| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Biophysics, Graduate School of Science, Kyoto University, Kyoto 606-8224;
Department of Immunobiology, Faculty of Applied Biological Science, Hiroshima University, Hiroshima 739; and
Department of Anatomy, Kumamoto University Medical School, Kumamoto 860, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The chicken B cell immune system is attractive because of its apparent simplicity. A single set of unique functional segments of both Ig heavy and light (L) chain genes is rearranged during early embryogenesis to generate a pool of B cell progenitors that will be diversified in the bursa by gene conversion, forming the preimmune repertoire (7, 8). Thus, every Ig rearrangement unbiased for particular specificities in a single GC can be analyzed by PCR. We have shown that postbursal B cells stimulated by Ags are able to generate somatic variants in splenic single GCs by both gene conversion and point mutations (9).
In chickens, the GC reaction has maximized by day 7 of primary
responses and has begun to wane by 14 days postimmunization (data not
shown). Since the earliest formation of GCs in chicken occurs 48 to
72 h after the injection of protein Ag (10), B cells in the GC at
day 7 have spent 
100 h in proliferation. To analyze the kinetics of
the accumulation of gene conversion and point mutations as evidence for
selection in the primary Ab response, we have chosen two time points,
at day 7 (100 h) and day 11 (200 h) postimmunization.
Here, we analyze the nucleotide sequences of the chicken Ig L-chain that are subjected to templated mutations by gene conversion from a restricted set of 25 pseudogene donors (7). We amplified L-chain genes in single GCs by PCR and used a computer-adaptable method for definitive assignments of gene conversion. This assignment separated base modifications brought by gene conversion from point mutations. Our studies show extensive diversification of B cells by gene conversion in the very early phase of GC development followed by a reduction in clonal complexity. This represents the oligoclonal development of the B cell repertoire generated by mutation and selection, which may lead to clonal evolution in each GC.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Eleven-week-old White Leghorn H-B 15 chickens (11) were immunized with a single i.v. injection of 100 or 500 µg of FITC-BSA. At 7 and 11 days after immunization, a single GC containing several thousand B cells was isolated as described (12) and transferred to a microcentrifuge tube, washed several times with PBS, and stored at -20°C.
DNA preparation, DNA amplification, and sequencing
DNA preparation, PCR amplification of VJ rearranged sequences, cloning in pUC 119, and sequencing were done as described previously (9) with a few modifications. To minimize PCR-introduced artificial mutations, we used Pfu polymerase having proofreading activity (13). The crude lysate (100 µl) was divided into five portions (20 µl) and independently subjected to two rounds of PCR using pairs of nested primers. The first round of amplification of 30 cycles was carried out in the same reaction tube in a 50-µl volume. The second round of amplification of 30 cycles was carried out in separate reactions using 2 µl of the first round reaction mixture in a 50-µl volume except 20 cycle amplification for clones 41201 to 41214 from GC 4 and 51210 to 51214 from GC 5. PCR reactions were monitored by the PCR product defined as a visible band of expected length on an ethidium bromide-stained gel. The PCR product reached its maximum after 30 cycles of amplification. Five simple crossovers during in vitro amplification were found in 98 DNA clones and eliminated from further studies.
Procedure of the gene conversion search
The putative precursor sequences shared by PCR sequences were aligned to the germline sequence, and PCR sequences to the precursors according to the method of Needleman and Wunsch (14). To assign the base modifications identified in these query sequences to templated gene conversion events, we developed a computer program, "conversion search." The Fortran 77 programs used in this study were written for the Unix workstation, SPARCstation 4 computer (Sun Microsystems, Palo Alto, CA). Since the smallest gene conversion event in chicken is no longer than 8 bp (15), homology search in the Ig L-chain pseudogene mini-database was performed with all possible sets of at least 8-bp query sequence containing the identified mutation in the range of 30 bp in both directions.
If this computer-based approach found more than two candidates as gene conversion donors, the longest one having at least two base modifications was selected as the final candidate of gene conversion donor. Thus, linked base modifications having counterparts in the pseudogene pool were assigned as templated gene conversions. Other single base changes were assigned as point mutations.
Calculations of the average number of base modifications, point mutations, and gene conversions in genealogic trees
The average number of base changes and gene conversion events among related sequences in genealogic trees was calculated according to the principle of unweighted pair-group method with arithmetic mean (16), as shown previously (9).
Construction of evolutionary trees
From PCR nucleotide sequences, the corresponding amino acid sequences were deduced. Optimal alignments of amino acid sequences were obtained by the methods of Needleman and Wunsch (14), and Berger and Munson (17), together with manual inspections. To construct the evolutionary trees, the evolutionary distance parameter k was calculated by the simple Poisson model correction as k = -ln (1 - K), where K represents the amino acid difference per residue between sequences compared (18). In this study, the evolutionary distance is shown as 100 k or corrected amino acid changes per 100 residues. The evolutionary trees were inferred by the neighbor-joining (NJ) method (19).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Eleven-week-old chickens were immunized i.v. with 100 or 500 µg of FITC coupled to BSA. On days 7 and 11 of primary immune responses, single GCs encased in a smooth connective tissue capsule were removed from the arterial tree of spleen using very fine forceps and analyzed as described (9, 12).
Since chickens have a single functional V
1 segment and 25
pseudogenes which serve as gene conversion donors for rearranged
L-chain loci (7), we used a PCR to amplify all V1-J fusions from
B cells isolated from individual GCs. To minimize PCR-introduced
artificial mutations we used Pfu polymerase, which has about
a 10-fold higher fidelity than Taq polymerase (13).
Amplified DNAs were cloned into plasmid vectors and the V-region
inserts of individual PCR clones were sequenced. Among a total of 93
L-chain rearrangements obtained from 7 single GCs, only 2 siblings of
repeated sequences were found (Table I
).
|
).
Sequences related by a unique VJ joint have likely originated from the
same precursor cell. Thus, the total number of related groups and
independent clones gives the minimum number of unique VJ junctions
present in single GCs. The percentage of unique VJ junctions present in
the analyzed cohort represents the clonal complexity of each GC (Table I
). As the immune response progressed from day 7 to day 11, the clonal
complexity decreased. This reduction may be largely due to interclonal
competition and intraclonal expansion.
|
To reveal the clonal expansions present in the early GCs, clonally
related L-chain sequences (group 4-I) obtained from the single day 7
GC, GC 4, were aligned with the germline sequence of the CB inbred
chicken (7) (Fig. 2). Putative precursor
segments shared by clonally related genes have been heavily mutated
(Fig. 2A). For the quantitative assignments of gene
conversion, a computer program "conversion search" was developed
for this study. Three conversion events were identified in the
precursor V1-J sequence; this number is comparable with an
average of four to six conversion events at the end of the bursal
maturation phase (20). No untemplated point mutations were observed in
this rearrangement. Therefore, this germinal center precursor cell
sequence (GCPC) is consistent with the sequence expected for a GC
progenitor B cell that has emigrated from the bursa. In fact,
pseudogenes used in this progenitor reflect the preferential bursal
usage of gene segments located in the inverted transcriptional
orientation with respect to the functional V1 gene (15).
|
B). The mutations in this collection of genes
included both templated gene conversions and point mutations. The
mutational patterns identified in GC 4 reflect an intraclonal
generation of Ab mutants during the expansion of individual B cell
clones and can be represented in the form of genealogic trees (Fig. 3). There was no biased usage for the
pseudogene segments used in gene conversion events.
|
Figure 4 depicts
clonally related sequences of L-chain (groups 5-I, 5-II, and 5-III)
obtained from a single day 11 GC, GC 5. Each GCPC carried many point
mutations (two for 5-I, six for 5-II, and three for 5-III) as well as
distinct conversions (five for 5-I, three for 5-II, and eight for
5-III) (Fig. 4, A, C, and E). Since
point mutations are common in splenic GCs (9) but very rare in bursal
follicles (7), these VJ segments may represent intermediates in the
clonal diversification ongoing in GCs. Pseudogene donor preference
differed between these precursors and the progenitor B cell of the
early GC with respect to the polarity. This suggests that these new
conversions were introduced in the GC. Thus, unique or shared
nucleotide changes of clonally related L-chain sequences obtained from
the day 11 single GCs may reflect the mutational patterns specified in
the late GC development (Fig. 4, B, D, and
F).
|
A. Clonally related L-chain
sequences obtained from other single GCs (GC 6, GC 7, and GC 8) are
also shown in the form of genealogic trees (Fig. 5, B–D).
Two precursor sequences of GC 6 were heavily mutated by point mutations
(at least nine for 6-I and four for 6-II) as well as gene conversions
(at least four for 6-I and six for 6-II).
|
The number of mutation events was calculated for each pair-group
of sequences shown in the genealogic trees (Figs. 3 and 5). The average
number of gene conversion events and point mutations in each pair-group
of clones (Fig. 3) is 4.6 and 6.2, respectively, for the day 7 L-chain
group 4-I (Table II). GCPC obtained from
the day 11 single GCs, GC 5 and GC 6, contained 5.2 conversion events
and 4.8 point mutations on average (Fig. 5). The 5.2 conversion events
of GCPC include 3 bursal events as shown for the GCPC of group 4-I
(Fig. 3). These estimations suggest that the day 11 GCPC is an
intermediate generated shortly before day 7. Thus, the mutational
events observed in each pair-group of these day 11 GC B-cells are
specified in the late stage of GC development up to day 11. As shown in
Table II, point mutations are common in both developmental stages of
the GC, but gene conversion events are strongly suppressed or
counterselected in the late stage.
|
Since proteins are molecules carrying out physiologic functions,
information on amino acid substitutions is useful for elucidating the
interclonal competition and clonal expansion based on the ability of
variant Ig to bind different epitopes of Ag. Thus, we translated PCR
sequences to the corresponding amino acid sequences. Lethal somatic
mutations were found in 3 of 44 IgL sequences from the 3 early GCs and
in 6 of 49 sequences from the 4 late GCs. We searched amino acid
sequence similarities between a pair of neighbor sequences and
constructed a unique evolutionary tree under the principle of minimum
evolution according to the NJ method (19) (Fig. 6). The NJ method provides not only the
topology but also the horizontal branch lengths representing
evolutionary distance of the final tree. Extensive polyclonal
diversification in the early GCs (Fig. 6A) and the
oligoclonal expansion and the decreasing clonal complexity in the late
GCs (Fig. 6B) are confirmed in accordance with the results
in Table I. On average, the IgL sequence was not different in
evolutionary distance from the germline but differed in the standard
deviation between the two developmental stages of GC: 22.9 ±
3.2% for the early GCs and 22.3 ± 6.8% for the late GCs. Both
bursal and GC diversifications contribute equally to the evolutionary
distance at the periphery. This suggests that the variant amino acid
sequences that define the Ag specificity are brought out to a similar
degree on each clone by gene conversions performed in the bursa and the
early GCs whereas fine tuning of mutations conferring higher affinity
is carried out by untemplated point mutations and phenotypic selection
in the late GCs. Frequent events of gene conversion may vitiate the Ab
paratope and not necessarily be utilized in the late GCs. Thus, the
divergence of sequence from the germline varied between individual
clones from the late GCs.
|
and 5) are also highly homologous in amino acid sequence
and assembled in neighboring branches of the evolutionary tree (Fig. 6). Since the primary structure responsible for Ag binding is located
in the complementarity-determining regions (CDRs) of each Ig chain, we
examined average evolutionary distances and standard errors between
clones of the different groups and those between clones of the same
group in the region corresponding to germline CDR of Ig L-chain (26
amino acid residues of CDR1, CDR2, and CDR3) as shown in Tables
III and IV.
Intraclonal diversity was confined to <22%, while interclonal
difference expanded to 33 to 88%. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
100 h in proliferation (10). Assuming that the doubling time
of GC B cells is the same as for bursal cells, 10 h (21), we
estimate that 10 cell divisions are cycled in both phases of induction,
early phase by 7 days and late phase by 11 days postimmunization.
Interestingly, the average number of base modifications of the Ig
L-chain genes present in the early GCs (day 7) was different from the
previous estimations with NP-BSA (9). Immunization with FITC-BSA
produced a fivefold higher frequency of base modifications (23.8 base
changes) (Table II) than did NP-BSA (4.5 base changes). The average
number of gene conversion events of Ig L-chain occurring during the
estimated 10 cell division cycles of early GC development was 4.6 in
the GCs induced with FITC-BSA (Table II) but only 1.3 after NP-BSA
immunization. The frequency of conversion events with NP-BSA was
comparable with that observed for bursal development alone (20). The
high frequency of base modification seen after FITC-BSA immunization is
brought about by frequent de novo gene conversion events. Since FITC is
a more complicated large molecule than NP, it is expected to have more
epitopes.
We defined the clonal complexity of GC populations as the percentage of
unique VJ junctions present in all VJ sequences analyzed. The average
clonal complexity for L-chain rearrangements in the early GC B cells
was as high as 86% in 3 GCs induced by FITC-BSA immunization (Table I)
and 63% in 4 GCs driven with NP-BSA (9). We found that each GC
contains 5000 B cells at 7 days after immunization (data not shown);
thus, each independent VJ clone may well be considered representative
of a minor population composed of hundreds of GC B cells. The number of
precursor cells homing to individual GCs is estimated at more than 20
from discrete V-J junctions (Table I). Polyclonal GC precursor cells
have been shown in human tonsils (22) and mouse spleen (5), although
these studies underestimate polyclonality by restricting analysis to
particular V to J rearrangements. In the late GCs, average clonal
complexity decreased by half, to 35% (Table I).
We were able to measure base modifications brought by gene conversion
and by point mutations separately; there are no uncertainties for
assignments of point mutations in L-chain genes since 1) every
pseudogene donor sequence is known (7), 2) every gene conversion event
is excluded by a newly developed computer program, and 3) frequencies
of PCR artifacts are made insignificant by using Pfu
polymerase. Thus, the rate of point mutation per base per generation is
estimated to be 1.7 x 10-3 during the early phase of
the GC reaction (6.2 base changes for the 355-bp region in 10 cycles)
(Table II). This mutation rate persisted during the late phase of
induction. These values are close to the somatic hypermutation rate of
10-3 during the course of murine immune responses
(2, 5, 23).
The gene conversion mechanism utilized for bursal diversification also
appears to be reactivated in GC B cells in the very early phase of Ag
stimulation (Table II). As shown in the GC B cells induced by NP-BSA
immunization (9), pseudogene donor segments used for postbursal
diversification after FITC-BSA immunization were scattered in the
L-chain locus and independent of the proximity to V1 and their
relative orientation. Thus pseudogene usage in GCs was distinct from
the bursal usage, suggesting the preference for L-chain expression.
Alternatively, another gene conversion mechanism exploiting an
extrachromosomal pseudogene donor may be involved (11). Interestingly,
recent studies in mice immunized with hapten carrier showed that RAG-1
and RAG-2 expression is reactivated and V(D)J recombination is induced
in GC B cells after encounter with Ag (24, 25). In the later phase of
the GC reaction, novel gene conversion events were very rare (Table II). This suggests that the conversion mechanism is down-regulated in
the mature GC environment or selected against an unwanted event.
We showed extensive diversification in the amino acid sequence of Ig
L-chains in the early GCs (Fig. 6A). In the late GCs, clonal
complexity decreased due to predominant interclonal competition for Ag,
and B cells bearing particular specificities characterized by different
CDR sequences showed oligoclonal development (Fig. 6B; Table IV). The oligoclonal development of B cells found in chicken single GCs
may represent an intermediate step of selection leading to the
acquisition of high affinity variants reactive to a single epitope.
Alternatively, oligoclonal B cells recognizing different epitopes of Ag
may develop in a single GC and possibly contribute to expanding immune
repertoire of memory B cells.
|
|
|
|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 GenBank accession numbers for the sequences reported in
this paper are AB003811 to AB003903.
3 Address correspondence and reprint requests to Dr. H. Yamagishi, Department of Biophysics, Graduate School of Science, Kyoto University, Kyoto 606-8224, Japan. E-mail address:
4 Abbreviations used in this paper: GC, germinal center; L, light; CDR, complementarity-determining region; GCPC, germinal center precursor cell sequence; NJ, neighbor-joining; NP, (4-hydroxy-3-nitrophenyl)acetyl.
Received for publication October 14, 1997. Accepted for publication January 5, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-J and novel joints of chimeric V pseudogenes on extrachromosomal circular DNA from chicken bursa. Eur. J. Immunol. 23:245.[Medline]
This article has been cited by other articles:
|
|
 |
|
  H. Arakawa and J.-M. Buerstedde Activation-induced cytidine deaminase-mediated hypermutation in the DT40 cell line Phil Trans R Soc B, March 12, 2009; 364(1517): 639 - 644. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Dibner, J. D. Richards, and C. D. Knight Microbial Imprinting in Gut Development and Health J. Appl. Poult. Res., January 1, 2008; 17(1): 174 - 188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. D'Avirro, D. Truong, B. Xu, and E. Selsing Sequence Transfers between Variable Regions in a Mouse Antibody Transgene Can Occur by Gene Conversion J. Immunol., December 15, 2005; 175(12): 8133 - 8137. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Mehr, H. Edelman, D. Sehgal, and R. Mage Analysis of Mutational Lineage Trees from Sites of Primary and Secondary Ig Gene Diversification in Rabbits and Chickens J. Immunol., April 15, 2004; 172(8): 4790 - 4796. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Arakawa, K.-i. Kuma, M. Yasuda, S. Ekino, A. Shimizu, and H. Yamagishi Effect of Environmental Antigens on the Ig Diversification and the Selection of Productive V-J Joints in the Bursa J. Immunol., July 15, 2002; 169(2): 818 - 828. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. L. Lundqvist, D. L. Middleton, S. Hazard, and G. W. Warr The Immunoglobulin Heavy Chain Locus of the Duck. GENOMIC ORGANIZATION AND EXPRESSION OF D, J, AND C REGION GENES J. Biol. Chem., December 7, 2001; 276(50): 46729 - 46736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Sonoda, M. Takata, Y. M. Yamashita, C. Morrison, and S. Takeda Homologous DNA recombination in vertebrate cells PNAS, July 17, 2001; 98(15): 8388 - 8394. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Winstead, S.-K. Zhai, P. Sethupathi, and K. L. Knight Antigen-Induced Somatic Diversification of Rabbit IgH Genes: Gene Conversion and Point Mutation J. Immunol., June 1, 1999; 162(11): 6602 - 6612. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Schiaffella, D. Sehgal, A. O. Anderson, and R. G. Mage Gene Conversion and Hypermutation During Diversification of VH Sequences in Developing Splenic Germinal Centers of Immunized Rabbits J. Immunol., April 1, 1999; 162(7): 3984 - 3995. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Diaz, A. S. Greenberg, and M. F. Flajnik Somatic hypermutation of the new antigen receptor gene (NAR) in the nurse shark does not generate the repertoire: Possible role in antigen-driven reactions in the absence of germinal centers PNAS, November 24, 1998; 95(24): 14343 - 14348. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
