| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Programa de Patología Molecular,* CentroNacional de Investigaciones Oncológicas, Madrid; and the HospitalVirgen de la Salud, Toledo, Spain
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
In addition to translocations, the BCL-6 gene experiences somatic mutations in the first intron, 100 bp downstream from the first noncoding exon, analogous to what occurs with IgVH genes, although at a lower frequency.7,8 BCL-6 somatic mutations have been described in normal and tumoral B lymphocytes, being more frequently observed in LBCL than in other tumoral types.8,9 BCL-6 mutations are considered to be markers of germinal center transit because they are absent in benign and malignant (mantle cell lymphoma) pregerminal center B cells and characteristically present in germinal center lymphocytes and derived tumors.
It has been suggested on several occasions that BCL-6 expression is probably deregulated as a consequence of somatic mutation.8 However, a relationship between the frequency or localization of these mutations and the expression of the protein has not been conclusively demonstrated.
Furthermore, consequences of the accumulation of BCL-6 mutations during the genesis and progression of lymphomas have not been clarified sufficiently. The only exception is the higher grade transformation of follicular lymphoma, in which the accumulation of new mutations has been described.10 However, this accumulation has not been shown to have any pathogenic association with the mechanism of the progression.
The mutation frequency of BCL-6 seems to be high in LBCL. An expression profile that includes the expression of BCL-6 and other germinal center markers has been found to be a reliable predictor of increased overall survival (OS) in LBCL.11 Moreover, it has been recently described that the expression of BCL-6 alone can be a reliable predictor of better survival in LBCL.12 For these reasons, we decided to determine whether BCL-6 mutations are associated with differences in the expression of the BCL-6 protein and the outcome of the patients in a group of LBCL cases. To this end, we analyzed mutations in the major mutation cluster (MMC) of BCL-6 in a series of 45 LBCL cases, relating them to the level of BCL-6 expression and clinical outcome. A group of 15 Burkitt’s lymphoma (BL) cases was also included to allow comparison of the effects of mutations in these different lymphoma types.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
We studied 45 LBCL and 15 BL patients who had been diagnosed according to the REAL classification.13 For most tests, each histological group was studied separately, making it possible to analyze molecular differences between them. Diagnostic pretreatment samples were analyzed in all cases. Tumoral and reactive specimens were selected from the Pathology Laboratory, Hospital Virgen de la Salud (Toledo, Spain), on the basis of the availability of paraffin-embedded and frozen tissue for molecular studies and the existence of an adequate clinical follow-up (except for four patients, whose clinical history was not available). All patients were treated with chemotherapy regimes including anthracyclines. A clinical follow-up was performed in all cases. Clinical follow-up of patients alive at the time of the study included periods ranging from 11 to 122 months (mean, 75.38 months). Age, clinical stage, performance status, levels of lactate dehydrogenase (LDH) in blood and the number of extranodal sites of the disease were used as outcome criteria, as set out by the International Prognostic Index (IPI).14 Patients with 0 to 2 unfavorable variables were considered as being low risk, and those with 3 to 5 variables as high risk. Complete remission was defined as the resolution of clinical and radiological evidence of disease for a minimum of 4 weeks.
Biopsy specimens of these 60 patients were divided into two parts: one was formalin-fixed and paraffin-embedded for morphological and immunohistochemical studies, while the other was embedded in Optimal Cutting Temperature Tissue-Tek (OCT; Sakura, Zoeterwoude, Netherlands) and preserved at -80°C for molecular studies.
DNA extraction from frozen sections was performed following the standard phenol-chloroform protocol. RNA was extracted from frozen sections. Before each DNA or RNA extraction, frozen sections of the specimens were analyzed to ensure that the specimen was representative of the tumor. The percentage of tumoral cells was greater than 80% in all cases.
To quantify BCL-6 mRNA expression, we extracted RNA from Raji cells. This cell line was obtained from the American Tissue Type Culture Collection (Rockville, MD), and maintained in RPMI 1640 (Sigma, St. Louis, MO) supplemented with 10% fetal calf serum, 2 mmol/L glutamine, and penicillin-streptomycin (Life Technologies, Inc., Grand Island, NY).
Immunostaining Techniques
BCL-6 protein was detected with monoclonal antibody for P6-B6p, a recombinant protein corresponding to amino acids 3 to 484 from DAKO, Glostrup, Denmark (1/10).15 The proliferation index was evaluated using nuclear antigen Ki67 expression, detected with the MIB1 antibody from Immunotech (Marseille, France).
Immunostaining techniques were performed in paraffin-embedded tissue sections. For antigen retrieval, before incubation with the antibodies (Ab), the slides were heated in a pressure cooker for 3 minutes in 0.01 mol/L of sodium citrate solution. Additionally, the slides were digested with proteinase K for 10 minutes at room temperature.
After incubation with the primary Ab, immunodetection was performed with biotinylated anti-mouse immunoglobulins, followed by peroxidase-labeled streptavidin (LSAB-DAKO, Denmark) and diaminobenzidine chromogen as substrate. All immunostaining was performed using the Techmate 500 (DAKO) automatic immunostaining device.
Incubations omitting the specific Ab, or with unrelated Abs, were performed to provide controls of the technique. The quality of the staining was checked on every slide, using BCL-6-positive reactive lymphocytes as an internal control.
Quantitative Studies
All scoring and interpretations of immunohistochemical results were performed independently by two of the authors (AIS and MJA) without knowledge of the clinical variables or the results of the molecular analysis.
High-magnification fields were chosen for the evaluation of BCL-6 and MIB1 expression, focusing on tumoral areas and counting up to 300 cells. All immunoreactive cells were considered to be positive. A manual cell-counting procedure was used so that all nontumoral subpopulations could be excluded on the basis of their cell morphology.
The intensity of BCL-6 immunostaining was classified into low, intermediate, or high groups when the signal was lower, similar, or higher, respectively, than reactive T lymphocytes in the tumor.
For survival analysis of BCL-6 expression, a cutoff of 50% of positive cells was used because this threshold divided the series into groups of similar size (median of percentage of BCL-6-positive cells: 50) and had the advantage of being easily reproducible.
Real-Time Polymerase Chain Reaction (PCR)
Total RNA extraction was performed using a protocol based on Trizol (Life Technologies, Inc.). cDNA was synthesized with avian myeloblastosis virus (AMV) retrotranscriptase (Promega, Madison, WI), according to manufacturer’s instructions.
Real-time PCR was developed using the TaqMan technology, in an ABI Prism 7700 Sequence Detector System (PE Applied Biosystems, Norwalk, CT). As a control of the quality and quantity of the RNA, GAPDH gene was amplified in parallel with that of BCL-6. The primers and probes used in this study have been described previously.12 Both probes were labeled at the 5' end with 6-carboxy-fluorescein phosphoramidite (FAM), and at the 3' end with 6-carboxy-tetramethyl-rhodamine (TAMRA) as quencher. The reactions were not multiplexed.
Each sample was measured from two different RNA extractions, each at two different dilutions, and each dilution in triplicate. To compare different experiments, in every PCR a standard curve (composed of different dilutions of a cDNA from Raji cells) and a calibrator (cDNA from a reactive tonsil) were derived, as suggested in the ABI 7700 User Bulletin 2 (PE Applied Biosystems). The conditions of the reactions were those recommended in the ABI 7700 User Bulletin 2 (PE Applied Biosystems).
Sequencing of BCL-6
A unique PCR product, 791 bp long, was amplified using 5'-CCGCTGCTCATGATCATTATTT-3' and 5'-TAGACACGATACTTCATCTCAT-3' primers. This fragment is located downstream of the first noncoding exon of BCL-6 and includes the entire MMC region.
The PCR reaction was performed in a 50-µl total volume containing 50 pmol of each primer, 0.1 mmol/L dNTP, 1.5 mmol/L MgCl2, and 2 U Taq Platinum (Life Technologies, Inc.). Conditions for amplification were as follows: 94°C, 5 minutes denaturation; 35 cycles of 30 seconds at 94°C, 30 seconds at 58°C, and 1 minute at 72°C; and a final extension step at 72°C for 10 minutes. PCR was performed in a Perkin Elmer 9700 GeneAmp PCR System (Norwalk, CT).
PCR products were purified by using the Microcon PCR kit (Millipore, Bedford, MA). Both strands were then directly sequenced, using the same primers as for the amplification and two additional internal oligonucleotides, in an ABI 370 (Perkin Elmer Applied Biosystems, Warrington, UK), following the manufacturer’s procedure. Mutations were identified by comparison with the BCL-6 germline sequence (GenBank accession number AF191831). To control for potential Taq errors, all PCR and sequencing procedures were performed twice.
Statistical Analysis
Statistical study of the correlation between distributions was performed using either Fisher’s exact test for categorical variables, the Kruskal-Wallis test for single-ranked data, and the Pearson correlation for double-ranked data.
The clinical variables analyzed in the survival studies were those
included in the IPI14
(measured as 0 to 2, 3 to 5), these
being: age (
60 versus >60 years), gender (female
versus male), clinical stage (I + II versus III +
IV), and LDH (normal versus >normal). Survival curves were
calculated by the Kaplan-Meier method and compared by the log-rank
test.16,17
Actuarial survival curves [OS and disease-free
survival (DFS)] were calculated using the Kaplan-Meier method.
Statistical significance was calculated using the log-rank test. Cox’s
proportional hazard univariate analysis18
was also
performed, providing estimates of the confidence interval and the
relative risk (RR) in terms of survival.
To identify the factors that might be of independent significance in influencing survival (OS and DFS), a Cox backward proportional hazard model was fitted.19 Variables included in the maximal models were IPI (0 to 2, 3 to 5) and presence of mutations inside the 423 to 443 cluster. The low-risk IPI and presence of mutations in the 423 to 443 region, found to be associated with higher survival probability, were taken as reference levels. All P values were two-sided, and values of 0.05 or less were considered to indicate statistical significance. SPSS 10.0 for Windows was used for all statistical analyses (SPSS Inc., Chicago, IL).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
An initial screening of BCL-6 protein expression was performed in
reactive lymphoid tissue. The pattern of expression was similar to that
described before: germinal center cells were variable although they
generally stained strongly for BCL-6. Mantle and interfollicular cells
were mostly negative, although some scattered cells were present in
these areas (Figure 1A)
.
|
.
In the other eight cases, the results could not be evaluated because of
the lack of adequate internal controls, which are usually provided by
BCL-6-positive reactive lymphocytes. Variation in signal intensity and
number of BCL-6-positive cells was observed and there was a strong
positive correlation between the two variables
(r = 0.437, P = 0.001, Spearman
correlation). Therefore, we only considered the number of positive
cells in the quantification of BCL-6 expression.
BCL-6 reactivity represented by fewer than 20% of positive cells was
found in 5 cases (all LBCL samples), whereas 13 cases (8 LBCL and 5 BL)
had more than 80% positive cells. Most cases (35 of 52) showed an
intermediate level of BCL-6 expression. Table 1
and Figure 1
summarize the
immunostaining results. There were only slight, nonsignificant
differences in the distribution of expression levels between LBCL and
BL cases.
|
. LBCL samples showed a good
correlation among protein and mRNA expression levels with the exception
of a discrepancy in five cases. The only striking finding was the
detection in MCL cases of mild, but distinct BCL-6 mRNA expression
levels, thus confirming the previous observations by Lossos and
co-workers.12 BCL-6 Mutations
Sequencing analysis was performed in 45 LBCL and 14 BL cases
identifying a wide variety of mutations that were present in the
majority of cases (80% LBCL and 71% BL were mutated). Table 2
shows a summary of the mutational
analysis. The frequency of mutations in our series was 3.7 x
10-3 mutations/bp (3.8 x
10-3 mutations/bp in LBCL cases and 3.7 x
10-3 mutations/bp in BL cases).
|
. The mutations present in the
series were mostly punctual substitutions. Two cases showed punctual
deletions, two cases had a punctual insertion, and one case featured a
long tandem duplication (31 bp) (See Table 2
for details). In agreement
with previous studies,7,20
there were two polymorphisms in
the MMC at positions 397 (C to G) and 520 (del T). Changes in these
nucleotides were observed in 42% of LBCL cases and 57% of BL cases.
Additionally, seven recurrent mutations were identified in more than
one patient (at positions 88, 112, 349, 350, 423, 432, 588). Moreover,
mutations to different nucleotides appeared at eight positions (110,
122, 125, 398, 448, 495, 514, 518) (recurrently mutated positions).
These positions can thus be considered to be mutational hot spots.
Hypothetical polymorphisms in these positions were excluded through the
study of nontumoral DNA for the changes in positions 88, 112, 423, 432,
448, 518, 588. Other possible polymorphisms had already been excluded
by previous studies.7,8,10,20-23
|
.
|
Because of its location in the first intron of the BCL-6 gene, MMC
could behave like a regulatory region, controlling the expression of
BCL-6. This prompted us to search for a relation between number of
mutations and protein expression. However, in our series of LBCL and BL
no such relation between these two variables was found
(r = 0.127, P = 0.37, Pearson
correlation coefficient) (Figure 3)
.
|
Distribution of Mutations along the Sequence of MMC: A Clustering Effect
Sequences of regulatory regions are not absolutely functional in
every base pair of their length. Usually, only small transcription
factor-binding regions are important for their regulatory physiological
effect. So we examined whether there were parts of the entire MMC where
mutations were clustered or had an effect on BCL-6 expression.
Distribution of mutations throughout the whole series is shown in
Figure 4
. In trying to identify small
regions of potential functional relevance, we focused on the existence
of mutational clusters associated with changes in BCL-6 protein
expression intensity. Thus, on the basis of the frequency of mutations
and/or variations in expression level, we defined two clusters,
spanning positions 106 to 127 and 423 to 443, that had a high mutation
frequency (1.7 x 10-2 mutations/bp for the
106 to 127 cluster, and 1.3 x 10-2
mutations/bp for the 423 to 443 cluster, versus 3.7 x
10-3 mutations/bp for the entire MMC). The
clustering distribution of BCL-6 mutations within MMC was statistically
significant (chi-square = 58.69, d.f. = 1 for cluster 106 to 127
and chi-square = 25.10, d.f. = 1 for cluster 423 to 443,
P < 0.001 for both clusters; Fisher’s exact test).
Three LBCL and one BL showed mutations in both clusters, six LBCL
showed mutations in 423 to 443 cluster, and eight LBCL and one BL
showed mutations in 106 to 127 cluster.
|
Because LBCLs and BLs are characterized by large differences in the
molecular mechanisms and clinical outcome, we analyzed the two groups
separately. The LBCL group yielded similar results to those obtained
from the entire series. However, BLs showed no clustering in the
BCL-6 mutation distribution, although the total number of BL cases
considered was relatively small (Figure 5)
. Additionally, no significant relation
was observed between the location of mutation and level of expression
in BL.
|
In the LBCL group, the Kaplan-Meier statistic obtained
indicated that patients with mutations within the 423 to 443 cluster
showed both an improved OS and DFS. Thus eight of nine LBCL patients
with mutations within the 423 to 443 cluster were still alive (Figure 6A
for OS) and in complete remission
(Figure 6B
for DFS) after a mean follow-up time of 110 months. This
contrasts with the OS and DFS observed in the rest of the series
(P = 0.011 for OS and P = 0.022
for DFS, log-rank test).
|
.
|
was found to be significantly associated with a longer OS
(P < 0.05). This effect was slightly greater
than that of IPI. These relationships were absent in the BL group.
Mutations within the 106 to 127 cluster had no significant effect on
survival probability. |
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Besides the distribution of mutations in clusters, we found some mutations appearing in more than one case (recurrent mutations), and positions where mutations gave rise to different base changes in several cases (recurrently mutated positions), as described previously.23 These positions are probably mutational hot spots, the possibility of them being polymorphisms having been excluded through the study of nontumoral DNA and/or in previous studies.7,8,10,20-23 Except for two cases, none of them coincided with the consensus sequence for mutations in IgH.
The functional relevance of mutations within an intronic region depends on its putative role as a regulatory region. This has been already demonstrated for the MMC in the BCL-6 gene, because it includes at least one region with a negative regulatory capacity.24 Mutations and other structural abnormalities within these regulatory regions may play a role in the deregulation of the expression of the BCL-6 gene in lymphomas.3 This deregulatory effect of BCL-6 mutations would be critically dependent on the precise location of the mutations, those involving small regions bound by transcription factors being of potential importance. This could explain why, in this and in previous studies, the overall frequency of BCL-6 mutations does not predict the clinical behavior or BCL-6 expression level. These variables may depend more on the precise location of these mutations.9 This consideration prompted us to analyze the relevance of specific redundant mutations or mutations involving discrete mutational hotspots, termed subclusters, within the MMC. Thus, we have focused on regions with an increased frequency of mutations and/or associated with marked alterations in the protein expression. We identified two regions, one defined by nucleotides 106 to 127 and another by nucleotides 423 to 443. Both regions contained a significantly higher proportion of mutations than in the MMC overall (five times greater than expected under a null hypothesis of random distribution; P < 0.001). These clusters have not been described in previous LBCL studies, but the analysis of the data provided by those authors seems to show similar clustering phenomena.9,23
Although mutations within the first subcluster were not associated with variations in the level of BCL-6 protein expression, cases with mutations inside the second subcluster had a higher level of BCL-6 protein expression than did the entire series (P = 0.005) that strongly suggests that they involve a region of importance in the regulation of BCL-6 protein expression. Additional arguments in favor of the relevance of this region in the BCL-6 regulation were provided by a logistic regression analysis showing that mutations in nucleotides 423 and 443, considered separately, tend to be associated with a high level of BCL-6 protein expression (data not shown). This clustering effect was observed in the LBCL group, whereas BL cases displayed no clear tendency toward clustering despite the presence of mutations in the MMC.
The regions identified here as being potentially involved in BCL-6 regulation do not coincide with those recently described by Kikuchi and co-workers,24 which could indicate the existence of different multiple regulatory regions within the BCL-6 MMC, identifiable through the use of alternative experimental approaches.
The importance of mutations within this cluster is further emphasized by the findings concerning clinical outcome in these patients. Thus, LBCL patients with mutations within this 423 to 443 subcluster had an increased OS probability compared with the remainder who did not have such mutations. Although this conclusion should be regarded with caution because of the small size of the series, the effect was strikingly absent in BL. Cases with mutations within the 106 to 127 subcluster showed no significant variations either in the level of expression of the protein or the clinical outcome.
These findings enable us to formulate a model in which mutations and deregulation of the level of expression of BCL-6 play a role in lymphoma genesis and/or progression. The relevance of the role of the BCL-6 gene is supported by recent data concerning BCL-6 down-regulated genes. Thus, the BCL-6 gene, described as a multifunctional regulator of lymphocyte differentiation and immune responses, has also been demonstrated to repress key genes in cell-cycle control, such as p27KIP1, blimp-1 (a repressor of c-myc), or cyclin D2,6 which could easily explain the clinical significance of BCL-6 mutations or alterations in the expression.
Alizadeh and co-workers,11 have shown that LBCL patients with germinal center profile expression (and hence BCL-6 positivity) had greater survival probability compared with LBCL patients with activated B-cell profile. Recently, the role of the expression of BCL-6 in survival has been clarified by Lossos and colleagues,12 showing that the high level of expression of this marker is associated with better clinical outcome. Our findings indicate a more subtle relationship between BCL-6 expression and clinical course, whereby only a high level of expression associated with mutations within the 423 to 443 cluster seems to be significantly related to the clinical course.
The physiological basis of this relationship between improved survival, increased BCL-6 expression, and clustering of BCL-6 mutations needs further investigation. Nevertheless, some recently published results show a relation between BCL-6 overexpression and increased sensitivity to apoptosis. Albagli and co-workers25 transfecting BCL-6 protein into SAOS2 cells observed cell-cycle arrest and an increased apoptotic rate. Moreover, a recent article, revealed that BCL-6, probably through the repression of blimp-1, and subsequent activation of c-myc, could also trigger apoptosis, a known effect of c-myc activation.6
Strikingly, in the BL cases analyzed here, the percentage of mutated cases was lower than in LBCL cases, although the mutation frequency was similar in both lymphoma types. Additionally, in this group, mutations within the MMC were not clearly clustered within the 107 to 126 and 423 to 443 subclusters, and neither were mutations within these regions associated with variations in the clinical course or the level of protein expression. Thus, it seems that the role of BCL-6 mutations in lymphoma is specifically dependent on histological type, probably reflecting the distinct molecular pathogenesis of each main lymphoma entity.
These results highlight the need for a more comprehensive analysis of the role of BCL-6 regulation in lymphoma genesis and progression, through the identification of putative transcription factors binding in the 423 to 443 region, and the analysis of the consequences of their binding for cell-cycle control, differentiation, and apoptosis.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by grants from the Fondo de Investigaciones Sanitarias (FIS 98/993), Ministerio de Sanidad y Consumo, and from the Comision Interministerial de Ciencia y Technologia (1FD97-0431), Spain.
Accepted for publication January 10, 2002.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Quijano, A. Lopez, A. Rasillo, S. Barrena, M. Luz Sanchez, J. Flores, C. Fernandez, J. M. Sayagues, C. S. Osuna, N. Fernandez, et al. Association between the proliferative rate of neoplastic B cells, their maturation stage, and underlying cytogenetic abnormalities in B-cell chronic lymphoproliferative disorders: analysis of a series of 432 patients Blood, May 15, 2008; 111(10): 5130 - 5141. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-W. Chen, X.-T. Hu, A. C. Liang, W.-Y. Au, C.-C. So, M. L. Wong, L. Shen, Q. Tao, K.-M. Chu, Y.-L. Kwong, et al. High BCL6 expression predicts better prognosis, independent of BCL6 translocation status, translocation partner, or BCL6-deregulating mutations, in gastric lymphoma Blood, October 1, 2006; 108(7): 2373 - 2383. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Margalit, H. Amram, N. Amariglio, A. J. Simon, S. Shaklai, G. Granot, N. Minsky, A. Shimoni, A. Harmelin, D. Givol, et al. BCL6 is regulated by p53 through a response element frequently disrupted in B-cell non-Hodgkin lymphoma Blood, February 15, 2006; 107(4): 1599 - 1607. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Jardin, C. Bastard, N. Contentin, F. Parmentier, J.-M. Picquenot, H. Tilly, F. K. Stevenson, and S. S. Sahota Intronic BCL-6 mutations are preferentially targeted to the translocated allele in t(3;14)(q27;q32) non-Hodgkin B-cell lymphoma Blood, September 1, 2003; 102(5): 1872 - 1876. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Niu, G. Cattoretti, and R. Dalla-Favera BCL6 Controls the Expression of the B7-1/CD80 Costimulatory Receptor in Germinal Center B Cells J. Exp. Med., July 21, 2003; 198(2): 211 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sanchez-Beato, A. Sanchez-Aguilera, and M. A. Piris Cell cycle deregulation in B-cell lymphomas Blood, February 15, 2003; 101(4): 1220 - 1235. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
