and comparison with mock ligations (data not shown). Presence of H chain and L chain inserts was verified using PCR (data not shown).
Sequence analysis of random clones
The rearranged H chain V genes of 51 primary transformants from
the library were sequenced, and the closest matching germline V, D, and
J gene segments were aligned. All rearrangements were inframe,
functional rearrangements, and all clones were of the IgE isotype (data
not shown). Of 51 rearrangements sequenced, 31 unique CDR3 were
observed. A total of 19 rearrangements had CDR3 sequences shared by
other rearrangements, but differed in the number or position of point
mutations across the V gene, and were therefore considered to be
independent sequences. Only two pairs of rearrangements (two
VH3 and two VH6
rearrangements) shared identical CDR3 and mutation pattern, and were
considered to be duplications of the same sequences. The two
duplications were not included in further analysis. There was a higher
degree of shared CDR3 sequences in rearrangements of the
VH6 family, as only 4 of 17 had unique CDR3
sequences. Table I shows the number of
unique CDRs, and the number of unique individual rearrangements.
|
H chain family usage was 16% (8 of 49) VH1,
39% (19 of 49) VH3, 12% (6 of 49)
VH4, and 33% (16 of 49)
VH6, and in total 11 different V gene segments
were used. Tables II and III show the
closest matching germline V, D, and J gene segment for each clone. In
all, 16 different IGHD genes were identified in the IgE rearrangements.
Interestingly, a number of the VH6 rearrangements
also used the IGHD7-27 D gene. There was no IGHD7-27 D gene usage in
VH1 or the VH4
rearrangements. No germline D gene segment could be aligned for clone
47. All J genes except for IGHJ1 were used. VH1,
VH3, and VH4 rearrangements
used a range of IGHJ genes including IGHJ6*02. In contrast,
VH6 rearrangements did not use the IGHJ6*02
segment and were dominated by IGHJ4*02. Clone 47 used the IGHJ5*01 gene
segment.
|
|
VH6 rearrangements demonstrated a low level
of N nucleotide addition at the D-J junction (2.06 ± 1.83;
mean ± SD) (Fig. 2). Using a
one-way ANOVA, this was found to be significantly different from N
addition at the D-J junction of VH1 (9.13 ±
7.88; p < 0.01), of VH3
(7.90 ± 4.15; p < 0.01), and
VH4 rearrangements (8.17 ± 4.49;
p < 0.05). The VH1,
VH3, and VH4 families
showed no significant difference in N additions at the D-J junction
when compared with each other.
|
The number of amino acids encoded by the CDR3 was determined for
each rearrangement, and the results are presented in Fig. 3. There was a significant difference
between VH6 rearrangements (11.9 ± 0.25;
mean ± SD) when compared with VH1
(17.5 ± 3.74; p < 0.001),
VH3 (14.2 ± 3.39; p <
0.001), and VH4 rearrangements (15.2 ±
4.17; p < 0.001). Statistical analysis was performed
using an unpaired, two-tailed t test.
|
All rearrangements except one VH6
rearrangement showed the presence of point mutations occurring across
the V gene. The percentage of identity with germline V genes for each
rearrangement is summarized in Tables II
and III. The degree of mutation was markedly
different between VH6 (1.36% ± 1.23; mean
± SD) rearrangements and rearrangements from other families. Fig. 4 represents a scattergraph of the
percentage of mutations for each family. The differences, with respect
to the VH6 family, were shown to be highly
significant using a one-way ANOVA, for VH1
(6.56% ± 2.08; p < 0.001), VH3
(4.92% ± 2.01; p < 0.001), and
VH4 rearrangements (7.42% ± 3.57;
p < 0.001).
|
The IgE Ab library was panned against immobilized ovomucoid over
five successive rounds. Ninety-six clones per round were screened for
ovomucoid binding in an ELISA. The clones were also tested for binding
to the blocking agent (skim milk), and all clones that bound ovomucoid
also demonstrated binding to skim milk. Antiovomucoid and skim milk
clones were observed in every panning round, peaking in number at round
3. Fig. 5 shows the binding of 12
positive clones from rounds 4 and 5, as well as a negative clone (4-85)
representative of the 468 negative clones. The 12 positive clones were
then subjected to further analyses. The original assay was repeated
(data not shown) and then clones were tested for binding to BSA in an
inhibition ELISA. The specificity of the binding was further explored
for the 12 clones in a number of inhibition assays with different
combinations of immobilized Ag, and Ag in the liquid phase. Binding to
immobilized ovomucoid could be inhibited by incubating with BSA and
skim milk protein (Fig. 6), and binding
to immobilized skim milk protein could be inhibited by incubating with
skim milk (data not shown). Binding to immobilized ovomucoid could be
inhibited by incubating with ovomucoid in supernatant. This assay was
performed on a limited number of clones (clones 5, 8, 37, and 38), and
the results are presented in Fig. 7.
Further investigations of Fab binding were prevented by our inability
to scale up cultures of the VH6 clones, perhaps
as a result of intracellular binding by highly reactive Fabs, leading
to cell death. Subsequent investigations of the binding were confined
to studies using cyclic peptides designed from the CDR3 regions of a
multiply reactive ovomucoid-binding and a control Fab.
|
|
|
The H chain V gene amino acid sequences are presented in Fig. 8 for each of the 12 ovomucoid-binding
clones. Sequence analysis showed that they all use the V gene segment
IGHV6-1, the only member of the VH6 family. Four
clones were identical. The 12 clones isolated by panning had CDR3
regions, and patterns of mutation previously seen in rearrangements
identified in the random sequencing experiment (data not shown). Nine
of the twelve clones were encoded by the IGHD7-27 D segment, and all
clones use the JH4b J gene (data not shown). All
of the polyreactive clones also had a consensus CDR3 sequence,
consisting of
DL(P)S(G)G(D)SYP(S)S(G)F(H)FDY, in
which amino acids common to all CDR3s are presented in boldface type.
In all clones, the CDR3 was 12 aa long. A variety of L chains was used
by the various polyreactive clones (data not shown).
|
To investigate the role of the H chain CDR3 in binding to
ovomucoid and other Ags, a cyclic peptide (CP-1) was designed from a
CDR3 sequence of several of the ovomucoid-binding IgE Fabs. A control
peptide (CP-N) was designed from the CDR3 of a
VH3-utilizing clone (clone 46). Using an ELISA,
the peptide CP-1 gave ODs at least 2-fold greater than control wells
(Ag free) when incubated with ovomucoid, tetanus toxoid, human IgG, and
human as well as bovine vWF (Figs. 9 and 10). The control peptide CP-N showed no
elevated ODs in the ELISA against any Ag specificity. Results of the
ELISA were confirmed in all cases by dot-blot assay (data not
shown).
|
|
).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The use of particular V genes has been noted in a range of diseases, including autoimmunity (44, 45, 46, 47, 48) and malignancy (49). IGHV6-1, for example, has recently been associated with idiopathic thrombocytopenic purpura (50). IGHV6-1, the most D-proximal VH gene segment (51), has also been found to be commonly expressed by fetal B cells (52, 53), and this expression pattern may be associated with its position on chromosome 14 (52).
This is not the first report of the use of the VH6 gene family in the IgE repertoire. IGHV6-1 has been identified in IgE rearrangements among 9 of 10 patients suffering from a variety of allergic diseases, including atopic dermatitis (54). The other small H chain V gene family (VH5) has also been observed encoding IgE Abs (22, 23, 24, 27). Like the VH6 gene family, the VH5 gene family (55) is rarely observed in the adult B cell repertoire (31). In one study, some V gene rearrangements lacked mutations and were in germline configuration (23) similar to results presented in the present study.
A number of D genes were identified in the IgE library. Most notable was the IGHD7-27 gene segment. This segment has been associated with Abs produced by B-1 cells, including fetal B cells (52, 56). The IGHD7-27 D gene has been shown to be present in up to 40% of first and second trimester fetal Ab transcripts; however, it is found in less than 1% of adult Ab transcripts (57). The bias for IGHD7-27 rearrangement among B-1 cells is thought to be due to its position, as it is the most J-proximal D gene (58). Given the low abundance of IGHD7-27-related Ab transcripts in adults, the frequent presence of the IGHD7-27 D gene in the IgE V genes reported in this work is striking. There was a strong correlation between IGHD7-27 D segment and IGHV6-1 gene usage, with 6 of the 16 unique VH6 clones using both segments. IGHD7-27 was also observed in 1 of 19 VH3-utilizing clones. Two studies have previously reported the existence of IGHD7-27 D genes encoding IgE H chains. One study observed only 1 of 29 rearrangements that used IGHD7-27 (28), while a second study observed 3 of 74 IgE rearrangements that used IGHD7-27 (22).
Further sequence analysis revealed that the VH6 rearrangements had other interesting features that were significantly different from the rearrangements involving other V gene families. The CDR3 lengths of the VH6 rearrangements were shorter than the CDR3 of VH1, VH3, and VH4 rearrangements. There are several possible factors that may have contributed to the short VH6 CDR3s: a lack of N addition; the frequent use of the short IGHD7-27 D gene; the lack of the long IGHJ6 J gene segment in any of the VH6 rearrangments. It is also possible that the CDR3 size observed reflects Ag selection for a particular CDR3 sequence, as all the VH6 CDR3s were highly related.
In this study, there was also a significantly lower number of point mutations observed in VH6 rearrangements compared with rearrangements from the other families. One VH6 rearrangement had no evidence of mutation, suggesting that affinity maturation by mutation and selection of mutants with higher affinities may not always be occurring in the IgE response of this individual.
The library of H and L chain V genes was expressed as phage Ab fusions and phage were selected on the basis of ovomucoid binding. Surprisingly, the ovomucoid-binding Fabs were also able to bind wells coated with skim milk protein and BSA, suggesting that these clones may be capable of binding to a number of dissimilar Ags.
Sequencing the V genes of each of the 12 ovomucoid/BSA-binding IgE Fabs showed that they were encoded by the IGHV6-1 V gene segment. Several of the 12 clones had rearranged V genes and H chain CDR3s that were identical to the VH6 rearrangements identified in the earlier random sequencing experiments. The IGHD7-27 D gene was found in 9 of 12 clones. All of the isolated IgE Fabs appeared to have a conserved motif within their CDR3 region and shared common amino acids.
The fact that all 12 ovomucoid-binding clones carried different L chains suggested that the L chain CDRs might not be participating in Ag binding. Polyreactivity has, in fact, been reported to sometimes arise solely from Ag interaction with the H chain CDR3. To investigate the contribution of the H chain CDR3s to ovomucoid binding, a cyclic peptide (CP-1) was designed based on the CDR3 sequence encoded by several of the ovomucoid-binding clones. In an ELISA, the peptide was shown to bind to a range of human and nonhuman Ags, including ovomucoid, while a control peptide (CP-N), based on a different CDR3 sequence, did not bind to any of the test Ags. Binding of CP-1 to BSA was not observed, despite the fact that the IgE Fabs bound BSA. This suggests BSA binding by IgE Fabs may require the other H chain CDRs and the L chain.
The results reported in this work are consistent with other studies, in which the polyreactivity of some Abs has been shown to be dependent upon the sequence and structure of the H chain CDR3. This has been shown for murine CD5+ B cells (59), mAbs derived from human CD5+ lymphocytes (60, 61, 62), leukemias of CD5+ B cells (63), and polyreactive recombinant Abs isolated from phage display libraries constructed from adult lymphocytes (64). The exact features of the H chain CDR3 that allow polyreactive binding are unknown; however, the short length of the CDR3 may be an important contributing feature.
It is perhaps significant that the IgE Fabs reported in this work were derived from a library constructed from the lymphocytes of a patient with multiple sensitivities. The patient reacted to many different allergens by radioallergosorbent assay, including egg white proteins, house dust mite, peanut extract, and numerous pollens. It is possible that polyreactive IgE may be, in part, responsible for the multiple reactivities observed. Certainly there is good reason to accept the relevance of H chain CDR3-binding properties to the patient’s egg sensitivity, for the cyclic peptide CP-1 was shown to inhibit serum IgE binding to ovomucoid. This suggests that the patient’s serum IgE Abs target the same epitopes as those that can be bound by the Fab-derived cyclic peptide.
No monoreactive IgE Fabs were isolated from this study, but their presence in the repertoire of this allergic patient cannot be fully excluded. A possible explanation as to why these Abs were not isolated could be due to competition for Ag during panning. Monoreactive IgE, if present in the library, could possibly have been out-competed by the large number of potentially polyreactive Abs. It is also possible that the conditions used in the panning of this library favored polyreactive rather than monoreactive Abs, and more stringent conditions may be needed to isolate high affinity monoreactive IgE.
The results presented in this study raise interesting issues regarding
B cell ontogeny and the genesis of the IgE repertoire. However, it may
also have important implications for the sensitization of mast cells
and basophils and the etiology of allergy. Mast cell and basophil
degranulation is caused by FcR1 aggregation that occurs when
cell-bound IgE binds to allergen. The interaction between surface IgE
and allergen is the key to unleashing the mast cell arsenal of
proinflammatory mediators. It has been shown in vitro, using the rat
basophil leukemia cell line, that degranulation is dependent upon the
valency of the Ag (65, 66). If polyreactive IgE Abs exist,
FcRI aggregation of mast cell- and basophil-bound IgE could more
easily occur. A much greater range of both self and nonself Ags could
also be responsible for degranulation. Polyreactive IgE may therefore
help explain why patients suffering from certain allergies, such as
atopic dermatitis, are often reactive to a wide range of allergens.
Further analysis of additional patients is now needed to clarify the
existence of polyreactive IgE Abs, the extent to which they contribute
to IgE production in different allergic disorders, and the source of
the B cell clones that produce them.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Andrew Collins, School of Microbiology and Immunology, University of New South Wales, Sydney, New South Wales, 2052 Australia. E-mail address: a.collins{at}unsw.edu.au
3 Abbreviations used in this paper: CDR, complementarity-determining region; vWF, von Willebrand factor.
Received for publication March 27, 2001. Accepted for publication April 19, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RI. Nature 402:B24.[Medline]
chain cDNA. Nucleic Acids Res. 11:3077. chain cDNA. Nucleic Acids Res. 11:719. VH5 transcripts from three unrelated patients with atopic dermatitis. J. Exp. Med. 177:99.. Science 246:1275. and heavy chain gene usage in early untreated systemic lupus erythematosus suggests intensive B cell stimulation. J. Immunol. 163:1027.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
