If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Hematology, University Hospital of Salamanca, Salamanca, SpainInstitute of Biomedical Research of Salamanca (IBSAL), Salamanca, SpainSpanish Lymphoma/Autologous Bone Marrow Transplant Study Group, University Hospital Marqués de Valdecilla, Santander, Spain
Department of Hematology, University Hospital of Salamanca, Salamanca, SpainServicio de Anatomía Patológica, University Hospital of Marqués de Valdecilla/IFIMAV, Santander, SpainSpanish Lymphoma/Autologous Bone Marrow Transplant Study Group, University Hospital Marqués de Valdecilla, Santander, Spain
Department of Hematology, University Hospital of Salamanca, Salamanca, SpainServicio de Anatomía Patológica, University Hospital of Marqués de Valdecilla/IFIMAV, Santander, SpainCancer Research Center-IBMCC (USAL-CSIC), Salamanca, Spain
Department of Hematology, University Hospital of Salamanca, Salamanca, SpainServicio de Anatomía Patológica, University Hospital of Marqués de Valdecilla/IFIMAV, Santander, SpainCancer Research Center-IBMCC (USAL-CSIC), Salamanca, Spain
Address reprint requests to Marcos González, M.D., Ph.D., Servicio de Hematología, Hospital Universitario de Salamanca, Paseo de San Vicente, 58-182, 37007, Salamanca, Spain
Department of Hematology, University Hospital of Salamanca, Salamanca, SpainServicio de Anatomía Patológica, University Hospital of Marqués de Valdecilla/IFIMAV, Santander, SpainCancer Research Center-IBMCC (USAL-CSIC), Salamanca, Spain
The pathogenesis of diffuse large B-cell lymphoma (DLBCL) remains partially unknown. The analysis of the B-cell receptor of the malignant cells could contribute to a better understanding of the DLBCL biology. We studied the molecular features of the immunoglobulin heavy chain (IGH) rearrangements in 165 patients diagnosed with DLBCL not otherwise specified. Clonal IGH rearrangements were amplified according to the BIOMED-2 protocol and PCR products were sequenced directly. We also analyzed the criteria for stereotyped patterns in all complete IGHV-IGHD-IGHJ (V-D-J) sequences. Complete V-D-J rearrangements were identified in 130 of 165 patients. Most cases (89%) were highly mutated, but 12 sequences were truly unmutated or minimally mutated. Three genes, IGHV4-34, IGHV3-23, and IGHV4-39, accounted for one third of the whole cohort, including an overrepresentation of IGHV4-34 (15.5% overall). Interestingly, all IGHV4-34 rearrangements and all unmutated sequences belonged to the nongerminal center B-cell–like (non-GCB) subtype. Overall, we found three cases following the current criteria for stereotyped heavy chain VH CDR3 sequences, two of them belonging to subsets previously described in CLL. IGHV gene repertoire is remarkably biased, implying an antigen-driven origin in DLBCL. The particular features in the sequence of the immunoglobulins suggest the existence of particular subgroups within the non-GCB subtype.
Diffuse large B-cell lymphoma (DLBCL) is a type of aggressive lymphoma that accounts for about 40% of B-cell non-Hodgkin lymphomas in adults,
Swerdlow S.H. Campo E. Harris N.L. Jaffe E.S. Pileri S.A. Stein H. Thiele J. Vardiman J.W. International Agency for Research on Cancer: WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. World Health Organization Press,
Geneva Switzerland2008
Swerdlow S.H. Campo E. Harris N.L. Jaffe E.S. Pileri S.A. Stein H. Thiele J. Vardiman J.W. International Agency for Research on Cancer: WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. World Health Organization Press,
Geneva Switzerland2008
Genome-wide expression profile technologies have allowed DLBCL NOS to be separated into two biological subtypes: germinal center B-cell–like (GCB) DLBCL and activated B-cell–like (ABC) DLBCL,
and have been proposed as feasible surrogates of genome-wide expression profiles.
DLBCL is constituted by large B cells harboring a clonal rearrangement of immunoglobulin (IG) genes. The rearrangement of the immunoglobulin heavy chain (IGH) gene occurs during B-cell ontogeny in an ordered fashion, leading to the assembly of distinct variable (IGHV), diversity (IGHD), and joining (IGHJ) genes.
During B-cell differentiation, rearrangements start with an IGHD gene joining to an IGHJ gene, forming a partial IGH-D-IGHJ (D-J) rearrangement, a process that can simultaneously occur in both alleles. This is followed by the joining of an IGHV gene to the partial D-J rearrangement, forming a completely functional IGHV-IGHD-IGHJ (V-D-J) rearrangement in one allele, but can also retain a partial nonfunctional rearrangement in the other allele. Once a rearrangement has become completely functional, the naive B cell can react to antigens in the germinal center (GC) of the secondary lymphoid organs, undergoing affinity maturation by the processes of somatic hypermutation (SHM) and class switch recombination.
The molecular features of IGH rearrangements may provide important information about the ontogeny of B cells. Thus, the intrinsic mutability of IGHV genes has been widely analyzed.
Since normal B cells undergo SHM as a response to antigen, classically within GC, the pattern of SHM is thought to provide clues about the mutation machinery and the role of antigen selection. Nonrandom, potentially antigen-driven SHM has been described in normal B cells, as well as in chronic lymphocytic leukemia (CLL) and classic hairy cell leukemia (HCL), which has been called canonical SHM.
The pattern and distribution of immunoglobulin VH gene mutations in chronic lymphocytic leukemia B cells are consistent with the canonical somatic hypermutation process.
canonical mutation criteria remain as an available method for establishing whether or not a set of sequences are affected by similar processes and selections.
Moreover, IGHV gene usage in IGH rearrangements has been analyzed in normal B cells,
Analysis of the human VH gene repertoire Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(−)/IgM+ B cells.
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
The involvement of particular antigens in lymphomagenesis may be supported by the identification of stereotyped groups with closely similar complementarity-determining region 3 (VH CDR3).
Stereotyped patterns of somatic hypermutation in subsets of patients with chronic lymphocytic leukemia: implications for the role of antigen selection in leukemogenesis.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
Chronic lymphocytic leukemias utilizing the VH3-21 gene display highly restricted Vlambda2-14 gene use and homologous CDR3s: implicating recognition of a common antigen epitope.
Although DLBCL is the most prevalent aggressive lymphoma, few studies have characterized the BCR in this entity, and most are based on small series of patients or in selected populations.
Diffuse large cell lymphomas are derived from mature B cells carrying V region genes with a high load of somatic mutation and evidence of selection for antibody expression.
Molecular analysis of immunoglobulin variable genes in human immunodeficiency virus-related non-Hodgkin's lymphoma reveals implications for disease pathogenesis and histogenesis.
In addition, partial D-J rearrangements, the pattern and distribution of mutations, and the existence of stereotyped sequences in DLBCL are still not well defined.
Molecular analysis of immunoglobulin variable genes in human immunodeficiency virus-related non-Hodgkin's lymphoma reveals implications for disease pathogenesis and histogenesis.
Here, we report a detailed immunogenetic analysis of the complete and partial IGH rearrangements in 165 DLBCL, the largest series to date, to improve our understanding of the biology of this disease.
Materials and Methods
Patient Characteristics and Immunohistochemistry
A total of 165 patients with DLBCL were selected retrospectively on the basis of a diagnosis of de novo DLBCL NOS according to the 2008 World Health Organization classification.
Swerdlow S.H. Campo E. Harris N.L. Jaffe E.S. Pileri S.A. Stein H. Thiele J. Vardiman J.W. International Agency for Research on Cancer: WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. World Health Organization Press,
Geneva Switzerland2008
The median age was 59 years, with a range from 15 to 86 years. Fifty-one percent of the patients were male, and 48% of the total series had an International Prognostic Index score ≥3.
A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma The Non-Hodgkin's Lymphoma Classification Project.
Accordingly, immunostaining with antibodies to BCL-6, CD10, and melanoma-associated antigen (mutated) 1 (MUM1/IRF4) was performed in 121 patients, which allowed them to be separated into GCB DLBCL (38%) and non-GCB (62%). A total of 159 patients received treatment with curative intent, 125 of them (81%) with rituximab-based regimens (Table 1).
Table 1Clinical-Biological Characteristics of DLBCL Patients at Diagnosis (n = 165)
Variable
% of Patients
Median follow-up, months (range)
27.8 (0–198)
Age [years, median (range)]
58.7 (15–86)
≤60
54
>60
46
Sex
Male
51
Female
49
Performance status
ECOG <2
70
ECOG ≥2
30
β2-microglobulin >3 mg/L
36
Albumin <3.5 g/dL
47
LDH elevated
59
IPI
IPI 0–1
24
2
28
3
24
IPI 4–5
24
Immunostaining
GCB
38
Non-GCB
62
Treatment
Rituximab-based
81
Others
19
ECOG, Eastern Cooperative Oncology Group; GCB, germinal center B-cell–like; IPI, International Prognostic Index; LDH, lactate dehydrogenase.
The study was approved by the local ethics review committee in accordance with Spanish legislation, and informed consent was obtained from all participants.
DNA Extraction
Tumor DNA was extracted from samples collected at the time of diagnosis. These involved lymph nodes (84%), bone marrow (10%), and other tumor tissues (6%). In fresh samples, high-molecular-weight DNA was isolated using the DNAzol reagent (MRC, Cincinnati, Ohio).
Heteroduplex analysis of VDJ amplified segments from rearranged IgH genes for clonality assessments in B-cell non-Hodgkin's lymphoma A comparison between different strategies.
DNA was extracted from formalin-fixed paraffin-embedded (FFPE) tissue in half of the lymph nodes using RecoverAll Total Nucleic Acid Isolation Kit (Ambion/Applied Biosystems, Foster City, California) or QuickExtract FFPE DNA Extraction Kit (Epicenter Biotechnologies, Madison, Wisconsin), with no differences in terms of recovery capacity.
Identification of Clonal Rearrangements
All samples were tested for the amplification of IGH rearrangements according to the BIOMED-2 Concerted Action protocols, where we actively participated in the standardization.
Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936.
In this strategy, complete V-D-J rearrangement amplification was performed by multiplex PCR with a set of family-specific primers of the framework region 1 (FR1) and one IGHJ consensus primer. In cases using the IGHV4-34 gene, a second PCR was performed with primers from the IGHV leader region. For the samples with no detectable amplification from FR1, PCR was performed from FR2. Additionally, partial D-J rearrangements were amplified in two different reactions using family-specific primers for the IGHD1-IGHD6 and IGHD7 families, respectively, together with the consensus IGHJ primer. The presence of the monoclonal rearrangement was then confirmed by GeneScan with an ABI 3130 DNA Sequencer (Applied Biosystems).
Sequencing and Identification of IGH Gene Rearrangements
PCR products were sequenced directly, including both forward and reverse reads, using Big-Dye terminators (Applied Biosystems).
Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936.
Germline IGHV, IGHD, and IGHJ genes from complete V-D-J rearrangements were identified using the IMGT/V-QUEST database (http://www.imgt.org, last accessed March 21, 2012). The following information was extracted: IGHV, IGHD, and IGHJ gene usage, the percentage of IGHV identity to the closest germline gene, and the VH CDR3 length and composition. The VH CDR3 length was determined by counting the number of amino acids from the cysteine at position 104 and the tryptophan at position 118, according to the IMGT unique numbering system.
Analysis of the human VH gene repertoire Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(−)/IgM+ B cells.
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
Percentage Identity with the Closest IGHV Germline
The use of 98% as the percentage identity cutoff value has customarily been used for the clinical consideration of a patient's sequence as mutated or unmutated based on the exclusion of potential genomic polymorphisms. This cutoff has a prognostic value in CLL.
Insight into the molecular pathogenesis of hairy cell leukaemia, hairy cell leukaemia variant and splenic marginal zone lymphoma, provided by the analysis of their IGH rearrangements and somatic hypermutation patterns.
: “truly unmutated” when no deviation from the germline was observed; “minimal/borderline mutated” when there was 97.0% to 99.9% similarity with the germline; and finally, “highly mutated” when the deviation from the germinal sequence was >97%.
SHM Characteristics
Nucleotide characteristics of IGHV mutations in DLBCL were studied to establish whether the pattern of distribution was consistent with the canonical SHM process.
The pattern and distribution of immunoglobulin VH gene mutations in chronic lymphocytic leukemia B cells are consistent with the canonical somatic hypermutation process.
This analysis was also performed separately in both the GCB and non-GCB subtypes, as well as in DLBCL expressing particular IGHV genes.
The defined characteristics of canonical SHM include: i) a higher ratio of replacement (R) versus silent (S) mutations in the CDRs than in the FRs; ii) base change bias for transitions; and iii) more than the expected percentage of mutations in the RGYW hotspots (R = A/G, Y = C/T, and W = A/T) relative to the total number of nucleotides located within these hotspots. In addition, to determine whether these characteristics were due to selection or to mutation machinery, the mutations at redundant wobble bases were analyzed as previously reported for CLL and HCL.
The pattern and distribution of immunoglobulin VH gene mutations in chronic lymphocytic leukemia B cells are consistent with the canonical somatic hypermutation process.
Wobble bases are the third nucleotide of redundant codons. Finally, mutations at A:T and G:C pairs were studied to determine whether the mutations were targeted with approximately the same frequency as in normal cells.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
to identify possible restrictions in the VH CDR3 amino acid composition of rearrangements. Briefly, the criteria for stereotyped patterns included: i) at least 50% amino acid identity; ii) at least 70% similarity between sequences; iii) IGHV usage of the same clan; iv) identical VH CDR3 length; and v) exact location of the sequence pattern within the VH CDR3 region. Analyses were performed using the ClustalW/ClustalX 2.0 multiple sequence alignment software.
those composed of only two sequences being considered “provisional.” A letter of the alphabet was assigned to the new subset. These were compared with previously published stereotyped patterns
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
The groups of patients were compared by the χ2 or Fisher exact tests using Graph-Pad 4.0 (GraphPad Software, San Diego, California) and SPSS 15.0 (SPSS, Chicago, Illinois).
Differences were considered to be statistically significant for values of P < 0.05.
Results
Detection of Clonal Rearrangements
The overall detection rate of monoclonality by amplifying V-D-J and/or D-J rearrangements was 90% (149 of 165). V-D-J rearrangements were detected in 130 of 165 patients (79%), and D-J rearrangements were seen in 68 of 165 patients (41%).
Complete IGHV-IGHD-IGHJ Rearrangements
A total of 109 of 130 complete V-D-J rearrangements were successfully sequenced. Reasons for not having a sequence in the remaining 21 cases were as follows: high polyclonal background (n = 2); double sequences, with no possible individual reading (n = 1); unspecific amplification, which means that a sequence could be read but the result did not fit with any possible rearrangement (n = 2); low fluorescence (intensity) of the peak rendering an amplification band too weak for accurate sequencing performance (n = 11); and partial sequence, enough to be identified as a V-D-J rearrangement but insufficient to be completely identified (n = 5).
A total of 103 were productive rearrangements, whereas 6 were nonproductive. Most sequences (82%) were obtained from FR1. Within the productive setting, IGHV gene repertoire analysis revealed that IGHV3 and IGHV4 were the predominant subgroups (Table 2). A total of 27 functional IGHV genes were identified, of which IGHV4-34 (15.5%), IGHV3-23 (9.7%), and IGHV4-39 (8.7%) were the most frequent. Regarding IGHD genes, a total of 25 were identified, with a predominance of IGHD3 (Table 2), especially IGHD3-22 and IGHD3-10. With respect to IGHJ gene usage, most cases used the IGHJ4 and IGHJ6 genes.
Table 2Distribution of Families in Productive V-D-J and Partial D-J Rearrangements
Sequence analysis was possible in 64 partial D-J rearrangements. The IGHD2 subgroup genes were most frequent (Table 2). In comparison with complete rearrangements, IGHD4 was significantly overrepresented in partial rearrangements (18.8% versus 5.8%, P < 0.05), whereas IGHD3 was underrepresented (14.1% versus 35.9%, P < 0.01). Overall, 20 IGHD genes were identified, of which IGHD2-2 was the most frequent. IGHD3-22 and IGHD3-3, two of the most frequent IGHD genes in V-D-J rearrangements, were completely absent in partial rearrangements, suggesting that the use of certain IGHD genes may favor the rearrangement process to obtain a final complete functional rearrangement. Finally, the distribution pattern of the IGHJ genes was similar to the complete rearrangements (Table 2).
Somatic Hypermutation
Percentage Identity to the Closest IGHV Germline
Somatic hypermutation (<98% of identity to the closest germline IGHV gene) was present in 94 of 103 (91%) patients who had productive complete V-D-J rearrangements (median: 10.8%; range: 0% to 26%). As expected, only 9% of the sequences were unmutated, similar to previously reported results.
4, 8, and 91 cases were “truly unmutated,” “borderline mutated,” and “highly mutated,” respectively. All truly unmutated and borderline mutated sequences were obtained from FR1.
Interestingly, all cases with the IGHV4-34 gene but one (15 of 16) were “highly mutated.” However, IGHV1-69, IGHV5-51, and IGHV3-21 were the most frequent genes used in “truly unmutated” and “borderline mutated” cases.
As expected from the physiopathological point of view, mutations were observed in a minority of cases (12%) of partial D-J rearrangements with a low mutation frequency (median: 4.3%, range: 3.5% to 11.1%).
Nucleotide Characteristics of Somatic Hypermutation
S mutations occurred randomly, with a relatively even distribution, whereas R mutations were mainly localized in the CDR1 and CDR2 regions, at the 5′ and 3′ ends of FR2 and the 3′ end of FR3 (see Supplemental Figure S1 at http://ajp.amjpathol.org).
Analysis of canonical SHM was possible in 96 DLBCL sequences with <100% identity with the germline. For the 2485 mutations analyzed, the R/S mutation ratio was 1.72, being greater in the CDR (2.56) than in the FR (1.42) regions (P < 0.01) (Table 3). The transition-to-transversion ratio was 1.16, significantly higher than the expected ratio of 0.5 if the frequencies of the types of mutation were random (P < 0.01), and there was no significant difference between the CDR and FR regions. Finally, 24.9% of the DLBCL mutations were located in RGYW hotspots, a value significantly greater than the expected percentage (14.0%, P < 0.01). Thus, SHM in DLBCL displayed all of the established characteristics of canonical SHM. These canonical characteristics persisted in the most common IGHV genes, except for IGHV1 (mainly IGHV1-69), which met these three criteria to a lesser extent (Table 3).
Table 3Characteristics of SHM in DLBCL (Global Series) and with Respect to Immunohistochemistry and the Most Common IGHV Families and Genes
Number of sequences
Number of mutations
R/S ratio
Transition/transversion ratio
% nucleotides in RGYW
% of mutations in RGYW
Total
CDR
FR
Total
CDR
FR
DLBCL
96
2485
1.69
2.56
1.42
1.161
1.13
1.18
14.0
24.9
GCB
31
931
1.71
2.77
1.42
1.155
0.97
1.25
13.3
22.9
Non-GCB
46
1037
1.56
2.05
1.39
1.216
1.30
1.18
14.2
25.2
IGHV1
15
631
1.42
1.45
1.42
1.246
1.29
1.24
17.9
26.7
IGHV3
42
1127
1.81
2.71
1.49
1.159
1.16
1.16
12.0
19.3
IGHV4
32
662
1.70
2.98
1.27
1.115
1.07
1.14
15.9
31.6
IGHV1-69
4
46
0.92
0.88
0.94
1.300
1.50
1.21
19.8
32.6
IGHV3-23
10
243
1.60
2.24
1.31
1.505
1.47
1.53
13.1
26.3
IGHV4-34
15
304
1.34
1.91
1.22
1.338
1.39
1.32
14.8
30.6
Figures in bold indicate different R/S ratios from that expected.
The bias for transitions over transversions was also reflected by the wobble base mutations (1.20, P < 0.01) (Figure 1), implying that this bias in DLBCL is due to the same mechanism of SHM.
Figure 1Nucleotides mutated in DLBCL. Relative frequency with which each nucleotide mutates to form each of the other three. The number of mutations occurring in each nucleotide is shown in parentheses. A: Total mutations in DLBCL. B: Mutations in redundant wobble bases.
Finally, G:C pairs were preferentially targeted for SHM relative to A:T (ratio: 1.86) (Figure 1). This observation was especially evident in the IGHV4-34 gene (ratio: 2.23).
Recurrent Amino Acid Changes
Recurrent amino acid changes were mainly found among IGHV4-34, IGHV4-39, and IGHV1-18 rearrangements. For example, a substitution at the FR3 codon 92 occurred in 13 of 16 (81%) cases of IGHV4-34; the substitution of S to T (6 of 16, 38%) and S to N (4 of 16, 25%) were the most frequent changes observed. This recurrent amino acid change contrasts with the adjacent codon, 93, which did not undergo any substitution in any of the 16 sequences. Although the IGHV4-34 genes were highly mutated, all 10 available IGHV4-34 sequences conserved the specific AVY motif at the end of FR1 codons 24 to 26, and the W at position 7 was intact in the 5 IGHV4-34 cases that could be analyzed from the leader region. By contrast, the N-glycosylation site in CDR2 of IGHV4-34 was lost in 7 of 16 (44%), and only one of the sequences that lost its natural site generated two novel sites in FR3. A summary list of recurrent amino acid substitutions for the predominant IGHV genes is given in Supplemental Table S1 (available at http://ajp.amjpathol.org). Some of these changes may indicate specific driven SHM in DLBCL.
VH CDR3 Composition and Stereotyped VH CDR3 Sequences
The median VH CDR3 length was 15 amino acids (range: 7 to 26). VH CDR3 was longer in the IGHV4 family than in the other families (16.8 versus 14.8, P = 0.01), and in “truly unmutated” and borderline cases as compared with “highly mutated” cases (18.2 versus 15.1, P = 0.01). This was particularly pronounced among the “truly unmutated” cases (21.5 versus 15.1, P < 0.01).
Three sequences among the 103 patients in our series met the criteria for stereotyped rearrangements and could be included into different subsets. Two sequences clustered within previously described CLL subsets (subsets 5 and 7D);
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
both contained the IGHV1-69 gene and were “truly unmutated.” The other case formed a probably realistic novel provisional subset (Table 4): subset A, sharing the V-D-J IGHV4-34-IGHD3-22-IGHJ2 conformation and with 79% identity of the amino acid sequence with a public DLBCL sequence.
Stereotyped rearrangements are compared with representative sequences from each subset. A complete list of sequences from subsets 5 and 7D is provided in Supplemental Table S2 at http://ajp.amjpathol.org. Sequences derived from the present study are remarked in bold.
IGHV usage differed between the immunohistochemical subtypes. IGHV1 (10 of 14) and IGHV4 (15 of 21) were more frequent in the non-GCB subtype, especially for IGHV1-69 (3 of 4) and IGHV4-34 (12 of 12). A higher percentage of SHM was observed in GCB (13.8%, range: 2.9% to 26.0%) than in non-GCB (10.1%, range: 0.4% to 23.6%, P < 0.01) cases. Moreover, 6 of 7 borderline or unmutated cases with available immunostaining were non-GCB, which could explain this difference. By contrast, all IGHV4-34 cases, despite being highly mutated, belonged to the non-GCB subtype. With respect to other molecular features, VH CDR3 length and the canonical characteristics for SHM showed no differences between the GCB and non-GCB subtypes (Table 3).
Comparative Analysis of IGH Rearrangements in DLBCLs with Other Normal B Cells and B-LPDs
To highlight the singular characteristics of DLBCLs, we compared our results with those reported for normal B cells and other B-LPDs. As shown in Table 5, the IGHV4-34 gene was significantly overrepresented in DLBCL relative to CD5−/IgM+ normal B cells (3% to 9%, P < 0.01) and to other B-LPDs, such as CLL (8.7%, P < 0.05), HCL (6.9%, P = 0.07), and especially with multiple myeloma and Waldeström macroglobulinemia (2.7% and 0%, respectively, P < 0.01). However, it was found in a similar proportion of patients with splenic marginal zone lymphoma and mantle cell lymphoma.
Analysis of the human VH gene repertoire Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(−)/IgM+ B cells.
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
Human immunoglobulin (Ig)M+IgD+ peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: cD27 as a general marker for somatically mutated (memory) B cells.
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
such as IGHV3-30, were completely absent from the present series (P < 0.01). Most complete V-D-J rearrangements were mutated, with a similar percentage of SHM (11.8%) to those previously reported in multiple myeloma (9%), but higher values than in CLL and Waldeström macroglobulinemia (2.4% to 6.6%).
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
Regarding SHM, G:C pairs were more often targeted for SHM than A:T pairs (ratio: 1.86) (Figure 1). This differs significantly from the distribution observed in CLL (ratio: 1.56, P < 0.01), HCL (ratio 0.96, P < 0.01), and in normal B cells (ratio: 1.00, P < 0.01).
The pattern and distribution of immunoglobulin VH gene mutations in chronic lymphocytic leukemia B cells are consistent with the canonical somatic hypermutation process.
Table 5Comparison of the IGHV, IGHD, and IGHJ Families and Significantly Different IGHV Genes in Our Series of DLBCL with Those Reported for B-LPDs and Normal B Cells in the Complete V-D-J Setting
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
Analysis of the human VH gene repertoire Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(−)/IgM+ B cells.
Partial D-J rearrangements were present in 41% of the cases in our series (Table 2), with a similar IGHD usage to that of other B-LPDs (data not shown).
Discussion
Although DLBCL is the most frequent variant of non-Hodgkin lymphoma, its ontogeny is still not well understood. In this study, we analyzed complete V-D-J rearrangements in 165 untreated DLBCL, which is the largest series reported to date. Additionally, we also characterized the partial D-J rearrangements. Our results, based on the characterization of the BCR in DLBCLs, with a biased use of certain IGHV genes, evidence of canonical SHM, and the findings of shared amino acid changes and stereotyped sequences, provide new information about the ontogeny of DLBCL. Our data suggest that certain particular antigens must be involved in the development of DLBCL by stimulating the proliferation of B cells that express surface IG encoded by certain IGHV genes. Moreover, DLBCLs show clear differences from other B-LPDs with respect to IGHV usage (Table 5) and SHM.
A strong bias at the level of IGHV gene usage was observed, with the IGHV4-34, IGHV3-23, and IGHV4-39 genes accounting for one third of the cohort. In addition, recurrent amino acid changes were found in certain IGHV genes, which may indicate specific driven SHM in DLBCL. These findings suggest that DLBCL could derive from certain B-cell populations specifically stimulated during the antigenic response. This hypothesis is even more plausible if the presence of some stereotyped sequences is taken into account. We identified three VH CDR3 stereotyped sequences in our series, which is a lower frequency than in CLL.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
The stereotyped sequences had a rearrangement involving the IGHV1-69 or IGHV4-34 genes. Moreover, two of them shared their IGHV1-69 rearrangement with subgroups previously described in CLL.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
It is well known that there is a relationship between CLL and DLBCL, since some cases of DLBCL occur on a CLL background (Richter transformation). However, the cases reported here are de novo DLBCL, indicating “common” stereotypes between DLBCL and CLL, in contrast to other B-cell malignancies.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
This highly restricted immunoglobulin gene repertoire with stereotyped VH CDR3 and common SHM targeting in DLBCL suggests a role for antigen selection in this entity, at least for some subsets of cases. Some of the antigens could be common to different B-LPDs but could ultimately lead to a different entity, depending on which population they stimulate. Although the results reported here support the notion that stereotyped BCRs in DLBCL do not seem to have as relevant implications as in CLL or other B-LPDs, the frequency of these stereotyped sequences and their possible biological and clinical significance in DLBCL need further analysis in larger and independent series.
The IGHV4-34 gene was overrepresented in our series (15.5% overall), appearing at a higher frequency than in normal B cells,
Analysis of the human VH gene repertoire Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(−)/IgM+ B cells.
Molecular analysis of immunoglobulin variable genes in human immunodeficiency virus-related non-Hodgkin's lymphoma reveals implications for disease pathogenesis and histogenesis.
However, there are other series in which this percentage was slightly smaller, although this discrepancy could be accounted for in terms of differences in the sample size or the CGB/non-GCB distribution.
Diffuse large cell lymphomas are derived from mature B cells carrying V region genes with a high load of somatic mutation and evidence of selection for antibody expression.
Interestingly, the patients showing IGHV4-34 rearrangements in our series shared several common features: i) all but one were highly mutated (15 of 16 > 2%, median: 9.5%, range: 1.7% to 23.6%); ii) all of them (12 of 12) belonged to the non-GCB subtype; and iii) all (10 of 10) conserved the AVY motif. Moreover, the critical W at position 7 was intact in the five cases that could be analyzed. This specific FR1 motif is thought to be responsible for binding the N-acetyllactosamine antigenic determinant, implying that these cells may retain the ability to bind to and be activated by superantigens, despite intense SHM activity.
The I binding specificity of human VH 4–34 (VH 4–21) encoded antibodies is determined by both VH framework region 1 and complementarity determining region 3.
Novel Ig V gene features of t(14;18) and t(3;14) de novo diffuse large B-cell lymphoma displaying germinal center-B cell like and non-germinal center-B cell like markers.
and we confirm, in a larger series, the strong association between DLBCL expressing the IGHV4-34 and non-GCB subtype. However, these results require future confirmation using genome-wide expression profiles as a tool for distinguishing between the various biological subtypes of DLBCL. In addition, we describe a new stereotyped VH CDR3 pattern in DLBCL using IGHV4-34. The molecular features of this subgroup differ from other B-LPDs using this gene. For example, in HCL IGHV4-34 is mutated in only 1 of 17 cases (6%)
Finally, mutations in codons 64 and 80 are found in IGHV4-34 sequences from DLBCL and from ocular adnexal extranodal marginal zone lymphoma, which is more frequent than in normal B cells or other B-LPDs.
Analysis of the human VH gene repertoire Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(−)/IgM+ B cells.
Human immunoglobulin (Ig)M+IgD+ peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: cD27 as a general marker for somatically mutated (memory) B cells.
Biased use of the IGHV4 family and evidence for antigen selection in Chlamydophila psittaci-negative ocular adnexal extranodal marginal zone lymphomas.
Novel Ig V gene features of t(14;18) and t(3;14) de novo diffuse large B-cell lymphoma displaying germinal center-B cell like and non-germinal center-B cell like markers.
observed ongoing mutations in non-GCB DLBCL, including IGHV4-34 cases, which would suggest an independent GC mechanism of SHM. Regardless of the GC dependence of these mutations, we observed that SHM of the non-GCB subtype met the canonical criteria (IGHV4-34 included), which do not support the need to invoke a mutational machinery different from the normal process for IGHV mutations.
In conclusion, on the basis of particular features in the sequence of the immunoglobulin genes, we found that different subgroups exist within the non-GCB subtype. Our data support the idea that DLBCL is characterized by a highly distinctive IG gene repertoire and, to the best of our knowledge, for the first time, we describe stereotyped sequences in DLBCL and “common” stereotypes between CLL and other B-LPD. These results indicate a role for antigen selection in DLBCL development and also open possibilities for future investigations into the biology of DLBCL.
Acknowledgments
We thank Alicia Antón, Felicitación García, Montserrat Hernández-Ruano (Hospital Universitario de Salamanca, Spain), and Mar López-Colomer (Centro Nacional de Investigaciones Oncológicas, Madrid, Spain) for their technical support. In addition, we thank all of the members of GELTAMO and the patients who participated in this study.
Distribution of replacement (A) and silent (B) mutations along the IGHV sequence. *Mutations that by themselves would have truncated the sequence. CDR, complementarity-determining region; FR, framework region.
References
Swerdlow S.H. Campo E. Harris N.L. Jaffe E.S. Pileri S.A. Stein H. Thiele J. Vardiman J.W. International Agency for Research on Cancer: WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. World Health Organization Press,
Geneva Switzerland2008
The pattern and distribution of immunoglobulin VH gene mutations in chronic lymphocytic leukemia B cells are consistent with the canonical somatic hypermutation process.
Molecular characterization of heavy chain immunoglobulin gene rearrangements in Waldenstrom's macroglobulinemia and IgM monoclonal gammopathy of undetermined significance.
Stereotyped patterns of somatic hypermutation in subsets of patients with chronic lymphocytic leukemia: implications for the role of antigen selection in leukemogenesis.
Stereotyped B-cell receptors in one third of chronic lymphocytic leukemia: towards a molecular classification with implications for targeted therapeutic interventions.
Chronic lymphocytic leukemias utilizing the VH3-21 gene display highly restricted Vlambda2-14 gene use and homologous CDR3s: implicating recognition of a common antigen epitope.
Diffuse large cell lymphomas are derived from mature B cells carrying V region genes with a high load of somatic mutation and evidence of selection for antibody expression.
Molecular analysis of immunoglobulin variable genes in human immunodeficiency virus-related non-Hodgkin's lymphoma reveals implications for disease pathogenesis and histogenesis.
Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936.
Insight into the molecular pathogenesis of hairy cell leukaemia, hairy cell leukaemia variant and splenic marginal zone lymphoma, provided by the analysis of their IGH rearrangements and somatic hypermutation patterns.
Human immunoglobulin (Ig)M+IgD+ peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: cD27 as a general marker for somatically mutated (memory) B cells.
The I binding specificity of human VH 4–34 (VH 4–21) encoded antibodies is determined by both VH framework region 1 and complementarity determining region 3.
Novel Ig V gene features of t(14;18) and t(3;14) de novo diffuse large B-cell lymphoma displaying germinal center-B cell like and non-germinal center-B cell like markers.
Biased use of the IGHV4 family and evidence for antigen selection in Chlamydophila psittaci-negative ocular adnexal extranodal marginal zone lymphomas.
Supported in part by grants from the Health Council of Castilla y León (GRS265/A/08) and the Health Research Program (PS0901382 and CM10/00078-Río Hortega, ISCIII).
P < 0.05.
00598-6&id=gr1.jpg){kind=link}