Search

To evaluate the utility of transcript profiling for prediction of protein expression levels, we compared profiles across the NCI-60 cancer cell panel, which represents nine tissues of origin. For that analysis, we present here two new NCI-60 transcript profile data sets (A based on Affymetrix HG-U95 and HG-U133A chips; Affymetrix, Santa Clara, CA) and one new protein profile data set (based on reverse-phase protein lysate arrays). The data sets are available online at http://discover.nci.nih.gov in the CellMiner program package. Using the new transcript data in combination with our previously published cDNA array and Affymetrix HU6800 data sets, we first developed a “consensus set” of transcript profiles based on the four different microarray platforms. Using that set, we found that 65% of the genes showed statistically significant transcript-protein correlation, and the correlations were generally higher than those reported previously for panels of mammalian cells. Using the predictive analysis of microarray nearest shrunken centroid algorithm for functional prediction of tissue of origin, we then found that (a) the consensus mRNA set did better than did data from any of the individual mRNA platforms and (b) the protein data seemed to do somewhat better (P = 0.027) on a gene-for-gene basis in this particular study than did the consensus mRNA data, but both did well. Analysis based on the Gene Ontology showed protein levels of structure-related genes to be well predicted by mRNA levels (mean r = 0.71). Because the transcript-based technologies are more mature and are currently able to assess larger numbers of genes at one time, they continue to be useful, even when the ultimate aim is information about proteins. [Mol Cancer Ther 2007;6(3):820–32]

The malignant Reed-Sternberg cell of Hodgkin disease is an aberrant B cell that persists in an immunolgically mediated inflammatory infiltrate. Despite its nonproductive immunoglobulin genes, the Reed-Sternberg cell avoids the usual apoptotic fate of defective immune cells through an unknown mechanism. A likely candidate is the surface receptor, CD40, consistently expressed by Reed-Sternberg cells, and the first link in the pathway to NF-κB activation, the central regulator of cytokine production and apoptosis. CD40 signaling in B lymphocytes coordinates the immune response, including immunoglobulin isotype switch and Fas-mediated apoptosis. CD40-induced NF-κB activation is mediated by adapter proteins, the TNF receptor (TNFR)-associated factors (TRAFs), especially TRAFs 2, 3, and 5. Using a Hodgkin cell line, this study demonstrates that CD40 activation of NF-κB is mediated by proteolysis of TRAF3. Results further demonstrate that the pathway can be blocked by treatment with pharmacologic doses of a specific protease inhibitor, pepstatin-A, even in the presence of a mutated NF-κB inhibitor, I-κBα. The stability of TRAF3 regulates CD40/NF-κB–mediated control of the immune response, which is central to the biologic activity of the Reed-Sternberg cell. Prevention of TRAF3 proteolysis may be an entry point for design of novel pharmaceuticals to treat Hodgkin disease and immune system disorders.

The malignant Reed-Sternberg cell of Hodgkin’s disease, first described a century ago, has resisted in-depth analysis due to its extreme rarity in lymphomatous tissue. To directly study its genome-wide gene expression, approximately 11,000,000 bases (27,518 cDNA sequences) of expressed gene sequence was determined from living single Reed-Sternberg cells, Hodgkin’s tissue, and cell lines. This approach increased the number of genes known to be expressed in Hodgkin’s disease by 20-fold to 2,666 named genes. The data here indicate that Reed-Sternberg cells from both nodular sclerosing and lymphocyte predominant Hodgkin’s disease were derived from an unusual B-cell lineage based on a comparison of their gene expression to approximately 40,000,000 bases (105 sequences) of expressed gene sequence from germinal center B cells (GCB) and dendritic cells. The data set of expressed genes, reported here and on the World Wide Web, forms a basis to understand the genes responsible for Hodgkin’s disease and develop novel diagnostic markers and therapies. This study of the rare Reed-Sternberg cell, concealed in its heterogenous cellular context, also provides a formidable test case to advance the limit of analysis of differential gene expression to the single disease cell.

Flow cytometric immunophenotypic analysis is critical in diagnosis and classification of acute leukemia and has been used after therapy to monitor for minimal residual disease. However, the presence of normal B-cell precursors, hematogones, particularly in the context of treated pediatric B-cell precursor acute lymphoblastic leukemia (BP-ALL), may confound such evaluation. In this study, the value of more specific genotypic markers (polymerase chain reaction evaluation of 2 antigen receptor genes) was assessed to resolve this issue. Flow cytometric analysis of enriched mononuclear cells revealed 1% to 20% precursor B cells (PBCs), based on expression of 1 or more pan-B cell antigens in addition to CD10, CD34, and terminal deoxynucleotidyl transferase in all 14 patients studied. Inasmuch as this mimicked the immunophenotype of the original leukemic clone, PBCs, in isolation, were considered suspicious for minimal residual disease. However, 11 of the 14 posttherapy specimens (79%) revealed no monoclonally rearranged antigen receptor genes, and 7 of these 11 patients had trackable genotypic markers at presentation. Accordingly, by PCR these 7 patients had complete molecular remission, supported by clinical follow up of 16 to 73 months. Among the remaining 4 patients with PCR-negative disease, 3 continue in remission, confirming the interpretation of false-positive flow cytometric analysis. In conclusion, flow cytometric monitoring of posttherapy bone marrow specimens from patients with BP-ALL may be misleading, if considered in isolation, in falsely suggesting the presence of minimal residual disease. Rather, PCR for antigen receptor gene rearrangements is a valuable and specific tool, helpful in differentiating hematogones from minimal residual disease in patients with treated BP-ALL whose bone marrow harbors increased PBCs.

Although Hodgkin's disease is highly responsive to treatments that cause apoptosis, it remains resistant to the physiological mechanisms intended to cause cell death. Presumably, the Reed-Sternberg cell defies endogenous apoptosis, persists, accumulates, and manifests the malignant disorder seen clinically. The Reed-Sternberg cell expresses several members of the tumor necrosis factor receptor superfamily. This family of receptors is involved in both activation and proliferation of cells, as well as either protection from or initiation of apoptosis in cells expressing these surface proteins. Signals from these receptors affect transcription. We reasoned that the activation state and resistance to apoptosis of Reed-Sternberg cells might be attributable to dysregulation of genes controling these processes. To determine gene expression by Reed-Sternberg cells, we developed a method of micromanipulation, global reverse transcription, and the reverse transcription-polymerase chain reaction and applied it to 51 single Reed-Sternberg cells and their variants from six cases of Hodgkin's disease. This report analyzes the gene expression of bcl-xs,bcl-xl, bax-α,bax-β, fadd, fas, fas ligand (fas L), ice,TNF-α, TNF-β,TNFR1, TNFR2, TRAF1,TRAF2, TRAF3, cIAP2, and tradd at the level of mRNA in the single Reed-Sternberg cells and their variants. The findings here suggest a molecular mechanism for the activated state and in vivo survival occurring in untreated Reed-Sternberg cells of Hodgkin's disease.

TGF-β1 inhibits the cell cycle progression of many types of cells by arresting them in the G1 phase. This cell cycle arrest has been attributed to the regulatory effects of TGF-β1 on both the levels and the activities of the G1 cyclins and their kinase partners. The activities of these kinases are negatively regulated by a number of proteins, such as p15INK4b, p21WAF1/Cip1, and p27Kip1, that physically associate with cyclins, cyclin-dependent kinases (Cdk), or cyclin-Cdk complexes. In epithelial cell lines, TGF-β1 was previously shown to inhibit cell cycle progression through down-regulation of Cdk4 and/or up-regulation of p15INK4b and/or p21WAF1/Cip1. However, TGF-β1 had little or no effect on the p27Kip1 mRNA and protein levels. In this report, we show that, in contrast to observations in epithelial cell lines, TGF-β1 increased the p27Kip1 mRNA and protein levels in the murine B cell lines CH31 and WEHI231. This TGF-β1-mediated induction of p27Kip1 also resulted in an increased association of p27Kip1 with Cdk2 and a decreased Cdk2 kinase activity. In contrast to epithelial cells, however, TGF-β1 had little or no effect on the Cdk4 and p21WAF1/Cip1 protein levels in these B cells. Finally, although several studies suggested a direct role of p53 in TGF-β1-mediated cell cycle arrest in epithelial cells, TGF-β1 inhibited cell cycle progression in CH31 even in the absence of wild-type p53. Taken together, these results suggest that TGF-β1 induces G1 arrest in B cells primarily through a p53-independent up-regulation of p27Kip1 protein.

The thymus is a major site of apoptosis where programmed cell death is involved in the deletion of self-reactive T cells. We have investigated the role of bcl-x in T cells by defining the expression of its two isoforms, bcl-xl and bcl-xs, in a series of human thymocyte cell lines and in human T lymphocytes using the ribonuclease protection assay. Bcl-xl was the predominant isoform expressed in T cell lines and in T lymphocytes, where expression was further enhanced by PMA/ionomycin. This broad expression supports a central role for bcl-x in thymocyte development perhaps through post-transcriptional regulation.

Most hematopoietic malignancies are widely disseminated even in their "early" stages and often do not have a well-defined localized phase. This makes them less amenable to conventional early screening methods such as imaging and observation. Furthermore, the staging systems for lymphomas are not particularly useful prognostically, with the possible exception of Hodgkin's disease. However, as currently compared with solid tumors, the extensively detailed understanding of the acquired (somatic) genetic lesions in leukemias and lymphomas provide useful molecular biomarkers for early detection. Moreover, well described high risk groups have been identified. These include individuals who are immunosuppressed, for example, iatrogenically following organ transplantation or those with AIDS. Also at high risk are patients treated with certain chemotherapeutic agents who are at risk for the development of acute non-lymphoblastic leukemia. Accordingly, these clinical settings might prove to be good models for evaluating molecular cancer risk markers and the possible introduction of chemoprevention. Here, we outline the biological basis for the application of biomarkers for the early detection of hematological neoplasia. These concepts may provide the stage for the creation of chemoprevention studies in leukemia and lymphoma.

The My4 antibody, one of a number of monoclonal antibodies that react with the CD14 antigen, was originally reported to weakly stain monocytes, macrophages, and granulocytes. However, recent studies have shown that the My4 antibody also stains normal peripheral blood B lymphocytes and some subtypes of B-cell non-Hodgkin's lymphoma. Thus, the authors have studied a large series of non-Hodgkin's lymphomas stained with the My4 antibody. In frozen sections of reactive lymph node biopsy specimens, the My4 antibody strongly stained mantle zone B lymphocytes and weakly reacted with dendritic reticulum cells and histiocytes. In a series of 245 non-Hodgkin's lymphomas, the My4 antibody stained 111 (45%) cases: 108 of 189 (57%) B-cell lymphomas, 3 of 50 (6%) T-cell lymphomas, and 0 of 6 null cell lymphomas. My4-positive B-cell lymphomas occurred in all histologic subtypes with the exception of small noncleaved cell lymphomas. Follicular lymphomas were most often My4 positive (82%). My4 antibody staining showed no correlation with Working Formulation grade. All three My4-positive T-cell lymphomas had a mature T-cell phenotype. Seventy-six of the 111 (68%) My4-positive lymphomas were also analyzed with at least one other anti-CD14 antibody, either Mo2 and/or Leu-M3. In all cases the antigens that react with Mo2 and Leu-M3 were not expressed. Thus, the staining of reactive and neoplastic B cells by My4 appears to be unique to this antibody and is not a feature of all anti-CD14 antibodies.

The absence of wild type retinoblastoma (Rb) gene expression in a wide variety of human solid tumors suggests an etiologic role for this tumor suppressor gene in human cancer. We have evaluated the involvement of Rb gene inactivation in the pathogenesis and progression of human lymphoma and leukemia. We examined the genomic configuration and transcription of the Rb gene in cultured cell lines and primary cases of T- and B-cell lymphomas and leukemias. By Southern analysis, abnormalities of the Rb locus were identified in 1 of 5 T-cell acute lymphoblastic lymphoma (T-ALL) cell lines, 1 of 26 primary cases of T- ALL, 1 of 40 primary cases of chronic lymphocytic lymphoma/well- differentiated lymphoma (CLL/WDL), and 1 of 15 primary cases of intermediately differentiated lymphoma (IDL). By Northern analysis, markedly reduced or abnormal expression of the Rb gene was identified in 2 of 5 T-ALL cell lines, 1 of 7 primary cases of T-ALL, 1 of 5 primary cases of CLL/WDL, and 1 of 6 primary cases of IDL. These findings show that Rb gene inactivation can be associated with a broad range of lymphoid neoplasms and that loss of the tumor suppressor function of Rb may influence the pathogenesis and progression of lymphoma/leukemia.

A major obstacle to investigations of Hodgkin's disease is the paucity of malignant cells, i.e., Reed-Sternberg cells and their variants, in tissues of patients with this disease. Consequently, the pathogenesis, cell of origin, and clonality of this relatively frequent lymphoma have remained unresolved. Results of recent studies suggest that in some instances Reed-Sternberg cells carry rearranged immunoglobulin heavy-chain joining region (JH) loci as well as chromosomal translocations involving band 14q32. Prompted by these findings, we sought to determine if the t(14;18) (q32;q21) translocation of follicular, non-Hodgkin's B-cell lymphoma was associated with Hodgkin's disease. To detect the possible t(14;18) (q32;q21) translocation within the rare malignant cells of Hodgkin's disease, we amplified sequences created by the t(14;18) translocation using the polymerase chain reaction (PCR). With this approach, DNA sequences carrying the direct fusion of the major breakpoint region of the candidate oncogene, bcl-2, derived from chromosome 18q21, with JH on chromosome 14q32 can be detected in as few as one in 10(5)-10(6) cells. In the present study, joined bcl-2/JH sequences were detected in tissues involved by Hodgkin's disease in 17 of 53 (32%) patients. The frequent association of bcl-2 translocation with Hodgkin's disease suggests that this oncogene has a role in the pathogenesis of Hodgkin's disease. That bcl-2 is involved in a major class of lymphoma in addition to follicular lymphoma implies a role for additional factors responsible for generating the two distinctive clinical and pathologic disease states.

The cancer/testis antigen (CTA) group of tumor-associated proteins have been reported to be expressed in various cancers and in adult testis but they are essentially not found in any other normal adult nonneoplastic tissues. Prompted by the frequent detection of SSX1 in a previous comprehensive expression profile of the Hodgkin's lymphoma (HL) cell line L428, we analyzed SSX expression by nonnested reverse-transcription polymerase chain reaction (RT-PCR) in 4 HL cell lines (L428, L540, HD-MY-Z, and KM-H2) and 32 tumor samples of HL. The cellular localization of SSX expression in the tumor samples was further analyzed by in situ hybridization (ISH). All 4 HL cell lines were positive by RT-PCR using SSX consensus primers. Using primers specific to individual SSX genes, all 4 cell lines expressed multiple SSX family members. Five tumor samples (15.6%) were positive by RT-PCR using SSX consensus primers and direct sequencing of the RT-PCR products showed that 4 of 5 expressed more than 1 SSX family member. ISH confirmed that SSX expression originated in HL cells in all 5 RT-PCR-positive tumor samples. Furthermore, ISH demonstrated SSX-positive HL cells in 6 of 11 cases (55%) that were negative by RT-PCR. Our results suggest that members of the SSX family of CTA are expressed in most HL. This subset of HL may be a candidate for immunotherapy approaches directed at SSX proteins.

Precursor B-lymphoblastic lymphoma (B-LBL) may present as a solitary bone tumor. Fewer than 10 cases with a proven precursor B-cell phenotype have been reported in the English literature. In this report, we describe four cases of B-lymphoblastic lymphoma presenting as a localized intraosseous mass, which clinically and histologically mimicked Ewing's sarcoma. Three tumors occurred in the tibia and one in the humerus. In all four cases, the initial diagnosis was either "Ewing's sarcoma" or "consistent with Ewing's sarcoma." All four patients were female. Three were children and one was an adult; mean age was 12.5 years (range, 4 to 31 years). All had extremity pain without significant constitutional symptoms. In three cases, the tumors were osteolytic on radiographic evaluation, and in one case, osteosclerotic. Immunohistochemical stains on paraffin-embedded tissue showed that the neoplastic cells expressed terminal deoxynucleotidyl transferase, CD43, vimentin, and CD99 (MIC2 gene product) in all cases. Three cases were negative for CD45. CD79a was positive in all four cases studied; however, CD20 (L26) was positive in only two of four cases. CD3 was negative in all cases. Two cases showed focal granular cytoplasmic staining for keratin. Two cases analyzed by polymerase chain reaction (PCR) revealed clonal rearrangement of the immunoglobulin heavy chain (IgH) gene. Follow-up revealed that the three pediatric patients, who received a high-dose multiagent chemotherapy regime for LBL, are disease free at follow-up intervals of more than 1, 11, and 12 years, respectively. The adult patient died two years after diagnosis with disseminated disease. Although rare, B-lymphoblastic lymphoma should be considered in the differential diagnosis of small round cell tumors of bone. A diagnosis of Ewing's sarcoma should be made only after complete immunophenotyping and, if necessary, molecular diagnostic tests to exclude lymphoblastic lymphoma. A limited panel of antibodies can lead to an erroneous diagnosis; B-lymphoblastic lymphoma may be negative for CD45 and CD20 but positive for CD99 and even for keratin, mimicking Ewing's sarcoma. Correct diagnosis is extremely important because LBL usually is curable in the pediatric age group with appropriate therapy.

We reviewed our experience with six T-cell-rich B-cell lymphomas (TRBL) presenting in skin. Immunohistochemical studies were performed on all biopsies. The lymphoid population consisted mainly of CD3 and/or UCHL-1 (CD45RO) positive T cells. 5 to 15% of the lymphoid cells stained for the B-cell marker L26 (CD20). Monoclonality of the B-cell component was demonstrated in all cases, utilizing either light chain restriction (5 cases) or clonal immunoglobulin heavy chain gene rearrangement by polymerase chain reaction (PCR) (2 cases). One case was confirmed to be monoclonal by both techniques. Additionally, no clonal rearrangements of the T-cell receptor gamma gene were observed. There was considerable morphological variety in these cases. In H&E stained sections, the differential diagnosis included pseudolymphoma, peripheral T-cell lymphoma, Hodgkin's disease, Lennert's lymphoma and a MALT lymphoma. A significant component of monoclonal plasma cells was present in 3 of 6 cases, suggesting a possible origin from cutaneous immunocytoma. In fact, one of our cases was a biphasic lymphoma displaying TRBL with a small focus of immunocytoma. We conclude that immunophenotypic analysis is necessary for the diagnosis of TRBL. Pathologists should be aware of this type of cutaneous B-cell lymphoma to avoid misinterpretation as a pseudolymphoma.

GATA-1 protein is thought to be a positive regulator of erythroid differentiation. However, ectopic expression of a conditional GATA- 2/estrogen receptor chimera was shown to inhibit erythroid differentiation in a hormone-dependent manner, suggesting the negative regulation of erythroid differentiation by GATA-2 protein. Accordingly, we reasoned that the quantitative balance of GATA-1 and GATA-2 protein might affect erythroid differentiation. In this report, we performed specific and quantitative measurements of GATA-1 and GATA-2 protein in a new erythroid cell line, SAM-1, after treatment with 12-O- tetradecanoylphorbol 13-acetate (TPA). On the basis of these measurements, we show that TPA-induced arrest of erythroid differentiation is coupled with the upregulation of GATA-2 protein, as well as the downregulation of GATA-1 protein. Our results suggest that it is the precise quantitative balance of GATA-1 and GATA-2 protein that regulates erythroid differentiation.

Despite intensive investigation, no clearly defined mechanism explaining human immunodeficiency virus (HIV)-induced cell killing has emerged. HIV-1 infection is initiated through a high-affinity interaction between the HIV-1 external envelope glycoprotein (gp120) and the CD4 receptor on T cells. Cell killing is a later event intimately linked by in vitro genetic analyses with the fusogenic properties of the HIV envelope glycoprotein gp120 and transmembrane glycoprotein gp41. In this report, we describe aberrancies in cell cycle regulatory proteins initiated by cell-cell contact between T cells expressing HIV-1 envelope glycoproteins and other T cells expressing CD4 receptors. Cells rapidly accumulate cyclin B protein and tyrosine-hyperphosphorylated p34cdc2 (cdk1) kinase, indicative of cell cycle arrest at G2 phase. Moreover, these cells continue to synthesize cyclin B protein, enlarge and display an abnormal ballooned morphology, and disappear from the cultures in a pattern previously described for cytotoxicity induced by DNA synthesis (S phase) inhibitors. Similar changes are observed in peripheral blood mononuclear cells infected in vitro with pathogenic primary isolates of HIV-1.

To determine efficiently the clonality of B-cell lymphoproliferative disorders, we modified an immunoglobulin heavy-chain (IGH) gene rearrangement polymerase chain reaction (PCR) assay that requires only a single primer germline variable (VH) and joining (JH) pair and does not involve nested priming, blot hybridization, radioactivity, or sequencing of the amplified PCR product. This simple PCR technique enabled detection of IGH gene rearrangements in as little as 10 pg (one cell equivalent) of DNA or when the clonal-to-polyclonal B-cell ratio was experimentally set at 1:1000. We detected IGH gene rearrangements in 83.5% (71 of 85) of clonal B-cell processes, a sensitivity approaching that of more cumbersome multiple primer and nested primer assays. Moreover, this technique is equally effective with fixed tissues, either B5 or formalin, and can be performed on minute samples, histologic sections, fine-needle aspirates, or cerebrospinal fluids. When compared with conventional Southern blot analysis using a genomic JH probe, the PCR assay demonstrated IGH gene rearrangements in 82% (37 of 45) of B-cell processes positive by Southern blot. No false-positive results were observed in 29 negative control tissues. We now use IGH gene PCR routinely in our laboratory for the detection of clonal B-cells in virtually any tissue sample as an aid in early diagnosis, staging, and monitoring, and the Southern blot procedure is reserved for only a minority of diagnostic cases. for only a minority of diagnostic cases.

Acquired immunodeficiency syndrome (AIDS) is a complex disease process induced by human immunodeficiency virus (HIV-1) infection.1 Although the linkage between HIV-1 infection and the development of AIDS has been established for a decade, 2 the molecular and biochemical basis for the profound and irreversible depletion of helper CD4+ T cells that follows HIV infection and paralyzes the immune system is not understood. A number of mechanisms have been proposed to account for CD4+ T killing by HIV, including the direct lysis of virally-infected cells, and the functional disruption of uninfected cells through an interaction with viral proteins.1,3,4 A recent hypothesis has proposed that, in HIV-infected individuals, there reemerges a cell death program normally utilized by immature T cells during development in response to specific stimuli accounting for both the early qualitative and late quantitative CD4+ T cell defects associated with AIDS.5

Deregulation of proto-oncogenes or inhibition of tumor suppressor genes results in the acquisition of a cellular growth advantage, usually manifested as increased proliferation [1]. Many of the mechanisms that mediate these functions have been elucidated, while the function of others still remains to be resolved. Nevertheless, there exists a broad stratification of genes whose alteration is mechanistic in the development of neoplasia. By contrast, bcl-2, which was originally identified in 1984 [2], appears to function by a mechanism independent of the above two categories, and it may be the first described member of a third broad class of genes whose deregulation plays a role in oncogenesis. It appears to act by inhibiting cell death, rather than by stimulating proliferation.

The origin of the Reed–Sternberg cell of Hodgkin's disease remained clouded in mystery for almost a century after its discovery in 1898. The major obstacle to its understanding is that, unlike other cancers, the malignant cell of Hodgkin's disease is vastly outnumbered by surrounding non-neoplastic cells at approximately 1000:1. We have devised several strategies to isolate Reed–Sternberg T-cells to determine their origin, global gene expression and, ultimately, their pathogenesis. This has increased the number of genes known to be expressed in Reed–Sternberg cells by >100-fold to over 12,000. Approaches such as density gradients, microdissection, and cell sorting help to enrich Reed–Sternberg cells for genomic DNA analysis. However, single-cell micromanipulation of living Reed–Sternberg cells was required to determine the genome-wide gene expression profile of these cells. Combined analysis of single cells and cell lines revealed the expression of 2666 named genes. Further analysis with high-density gene expression microarrays has demonstrated the expression of approximately 12,000 genes by Reed–Sternberg cells. The gene expression profile is that of an aberrant germinal center B-lymphocyte that resists apoptosis through CD40 signaling and NFkB activation. Gene expression analysis of Hodgkin's disease is an extreme test case demonstrating the application of high-throughput gene expression studies even to individual cells from clinical samples.

The inactivation or loss of tumor suppressor genes (anti-oncogenes) has been implicated as a mechanism central to the pathogenesis of many solid tumors. More recently, we and others have identified a role of one rumor suppressor gene, the retinoblastoma gene, in the development of human lymphoid lymphoma and leukemia. Here we review the involvement of the retinoblastoma gene in the control of normal lymphocyte cell division and the consequences of inactivation of the retinoblastoma gene for the development of lymphoid neoplasia. Our survey has disclosed a broad involvement of retinoblastoma gene inactivation in a wide variety of non-Hodgkin's lymphomas and lymphocytic leukemia. Based on these early findings, it appears likely that tumor suppressor genes may well be involved in many hematopoietic neoplasma.

The demonstration of immunoglobulin or T-cell receptor gene rearrangements in human lymphoproliferative processes with the use of DNA hybridization has gained great popularity as a sensitive laboratory adjunct to diagnostic hematopathology. The fact that nearly all B- or T-cell malignant lymphomas and leukemias have one or more rearranged antigen receptor genes provides a biologic basis for a diagnostic test. To formally analyze the sensitivity, specificity, and reproducibility of gene rearrangements in the diagnosis of human lymphoproliferative disease, the authors conducted a large, multiinstitutional study. Through a blinded, controlled approach, gene rearrangement analysis of 275 cases was shown to carry a high correlation with conventional phenotyping and histologic diagnosis, with only minor false-positive and false-negative rates. Significantly, no rearrangements were detected in normal lymphoid tissues or carcinomas, sarcomas, or melanomas. In a randomized study of 50 cases, laboratory results showed a high rate of interlaboratory agreement, regardless of the level of previous experience. Furthermore, the reproducibility of interpretation of data (Southern blot autoradiograms) of 192 cases showed high concordance among 11 observers from multiple laboratories. Based on these findings, the authors propose a set of guidelines for interpretation of gene rearrangement analysis that, if carefully followed, renders this a highly reproducible, safe, and accurate addition to the diagnostic regimen for human lymphoproliferative processes.

One hundred ninety-three patients with stage II, III, or IV follicular large-cell, diffuse large-cell, diffuse mixed, immunoblastic, or diffuse small noncleaved-cell (non-Burkitt's) lymphoma were randomized to receive either cyclophosphamide 650 mg/m2 intravenously (IV), doxorubicin 25 mg/m2 IV, etoposide 120 mg/m2 IV on day 1, mechlorethamine 6 mg/m2 IV, vincristine 1.4 mg/m2 (no cap at 2 mg total dose) IV on day 8, prednisone 60 mg/m2 orally daily days 1 through 14, procarbazine 100 mg/m2 orally daily days 8 through 14, and methotrexate 500 mg/m2 IV on day 15 with leucovorin 50 mg/m2 orally every 6 hours for four doses beginning 24 hours after methotrexate with cycles repeated every 28 days (ProMACE-MOPP) or same day-1 treatment as ProMACE-MOPP plus cytarabine 300 mg/m2 IV, bleomycin 5 U/m2 IV, vincristine 1.4 mg/m2 (no cap at 2 mg total dose) IV, and methotrexate 120 mg/m2 IV on day 8, leucovorin 25 mg/m2 orally every 6 hours for four doses beginning 24 hours after methotrexate, and prednisone 60 mg/m2 orally daily days 1 through 14 with cycles repeated every 21 days (ProMACE-CytaBOM). Co-trimoxazole two double-strength tablets orally twice daily throughout the period of treatment was added to the ProMACE-CytaBOM regimen when an increased risk of Pneumocystis carinii pneumonia was found in the first 35 patients receiving this combination. Median follow-up is 5 years. Among the 99 patients treated with ProMACE-MOPP, 73 achieved a complete remission (CR) (74%), 30 complete responders have relapsed (41%), and 45 patients have died (45%), including two (2%) of treatment-related causes. Among the 94 patients treated with ProMACE-CytaBOM, 81 achieved a CR (86%), 22 complete responders have relapsed (27%), and 31 patients have died (33%). The complete response rate (P2 = .048) and survival (P2 = .046) were significantly higher for patients treated with ProMACE-CytaBOM. The mortality of ProMACE-CytaBOM treatment overall was six of 94 patients (6.4%). There was no treatment-related mortality among patients treated with prophylactic co-trimoxazole (n = 59). ProMACE-CytaBOM combination chemotherapy with co-trimoxazole prophylaxis is a safe and effective treatment for patients with aggressive histology malignant lymphoma and is superior to ProMACE-MOPP.

A fusion complementary DNA in the T cell line HSB-2 elucidates a provocative mechanism for the disruption of the putative hematopoietic transcription factor SCL. The fusion cDNA results from an interstitial deletion between a previously unknown locus, SIL (SCL interrupting locus), and the 5' untranslated region of SCL. Similar to 1;14 translocations, this deletion disrupts the SCL 5' regulatory region. This event is probably mediated by V-(D)-J recombinase activity, although neither locus is an immunoglobulin or a T cell receptor. Two other T cell lines, CEM and RPMI 8402, have essentially identical deletions. Thus, in lymphocytes, growth-affecting genes other than immune receptors risk rearrangements.

The rearranging antigen receptor genes of lymphoid cells serve as unique clonal markers of lymphoid neoplasms. Gene rearrangement analysis is a highly sensitive and reproducible tool which is useful in the diagnosis and classification of malignant lymphoma/leukemia. Although clonality can often be determined among B cell neoplasms by virtue of immunoglobulin isotype analysis, no such phenotypic marker of clonality exists for T cells. Therefore, clonality of T lymphoproliferative processes is most readily determined by rearrangement analysis of the T cell antigen receptor genes. The alpha, beta, gamma, and delta genes of the T cell receptor gene family encode heterodimeric surface antigen receptors and undergo rearrangement early in T cell differentiation. Identification of rearrangement of T cell antigen receptor genes provides valuable diagnostic information concerning cellular lineage, clonality and classification of T cell neoplasms. This molecular approach is applicable to the diagnosis of occult disease, relapse, and resolution of diagnostic dilemmas in any type of tissue sample including fluids and needle aspirations.

Discovery of 830 candidate therapeutic targets and diagnostic markers for breast cancer using oligonucleotide microarray technology

We have determined the genome-wide gene expression profile of Hodgkin disease, the most common cancer of young adult Americans. Sequences were obtained from approximately 40,000 ESTs (20,000,000 bases) from Hodgkin disease sources. These included cell lines, primary tissue and living, single Reed-Sternberg cells — the exceedingly rare but biologically active malignant cell of Hodgkin lymphoma. Results of the high throughput sequencing increased the number of named, expressed genes in Hodgkin disease from 100 to 2666. We devised an algorithm to globally compare the gene expression profile of Hodgkin's disease to 100,000 ESTs from different hematopoetic cell types. The analysis supports a B cell lineage of the elusive Reed-Sternberg cell and provides a public database of the gene expression profile of Hodgkin lymphoma1 (http://www.hodgkins.georgetown.edu).

Table of the report entitled "Superiority of ProMACE-CytaBOM Over ProMACE-MOPP in the Treatment of Advanced Diffuse Aggressive Lymphoma: Results of a Prospective Randomized Trial" by Longo et al (J Clin Oncol 9:25–38, 1991) contained several errors. It is reprinted here correctly in its entirety: Please see the PDF for Table.

The immunologic phenotypes of 59 cases of diffuse, aggressive, non-Hodgkin’s lymphomas were determined using a battery of immunologic and cytochemical techniques. Included were cases of diffuse, large cell “histiocytic,” mixed cell, and undifferentiated non-Burkitt’s. Burkitt’s lymphoma, lymphoblastic lymphoma, and mycosis fungoides/SCzary’s syndrome were excluded from this study since these are distinct clinicopathologic entities with well-recognized immunologic phenotypes. The immunotype could be determined in 57/59 (97%) cases tested: 31 of 59 cases (53%) were B-cell type, 25 of 59 (42%) were peripheral T-cell type, and one was true histiocytic. Two cases had no detectable markers and were called “null cell.” This relatively high frequency of peripheral T-cell lymphomas in an American series previously has not been observed and may be a result of progressive improvements in immunologic techniques. Monoclonal anti-T cell antibody staining was performed in 11 T-cell cases and corroborated the findings using spontaneous E-rosette formation. Eight of the T-cell lymphomas had a helper cell phenotype whereas one had a suppressor cell phenotype and two could not be subclassified. All B-cell lymphomas in this series possessed monoclonal surface immunoglobulin detected by direct immunofluorescence of viable cells. Enzyme cytochemistry profiles only partially correlated with immunotype and were not believed to be helpful in the determination of specific phenotypes. There were no significant differences between the B-cell and T-cell diffuse aggressive lymphomas with respect to sex, constitutional symptoms, stage, sites of extranodal involvement, complete remission rate, or survival when they were studied prior to the initiation of aggressive therapy. Although immunotyping can be successfully performed in essentially all cases of diffuse, aggressive non-Hodgkin’s lymphomas, to date, the authors have been unable to demonstrate that immunotype alone has an independent prognostic effect. Cancer 541310-1317, 1984. LTHOUGH THE DIFFUSE, aggressive, non-Hodgkin’s A lymphomas are potentially curable,’ nearly 50% of patients with these lymphomas have failed to achieve complete remission and 40% die within the first year following initial diagnosis.’ In a previous study, we have shown that histologic subclassification of these lymphomas does not distinguish between patients with long-term maintained remissions and those who die within the first year.2 The lack of correlation of patient survival and histologic type was true for both the classification of Rap

Molecular genetic analysis of rearranging antigen-receptor genes in non-Hodgkin's lymphomas has revealed the clonality and lineage in the majority of cases. In an analogous approach, we sought to apply gene rearrangement analysis to Hodgkin's disease to understand better the clonality and origin of this disorder. However, the putative neoplastic cell of Hodgkin's disease, the Reed-Sternberg cell and its variants, is extraordinarily rare in most cases of Hodgkin's disease. On the average, Reed-Sternberg cells and variants represent 0.1 per cent of total cell suspensions of nodular sclerosing Hodgkin's disease. As this frequency is below the minimum threshold of sensitivity of the Southern blot assay, we attempted to enrich for Reed-Sternberg cells before DNA extraction and analysis. Using either elutrition or Percoll density gradient centrifugation, we were able to enrich the percentage of Reed-Sternberg cells and variants to above 1 per cent in five cases of nodular sclerosing Hodgkin's disease. In three of these cases, immunoglobulin gene rearrangements were identified, but no T cell receptor gene rearrangements were seen. No rearrangements were detected in unseparated cells or in the Reed-Sternberg cell-depleted fractions. In addition, the L428 Hodgkin's disease cell line was found to have one rearranged and one deleted heavy-chain gene, a rearranged kappa gene, a rearranged lambda gene, and a single rearranged beta allele. No rearrangements of the T gamma gene were found in L428. Taken together, these findings indicate that clonal cell populations are present in Hodgkin's disease and suggest the possibility of a clonally expanded lymphoid cell in this disorder.

A small (2.8-kilobase, kb) major breakpoint region localized to segment 18q21 rearranges in greater than 70% of t(14;18)(q32;q21) lymphomas. This rearrangement interrupts the Bcl-2 gene and introduces it into the Ig locus at 14q32. The rearrangement between the joining region (JH) of Ig on chromosome 14 and the 18q21 region creates a translocation- specific DNA rearrangement. We generated probes that distinguish the 14;18 juncture on the derivative (der) 14 and der (18) chromosomes, providing a molecular approach to t(14;18) identification. Approximately 60% of unselected follicular lymphomas, 20% of diffuse large cell lymphomas, and 50% of adult undifferentiated non-Burkitt lymphomas demonstrated 14;18 rearrangements within the major breakpoint region. Examination of DNA for 14;18 rearrangements resolved the identity of 14q+ chromosomes in two patient's cells that lacked an obvious reciprocal partner. Identification of the exact restriction fragments that mediate translocations complements routine cytogenetics. The detection of DNA rearrangements does not require dividing cells or the presence of an identifiable reciprocal partner and can detect clonal translocation rearrangements when the neoplastic cells are only a minority of all cells present.

B cell chronic lymphocytic leukemia (CLL) cells appear to be arrested in their differentiation so that little immunoglobulin is secreted in most cases. To determine their capacity for further differentiation we stimulated cells from a series of 10 cases of CLL with a phorbol ester and assayed for production of immunoglobulin protein, accumulation of immunoglobulin mRNA, and alterations in cell surface markers. We found that cells from all cases were induced to secret monoclonal immunoglobulin of the same heavy and light chain type as the surface membrane immunoglobulin type. Immunoglobulin secretion was preceded by a rapid increase in the levels of mRNA coding for IgM, predominantly the secretory form, mu s-mRNA, rather than the membrane form, mu m-mRNA. A similar selection of mu s- over mu m-mRNA is known to occur in plasma cells by a mechanism of differential processing of mRNA from a single mu-chain gene. Except for a decline in the expression of surface IgD, cell surface determinants remained unaffected both in terms of the percentage of positive cells and the relative number of sites per cell. In contrast to previous studies, these results indicate that CLL cells consistently retain the capacity to further differentiate toward plasma cells and secrete immunoglobulin. The immunoglobulin secretion is mediated, at least in part, by a developmentally regulated increment in mu s-mRNA.

Acquisition of mature T cell function and the T cell antigen receptor repertoire occur in the thymus. In an effort to delineate the cascade of events leading to T cell maturation, we analyzed a series of clonal human precursor T cell neoplasms representing early, middle, and late stages of intrathymic differentiation. Rearrangements of the T cell receptor beta and gamma genes appear concurrently and are preceded by surface expression of the 3A1 (CD7) molecule. Subsequent transcription of the beta gene is coordinated with surface expression of T1 (CD5) and T11 (CD2), transcription of T3 delta mRNA, and the appearance of intracellular T3 (CD3) protein. As late events, T alpha gene transcripts appear and, finally, T3, the multichain complex linked to the T cell receptor, is presented on the cell surface. Findings reported here provide a model of the developmental orchestration of genes encoding antigen recognition in human T cells.

The phenotypes of early stages of T cell maturation are reflected by precursor T (lymphoblastic) neoplasms. In the present study, a series of such neoplasms was analyzed to reveal the developmental association of the expression of stage-related cell surface markers and T cell receptor gene rearrangement. Rearrangements of the T cell receptor beta- chain (T beta) gene were found in most, but not all, cases (88%) of T cell lymphoblastic neoplasms. T beta gene rearrangement preceded surface expression of the T cell receptor-linked molecular complex T3. Of all monoclonal anti-T cell antibodies tested, only antibody 3A1 was capable of reacting with neoplastic cells from all cases irrespective of the occurrence of T cell receptor gene rearrangements. In contrast, markers T1 and T11, normally expressed by mature T cells, were absent from the neoplastic cells in many cases (73% and 60% positive cases, respectively). Thus, antibody 3A1 is a valuable probe for the identification of T lymphoblastic neoplasms since its target antigen is consistently expressed and does not require prior T beta gene rearrangement. Furthermore, expression of 3A1 prior to T beta gene rearrangement suggests that it may be a cell surface protein that participates in the triggering of T cell receptor gene rearrangement and expression. It is concluded that precursor T cell neoplasms display an early T cell development hierarchy that, in sequence, consists of 3A1 expression, T beta gene rearrangements, and surface T3 expression.

Expression of class II major histocompatibility (MHC) antigens by peripheral blood monocytes from 12 patients with Hodgkin's disease (HD) were studied employing antiserum and complement-mediated cytotoxicity. Overall, the expression of class II MHC antigens was significantly decreased on HD monocytes (cytotoxicity index [C.I.] = 60.2 ± 5.8% vs 77.6 ± 2.8%, P < .02). This decrease was most marked in patients with more severe disease. In fact, mean alloantigen expression for patients with advanced stages of disease was 58% of that observed in controls. The number of human lymphocyte antigen (HLA)-DR antigenic sites per cell was also reduced as determined by monoclonal anti-DR antibody and FACS analysis. There was 38% more HLA-DR per cell in normal controls than in moderately advanced Hodgkin's patients. Class II MHC antigen expression on HD monocytes were increased partially by an IFN-7 containing concanavalin A-stimulated human mononuclear cell culture supernatants (Con A sup), although remaining subnormal. When monocytes were cultured with Con A sup and indomethacin, alloantigen expression was increased in HD and control monocytes, but indomethacin failed to normalize class II MHC antigen expression on HD monocytes (C.I. = 72.3 ± 4.7% vs 90.2 ± 1.8%, P < .01). We conclude that PGE accounts only inpart for the decreased alloantigen expression by HD monocytes. Interleukin (IL) 1 production by patients' monocytes was not reduced as compared to normals and therefore does not contribute to the decreased MHC II antigen expression. Decreases in alloantigen expression may be an important determinant of the T cell-mediated immune abnormalities in Hodgkin's disease.

Monoclonal antibodies J5, VIL-A1, and BA-3, known to react with the common acute lymphoblastic leukemia antigen (CALLA) were found to specifically stain normal human polymorphonuclear neutrophils (PMN). The antigen detected on PMN had a molecular weight (95,000-110,000 mol wt) close to that of CALLA (95,000-100,000 mol wt) and thus these surface membrane antigens are likely related, if not identical. The fluorescent staining intensity of PMN is comparable to that of CALLA-positive leukemic cells and the presence of PMN in patient samples could potentially produce false-positive results in diagnosis.

Clinical and histopathologic material from 151 cases of diffuse mixed, diffuse histiocytic, and diffuse undifferentiated non-Burkitt's lymphomas have been reviewed to determine the factors that predict long- term survival. Median survival of all patients was 34 mo with 43% alive at 70 mo. Factors associated with a poor prognosis include: male sex, constitutional symptoms, advanced stage, bone marrow involvement, huge (greater than 10 cm) abdominal masses with gastrointestinal involvement, hepatic involvement, hemoglobin greater than 12 g/dl, or serum LDH greater than 250 U. The best prediction of a given patient's survival was defined by a set of four variables, which includes sex, symptoms, bone marrow status, and the presence or absence of a huge abdominal mass with gastrointestinal involvement. In contrast, classification of these patients according to the histopathologic categories of Rappaport of Strauchen did not define patient groups with significant differences in survival, nor did these categories correlate with the previously described clinical factors. Knowledge of the distribution of these prognostic factors in any clinical trial is needed before therapeutic results can be compared. In addition, such data may define subsets of patients for whom current therapy is inadequate and conversely those patients for whom current therapy yields excellent long-term survival.

Recombinant DNA probes were cloned for the areas flanking the breakpoint on chromosome 18 in cells from a patient with acute lymphocytic leukemia of the B-cell type; cells of this line carry the t(14;18) chromosomal translocation. Two of the probes detected DNA rearrangements in approximately 60 percent of the cases of follicular lymphoma screened. In follicular lymphoma, most of the breakpoints in band q21 of chromosome 18 were clustered within a short stretch of DNA, approximately 2.1 kilobases in length. Chromosome 18-specific DNA probes for the areas flanking the breakpoints also detected RNA transcripts 6 kilobases in length in various cell types. The gene coding for these transcript (the bcl-2 gene) seems to be interrupted in most cases of follicular lymphomas carrying the t(14;18) chromosomal translocation.

Many cancers have been cured by chemotherapeutic agents. However, other cancers are intrinsically drug resistant, and some acquire resistance following chemotherapy. Cloning of the cDNA for the human MDR1 gene (also known as PGY1), which encodes the multidrug efflux protein P-glycoprotein, has made it possible to measure levels of MDR1 RNA in human cancers. We report the levels of MDR1 RNA in greater than 400 human cancers. MDR1 RNA levels were usually elevated in untreated, intrinsically drug-resistant tumors, including those derived from the colon, kidney, adrenal gland, liver, and pancreas, as well as in carcinoid tumors, chronic myelogenous leukemia in blast crisis, and cell lines of non-small cell carcinoma of the lung (NSCLC) with neuroendocrine properties. MDR1 RNA levels were occasionally elevated in other untreated cancers, including neuroblastoma, acute lymphocytic leukemia (ALL) in adults, acute nonlymphocytic leukemia (ANLL) in adults, and indolent non-Hodgkin's lymphoma. MDR1 RNA levels were also increased in some cancers at relapse after chemotherapy, including ALL, ANLL, breast cancer, neuroblastoma, pheochromocytoma, and nodular, poorly differentiated lymphoma. Many types of drug-sensitive and drug-resistant tumors, including NSCLC and melanoma, contained undetectable or low levels of MDR1 RNA. The consistent association of MDR1 expression with several intrinsically resistant cancers and the increased expression of the MDR1 gene in certain cancers with acquired drug resistance indicate that the MDR1 gene contributes to multidrug resistance in many human cancers. Thus, evaluation of MDR1 gene expression may prove to be a valuable tool in the identification of individuals whose cancers are resistant to specific agents. The information may be useful in designing or altering chemotherapeutic protocols in these patients.

The human T-cell leukemia/lymphoma virus (HTLV) is a novel type-C retrovirus isolated from patients with T-lymphoproliferative malignancies. Thirteen cases of HTLV-associated malignancy from US centers were studied in detail. Ten of these cases share common clinical features and define a typical virus-associated adult T-cell leukemia/lymphoma (ATL). All ten patients presented with Ann Arbor stage IV lymphoma because of skin involvement, bone marrow involvement, or lymphomatous leptomeningitis. Lymphadenopathy occurred in 7 of 10 patients at presentation, and the malignant cells were cytologically pleomorphic. Leukemia occurred in 60% of the patients at presentation. Hypercalcemia was found initially in two-thirds of the patients, with lytic bone lesions or positive bone scans in 7 of 10. Complete remission occurred in 40%, but all have relapsed. These cases closely resemble those virus-positive cases of adult T-cell leukemia/lymphoma (ATL) reported from Japan and the Caribbean. Three additional virus- positive patients had atypical presentations and diagnoses (acute lymphocytic leukemia, Sezary's syndrome, leukemic reticuloendotheliosis), usually with less aggressive clinical courses and atypical demographic and laboratory features. Presence of HTLV serum antibodies in cases of ATL (with hypercalcemia and circulating malignant cells) appears to define a distinct clinicopathologic entity that may occur in geographic clusters.

We studied the configuration and expression of the gene encoding the beta chain of the T cell receptor (TCR beta) in cell lines and primary tumor cells infected by the human T cell leukemia/lymphoma (lymphotrophic) virus type I (HTLV-I). Most of the cell lines and all the primary tumor cells showed rearrangement of the TCR beta gene, and in each case the rearrangement was distinct. The majority of cases examined were clonal with respect to a particular TCR beta gene rearrangement. Primary tumor cells from one case (SD) were found to have a tandem duplication of a portion of chromosome 7; this appears to have resulted in the presence of three alleles of the TCR beta gene, each of which is arranged differently. This suggests that the chromosomal abnormality, and possibly infection by HTLV-I, occurred before TCR beta gene rearrangement. Cell lines infected by HTLV-I express levels of TCR beta mRNA similar to PHA stimulated lymphocytes, suggesting that this gene is not transcriptionally activated as a result of infection by HTLV-I. Cloned T cells of known antigen specificity that are infected by HTLV-I in vitro show impairment of immune function, including loss of antigen-specific responsiveness and the acquisition of alloreactivity. Comparison of the configuration of the TCR beta gene before and after infection revealed no changes detectable by Southern blot analysis. Levels of expression of the TCR beta gene at the mRNA level and surface expression of the T3 complex were also not significantly altered, suggesting that changes in immune function cannot be attributed to quantitative changes in the TCR molecule. The configuration of the TCR beta gene in primary tumor cells infected by HTLV-I was compared with that in the derived cell lines. In all pairs examined, the configuration in the primary tumor cells was different from that in the cell lines, strongly suggesting that the cells that grow in culture are not the original neoplastic cells.

To elucidate the mechanism of the t(14;18)(q32;q21) chromosomal translocation found in follicular lymphoma, we examined the structure of both derivative (der) chromosomal breakpoints as well as their germ-line predecessors. We noted that chromosome segment 18q21 was juxtaposed with immunoglobulin heavy (H) chain gene diversity (DH) regions on all five der(18) chromosomes we examined, and we confirmed the juncture with immunoglobulin H-chain gene joining (JH) regions on the der(14) chromosome. However, the t(14;18) was not fully reciprocal in that chromosome 14 DNA between the DH and JH regions was deleted. Furthermore, extra nucleotides, reminiscent of "N" segments, were present at the der(14) and possibly der(18) junctions. This indicates that despite the mature B-cell phenotype of follicular lymphoma, the t(14;18) occurs during attempted DH-JH joining, the earliest event in immunoglobulin rearrangement in a pre-B-cell. Our detailed analysis of the germ-line 18q21 region indicated that most breakpoints clustered within a 150-base-pair major breakpoint region. However, we found no evidence for evolutionarily conserved immunoglobulin-like recombinational signals at 18q21, arguing against a role for immunoglobulin recombinase in chromosome 18 breakage. Instead, a direct repeat duplication of chromosome 18 sequences was discovered at both chromosomal junctures, typical of the repair of a naturally occurring staggered double-stranded DNA break. These results prompt a translocation model with illegitimate pairing of a staggered double-stranded DNA break at 18q21 and an immunoglobulin endonuclease-mediated break at 14q32 and with N-segment addition, repair, and ligation to generate der(14) and der(18) chromosomes.

Routine morphologic and cytochemical study of the leukemic cells of a 13-month-old child did not permit definitive determination of either a lymphoid or a myeloid cell origin. However, ultrastructural cytochemical analysis revealed platelet peroxidase (PPO) reactivity. The malignant cells expressed PPO in both unfixed and fixed preparations. PPO was observed in the perinuclear space and reticulum but not in the Golgi saccules. In a 1% glutaraldehyde, 1% formaldehyde solution known to inhibit PPO but not myeloperoxidase, no reaction product was observed. The demonstration of PPO activity in these undifferentiated cells established their megakaryoblastic lineage. Further studies including DNA flow cytometry confirmed that these cells were megakaryoblasts. This study demonstrates that megakaryoblastic malignancy may occur as an acute leukemia of childhood and emphasizes the value of ultrastructural cytochemistry in cases of acute leukemia which defy routine classification. Abnormalities of chromosome 21 were present in our patient as in four previous cases of megakaryoblastic leukemia in childhood. The nonrandomness of chromosome 21 involvement in this disease should therefore be considered.

Despite intensive efforts using a wide variety of approaches, the cellular lineage and clonality of the abnormal cells of Hodgkin's disease have remained an enigma. In the present study, cell separation techniques that enriched for Reed-Sternberg cells and their variants were used to generate sufficient percentages of abnormal cells to allow detection of rearrangements in these cell fractions. DNA from the involved tissues of eight Hodgkin's disease patients was subjected to Southern blot analysis to detect rearrangements of T cell antigen receptor genes and immunoglobulin genes. Immunoglobulin gene rearrangements were found in three of five cases in which Reed- Sternberg cells and their variants were enriched by cell separation techniques to cell frequencies greater than 1%. Rearrangements of immunoglobulin heavy chain genes occurred in two cases, and a lambda light chain gene rearrangement occurred in a third case. Rearrangements were not detected in lymphocyte fractions or in unseparated cells prepared from the same tissues. The putative Hodgkin's cell line, L428, also contained rearrangements of immunoglobulin heavy and kappa and lambda light chain genes and, in addition, harbored a single T cell receptor beta gene rearrangement. These findings indicate that Reed- Sternberg cell-enriched fractions contain clonal cell populations and provide a lead, at the molecular genetic level, to a possible lymphoid derivation of the Reed-Sternberg cell.

Immune complexes of BSA-anti BSA (IgG F(ab′)2) will activate human complement by the alternate pathway. The C3 bound to these complexes binds to human peripheral blood mononuclear cells which bear complement receptors. This interaction is visualized by fluoresceinated antigen (BSA-FITC) or fluorescein conjugated antisera directed at the C3 component of complement. The assay appears to be more sensitive than the usual IgMEAC rosette technique and correlates well with the rosette method.