September 8, 2008

Recurrent Inactivation of the PRDM1 Gene in Primary Central Nervous System Lymphoma

By Courts, Cornelius Montesinos-Rongen, Manuel; Brunn, Anna; Bug, Stefanie; Siemer, Dorte; Hans, Volkmar; Blumcke, Ingmar; Klapper, Wolfram; Schaller, Carlo; Wiestler, Otmar D; Kuppers, Ralf; Siebert, Reiner; Deckert, Martina

Abstract Primary lymphomas of the CNS (PCNSLs) show molecular features of the late germinal center exit B-cell phenotype and are impaired in their terminal differentiation as indicated by a lack of immunoglobulin class switching. Because the positive regulatory domain I protein with ZNF domain (PRDM1/BLIMP1) is a master regulator of terminal B-cell differentiation into plasma cells, we investigated a series of 21 PCNSLs for the presence of mutations in the PRDM1 gene and alterations in the expression pattern of the PRDM1 protein. Direct sequencing of all coding exons of the PRDM1 gene identified deleterious mutations associated with abrogation of PRDM1 protein expression in 4 of 21 (19%) PCNSLs. Thus, similar to systemic diffuse large B-cell lymphomas, PRDM1 may be a tumor suppressor in some PCNSL and contribute to lymphomagenesis by impairing terminal differentiation.

Key Words: BLIMP1, Central nervous system, Diffuse large B-cell lymphoma, Primary central nervous system lymphoma, PRDM1.


Primary central nervous system lymphomas (PCNSLs) are high-grade non-Hodgkin lymphomas of the diffuse large B-cell type (DLBCL) that are confined to the CNS (1). The pathogenesis and the reasons why PCNSLs have a worse prognosis than systemic DLBCLs (2) are largely unknown. Using gene expression profiling, we have recently shown that PCNSLs are more closely related to memory B cells than to germinal center (GC) B cells (3). Thus, late GC exit phenotype B cells are very likely the cells of origin of PCNSLs. Primary central nervous system lymphomas are characterized by active somatic hypermutation (4, 5), absence of class switch recombination (6), and expression of the BCL-6 and IRF-4/MUM-1 proteins. These features suggest that the tumor cells are impaired in their terminal differentiation.

The positive regulatory domain I protein with ZNF domain (PRDM1), the human homolog of murine BLIMP1, is a master regulator of terminal B-cell differentiation (7). In nonneoplastic lymphoid tissue, PRDM1 is expressed in subsets of GC B cells in human tonsils, in proliferating cells bearing GC markers, and in precursors of mature plasma cells that show cytoplasmic immunoglobulin and CD138 expression; it is, however, absent from memory B cells (8). PRDM1 promotes terminal differentiation of B cells along the plasma cell lineage by blocking the expression of genes implicated in B-cell receptor signaling and cell proliferation (9-11).

The PRDM1 gene maps to chromosome region 6q21 and consists of 8 exons (12, 13). There are 2 major splice variants: PRDM1-alpha, which encodes the longer isoform, and PRDM1-beta, which contains a distinct 5'-untranslated region and lacks an in-frame portion of the 5' coding region, resulting in a shorter N-terminus. PRDM1-beta lacks repressive function on multiple targets but maintains normal DNA binding activity and nuclear localization (14).

Systemic DLBCLs have been shown to harbor alterations of the PRDM1 gene (15-17). Thus, in DLBCLs, the PRDM1 gene may function as a tumor suppressor, and it is assumed that PRDM1 contributes to lymphomagenesis by blocking B-cell differentiation into plasma cells. Interestingly, PRDM1 alterations were confined to the activated B-cell (ABC) type of DLBCL and were not detected in GC B- cell-like and non-ABC/non-GC B-cell-type DLBCL (15). Because a subset of PCNSLs also assigns to the ABC type (3), we analyzed a series of 21 PCNSLs for alterations of the PRDM1 gene. Our studies demonstrate mutational inactivation of the PRDM1 gene and abrogation of its protein in 4 of 21 (19%) PCNSLs, similarly suggesting a tumor suppressor function of PRDM1 in PCNSLs.


Patients and Diagnosis

Stereotactic biopsies containing at least 80% of tumor cells were obtained from 21 immunocompetent patients (9 men, 12 women; mean age, 66 years; range, 28-82 years). The histopathologic diagnosis of PCNSL of the DLBCL type was confirmed using routine histology and immunohistochemistry, as previously described (Table 1) (18). All studies were approved by local ethics committees, and informed consent was provided according to the Declaration of Helsinki. Systemic lymphoma manifestation was excluded by extensive staging. For comparison with lymphoma samples, tonsils were obtained with written consent from 2 patients who underwent tonsillectomy for nonneoplastic lesions.


Immunohistochemistry for CD20, BCL-6, and MIB1 was performed according to standard methods (18). To demonstrate BLIMP1 protein, the monoclonal mouse antihuman antibody to PRDM1/Blimp-1 (clone 3H2- E8; Novus Biologicals, Littleton, CO; dilution, 1:50) was used with the DetectionLine System (DCS, Hamburg, Germany). 3,3'- diaminobenzidine (DCS) as chromogen with H^sub 2^O^sub 2^ (Merck, Darmstadt, Germany) as cosubstrate were used on formalin-fixed, paraffin-embedded or frozen sections in 18 of 21 cases. In 3 patients, the tiny size of the biopsy precluded PRDM1 staining. To score for the fraction of PRDM1-expressing tumor cells, a semiquantitative grading system was applied as follows: (-), negative; (+), less than 10% of immunoreactive tumor cells; (++), 10% to 50% of immunoreactive tumor cells; and (+++), greater than 50% of immunoreactive tumor cells. Sections of a case of multiple myeloma were used as a positive control for PRDM1 expression.

TABLE 1. Patients' Data and PCNSL Localization

Isolation of GC B Cells

Human tonsils were minced and mononuclear cells were obtained using a Ficoll density gradient (Amersham Biosciences, Freiburg, Germany). Germinal center B cells were stained with monoclonal mouse anti-human CD77-fluorescein isothiocyanate (Becton-Dickinson, Heidelberg, Germany) and isolated with anti-fluorescein isothiocyanate-coupled magnetic beads (Miltenyi, Bergisch-Gladbach, Germany). B cells were enriched with CD19-coupled magnetic beads (Miltenyi). Thereafter, GC B cells were stained with monoclonal mouse anti-human CD38-PE (Becton-Dickinson) and CD77-fluorescein isothiocyanate (Becton-Dickinson). Centroblasts were sorted as CD38^sup +^CD77^sup +^ cells using a Vantage fluorescence-activated cell sorter (Becton-Dickinson).

Isolation of CD19^sup +^CD27^sup +^ B Cells

Human peripheral blood mononuclear cells were obtained using a Ficoll density gradients (Amersham Biosciences). B cells were enriched with CD19-coupled magnetic beads. Thereafter, CD19^sup +^ B cells were stained with monoclonal CD27-PE (Becton-Dickinson) and sorted as CD27^sup +^ cells by use of a IACS Vantage cell sorter (BectonDickinson).

DNA Extraction

DNA was extracted from frozen tumor biopsies, blood, and 2 x 10^sup 6^ JJN3 cells of the myeloma cell line (Deutsche Sammlung von Mikroorganismen und Zellkulturen, Heidelberg, Germany; cultured under standard conditions) according to the manufacturer's instructions using the Puregene DNA purification kit (Gentra Systems, Minneapolis, MN). DNA concentration was measured fluorometrically (Qbit; Invitrogen, Karlsruhe, Germany).

Polymerase Chain Reaction and DNA Sequencing

All exons of the PRDM1 gene were amplified by polymerase chain reaction (PCR) using hot Start I Taq Polymerase (USB Corp., Cleveland, OH). Primer sequences and cycling conditions are available upon request. Amplicons were separated by agarose gel electrophoresis and subsequently cleaned up using ExoSAP-I (USB Corp.) to remove excess nucleotides and primers. The cleaned PCR products were directly sequenced from both sides on an ABI 3730 capillary sequencer (Applied Biosystems, Weiterstadt, Germany) using the same primers as for PCR.

Sequence and mutation analyses were performed using Mutation Surveyor (SoftGenetics LLC, State College, PA) and Laser Gene software (DNAStar, Madison, WI) with the reference sequences NC_000006 (Region, 106640888_ 106664506; genomic PRDM1 sequence), NM_001198 (PRDM1-alpha transcript), and NM_182907 (PRDM1-beta transcript).

Cloning and DNA Sequencing

To analyze the splice site mutations in Cases 2 and 3, exon- spanning PCR was performed (primers available on request). In Case 3, PCR products were separated via gel electrophoresis, isolated by gel excision, and cloned into the pCR4-TOPO vector (Invitrogen) using the Topo TA cloning Kit (Invitrogen). The resulting plasmid vectors were used to transform competent Escherichia coli cells (Invitrogen). After bacterial amplification, plasmid vectors were harvested by performing Miniprep using the Qiaprep Miniprep Kit (Qiagen) and linearized by PstI digest (Fermentas, St. Leon-Rot, Germany). The linearized plasmid vectors were then purified by phenol/chloroform extraction and subsequently sequenced bidirectionally. In Case 2, PCR products were cleaned up using ExoSAP-I and directly sequenced.

RNA Extraction and Reverse Transcription

RNA was extracted from frozen tumor biopsies, 5 x 10^sup 6^ JJN3 myeloma cells, and 1.8 x 10^sup 7^ GC B cells pooled from the 2 tonsils using the TRIzol/chloroform method (Invitrogen) according to the manufacturer's instructions. RNA concentration was measured fluorometrically. First-strand cDNA was synthesized using the high- capacity cDNA reverse transcription kit (Applied Biosystems).Quantitative Reverse-Transcriptase-PCR Differential expression of the 2 PRDM1 splice variants (i.e. PRDM1-alpha and PRDM1-beta) in 21 PCNSLs was analyzed by quantitative PCR using the cDNA-specific hydrolysis probe-based gene expression assays Hs00153357_ml (detecting PRDM1-alpha and PRDM1-beta) and Hs01068508_ml (detecting PRDM1-alpha) on an ABI 7300 sequence detection system (Applied Biosystems). Water substituted for cDNA was used as a negative control. JJN3 cDNA served as positive control. All samples were studied in triplicate. For normalization, TATA-box-binding protein (Hs00427620_ml) was selected as reference gene; it had been determined to be suitable for this in previous experiments (data not shown). The normalized data were calibrated by comparison to centroblast expression values using the DeltaDeltaCt- Method (19).

FIGURE 1. Schematic representation of the positive regulatory domain I protein with ZNF domain (PRDM1) genetic locus, mRNA, and protein with mutations detected in this series of 21 primary central nervous system lymphomas. Exons are indicated by open boxes. The filled box indicates the site of alternative splicing yielding the PRDM1-beta isoform. Ac, acidic domain; PR, positive regulatory domain; Pro, proline-rich domain; ZF, domain containing 5 zinc fingers; filled triangle, frameshift mutation; open triangle, splice site mutation.

Denaturing High-Performance Liquid Chromatography

Case 7 with the observed heterozygous G-to-A exchange was screened against 43 controls to identify whether this base pair change was polymorphic or disease-related. Primers flanking the region of interest were 5'-CATTCCATCCTCCACCACTC-3' (sense) and 5'- GGGTAGGAGCCCAAACCTT-3' (anti-sense), leading to a PCR product of 186 bp. Amplified DNA was preheated to 95[degrees]C and subsequently cooled slowly to room temperature to force heteroduplex formation. Five microliters of each probe was injected into a WAVE 4500 DNA fragment analysis system (Transgenomic, San Jose, CA) and analyzed at a temperature of 64.1[degrees]C using a linear gradient of buffer A (0.1 mol/L of triethylammonium acetate) and buffer B (0.1 mol/L of triethylammonium acetate, 25% acetonitrile) at a flow rate of 0.9 ml/ minute. Amplicons displaying aberrant chromatograms were sequenced on an ABI 310 genetic analyzer.


Alterations of the PRDM1 Gene

All exons, including exon-intron boundaries of the PRDM1 gene, were amplified and sequenced in a series of 21 cases of PCNSL. Functionally relevant mutations of the PRDM1 gene were detected in 4 of these PCNSLs (PCNSLs 1-4; Fig. 1, Table 2A). Exon 1 (PCNSL 1) harbored a G[arrow right]A transition at position 299 in the 5'- untranslated region. The functional consequences of this point mutation are unknown.

Splice site mutations were detected in PCNSLs 2 and 3 (Figs. 2B, C). In PCNSL 2, a G[arrow right]A transition at the last nucleotide of exon 2 abrogated normal splicing at the mRNA level. This led to a 100-bp insertion through activation of an alternative splice site located downstream, thereby causing a frameshift (Figs. 2B, 3A).

In PCNSL 3, the mutation consisted of a bp exchange G[arrow right]C at the exon 4-intron 4 border. This compromised splicing between exons 4 and 5, resulting in a premature stop codon. Consequently, at the mRNA level, a 68-bp deletion directly preceded the mutated splice site and caused a frameshift (Figs. 2C, 3B).

TABLE 2. Genetic Alterations of the PRDM1 Gene, Predicted Consequences, mRNA Transcription, and Protein Expression in PCNSL

FIGURE 2. Pathogenic mutations of the positive regulatory domain I protein with ZNF domain (PRDM1) gene in primary central nervous system lymphoma (PCNSL). (A) In PCNSL 1, the arrow points to a C[arrow right]G exchange in exon 2 (452), which is followed by a G duplication. (B) In PCNSL 2, a G[arrow right]A nucleotide exchange is indicated by the arrow. This exchange is located at the last nucleotide of exon 2 (bp 525), thereby leading to a splice site mutation with subsequent frameshift. The dotted line represents the exon-intron boundary. (C) In PCNSL 3, a G[arrow right]C transition is shown (arrow). This nucleotide exchange affected the splice site (bp 898) of exon 4 and resulted in a frameshift. The dotted line represents the exon-intron boundary. (D) In PCNSL 4, a C has been inserted at bp 1164 of exon 5 (arrow), causing a frameshift. (A-D) In all tumors, the lack of double peaks indicates the absence of the wild-type allele. NM_001198 served as reference sequence.

In PCNSLs 1 and 4, a nucleotide duplication was detected (Figs. 2A, D). In PCNSL 1, exon 2 was affected at position 452, where a duplication of a G preceding a C[arrow right]G transversion led to a frameshift and produced a premature stop codon (Fig. 2A). In PCNSL 4, exon 5 was affected at position 1164, where a duplication of a C led to a frameshift introducing a premature stop codon (Fig. 2D). At the mRNA level, all mutations were homozygous, as indicated by the absence of double peaks in the sequences (Figs. 2A-D).

FIGURE 3. Insertion and deletion of the positive regulatory domain I protein with ZNF domain (PRDM1) locus in 2 cases of primary central nervous system lymphoma (PCNSL). (A) The exon 2-intron 2 splice site was amplified by reverse-transcriptase-polymerase chain reaction (RT-PCR) using cDNA of PCNSL 2. Although RT-PCR using cDNA of the JJN3 myeloma cell line, which served as control, yielded the expected PCR product of 263-bp length, 100 bp have been inserted in the PCNSL. (B) The exon 4-intron 4 splice site was amplified by RT- PCR using cDNA of PCNSL 3. Although RT-PCR using cDNA of the JJN3 myeloma cell line, which was used as control, resulted in the expected 395-bp product, 68 bp have been deleted in the PCNSL. As further control, genomic DNA of the JJN3 myeloma cell line was amplified, yielding the normal band of 5.6 kb size. (A, B) Agarose gel. M, molecular marker.

In addition to pathogenic mutations, a series of polymorphisms were detected; these are described in Table 2B.

PRDM1 mRNA Transcription and Protein Expression in PCNSL

To analyze whether mutations of the PRDM1 gene had an impact on the level of mRNA transcription and protein expression, quantitative reverse-transcriptase-PCR and immunohistochemical analyses were performed. Positive regulatory domain I protein with ZNF domain mRNA transcripts were detected in all 21 PCNSLs. In all but 1 PCNSL (20/ 21; 95%), mRNA levels exceeded levels of tonsillar GC centroblasts (Table 2). Positive regulatory domain I protein with ZNF domain mRNA levels did not correlate with the presence or absence of mutations of the PRDM1 gene.

Immunohistochemistry for PRDM1 expression was carried out in 18 cases for which sufficient material was available. Positive regulatory domain I protein with ZNF domain protein was expressed by the tumor cells in 13 of 18 PCNSLs (72%; Table 2). In all PCNSLs with pathogenic mutations (PCNSLs 1-4), the tumor cells did not express PRDM1 protein (Fig. 4B). In PCNSLs 17 and 19, which did not harbor a PRDM1 mutation and only a low transcription of PRDM1-alpha mRNA, the tumor cells also did not express PRDM1 protein (Table 2). Moreover, in PCNSLs with PRDM1 expression, the protein was confined to the nucleus of the tumor cells. In general, PRDM1 protein expression was rather low in most positive cases. Indeed, in 4 and 6 cases, less than 10% and 10% to 50% of the tumor cells expressed PRDM1 protein, respectively (Fig. 4A, Table 2).


This study reports recurrent inactivation of the PRDM1 gene in a fraction of PCNSLs due to clonal deleterious mutations. The mutations consisted of single-nucleotide exchanges and insertions that targeted splice donor sites. They resulted in a frameshift and, ultimately, in abrogation of PRDM1 protein expression by the tumor cells. Thus, structural alterations of the PRDM1 gene are associated with a severe functional impairment of this tumor suppressor gene and may contribute to the pathogenesis of PCNSL.

The observed frequency of 4 of 21 (19%) PCNSLs carrying PRDM1 mutations appears to be increased as compared to systemic DLBCLs, which have been reported to have a mutated PRDM1 gene in 24% of DLBCL of the ABC type only (15). The tiny sizes of the biopsies in the present study precluded gene expression profiling to assign them to 1 of the subtypes defined for systemic DLBCL. Based on our previous data, which showed that PCNSL display a gene expression profile of the ABC type in 24% (3), the overall PRDM1 mutation frequency is calculated to exceed that of systemic DLBCL. The latter can be calculated as 9%, that is, 8 cases of the 92 DLBCL cases that have been subtyped by gene expression profiling (15). This estimate is further corroborated by Tate et al (17), who observed PRDM1 gene mutations in 1 of 13 (8%) systemic DLBCLs. Nevertheless, the limited number of PCNSLs investigated to date has to be taken into account.

Functionally important, in all 4 PCNSLs with deleterious mutations, the changes lead to frameshifts. These frameshift mutations targeted exons 2, 4, and 5 of the PRDM1 gene. A nucleotide duplication was present in 2 PCNSLs; none of these frameshift mutations have been reported in systemic DLBCLs.

FIGURE 4. Positive regulatory domain I protein with ZNF domain (PRDM1) protein expression in primary central nervous system lymphoma (PCNSL). (A, B) Whereas approximately 10% to 50% of the tumor cells of a PCNSL (14) without a mutation of the PRDM1 gene express PRDM1 protein in their nucleus (A), tumor cells of a PCNSL (1) with a frameshift mutation are PRDM1^sup -^ (B). (C) As positive control, a multiple myeloma is shown the tumor cells of which express PRDM1 in their nucleus. Anti-PRDM1 immunohistochemistry, slight counterstaining with hem alum, original magnification: 400 x .

Interestingly, PCNSL 2 harbored a nucleotide exchange at the splice site that affects the last nucleotide of exon 2 (bp 525), leading to a 100-bp insertion at the mRNA level, and the same position of the PRDM1 gene has been identified as a mutational hot spot in systemic DLBCLs (15, 16). Theoretically, all mutations reported at this hot spot in systemic DLBCLs and those identified here in PCNSLs are predicted to lead to a truncated PRDM1 protein with a corrupt positive regulatory domain, which is indispensable for normal PRDM1 function (14, 20). The lack of PRDM1 protein expression, as revealed by immunohistochemistry, however, suggests that nonsense-mediated decay might account for a complete inactivation of the protein through this mutation. Mutations of the PRDM1 gene did not correlate with PRDM1 mRNA transcription in PCNSL; this is in accordance with observations in systemic DLBCL (15). In this regard, the presence of nonmalignant T lymphocytes, which are part of the reactive inflammatory infiltrate in PCNSL (1) and may also express PRDM1 (21), should be taken into account. More importantly, immunohistochemistry revealed the absence of PRDM1 protein in the tumor cells of all PCNSLs with frameshift mutations. In contrast, 78% (11/14) of PCNSLs without inactivating PRDM1 mutations expressed the PRDM1 protein. Thus, PCNSLs seem to differ from systemic DLBCLs in this regard. For systemic DLBCL, wide variations in PRDM1 expression have been reported with expression rates of 43% (101/134) and 5% (4/79) (21, 22). Functionally relevant, systemic DLBCL with inactivating PRDM1 mutations also lacked PRDM1 protein (15). Interestingly, whereas the overall survival and disease-free survival rates of patients with PRDM1- positive systemic DLBCLs were not significantly different from those of PRDM1-negative cases, PRDM1-positive patients had a significantly shorter failure-free survival rate (22). Whether PRDM1 may serve as a prognostic biomarker in PCNSLs requires further investigation.

In systemic DLBCLs, PRDM1-inactivating mutations were restricted to BCL6^sup +^CD10^sup -^MUM1^sup +^ DLBCL. All of the PCNSLs in the present series were BCL6^sup +^CD10^sup -^MUM1^sup +^ and are therefore immunophenotypically comparable to systemic DLBCLs harboring inactivating mutations of the PRDM1 gene.

In PCNSLs harboring inactivating mutations, sequence analysis clearly demonstrated the mutations to be homozygous. Together with the lack of PRDM1 protein expression, this observation strongly arguments for lack of the normal PRDM1 allele in these tumors and provides evidence for biallelic inactivation of the gene. This may be due either to a deletion of the second allele or to uniparental disomy.

Functionally important, PRDM1 may also act as tumor suppressor gene in PCNSLs, as has been suggested for systemic DLBCLs. Because PRDM1 is critical for the induction of terminal B-cell differentiation into plasma cells (7, 9), PRDM1 mutations might at least in part explain the impaired differentiation of the tumor cells of PCNSL.

In normal GC B cells, PRDM1 and BCL6 expression is mutually exclusive because BCL6 is a powerful inhibitor of PRDM1. In PCNSLs, which harbor BCL6 translocations in a significant fraction (23, 24), however, PRDM1 and BCL6 proteins are expressed by malignant cells within the same tumor, possibly concomitantly, and may even cooperate in the impairment of terminal differentiation of the tumor cells and thereby contribute to lymphomagenesis.


The authors thank Elena Fischer and Irmgard Henke for excellent technical assistance.


1. Deckert M, Paulus W, Louis DN, Ohgaki H, Wiestier OD, Cavenee WK. WHO Classification of Tumors Pathology and Genetics of Tumours of the Nervous System. Malignant Lymphomas. Lyon, France: IRAC, 2007: 188-92

2. Schlegel U, Schmidt-Wolf IG, Decken M. Primary CNS lymphoma: Clinical presentation, pathological classification, molecular pathogenesis and treatment. J Neurol Sci 2000;181:1-12

3. Montesinos-Rongen M, Brunn A, Bentink S, et al. Gene expression profiling suggests primary central nervous system lymphomas to be derived from a late germinal center B cell. Leukemia 2008;22: 400-5

4. Montesinos-Rongen M, Kuppers R, Schluter D, et al. Primary central nervous system lymphomas are derived from germinal-center B cells and show a preferential usage of the V4-34 gene segment. Am J Pathol 1999;155:2077-86

5. Montesinos-Rongen M, Van Roost D, Schaller C, Wiestler OD, Deckert M. Primary diffuse large B-cell lymphomas of the central nervous system are targeted by aberrant somatic hypermutation. Blood 2004;103: 1869-75

6. Montesinos-Rongen M, Schmitz R, Courts C, et al. Absence of immunoglobulin class switch in primary lymphomas of the central nervous system. Am J Pathol 2005;166:1773-79

7. Kallies A, Nutt SL. Terminal differentiation of lymphocytes depends on Blimp-1. Curr Opin Immunol 2007;19:156-62

8. Angelin-Duclos C, Cattoretti G, Lin KI, Calame K. Commitment of B lymphocytes to a plasma cell fate is associated with Blimp-1 expression in vivo. J Immunol 2000;165:5462-71

9. Shaffer AL, Lin KI, Koul TC, et al. Blimp-1 orchestrates plasma cell differentiation by extinguishing the mature B cell gene expression program. Immunity 2002e;17:51-62

10. Shapiro-Shelf M, Lin KI, Sisisky D, Liao J, Calame K. Blimp- 1 is required for maintenance of long-lived plasma cells in the bone marrow. J Exp Med 2005;202:1471-76

11. Turner CA, Jr, Mack DH, Davis MM. Blimp-1, a novel zinc finger-containing protein that can drive the maturation of B lymphocytes into immunoglobulin-secreting cells. Cell 1994;77:297- 306

12. Keller AD, Maniatis T. Identification and characterization of a novel represser of beta-interferon gene expression. Genes Dev 1991;5: 868-79

13. Mock BA, Liu L, LePaslier D, Huang S. The B-lymphocyte maturation promoting transcription factor BLIMP1/PRDI-BF1 maps to D6S447 on human chromosome 6q21-q22.1 and the syntenic region of mouse chromosome 10. Genomics 1996;37:24-28

14. Gyory I, Fejer G, Ghosh N, Seto E, Wright KL. Identification of a functionally impaired positive regulatory domain I binding factor 1 transcription represser in myeloma cell lines. J Immunol 2003;170: 3125-33

15. Pasqualucci L, Compagno M, Houldsworth J, et al. Inactivation of the PRDM1/BLIMP1 gene in diffuse large B cell lymphoma. J Exp Med 2006;203:311-17

16. Tam W, Gomez M, Chadburn A, Lee JW, Chan WC, Knowles DM. Mutational analysis of PRDM1 indicates a tumor-suppressor role in diffuse large B-cell lymphomas. Blood 2006;107:4090-100

17. Tate G, Hirayama-Ohashi Y, Kishimoto K, Mitsuya T. Novel BLIMP1/PRDM1 gene mutations in B-cell lymphoma. Cancer Genet Cytogenet 2007;172:151-53

18. Brunn A, Montesinos-Rongen M, Strack A, et al. Expression pattern and cellular sources of chemokines in primary central nervous system lymphoma. Acta Neuropathol (Berl) 2007;114:271-76

19. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 2001;25:402-8

20. Ghosh N, Gyory I, Wright G, Wood J, Wright KL. Positive regulatory domain I binding factor 1 silences class II transactivator expression in multiple myeloma cells. J Biol Chem 2001;276:15264-68

21. Cattoretti G, Angelin-Duclos C, Shaknovich R, Zhou H, Wang D, Alobeid B. PRDM1/Blimp-1 is expressed in human B-lymphocytes committed to the plasma cell lineage. J Pathol 2005;206: 76-86

22. Garcia JF, Roncador G, Garcia JF, et al. PRDM1/BLIMP-1 expression in multiple B and T-cell lymphoma. Haematologica 2006;91: 467-74

23. Montesinos-Rongen M, Zuhlke-Jenisch R, Gesk S, et al. Interphase cytogenetic analysis of lymphoma-associated chromosomal breakpoints in primary diffuse large B-cell lymphomas of the central nervous system. J Neuropathol Exp Neurol 2002;61:926-33

24. Schmidt H, Alaska T, Zuhlke-Jenisch R, et al. Chromosomal translocations fusing the BCL6 gene to different partner loci are recurrent in primary central nervous system lymphoma and may be associated with aberrant somatic hypermutation or defective class switch recombination. J Neuropathol Exp Neurol 2006;65:776-82

Cornelius Courts, PhD, Manuel Montesinos-Rongen, PhD, Anna Brunn, MD, Stefanie Bug, MD, Dorte Siemer, PhD, Volkmar Hans, MD, Ingmar Blumcke, MD, Wolfram Klapper, MD, Carlo Schaller, MD, Otmar D. Wiestier, MD, Ralf Kuppers, PhD, Reiner Siebert, MD, and Martina Deckert, MD

From the Department of Neuropathology (CC, MM-R, AB, MD), University Hospital of Cologne, Cologne; Institute of Human Genetics (SB, RS), Christian-Albrechts University Kiel, Kiel; Institute for Cell Biology (Tumor Research) (DS, RK), University of Duisburg- Essen, Medical School, Essen; Department of Neuropathology (VH), Evangelisches Krankenhaus, Bielefeld; Department of Neuropathology (IB), Friedrich-Alexander-University, Erlangen; Institute of Pathology (WK), Christian-Albrechts University Kiel, Kiel; Department of Neurosurgery (CS), University Hospital Bonn, Bonn; and German Cancer Research Center (ODW), Heidelberg, Germany.

Send correspondence and reprint requests to: Martina Deckert, MD, Abteilung fur Neuropathologie, Universitatsklinikum Koln, Kerpener Str. 62, D-50937 Koln, Germany; E-mail: [email protected]

Drs. Courts and Montesinos-Rongen contributed equally to the work.

This study was supported by Grant No. 107733 from the Deutsche Krebshilfe/Dr. Mildred Scheel-Stiftung fur Krebsforschung.

Copyright Lippincott Williams & Wilkins Jul 2008

(c) 2008 Journal of Neuropathology and Experimental Neurology. Provided by ProQuest LLC. All rights Reserved.