| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Guy’s, King’s and St. Thomas’ Medical School, Department of Histopathology, St. Thomas’ Campus,
Guy’s, King’s and St. Thomas’ Medical School, The Rayne Institute, St. Thomas’ Campus, and
Guy’s, King’s and St. Thomas’ School of Dentistry, Department of Oral Pathology, Guy’s Campus, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The diversity of cells of B lineage can be investigated by analyzing their Ig heavy chain genes (IgH)3 (10, 11). The diversity of human Ig genes is achieved through the processes of gene rearrangement and somatic hypermutation. As Ig variable (IgVH), diversity (DH), and junctional (JH) gene segments rearrange in the bone marrow, additional nucleotides (N nucleotides) are added randomly in the junctions by the enzyme TdT. In this way, a unique junctional complementarity-determining region (CDR) sequence, CDR3, is created. Therefore, cells sharing the same CDR3 sequence are considered to be progeny of a single clone, which may have been amplified during a germinal center response (12). IgH genes of B cells can also undergo somatic hypermutation after activation by Ag, thus naive B cells have germline IgH gene sequences, whereas mutated Ig gene sequences are seen in cells that have been through a germinal center reaction (13, 14). The mean number of mutations seen in postgerminal center cells of the tonsil, spleen, and lymph node vary between 5.9 and 7.5 mutations per IgVH gene (15). IgA genes generally have higher numbers of mutations than IgG genes, which in turn have a higher number of mutations than IgM genes (16, 17, 18). We have previously found, in humans, that B cells and plasma cells from the duodenum, ileum, and colon carry IgH genes that have a much higher level of mutation (10, 11, 17, 19). The reasons for the higher level of mutation are not yet known.
We have used RT-PCR, cloning, and sequencing to obtain sequences of IgM and IgA genes from plasma cells in human salivary gland. IgM and IgA genes were also obtained from plasma cells in human duodenal lamina propria and splenic red pulp, as examples of mucosal and systemic populations, for comparison. To examine the local diversity of Ab-forming cells, separate samples were taken from each tissue, and CDR3 regions were compared with each other, within and between isotypes, to identify related cells. The sequences were compared with published germline sequences to determine the extent of somatic hypermutation, and the CDR3 junctional regions of each gene rearrangement were analyzed for JH and DH usage and the number of N additions. Our results show that the IgH genes from the salivary gland have more in common with the IgH genes from duodenal plasma cells than the IgH genes from splenic plasma cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
|
1-mm3 quantities of each
tissue, prepared by chipping the block while still frozen. Efforts were
made to avoid sampling any organized lymphoid tissue; prior
histological examination of sections from the same block for duodenum
and salivary gland and selection of red pulp only for spleen. cDNA was
prepared using Moloney murine leukemia virus reverse transcriptase and
oligo(dT) primers. IgVH4–34 genes were PCR
amplified using a seminested strategy as previously described
(17). Briefly, 5' primers specific for
VH4 family leader region
(5'-ATGAAACACCTGTGGTTCTT-3') and VH4–34 gene
(5'-AGCTACAGCAGTGGGGCG-3') were used in conjunction with a 3'
primer specific either for Cµ, (5'-CAGGAGACGAGGGGGAA-3') or for C
(5'-GGAAGAAGCCCTGGACCAGGC-3') and the PCR products were purified
and cloned into pGEM-T vectors (Promega, Madison, WI) for
sequencing.
Sequencing was either done by Qiagen (Dorking, U.K.) or conducted using
an Applied Biosystems 377 sequencer (Foster City, CA) and dye
terminator cycle sequencing kits according to manufacturer’s
protocols. Analysis of sequences was done with the aid of Genejockey II
sequencing software and the V Base sequencing directory (Medical
Research Council Center for Protein Engineering, Cambridge, U.K.).
Sequences were considered to be related if they shared the same CDR3
regions but had differences in the number of point mutations. When the
frequency of mutations per Ig gene was compared between groups, the
mean number of mutations for a family of sequences was counted once.
DH regions were identified using the criteria of
a minimum of 5-bp identity over a 6-bp length of sequence. Where a
sequence showed identity with more than one
DH region, all possible
DH regions in the CDR3 were identified, and the
ones that could be assigned without overlap, and with the minimum
number of nucleotides in between, were used. Sequences are available
from GenBank/EMBL/DDBJ databases under the accession numbers AJ253020
to AJ253112. Comparisons of DH and
JH usage were done using
2 analysis. All other data was tested for
normality using Microstat software. Normally distributed arrays were
then compared using Student’s t test. Data that was not
normally distributed was analyzed by the Mann-Whitney U
test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), indicating that there is
dissemination of related plasma cells or their precursors. Examples of
this were seen in all three tissue types. In the salivary gland,
related sequences associated with both C and Cµ were identified in
the different sites studied within each sample. All three of these
genes were highly mutated, indicating that class switching can occur
after diversification of the gene by somatic hypermutation.
|
Fig. 2 shows the number of mutations
in each gene of IgA and IgM isotype in all three tissues. There is a
significant increase in the number of mutations in Ig genes that have
switched to IgA, compared with the number of mutations in IgM genes.
When comparing the frequency of mutation between genes from different
tissues, but the same isotype, it can be seen that the mean number of
mutations in IgM genes from spleen (6.7) is lower than that from
salivary gland (18.3) and the duodenum (16.2). These differences are
significant (p = 0.006 and p =
0.005, respectively). Similarly, IgA genes from spleen had a lower mean
number of mutations (14.9) than IgA genes in salivary gland (28.9) and
duodenum (21.7), although only the former difference reached
significance (p = 0.0002). There were no
significant differences in the number of mutations between
isotype-matched IgVH genes from salivary gland
and duodenum.
|
There was no difference in the mean length of CDR3 regions in
IgVH genes of the two different isotypes in any
tissue. Neither was there any difference in the mean CDR3 length of
IgVH genes from salivary gland and duodenum.
However, the mean CDR3 lengths of IgVH genes from
both salivary gland and duodenum, for both isotypes, were significantly
shorter than that of the spleen (Fig. 3).
|
shows the results of analysis of
all CDR3 regions in terms of length of JH and
DH regions, number of N additions, and number of
DH regions used. The increase in CDR3 size in
spleen cannot be accounted for by an increase in N additions (Fig. 4a) because the mean number of N additions per Ig gene in
the spleen is 6.0 bp, compared with 6.6 bp and 6.4 bp for Ig genes of
duodenum and salivary gland, respectively. The mean sizes of
JH region for duodenum, salivary gland, and
spleen are 13.9, 11.4, and 15.6 bp, respectively (Fig. 4b).
Only the difference between spleen and salivary gland is significant
(p = 0.02). There is an increased usage of the
largest JH segment, JH6, in
the spleen, which may account for this, although the difference is not
significant (Table II). In contrast, the
mean total length of DH gene segments per CDR3
region is increased in IgVH genes from spleen
(26.3 bp) compared with those from duodenum (19.6 bp) and salivary
gland (19 bp). These differences are significant
(p = 0.004 and p = 0.003,
respectively) and are illustrated in Fig. 4c. Some, but
not all, of this difference can be accounted for by the use of a
slightly larger number of DH segments in CDR3
regions of Ig genes from spleen (mean number of 2.4
DH segments per gene) compared with duodenum or
salivary gland (mean numbers of 2.1 and 1.8 DH
segments per gene, respectively), as shown in Fig. 4d. Some
DH segments were incorporated into the CDR3 in
the reverse orientation; the proportion of DH
segments that were reversed was the same in all three groups.
|
|
. The DH segment
D4 was significantly more common in spleen than gut
(p = 0.03). The D4 DH
segments tended to be larger, which contributed to an overall greater
average DH length in the spleen. The difference
in D4 usage, the higher average number of DH
regions used per gene, and the bias toward the use of
JH6 are major factors resulting in the
significantly longer CDR3 in the sequences isolated from splenic
cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
5–10 mm apart. Because there was no evidence
of organized lymphoid tissue within the salivary glands used, it is
presumed that the clonal expansion occurred elsewhere and related
plasma cell precursors migrated to the same site in the salivary gland.
If this is the case, then the plasma cell population must be widespread
but be of limited diversity. An alternative explanation is that
expansion, hypermutation, and isotype switching may occur within the
salivary gland, outside the germinal center environment.
Related clones, which are highly mutated but switched to different
isotypes (Fig. 1
a), indicate that isotype switching can
occur after the onset of hypermutation. This has previously been
observed in duodenum, spleen, and tonsil (16, 17, 18) and
suggests that Abs of the same specificity can be associated with
different isotypes. It is interesting to observe a highly mutated
sequence associated with both IgA and IgM, where there are more
mutations in the IgM gene, because data in this paper and others
(16, 17, 18) show that IgA genes generally have a higher
number of mutations. Therefore, it is possible for a sequence to become
highly mutated before it switches from IgM to IgA and for the IgM gene
to continue to accumulate more mutations than the IgA gene, though this
is not likely to be a common event.
Previous studies have indicated that the number of mutations in IgH genes of gastrointestinal plasma cells is high (10, 11, 17, 19) compared with published reports of the level of mutation in spleen, tonsil, and peripheral blood (16, 20, 21, 22, 23, 24, 25). However, we have shown that the number of mutations in IgA genes in lamina propria plasma cells of the duodenum is higher than the number in IgM genes (17). Because the ratio of IgA:IgM plasma cells in the gastrointestinal tract is much higher than in the periphery (26), a random sampling of cells from either tissue would result in a bias in the overall frequency of mutation observed. In this study, we have isolated RNA and amplified IgH genes from cDNA using primers specific for IgA and IgM genes to compare isotype-matched genes from the different tissues. The IgH genes from both IgM and IgA plasma cells of the salivary gland carry a surprisingly high number of mutations per gene, considering that polyspecificity is often encoded by unmutated Ig genes (27), and Abs found in saliva may be polyspecific. The frequency of mutation of IgM VH genes in the spleen is comparable with the previously published data on frequency of mutation in IgM and IgG of tonsil (16, 21), splenic memory B cells (25), splenic marginal zone B cells (13, 24), and lymph node marginal zone B cells (24). The frequency of mutation of IgVH genes from salivary gland is highest of all groups, and is comparable with that of the duodenum, rather than the spleen. This observation is consistent with the salivary gland being part of the common mucosal immune system, which is separate from the peripheral immune system. It has been shown that in rabbits the tonsil can be an inductive site producing the effector cells of the salivary gland (5). Previously published data on the number of mutations in genes from human tonsil (16) has shown that there are some highly mutated IgA VH genes in tonsil, but the overall frequency of mutation of IgA VH genes in tonsil is still significantly less than that seen for the salivary gland IgA VH genes. It is difficult to reconcile such a large discrepancy in the numbers of mutations in salivary gland IgH genes compared with those of tonsil with the role of tonsils as the primary inductive site for salivary gland. While some salivary Abs may arise as a result of Ag challenge to the tonsil, it is likely that other inductive sites, within the mucosal immune system, generate the bulk of the plasma cells in salivary gland.
Because the same gene rearrangement can be associated with different
isotypes, it is not surprising that there is no significant difference
between the size of the CDR3 region in IgA compared with IgM genes.
However, the CDR3 sizes of Ig genes from duodenum and salivary gland
were significantly smaller than those of spleen for both isotypes (Fig. 3). Comprehensive analysis of the CDR3 regions revealed that the
increased CDR3 size in splenic Ig sequences was mainly due to an
increased overall contribution by DH segments
(Fig. 4c), which in turn was partly due to an increased
number of DH segments in each CDR3 region and a
difference in D4 usage (Fig. 4d and Table III). N
nucleotides did not contribute to the increased CDR3 size in splenic Ig
sequences, because there was no significant difference between N
additions in splenic sequences and those in sequences from duodenum and
salivary gland. The possibility of a correlation between the extent of
hypermutation and the length of CDR3 was considered. The splenic
sequences have the lowest overall mutation rate in
VH, with the largest overall CDR3 region size,
and vice versa for the salivary gland sequences. However, when the
number of mutations was considered alongside the size of CDR3 region
for each individual gene, it was apparent that there was no correlation
between these two factors (data not shown).
JH6 is much larger than all the other
JH segments and can often account for differences
in CDR3 size. However, although there is a slight increase in the
number of JH6 segments used in Ig genes from
spleen, this is not significant (Table II). A contributing factor may
be the length of sequence deleted from the JH
segment, which seems to be lower in splenic genes than in those from
duodenum and salivary gland, although these numbers are too small for
statistical analysis (data not shown).
Compared with other studies (20, 22, 23), we observed a relatively high number of DH segments per Ig gene. The method of assignment used in the study was the same as that applied by Brezinschek et al. (20). We observed the same tendency to use more DH regions when the more stringent criteria of Brezinschek et al. (22) were applied (data not shown). Levels of somatic hypermutation are likely to affect the assignment of DH regions, a higher level of mutation resulting in the identification of more DH regions. However, it is not thought that this would have caused these results, because more DH regions were identified in the spleen where the frequency of somatic hypermutation was lowest. The basis for the difference between our data and previously published data is unclear. However, we have focussed on the properties of plasma cell populations that have not been the subject of any similar investigations.
We have shown here that the characteristics of
IgH genes isolated from the mucosal tissues of
the duodenum and salivary gland are different from those of
IgH genes in the spleen. Splenic
IgH genes sequenced in this study have the same
characteristics as IgH genes previously reported
in other peripheral sites (13, 16, 21, 24, 25). It is also
interesting to note that in this group of 32 splenic Ig genes there are
two that use JH1 segments, whereas no sequences
from duodenum, salivary gland, or over 200 previously reported Ig genes
from lamina propria plasma cells (10, 11, 17, 19) were
found that used JH1. Because
JH1 is so rarely used anyway, a much larger
sampling from spleen than this one would be required to determine
whether this is significant. Therefore, we propose that
IgH genes in a mucosal plasma cell population
have characteristics distinguishable from those seen in the peripheral
immune system. The fact that any differences can be seen at all
supports the theory that the common mucosal immune system is separate
from the peripheral system. Mucosal IgH genes
generally have a higher number of somatic mutations in their
VH regions and a shorter CDR3 region. Due to the
high variance for these parameters in the populations studied, it would
be difficult to be able to use them to distinguish between cells
originating in the peripheral immune system and not the mucosal immune
system. However, there does seem to be a "cut off" point for the
number of mutations found in IgM genes in the periphery. We have not
seen any IgM genes with more than 24 mutations isolated from the
periphery, while we have quite often seen IgM genes with a larger
number of mutations in mucosal tissues (Fig. 2 and Ref.
17).
By studying the IgM and IgA isotypes separately, we have shown that the differences in characteristics between different tissues are not due to variations in isotype. The reasons for these differences are not clear. It has previously been suggested that the highly antigenic nature of mucosal surfaces may result in the constant stimulation of B cells to enter germinal centers and accumulate a higher number of mutations (11). If overstimulation were the cause of the high numbers of mutations seen in mucosal tissues, then a similar level of mutation would be expected in the peripheral tissues as a result of chronic systemic Ag challenge. Patient 5 had a perforated bowel and the spleen was histologically normal but showed signs of mild acute inflammation consistent with challenge by luminal Ag. Even so, the IgH genes from this spleen were still less mutated than those seen in duodenum and salivary gland. These data suggest that there are fundamental differences in the mechanisms of mutation and/or selection in the peripheral and mucosal compartments.
The significance of the difference in CDR3 length and DH region usage between the mucosal and peripheral IgH genes is unclear. Differences in CDR3 region length could arise as a result of codon deletions arising during somatic hypermutation or some other process, Ag selection, or origin of precursor B cells from a population that has a slightly different mechanism of Ig gene rearrangement. Although it has been documented that deletions and insertions can occur during somatic hypermutation, it is unlikely that the change in CDR3 length is due to this phenomenon, because we saw no evidence of deletions within DH segments in any Ig gene studied. Because this study sequenced cDNA, the Ig genes are expressed and therefore may have been selected by Ag. It is possible that the Abs encoded by IgH genes with shorter CDR3 lengths have higher affinity for Ags prevalent in the mucosal immune system. However, it is difficult to envisage such a distinction of Ag between the two compartments, especially in the light of the medical history of patient 5 as discussed above. The remaining possibility is of distinct precursor B cell populations for the two compartments (28), though no evidence exists as yet for this in humans.
In summary, we have shown that the plasma cells of the salivary gland are highly mutated and have a shorter mean CDR3 length, mainly due to differences in DH region genes. Related plasma cells can be seen in this effector site, implying that the plasma cell population using the IgVH4–34 gene must be widespread, but with limited diversity. The characteristics of salivary gland plasma cells are similar to those of the duodenum, but different from those in the spleen. Because previous reports of the number of mutations in tonsillar IgH genes show they are not highly mutated, there must be some other inductive site (presumably within the mucosal immune system) seeding the salivary gland with plasma cells.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jo Spencer, Guy’s, King’s and St. Thomas’ Medical School, Department of Histopathology, St. Thomas’ Campus, Lambeth Palace Road, London SE1 7EH, U.K. E-mail address:
3 Abbreviations used in this paper: IgH, Ig heavy chain gene; CDR, complementarity-determining region.
Received for publication August 5, 1999. Accepted for publication November 16, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Yuvaraj, G. Dijkstra, J. G. M. Burgerhof, P. M. Dammers, M. Stoel, A. Visser, F. G. M. Kroese, and N. A. Bos Evidence for Local Expansion of IgA Plasma Cell Precursors in Human Ileum J. Immunol., October 15, 2009; 183(8): 4871 - 4878. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Su, J. N. Gordon, F. Barone, L. Boursier, W. Turnbull, S. Mendis, D. K. Dunn-Walters, and J. Spencer Lambda Light Chain Revision in the Human Intestinal IgA Response J. Immunol., July 15, 2008; 181(2): 1264 - 1271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Planque, Y. Bangale, X.-T. Song, S. Karle, H. Taguchi, B. Poindexter, R. Bick, A. Edmundson, Y. Nishiyama, and S. Paul Ontogeny of Proteolytic Immunity: IgM SERINE PROTEASES J. Biol. Chem., April 2, 2004; 279(14): 14024 - 14032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Collins, M. Ikutani, D. Puiu, G. A. Buck, A. Nadkarni, and B. Gaeta Partitioning of Rearranged Ig Genes by Mutation Analysis Demonstrates D-D Fusion and V Gene Replacement in the Expressed Human Repertoire J. Immunol., January 1, 2004; 172(1): 340 - 348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Meek, J. W. Back, V. N. A. Klaren, D. Speijer, and R. Peek Conserved regions of protein disulfide isomerase are targeted by natural IgA antibodies in humans Int. Immunol., November 1, 2002; 14(11): 1291 - 1301. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S J C Golby and J Spencer Where do IgA plasma cells in the gut come from? Gut, August 1, 2002; 51(2): 150 - 151. [Full Text] [PDF]
|
|
|
|
|
|
V C Thoree, S J C Golby, L Boursier, M Hackett, D K Dunn-Walters, J D Sanderson, and J Spencer Related IgA1 and IgG producing cells in blood and diseased mucosa in ulcerative colitis Gut, July 1, 2002; 51(1): 44 - 50. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
