| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Light Chain Rearrangements in Human Intestinal Plasma Cells 1
* Department of Histopathology, GKT Medical School, St. Thomas’ Campus, London, United Kingdom; and
Department of Gastroenterology, St. Thomas’ Hospital, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
J rearrangements in the plasma cell and B cell populations. Rearrangements that were either in-frame or out-of-frame between V and J were compared. Usage of IgV families in the in-frame rearrangements from the plasma cells resembled that observed in the mantle cells, suggesting that antigenic selection for cellular specificity does not dramatically favor any particular V segment. However, in marked contrast, out-of-frame rearrangements involving V1 and V2 families are rarely observed in intestinal plasma cells, whereas rearrangements involving V5 are increased. This resulted in significantly biased ratios of in-frame:out-of-frame rearrangements in these V families. Out-of-frame rearrangements of IgVJ from plasma cells, including those involving the V5 family, have a significant tendency not to involve J1, consistent with the hypothesis that this population includes rearrangements generated by secondary recombination events. We propose that modification of out-of-frame rearrangements of IgVJ exists, probably a consequence of secondary rearrangements. This may be a mechanism to avoid translocations to susceptible out-of-frame IgVJ rearrangements during somatic hypermutation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
L chain locus is initiated. If a productive rearrangement is achieved, the cell will mature to express H chain with L chain. If rearrangement is unsuccessful on both alleles, a DNA deletion event mediated by the -deleting element inactivates the locus and rearrangement of the L chain alleles proceeds (4). Sequential accessibility of and L chains for rearrangement to ensure isotypic exclusion is thought to be regulated by differential transcription of the germline target sequences under the control of the locus-specific enhancers (5, 6). Studies of human lymphomas and single human B cells have confirmed that this chronological sequence of rearrangement events occurs in the majority of developing human B cells (7, 8, 9). Rearrangement at each locus is imprecise and only a third of the rearrangements are likely to be in the correct reading frame between V and J. Since rearrangement of a second allele can occur if a first allele is not successfully rearranged, a final percentage of in-frame rearrangements of 71.4% would be expected, if no other factors influence the outcome. Distortion of the expected 71.4% of in-frame rearrangements has been reported. In chickens, although a third of the L chain rearrangements are in-frame from days 10–12 of development, at successive later stages of development in the bursa an increasing proportion of these genes become in-frame (10). In pigs, the frequency of in-frame H chain rearrangements can reach almost 100%, depending on the developmental stage of the pig and anatomic location of cells undergoing rearrangement (11). The mechanisms generating the bias toward in-frame rearrangements in these species are not yet understood and may not be the same, since the Ig loci are organized differently.
Out-of-frame genes are not under any selective pressure because they are not expressed. Therefore, the relative frequency with which V region families occur in the out-of-frame rearrangements has been considered to reflect the biases of rearrangement in the absence of selection for either structural or functional viability or antigenic specificity (12). Out-of-frame Ig rearrangements can be considered to be a liability to the cells carrying them because lymphomagenesis commonly involves translocation of oncogenes to unused Ig rearrangements (13). Although some translocations are clearly the result of aberrant recombination events, somatic hypermutation, which occurs in both used and unused rearrangements and which can potentially generate DNA double-strand breaks in the Ig hypermutation domain, appears to facilitate the translocation to unused rearrangements in some cells (13). In Burkitt’s lymphoma for example, the c-myc translocations occur predominantly to IgH within the hypermutation domain (14). It is interesting that translocations to unused H chain rearrangements exceed translocations to unused L chain rearrangements (13, 15). In our studies of highly mutated IgH and IgL rearrangements from intestinal plasma cells, IgH and IgL appear to be targeted equivalently by the hypermutation mechanism (16). This suggests that mechanisms exist that minimize translocations to unused L chain rearrangements. The -deleting element inactivates most unused alleles by deletion of the 5' enhancer and most out-of-frame rearrangements do not undergo somatic hypermutation (16).
In human intestine, B cells and plasma cells occupy different histological niches. B cells (naive and memory B cells) are present in the organized lymphoid tissue of the Peyer’s patches and plasma cells are diffusely distributed in the lamina propria. It is thought that these populations are developmentally related in humans; intestinal plasma cells being derived from immune responses induced in the Peyer’s patches (17, 18). IgVH and VL genes used by gut plasma cells have a characteristically high load of somatic mutations consistent with origin from germinal centers in a site of chronic Ag challenge (16). As yet there is no evidence of B1 cell precursors of intestinal plasma cells in humans as described in mice (19, 20). In this study, we report dramatic biases in the out-of-frame repertoire of VJ rearrangements from intestinal plasma cells that are not apparent in their putative precursor populations in Peyer’s patches. The data strongly suggest that the human L chain locus undergoes modification of out-of-frame rearrangements in human intestinal lymphoid tissue, probably as a consequence of secondary rearrangements. This may be another mechanism to avoid potentially lethal translocations to the hypermutation domain in IgL.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Samples of human intestine from three patients (two male, one female) ages 72, 76, and 81 years were studied. Tissue from the same patients had been used for earlier study of IgH chain genes and L chain genes (16). All specimens were from right hemicolectomy for Duke’s stage B or C adenocarcinoma. The tissue was taken at a distance from the tumor and snap frozen in liquid nitrogen within 1 h of surgery and kept in an anonymized tissue bank under liquid N2 until required. Tissues were used with the approval of the local ethics committee. All tissues studied were histologically normal.
Immunochemistry and microdissection
Frozen sections (7 µm) were cut from the stored frozen samples and put on poly-L-lysine-coated slides and kept at -20°C until required. For samples of colon, the distribution of B cells was identified using mAb CD20 (clone L26; DAKO, Ely, U.K.) and an indirect immunoperoxidase method with rabbit anti-mouse peroxidase conjugate and diaminobenzidine substrate. Sections were left without mountant or coverslips to dry overnight. Areas of colonic lamina propria devoid of CD20-positive cells (most of the lamina propria is CD20 negative) were dissected by hand under the microscope by scraping the dry section with a 25-gauge needle and subsequently washing it into TaqDNA polymerase 1x reaction buffer containing 0.25 mg/ml proteinase K (Sigma-Aldrich, Poole, U.K.). The DNA was released for PCR using proteinase K as previously described (21).
Samples of terminal ileum were stained with mAb IgD (anti-IgD; DAKO) using the same staining method to identify mantle zone B cells in Peyer’s patches. Mantle zone and marginal zone B cells were microdissected separately using a Leitz micromanipulator as described previously (17). DNA was prepared as described above.
Amplification of VJ gene rearrangements by multiplex PCR
Ten V families were simultaneously amplified using a multiplex PCR method by putting all external V and J primers in the first round and all internal V and J primers in second round. The 5' V primer sequences were as described by Farner et al. (12) and 3' J primers as described by Brauninger et al. (22). Alignments of J regions and J primers are shown in Table I. The PCR was conducted in 50-µl volumes containing 10-µl DNA samples, 50 ng of each 5' primers, 100 ng of each 3' primers, 200 µM of each dNTP, and 2.5 mM MgCl2 in TaqDNA polymerase 1x reaction buffer (Promega, Southampton, U.K.). Five microliters of first-round PCR product was used for the second round. A touchdown PCR program was used. After hot start at 95°C for 10 min, 1 U TaqDNA polymerase (Promega) was added in each reaction at 60°C. This was followed by seven cycles of; 95°C for 1 min (66°C for 2 min, decreased by 2°C in each of the seven cycles), and 72°C for 45 s. This was followed by 30 cycles of 95°C for 30 s, 52°C for 30 s, and 72°C for 1.5 min. Finally, the primer extension was conducted at 72°C for 5 min. The second round was performed in the same way as the first round, except that the annealing temperature was 68°C during first seven cycles and 56°C for the following 30 cycles. All reactions were performed using a Px2 thermal cycler (Thermo Hybaid, Ashford, U.K.). PCR products were separated in 3.5% agarose gel and PCR products were randomly selected for DNA sequencing.
|
J gene rearrangements using V family-specific PCR
VJ gene rearrangements using each of the 10 IgV families were amplified individually with family-specific primers and nested PCR protocols. The PCR was conducted in a 50-µl volume containing 100 ng of 5' family-specific V primers as used in the multiplex PCR and 100 ng of an alternative set of J primers, as in Table I (12). Ten microliters of the DNA sample was used for the first round and 2 µl of first-round PCR products for the second round. All reactions contained 200 µM of each dNTP and 2.5 mM MgCl2 in TaqDNA polymerase 1x reaction buffer. After hot start at 95°C for 10 min, 1 U TaqDNA polymerase (Promega) was added to each reaction at 60°C and then 35 cycles of 95°C for 30 s (family-specific temperature ranging between 50 and 56°C) for 30 s and 72°C for 1.5 min. Finally, the primer extension was carried on at 72°C for 5 min. PCR products were separated in 3.5% agarose gel. The product bands at 
350 bp were selected for cloning and sequencing.
Cloning and DNA sequencing
PCR products were cloned into the pGEM-T vector (Promega) and were sequenced by Qiagen Sequencing Services (Qiagen, Hilden, Germany). V and J gene sequences were analyzed using Genejockey II software and alignments were made using the V BASE sequence directory and the IMGT database (http://imgt.cines.fr). J segments were assigned manually. A sequence was considered to be in-frame if the J segment was in the same reading frame as V across the VJ junction. If stop codons, insertions, or deletions affecting the reading frame and rendering the rearrangement nonproductive were introduced as a consequence of somatic hypermutation, the rearrangements were still considered to be in-frame if J was in the same reading frame as V. Changing this designation, which only affected a small number of samples, did not influence the final data, mostly because the sequences with "destructive" mutations in V were a subset of the cells that were already designated out-of-frame between V and J. A total of 348 bacterial clones were sequenced. Some rearrangements with the same CDR3 were sequenced more than once and were classified as a single rearrangement in this study. In total, we identified 300 different rearrangements of IgVJ from the different individuals and sites studies, as described in Table II. Mutations from germline V sequence were counted and presented as percentage of total nucleotides that are mutated to allow comparison between different V families, since these differ in the length of the V segment. Sequences analyzed are accessible from European Molecular Biology Laboratory/GenBank under accession numbers AJ578180– J578400, AJ576317– AJ576319, AJ535508–AJ535516, AJ400000, AJ493281–AJ493291, AJ508940–AJ508950, AJ535510, AJ399965–AJ399973, AJ399975–AJ399979, AJ399981–AJ399983, AJ399986–AJ399990, AJ399992–AJ399994, AJ399996– AJ399999, and AJ582974–AJ582988.
|
Numbers of in-frame and out-of-frame rearrangements in individual families were compared using 
2 to test the hypothesis that there was no difference between different groups. Microstat software was used to test the normality of the number of mutations in the groups of sequences. Normally distributed data were then compared using Student’s t test. Data that were not normally distributed were compared using a Mann-Whitney U test. Observed differences were considered to be statistically significant at p 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
J rearrangements
The repertoire of IgVJ rearrangements in intestinal lamina propria plasma cells and mantle zone cells from Peyer’s patches was studied using a multiplex PCR method with primers to amplify rearrangements involving all V families and J segments. This method may generate some biases in V family and J usage. However, we consider that this approach is justified here because if any such biases exist, they are likely to operate equally on in-frame and out-of-frame rearrangements of both plasma cells and mantle zone cells. The in-frame repertoire of IgVJ rearrangements isolated from the plasma cells was not significantly different than the in-frame repertoire of IgVJ rearrangements isolated from the mantle zone population, with the exception of rearrangements involving V3 which were significantly less common in the plasma cells (Table III). However, there were significant biases in the out-of-frame repertoire of IgVJ rearrangements in the plasma cells. Out-of-frame rearrangements involving V1 and V2 were significantly absent. In contrast, out-of-frame rearrangements involving V5 were increased so that they were more abundant than the in-frame rearrangements. The distortion of the repertoire of out-of-frame genes resulted in a skewed ratio of in-frame:out-of-frame rearrangements in V1, V2, V4, and V5 which was significantly different than that observed in the mantle zone cells. To ensure that this distortion was not generated by the methodology, we amplified each family separately using V family-specific PCR primers and a different set of J region primers to that used with the multiplex method. As shown in Table IV, the biases in the ratio of in-frame:out-of-frame rearrangements in Table III are equally apparent using V family-specific PCR.
|
|
We considered it possible that some families of IgV may be more heavily targeted by somatic hypermutation than others, so that out-of-frame rearrangements, which are not constrained by need to encode viable Igs for survival, may no longer bind primers and be absent from the data. As shown in Fig. 1, there is no significant difference between frequencies of mutation observed in the IgV families either in-frame or out-of-frame when these are present, and this is not likely to be the origin of the biases observed.
|
DNA from marginal zone and mantle zone cells microdissected from Peyer’s patches was used in V2 family-specific PCR. The numbers of in-frame and out-of frame V2 rearrangements from Peyer’s patch B cells that had 4 or >4 mutations in V2 were compared with the number of in-frame and out-of-frame V2 rearrangements obtained from plasma cells. The memory cell population did not show the bias toward in-frame rearrangements seen in the plasma cells (Table V). The frequency of point mutations in the memory cell population (average 13.6 mutations per V2 gene sequenced) is significantly lower than that observed in the plasma cells (average 20.0 mutations per V2 gene sequenced) using the Mann-Whitney U test (p = 0.02).
|
1and V gene segments in in-frame and out-of-frame rearrangements
Frequency of J segments was compared in IgVJ rearrangements. In-frame rearrangements from plasma cells studied either using family-specific primers or multiplex PCR and both in-frame and out-of-frame rearrangements from the mantle zone had similar frequency of J1 incorporation. However, J1, the most 5' J region in the human J–V cluster, was significantly reduced in the out-of-frame rearrangements from gut plasma cells (Fig. 2). This was true when individual families that included out-of-frame rearrangements were studied independently (Table VI). IgV segments occur in three clusters, A, B, and C, where A is the most 3' J proximal cluster (http://imgt.cines.fr). Analysis of IgV clusters in-frame and out-of-frame rearrangements of mantle zone and plasma cells generated by the multiplex PCR method shows that there is no difference in V cluster or J usage when in-frame rearrangements in gut plasma cells and mantle cells are compared. However, there is a bias against rearrangements to the most 3' V cluster, cluster A in the out-of-frame rearrangements from gut plasma cells (Fig. 3). There is a dominance of J1 in rearrangements involving cluster A and overall dominance of rearrangements involving cluster C in the out-of-frame rearrangements from mantle cells.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
J rearrangements in human intestinal plasma cells
We have observed significant biases in the ratio of in-frame:out-of-frame IgVJ rearrangements in human intestinal plasma cells involving some V families. V1 and V2 were almost exclusively in-frame, while V5 had a lower percentage of in-frame rearrangements than expected. The biases were apparent when these rearrangements were compared with either expected frequencies of in-frame rearrangements or a control population of cells from the mantle zone from one of the patients studied. We excluded the possibility that the biases were generated by methodology or hypermutation. Use of surgical right hemicolectomy specimens allows careful selection of sites to be sampled and good tissue morphology. However, these specimens tend to be from older individuals and inclusion of mantle cells from one of these patients was considered to be an essential control showing that biases were not due to age.
The biases in in-frame:out-of-frame rearrangements involving some IgV families in plasma cells are due to distortions in the population of out-of-frame rearrangements. There were highly significant differences in the representation of V families in the out-of-frame IgVJ rearrangements isolated from the plasma cells compared with all other groups represented as columns in Table III. Most notably, V1 and V2 were absent from the out-of-frame IgVJ rearrangements and there were more rearrangements involving V5. The out-of-frame repertoire is not expressed and is unselected. This distortion may represent modification of the out-of-frame rearrangements by recombination to remove predominantly out-of-frame V1, V2, and possibly generating new rearrangements including those to V5 as a consequence.
Occurrence of IgV segments in the in-frame and out-of-frame rearrangements of IgVJ in mantle cells was not significantly different than that seen in the blood in other studies (12). The repertoires of in-frame rearrangements isolated from the plasma cells and mantle zone cells by multiplex PCR were similar in their usage of V families, V clusters, and J regions, although there was a tendency not to use V3 in the plasma cell population. Within the families, usage of family members was also similar in the in-frame rearrangements from plasma cells and mantle cells (data not shown). This suggests that despite the heavy bacterial load and potential encounter with food Ags in the gut, there is no apparent bias in the usage of IgV or J segments in the intestinal plasma cell response.
Biases in J rearrangement
If out-of-frame rearrangements involving V1 and V2 are lost and some rearrangements of V5 are acquired as a result of DNA recombination events, then the most 5' J segment, J1, would be expected to occur at a lower frequency in V5 rearrangements. Lack of J1 in out-of-frame rearrangements involving V5 is apparent and statistically significant when V5 rearrangements are amplified using either family-specific or multiplex PCR. Interestingly, other out-of-frame rearrangements from plasma cells primarily involving V genes in clusters B and C also lack J1, suggesting that some of these may arise as a consequence of secondary rearrangements. The lack of J1 in the out-of-frame alleles is not an apparent difference due to expansion of J1 in the in-frame rearrangements, since J1 usage in plasma cells and mantle cells is similar.
In an earlier detailed analysis of VJ rearrangements in single B cells from human blood, rearrangements involving J7 were commonly observed (12). In contrast, these were rare in the three patients studied here and another study of single human B cells (23). To ensure that our data are not artifactually skewed in this respect, we used two different sets of J primers (12, 22) with the same result. We also investigated blood and intestinal cells from younger individuals and saw little or no J7 (data not shown). The reason for the difference is not clear. However, this does not affect the interpretation of the results presented here since the studies are internally controlled and consistent with the findings of another group (23). We consider the lack of J1 in the out-of-frame repertoire of plasma cells to be a significant finding consistent with the hypothesis that some observed out-of-frame rearrangements may be derived from secondary rearrangements that could possibly eliminate other out-of-frame genes.
Secondary rearrangements of IgL chain genes
Rearranged L chain loci may retain unrearranged 5' V and 3' J regions with correctly orientated recombination signal sequences, providing a theoretical opportunity for secondary rearrangements. Secondary rearrangements at the L chain loci during B cell development to avoid generation of autoreactive cells by receptor editing have been described in a number of mouse models (24, 25). There is also evidence from studies of murine and human cells that L chain revision can modify the mature B cell repertoire, possibly through expression of RAG genes in germinal centers (26, 27, 28, 29, 30). However, it is unlikely that secondary rearrangements shape the mature B cell repertoire once B cells have bound specific Ag and following the initiation of hypermutation (31, 32). Removal of out-of-frame IgVJ rearrangements has not been described, but could theoretically occur during the initiation of germinal center responses when RAG genes are expressed (26, 27, 28, 29, 30) or even at the plasma cell stage (33). We saw no evidence of RAG or Tdt activity in lamina propria cells by RT-PCR (data not shown). However, since it is not clear at which stage during the life of the plasma cell such modification might occur, this is inconclusive.
Marginal zone B cells have mutated IgVJ but do not have a biased percentage of in-frame rearrangements to V2
Memory cells from Peyer’s patches have out-of-frame rearrangements involving V2 and therefore do not resemble the plasma cells in this respect. The relationship between the Peyer’s patch marginal zone B cell population and the germinal center is not clearly understood. We have observed proliferation of human Peyer’s patch marginal zone B cells in situ using immunohistochemistry (data not shown) and it is possible that they may be involved in T-independent responses outside the germinal centers that are potential sites of RAG re-expression (26, 27, 28, 29, 30, 34). It is also possible that a significant subset of intestinal plasma cells is unrelated to the Peyer’s patches in humans as in mice (19) and that the biases are generated during the development of the precursors of intestinal plasma cells independent of germinal centers. It has also been demonstrated using a cell line that human plasma cells may be able to revise L chains independently (33), although the biological significance of this is as yet unclear.
Summary
We have shown that rearrangements to V1 and V2 are rare in the out-of-frame repertoire of intestinal plasma cells, whereas rearrangements involving V5 are increased. Out-of-frame rearrangements of IgVJ from plasma cells have a reduced frequency of J1 and V cluster A, consistent with the hypothesis that these genes are modified by secondary recombination events. This may be a mechanism to avoid translocations to out-of-frame IgVJ rearrangements during somatic hypermutation.
|
Footnotes |
|---|
2 Address correspondence and reprint requests Dr. Jo Spencer, Department of Immunobiology, Guy’s, King’s and St. Thomas’ Medical School, New Guy’s House, Guy’s Campus, London Bridge, London SE1 9RT, U.K. E-mail address: jo.spencer{at}kcl.ac.uk
3 Abbreviation used in this paper: RAG, recombination-activating gene.
Received for publication July 1, 2003. Accepted for publication December 8, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
genes in human B cells. Nature 316:260.[Medline]
1 light chain gene is developmentally regulated and independent of Ig rearrangements. J. Exp. Med. 197:1165. genes. Cell 110:575.[Medline]
light-chain genes are deleted or rearranged in -producing B cells. Nature 290:368.[Medline]
J repertoire. J. Immunol. 162:2137. genes are almost all in-frame and highly mutated and most genes are highly mutated when in-frame and minimally mutated when out-of-frame. Eur. J. Immunol. 30:2908.[Medline]
repertoire. J. Mol. Biol. 268:69.[Medline]
light chain genes in mature B cells in vitro and in vivo: function of re-expressed recombination-activating gene (RAG) products. J. Exp. Med. 187:795.J repertoire in human fetal spleen: evidence for positive selection and extensive receptor editing. J. Immunol. 165:6322.This article has been cited by other articles:
|
|
 |
|
  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]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), plasma cells studied by multiplex PCR (MP; 
), and mantle zone cells amplified by multiplex PCR (MP; 
). The difference in J