Loss of PRDM1/BLIMP-1 function contributes to poor prognosis of activated B-cell-like diffuse large B-cell lymphoma.

PRDM1/BLIMP-1, a master regulator of plasma-cell differentiation, is frequently inactivated in activated B-cell-like (ABC) diffuse large B-cell lymphoma (DLBCL) patients. Little is known about its genetic aberrations and relevant clinical implications. A large series of patients with de novo DLBCL was effectively evaluated for PRDM1/BLIMP-1 deletion, mutation, and protein expression. BLIMP-1 expression was frequently associated with the ABC phenotype and plasmablastic morphologic subtype of DLBCL, yet 63% of the ABC-DLBCL patients were negative for BLIMP-1 protein expression. In these patients, loss of BLIMP-1 was associated with Myc overexpression and decreased expression of p53 pathway molecules. In addition, homozygous PRDM1 deletions and PRDM1 mutations within exons 1 and 2, which encode for domains crucial for transcriptional repression, were found to show a poor prognostic impact in patients with ABC-DLBCL but not in those with germinal center B-cell-like DLBCL (GCB-DLBCL). Gene expression profiling revealed that loss of PRDM1/BLIMP-1 expression correlated with a decreased plasma-cell differentiation signature and upregulation of genes involved in B-cell receptor signaling and tumor-cell proliferation. In conclusion, these results provide novel clinical and biological insight into the tumor-suppressive role of PRDM1/BLIMP-1 in ABC-DLBCL patients and suggest that loss of PRDM1/BLIMP-1 function contributes to the overall poor prognosis of ABC-DLBCL patients.

Protein expression analysis of chromosome 12 candidate genes in chronic lymphocytic leukemia (CLL).

The pathogenic role of trisomy 12 in chronic lymphocytic leukemia (CLL) remains unresolved, but recently an upregulated RNA expression level has been observed for chromosome 12 candidate genes. In the current study, the protein expression of chromosome 12 candidate genes was characterized by comparing CLL cases with (n=58) or without (n=16) trisomy 12, CD19+-B-cells and cell lines (JVM-2, EHEB, JURKAT). Immunoblotting was performed to quantify the levels of AID, APAF-1, ARF3, CCND2, CDK2, CKD4, GLI, MDM-2, p27, Smac/DIABLO and STAT6 (signal transducer and activator of transcription 6). The cell lines showed distinct expression patterns for CCND2, MDM-2, p27, Smac/DIABLO and STAT6, and displayed higher levels of CDK2 and CDK4 than the CLL cases. JURKAT and the CLL cases expressed uniformly high levels of p27, but low levels of CCND2. AID expression in the CLL cases was weak with slight variations regardless of the subgroup affiliation. The expression of the investigated proteins was independent of the trisomy 12 status as well as of the VH mutation status. The comparison of CD19+-B-cells with CLL revealed higher protein levels in CLL for CDK4, p27, Smac/DIABLO and STAT6. Further studies including protein expression experiments in genetic high-risk subgroups of CLL have to elucidate whether these proteins qualify as candidates for targeted CLL therapies.

New approaches to lymphoma diagnosis.

Recent years have brought an explosion of new diagnostic tools to the pathology of lymphomas, which have permitted more precise disease definition and recognition of factors that can predict prognosis and response to treatment. These new methods exploit both the biological features of normal lymphocytes as they progress through differentiation pathways and the genetic abnormalities that characterize malignant transformation. These features can be assessed in individual tumors with techniques that detect proteins (immunophenotyping), messenger RNA (in-situ hybridization), or changes in DNA [Southern blot, PCR, fluorescence in-situ hybridization (FISH), and gene sequencing]. Recently, the novel technology of "gene chips" or DNA microarrays has greatly enhanced the efficiency of analyzing expression of many genes simultaneously at the RNA level. Understanding the relationship of lymphoid neoplasms to their normal counterparts and the genetic events that lead to malignant transformation in lymphoid cells are essential for physicians caring for patients with lymphoma, since these are the basis of modern classification, diagnosis, and prognosis prediction. Although microarray technology is not ready for prime time in the daily diagnosis of lymphoma, practitioners should understand its potential and limitations. The vast majority of lymphoid neoplasms worldwide are derived from B lymphocytes at various stages of differentiation. The review by Harald Stein and colleagues present the events of normal B-cell differentiation that are relevant to understanding the biology of B-cell neoplasia. These include antigen receptor [immunoglobulin (Ig)] gene rearrangement, somatic mutations of the Ig variable region genes, receptor editing, Ig heavy chain class switch, and differential expression of a variety of adhesion molecules and receptor proteins as the cell progresses from a precursor B cell to a mature plasma cell. Most lymphoid neoplasms have genetic abnormalities, many of which appear to occur during the gene rearrangements and mutations that characterize normal B-cell differentiation. Dr. Riccardo Dalla Favera reviews the mechanisms of these translocations and other abnormalities, and their consequences for lymphocyte biology. The association of specific abnormalities with individual lymphomas is reviewed. Dr. Wing C. Chan reviews the technology and applications of DNA microarray analysis, its promises and pitfalls, and what it has already told us about the biology of lymphomas. Finally, what does this all mean? The applications, both current and future, of these discoveries to the diagnosis and treatment of patients with lymphoma are discussed by Dr. Nancy Lee Harris.

Hypermutation of multiple proto-oncogenes in B-cell diffuse large-cell lymphomas.

Genomic instability promotes tumorigenesis and can occur through various mechanisms, including defective segregation of chromosomes or inactivation of DNA mismatch repair. Although B-cell lymphomas are associated with chromosomal translocations that deregulate oncogene expression, a mechanism for genome-wide instability during lymphomagenesis has not been described. During B-cell development, the immunoglobulin variable (V) region genes are subject to somatic hypermutation in germinal-centre B cells. Here we report that an aberrant hypermutation activity targets multiple loci, including the proto-oncogenes PIM1, MYC, RhoH/TTF (ARHH) and PAX5, in more than 50% of diffuse large-cell lymphomas (DLCLs), which are tumours derived from germinal centres. Mutations are distributed in the 5' untranslated or coding sequences, are independent of chromosomal translocations, and share features typical of V-region-associated somatic hypermutation. In contrast to mutations in V regions, however, these mutations are not detectable in normal germinal-centre B cells or in other germinal-centre-derived lymphomas, suggesting a DLCL-associated malfunction of somatic hypermutation. Intriguingly, the four hypermutable genes are susceptible to chromosomal translocations in the same region, consistent with a role for hypermutation in generating translocations by DNA double-strand breaks. By mutating multiple genes, and possibly by favouring chromosomal translocations, aberrant hypermutation may represent the major contributor to lymphomagenesis.

BCL-6 mutations are associated with immunoglobulin variable heavy chain mutations in B-cell chronic lymphocytic leukemia.

The cell of origin of B-cell chronic lymphocytic leukemia (B-CLL) is still uncertain. Recent studies have indicated that a fraction of B-CLL displays somatically mutated immunoglobulin variable heavy chain (IgV(H)) genes, which suggests an origin from a post-germinal center (GC) B cell. It has been shown that the 5' noncoding region of the BCL-6 proto-oncogene is affected by mutations in normal GC B-lymphocytes and in lymphoid malignancies displaying GC/post-GC phenotype. To further explore the cellular origin of B-CLL, we have analyzed 34 cases for mutations in the BCL-6 5' noncoding region and in the IgV(H) genes. We found somatically mutated IgV(H) genes in 24 (73%) of 33 samples (average frequency, 6.5 x 10(-2)/bp) and BCL-6 mutations in 8 (24%) of 34 cases (average frequency, 0.14 x 10(-2)/bp in the mutated cases). The occurrence of BCL-6 mutations was restricted to those cases displaying IgV(H) mutations. Analysis of BCL-6 protein expression as a marker of GC phenotype showed that, regardless of the presence of IgV(H) or BCL-6 mutations, B-CLLs express BCL-6 at levels clearly below those found in normal or transformed GC B cells. These results indicate that a subset of B-CLL derives from a cell that has been exposed to the somatic hypermutation mechanism and support the hypothesis that BCL-6 mutations result from the same process that targets immunoglobulin genes.

Distribution and pattern of BCL-6 mutations throughout the spectrum of B-cell neoplasia.

BCL-6 mutations are accumulated during B-cell transit through the germinal center (GC) and provide a histogenetic marker for B-cell tumors. On the basis of a comprehensive analysis of 308 B-cell neoplasms, we (1) expand the spectrum of tumors associated with BCL-6 mutations; (2) corroborate the notion that mutations cluster with GC and post-GC B-cell neoplasms; and (3) identify heterogeneous mutation frequency among B-lineage diffuse large cell lymphoma (B-DLCL) subsets. Mutations are virtually absent in acute lymphoblastic leukemia (P <.001) and mantle cell lymphoma (P <.05), whereas they occur frequently in GC or post-GC neoplasms, including lymphoplasmacytoid lymphoma, follicular lymphoma, MALT lymphomas, B-DLCL and Burkitt lymphoma. Among B-DLCL, mutations occur frequently in systemic nodal B-DLCL, primary extranodal B-DLCL, CD5(+) B-DLCL, CD30(+) B-DLCL, and primary splenic B-DLCL, suggesting a similar histogenesis of these B-DLCL subsets. Conversely, mutations are rare in primary mediastinal B-DLCL with sclerosis (10.0%; P <.01), supporting a distinct histogenesis for this lymphoma. Longitudinal follow-up of B-DLCL transformed from follicular lymphoma shows that they BCL-6 mutations may accumulate during histologic progression. Mutations also occur in some B-cell chronic lymphocytic leukemias, small lymphocytic lymphomas, and hairy cell leukemias, consistent with the hypothesis that a fraction of these lymphoproliferations are related to GC-like cells. Finally, the molecular pattern of 193 mutational events reinforces the hypothesis that mutations of BCL-6 and immunoglobulin genes are caused by similar mechanisms. (Blood. 2000;95:651-659)

BCL-6 mutations in normal germinal center B cells: evidence of somatic hypermutation acting outside Ig loci.

The molecular mechanism involved in the process of antigen-driven somatic hypermutation of Ig genes is unknown, but it is commonly believed that this mechanism is restricted to the Ig loci. B cell lymphomas commonly display multiple somatic mutations clustering in the 5'-regulatory region of BCL-6, a proto-oncogene encoding for a POZ/Zinc finger transcriptional repressor expressed in germinal center (GC) B cells and required for GC formation. To determine whether BCL-6 mutations represent a tumor-associated phenomenon or reflect a physiologic mechanism, we screened single human tonsillar GC B cells for mutations occurring in the BCL-6 5'-noncoding region and in the Ig variable heavy chain sequences. Thirty percent of GC B cells, but not naive B cells, displayed mutations in the 742 bp region analyzed within the first intron of BCL-6 (overall frequency: 5 x 10(-4)/bp). Accordingly, an expanded survey in lymphoid malignancies showed that BCL-6 mutations are restricted to B cell tumors displaying GC or post-GC phenotype and carrying mutated Ig variable heavy chain sequences. These results indicate that the somatic hypermutation mechanism active in GC B cells physiologically targets non-Ig sequences.

BCL-6 in aids-related lymphomas: pathogenetic and histogenetic implications.

AIDS-related non-Hodgkin's lymphomas (AIDS-NHL) are classified into Burkitt's lymphoma, diffuse large cell lymphoma (DLCL), and body cavity based lymphoma. The molecular pathogenesis of AIDS-NHL is complex and involves the genetic alteration of several cancer related genes, including the BCL-6 proto-oncogene. BCL-6 encodes a zinc finger transcription factor which is selectively expressed by germinal center (GC) B-cells, but not by pre-GC or post-GC B-cells. Genetic alterations of BCL-6 occur frequently among B-cell NHL and comprise gross rearrangements as well as small mutations of the 5' noncoding region of the gene. Gross rearrangements of BCL-6 among AIDS-NHL cluster with 20% AIDS-DLCL. Conversely, mutations of the 5' noncoding region of BCL-6 occur at sustained frequency throughout the clinico-pathologic spectrum of AIDS-NHL and represent the most common genetic alteration presently detectable in these lymphomas. The frequency of BCL-6 mutations, as well as their location in the proximity of the BCL-6 regulatory regions, suggest that they may play a pathogenetic role in AIDS-related lymphomagenesis. Beside their pathogenetic implications, the occurrence of BCL-6 mutations among AIDS-NHL bears histogenetic relevance because BCL-6 mutations are regarded as a marker of B-cell transition through the GC. Thus, it is conceivable that a large fraction of AIDS-NHL is histogenetically related to GC or post-GC B-cells. This notion is further confirmed by the observation that AIDS-NHL frequently express the BCL-6 protein, which stains selectively GC B-cells throughout B-cell differentiation.

Evaluation of immunotoxins containing single-chain ribosome-inactivating proteins and an anti-CD22 monoclonal antibody (OM124): in vitro and in vivo studies.

Immunotoxins were prepared with three ribosome-inactivating proteins (RIP), momordin, pokeweed antiviral protein from seeds (PAP-S) and saporin-S6, linked to the anti-CD22 monoclonal antibody OM124. These immunotoxins inhibited protein synthesis by CD22-expressing cell lines Daudi, EHM, BJAB, Raji and BM21 with IC50 (concentration causing 50% inhibition) ranging from < 5 x 10(-15) to 7.6 x 10(-11) M as RIP, and IC90 (concentration causing 90% inhibition) ranging from 5 x 10(-14) to 5 x 10(-8)M, with no effect on a CD22-negative HL60 cell line at the highest concentration tested (5 x 10[-8] M). Apoptosis was induced in sensitive cells. The formation of bone marrow colonies was inhibited by no more than 40% by the immunotoxins at concentrations up to 10(-9) M. Treatment with the immunotoxins, alone or in combination, significantly extended the survival time of mice bearing transplanted Daudi cells. A treatment with cyclophosphamide and OM124/saporin immunotoxin was particularly effective in SCID mice transplanted with a low number of cells (3 x 10[-6]), when 60% of the animals remained tumour-free.

Analysis of a 69-kb contiguous genomic sequence at a putative tumor suppressor gene locus on human chromosome 6q27.

Multiple neoplasias including B-cell non-Hodgkin's lymphoma, breast carcinoma, and ovarian carcinoma, have been associated with frequent deletions of the distal region on the long arm of human chromosome 6, suggesting the presence of one or more tumor suppressor gene(s) at this locus. Loss of heterozygosity analysis of breast and ovarian tumors has further restricted the minimal region of loss within 6q27. To further characterize this genomic region for gene content including putative tumor suppressor genes as well as other elements that may contribute to tumorigenesis, a 68940-bp contiguous sequence, encompassing markers D6S193 and D6S297, was generated by random shotgun sequencing of a cosmid, P1, and PAC contig. In addition, exon trapping was performed utilizing a subset of these clones. Sixteen trapped exons, ranging in size from 44 to 399 bp, span this approximately 69-kb region. Many other putative exons have been identified computationally. Further analysis has identified 13 potential promoters and 13 putative polyadenylation sites in the region. Northern analysis identified a transcript mapping within this interval that is expressed in ovarian, breast, and lymphoid-derived tumor cell lines. Consideration of these data, together with the demonstration of several regions of high CpG content, suggests the possibility of several genes at this locus.

Immunocytochemical diagnosis of acute promyelocytic leukemia (M3) with the monoclonal antibody PG-M3 (anti-PML).

Acute promyelocytic leukemia (APL) is characterized by a reciprocal 15; 17 chromosomal translocation, which fuses the promyelocytic leukemia (PML) and retinoic acid receptor alpha (RARalpha) genes, leading to the expression of the PML/RARalpha fusion oncoprotein. Immunocytochemical labeling of the wild-type PML protein with the PG-M3 monoclonal antibody (MoAb) directed against the amino terminal portion of the human PML gene product, produces a characteristic nuclear speckled pattern that is due to localization of the protein into discrete dots (5 to 20 per nucleus), named PML nuclear bodies. The architecture of PML nuclear bodies appears to be disrupted in APL cells that bear the t(15; 17), thus resulting in a change of the nuclear staining pattern from speckled (wild-type PML protein) to microgranular (PML-RARalpha fusion protein). To assess whether the PG-M3 MoAb could assist in the diagnosis of APL (M3), bone marrow and/or peripheral blood samples from 100 cases of acute nonlymphoid leukemias of different subtypes were blindly immunostained with the PG-M3 MoAb, using the immunoalkaline phosphatase (APAAP) or immunofluorescence technique as detection system. Notably, the abnormal (micropunctate) pattern of the PML/RARalpha fusion protein (usually >/=50 small granules/per nucleus) was observed in APL (M3) samples, but not in other types of acute nonlymphoid leukemias. Immunocytochemical labeling with PG-M3 was particularly useful in the diagnosis of microgranular variant of APL (M3V) (three cases misdiagnosed as M4 and M5), and also to exclude a morphologic misdiagnosis of APL (six of 78 cases). In all cases investigated, immunocytochemical results were in agreement with those of reverse transcription-polymerase chain reaction (RT-PCR) for PML/RARalpha. Because the epitope identified by PG-M3 is located in the aminoterminal portion of PML (AA 37 to 51), the antibody was suitable for recognizing APL cases characterized by breakpoint occurring at different sites of PML (bcr 1, bcr 2 and bcr 3). In conclusion, immunocytochemical labeling with PG-M3 represents a rapid, sensitive, and highly-specific test for the diagnosis of APL that bears the t(15; 17). This should allow an easy and correct diagnosis of this subtype of acute leukemia to any laboratory provided with a minimal equipment for immunocytochemistry work.

Bcl-6 protein expression in normal and neoplastic lymphoid tissues.

The human bcl-6 gene, which is rearranged in about 30% of diffuse large B-cell lymphomas (DLCL-B), encodes for a Kruppel-type zinc finger protein of 706 amino acids. In order to investigate the expression of the bcl-6 gene at the protein level, two monoclonal antibodies (PG-B6a and PG-B6p) directed against the human bcl-6 protein were generated by immunizing Balb/c mice with a recombinant protein corresponding to the amino-terminal region (amino acids 3-484) of bcl-6. PG-B6a (a = avian) recognized the most conserved bcl-6 epitope (expressed in many animal species, including avian). PG-B6p (p = paraffin) reacted with an epitope of bcl-6 partially resistant to fixatives and detectable on microwave-heated paraffin sections. At immunocytochemistry, bcl-6 localized in the nucleus with a microgranular or diffuse pattern. Strong nuclear expression of bcl-6 was mainly detected in normal germinal-center B-cells, whereas mantle- and marginal-zone B cells, as well as plasma cells and marrow B-cell precursors, did not express bcl-6. These immunohistological findings strongly suggest that bcl-6 may play a role as a regulator of germinal-center related functions. All MoAbs stained neoplastic cells of follicular lymphomas, DLCL-B, and Burkitt's lymphomas. In DLCL-B, bcl-6 expression was independent of bcl-6 gene rearrangements and did not correlate with expression of other markers or the proliferation index. Among low-grade B-cell lymphomas, immunostaining for bcl-6 proved useful for differentiating proliferation centers in B-CLL (bcl-2+/bcl-6-) from trapped germinal centers in mantle-cell lymphomas (bcl-2-/bcl-6+). Strong nuclear positivity for bcl-6 was consistently detected in tumor (L&H) cells of nodular, lymphocyte-predominant Hodgkin's disease (NLPHD). These results further support the concept that NLPHD is a histogenetically distinct (germinal-center derived) subtype of HD. Notably, the nuclei of reactive CD3+/ CD4+ T cells near to and rosetting around L&H cells in NLPHD were also strongly bcl-6+, but lacked CD40 ligand (CD40L) expression. This staining pattern clearly differed from that of classic HD, whose cellular background was made up of CD3+/CD4+ T cells showing the bcl-6-/CD40L+ phenotype. The above immunohistological findings suggest that (a) bcl-6 may play a role in regulating B-cell differentiation step(s) occurring within germinal centers; (b) deregulated bcl-6 expression caused by rearrangements may contribute to B-lymphomagenesis; (c) bcl-6 is possibly involved in the pathogenesis of NLPHD.

Targeted antibodies in the treatment of lymphomas.

Monoclonal antibodies coupled to drugs and toxic agents (immunotoxins) or radionuclides (radioimmunoconjugates) represent new tools for immunotherapy of haematological malignancies. Immunotoxins constructed with toxins of either plant or bacterial origin have shown a powerful antitumor activity both in vitro and in mice with severe combined immunodeficiency bearing various kinds of leukaemias and lymphomas. Preliminary clinical trials have shown an activity of these compounds at least in a proportion of patients. However, tumour responses have generally been partial and transient. The main problems with immunotoxin therapy remain the inability of immunotoxins to target tumour cells in the presence of a high burden of disease, the host immune response against both the antibody and the toxin moieties, which precludes repeated administration of immunotoxins, and the vascular leak syndrome. Targeting of tumour cells with specific antibodies armed with radionuclides (usually iodine-131 or yttrium-90) appears to be an even more attractive approach. Preliminary clinical studies have clearly demonstrated the ability of radioimmunoconjugates, especially when administered at high dose followed by bone marrow rescue, to induce durable complete remission in patients with non-Hodgkin's lymphomas refractory to conventional therapies. Radioimmunotherapy also overcomes the antigenic heterogeneity of the tumour cell population, since antigen negative tumour cells will be irradiated by the nearby targeted antigen-positive cells. Efforts should now be focused on defining more precisely the optimal clinical setting for administration of immunotoxin and radioimmunoconjugates (e.g. minimal residual disease), to reduce the immunogenicity of these compounds and solve the problem of vascular leak syndrome.

Heterogeneous nuclear expression of the promyelocytic leukemia (PML) protein in normal and neoplastic human tissues.

The RING-finger promyelocytic leukemia (PML) protein is the product of the PML gene that fuses with the retinoic acid receptor-alpha gene in the t(15; 17) translocation of acute promyelocytic leukemia. Wild-type PML localizes in the nucleus with a typical speckled pattern that is a consequence of the concentration of the protein within discrete subnuclear domains known as nuclear bodies. Delocalization of PML from nuclear bodies has been documented in acute promyelocytic leukemia cells and suggested to contribute to leukemogenesis. In an attempt to get new insights into the function of the wild-type PML protein and to investigate whether it displays an altered expression pattern in neoplasms other than acute promyelocytic leukemia, we stained a large number of normal and neoplastic human tissues with a new murine monoclonal antibody (PG-M3) directed against the amino-terminal region of PML. As the PG-M3 epitope is partially resistant to fixatives, only cells that overexpress PML are detected by the antibody in microwave-heated paraffin sections. Among normal tissues, PML was characteristically up-regulated in activated epithelioid histiocytes and fibroblasts in a variety of pathological conditions, columnar epithelium in small active thyroid follicles, well differentiated foamy cells in the center of sebaceous glands, and hypersecretory endometria (Arias-Stella). Interferons, the PML of which is a primary target gene, and estrogens are likely to represent some of the cytokines and/or hormones that may be involved in the up-regulation of PML under these circumstances. In keeping with this concept, we found that PML is frequently overexpressed in Hodgkin and Reed-Sternberg cells of Hodgkin's disease, a tumor of cytokine-producing cells. Among solid tumors, overexpression of PML was frequently found in carcinomas of larynx and thyroid (papillary), epithelial thymomas, and Kaposi's sarcoma, whereas carcinomas of the lung, thyroid (follicular), breast, and colon were frequently negative or weakly PML+. We did not observe any changes in the levels of PML expression as the lesion progressed from benign dysplasia to carcinoma. Our immunohistological data are consistent with the hypothesized growth suppressor function of PML and strongly suggest that PML expression levels are likely to be modulated by a variety of stimuli, including cytokines and hormones.