Lymphoid Leukemia Cells Research Articles

Myeloperoxidase: expression and modulation in a large panel of human leukemia-lymphoma cell lines

Myeloperoxidase (MPO) is found exclusively in the azurophilic granules (primary lysosomes) of normal myelomonocytic cells. Cytochemical staining for MPO activity is used clinically to distinguish myeloid from lymphoid leukemias. We studied the expression of MPO at the RNA and protein level in 140 continuous human leukemia-lymphoma cell lines using classical cytochemistry, immunofluorescent staining with a specific monoclonal antibody, Northern blot analysis, and a reverse transcription-polymerase chain reaction (RT-PCR) amplification assay. Seventy-eight lymphoid leukemia, myeloma, and lymphoma cell lines were negative; only 3 pre-B-acute lymphoblastic leukemia (ALL) cell lines were MPO-positive. Two of these MPO-positive pre-B-ALL cell lines showed a trace expression after RT-PCR and Southern blotting corresponding to 4% to 6% of the transcripts found in other positive myeloid cell lines. The third pre-B-ALL cell line was positive in Northern blots and cytochemical/immunofluorescent staining; however, only few cells were weakly positive in the latter assay. Although 15 of 59 cell lines assigned to the myeloid, monocytic, megakaryocytic, or erythroid lineages were MPO-positive in Northern blots, those 15 and 13 additional cell lines showed bands of mRNA after RT-PCR. MPO protein was detected in all 16 Northern-positive cell lines; on the other hand, there were 4 cell lines that were protein-positive, but Northern- negative. Differentiation induced by protein kinase C activators 12-O- tetradecanoylphorbol 13-acetate and Bryostatin 1 or by all-trans retinoic acid was associated with a decrease in MPO mRNA in all 7 initially positive cell lines studied, even leading to the complete absence of transcripts, but the enzymatic activity of the differentiated cells was only slightly less than that of unstimulated cells. MPO expression could not be induced in 10 initially negative cell lines. The half-life of MPO mRNA was found to be about 6 hours and was not shortened by prior exposure of the cells to the differentiation- inducing agents. These results confirm that MPO expression is mainly associated with myelomonocytic cells, but also underline the notion that MPO cannot be used as an absolutely lineage-specific marker for the distinction of leukemic cells. MPO can be used as an excellent parameter to characterize the various stages of normal and induced differentiation.

PRODUCTION, CHARACTERIZATION, AND FUNCTIONAL STUDIES ON HIM35: A CD15-LIKE MoAb WITH SOME BIOLOGICAL ACTIVITIES

The MoAb HIM35 (IgM) has been found to react with 100% granulocytes and 20%~30% monocytes from peripheral blood; 6/6 cases of CML cells; some (14/34 cases) AML cells and most myelomonocytic cell lines. However, it did not react with normal lymphocytes, thymocytes, platelets, red blood cells, lymphoid leukemia cells, and any tested T, B leukemic cell lines. The molecular weight of the antigen recognized by HIM35 was 220000. The binding of HIM35 to its target cells could be competitively inhibited by previously characterized anti-CD 15 MoAbs. Functional studies showed that HIM35 was able to inhibit phagocytosis by granulocytes and to stimulate CFU-GM growth synergistically with IL-3 or GM-CSF. Key words: Antibody, monoclonal; biologic characteristics

Expression of the MuLV-tumor-associated antigen is restricted to MuLV-transformed cells.

Immunization of mice with Moloney murine leukemia virus (MoMuLV) induces the generation of a population of CTL which recognizes a non-viral, tumor-associated antigen (TAA) expressed on MuLV-induced tumors. To determine whether this TAA could be used as a pre-leukemic or leukemic cell marker, CTL clones directed against Moloney viral and TAA antigen were used to analyze viral and TAA antigen expression on chronically infected and leukemic lymphoid cells obtained from mice inoculated neonatally with MoMuLV. Although both sets of cells could be recognized and lysed by viral antigen specific CTL, they are not recognized by TAA-specific CTL. Only after transformed cell lines were established from leukemic spleen cells could susceptibility to TAA-specific CTL be observed. Thus, the appearance of the MoMuLV-TAA was restricted to MoMuLV-transformed cells.

The cytotoxicity of copper(II) complexes of heterocyclic thiosemicarbazones and 2-substituted pyridine N-oxides

Thiosemicarbazone complexes of copper(II) were shown to be potent cytotoxic/antineoplastic agents against the growth of murine and human tumor cells. Selectivity of some agents was demonstrated against specific solid tumor growth. In L1210 lymphoid leukemia cells the copper complexes preferentially inhibited DNA synthesis with their major effects on the purine de novo pathway at PRPP amido transferase, IMP dehydrogenase and dihydrofolate reductase. The reductions of purines correlated positively with inhibition of DNA synthesis and cytotoxicity of the agents tested. DNA itself was fragmented after incubation with the drug; however, no binding of the agent to nucleotide bases or intercalation between base pairs was evident.

Expression of selected human HOX-2 genes in B/T acute lymphoid leukemia and interleukin-2/interleukin-1 beta-stimulated natural killer lymphocytes

Although the key role of human homeobox (HOX) genes in development is well established, their function in adult cells is still under scrutiny. We have analyzed, in normal adult blood cell subpopulations, acute lymphoid leukemia (ALL) cells lines, and primary blasts, the RNA expression of all HOX-2 cluster genes (5'-2.5, 2.4, 2.3, 2.2, 2.1, 2.6, 2.7, 2.8, 2.9, 3') and nine genes in the HOX-1, -3, and -4 cluster by Northern blotting, RNAse protection, and/or reverse transcriptase polymerase chain reaction (RT-PCR). The analyzed HOX-1, -3, and -4 genes were never expressed in all tested cell populations. Natural killer (NK) cells activated in interleukin-2 (IL-2)/IL-1 beta-treated cultures exhibit a gradually increasing, abundant expression of three HOX-2 genes (2.2, 2.6, 2.8), while three other genes (2.3, 2.1, 2.7) are expressed at a lower level at late culture times. However, no HOX-2 gene is expressed in quiescent lymphocytes (NK, B and T [T-cell receptor (TCR) alpha/beta, gamma/delta lymphocytes, thymocytes] cells), granulocytes, and monocytes. In B- and T-ALL cell lines, HOX-2 genes are expressed according to different patterns: (1) widespread transcription (seven of nine genes, including 2.3 and 2.6) in the Peer line bearing the TCR gamma/delta; (2) expression of 2.5, 2.2, and 2.6 in the SEZ 627 line, which derives from an HTLV-1+ T-helper leukemia; (3) transcription of 2.3 and 2.6 in both the T-ALL CEM line and four B-ALL lines (interestingly, CALLA- B-ALL lines are constantly 2.3/2.6 RNA+); (4) no HOX-2 gene expression was detected in one T- and two B-ALL lines. Primary blasts from five T- and five pre-B-ALL showed selective expression of one or more HOX-2 genes, namely 2.5, 2.2, 2.6, and 2.7. Our data are compatible with the hypothesis that selected HOX-2 genes play a role in the IL-2/IL-1 beta-induced activation and/or proliferation of normal NK lymphocytes and possibly in the oncogenetic process of some T- and B-ALL.

Expression of Selected Human HOX-2 Genes in B/T Acute Lymphoid Leukemia and Interleukin-2/Interleukin-l β-Stimulated Natural Killer Lymphocytes

Abstract Although the key role of human homeobox (HOX) genes in development is well established, their function in adult cells is still under scrutiny. We have analyzed, in normal adult blood cell subpopulations, acute lymphoid leukemia (ALL) cells lines, and primary blasts, the RNA expression of all HOX-2 cluster genes (5′-2.5, 2.4, 2.3, 2.2, 2.1, 2.6, 2.7, 2.8, 2.9, 3′) and nine genes in the HOX-1, -3, and -4 cluster by Northern blotting, RNAse protection, and/or reverse transcriptase polymerase chain reaction (RT-PCR). The analyzed HOX-1, -3, and -4 genes were never expressed in all tested cell populations. Natural killer (NK) cells activated in interleukin-2 (IL-2)/IL-1 beta-treated cultures exhibit a gradually increasing, abundant expression of three HOX-2 genes (2.2, 2.6, 2.8), while three other genes (2.3, 2.1, 2.7) are expressed at a lower level at late culture times. However, no HOX-2 gene is expressed in quiescent lymphocytes (NK, B and T [T-cell receptor (TCR) alpha/beta, gamma/delta lymphocytes, thymocytes] cells), granulocytes, and monocytes. In B- and T-ALL cell lines, HOX-2 genes are expressed according to different patterns: (1) widespread transcription (seven of nine genes, including 2.3 and 2.6) in the Peer line bearing the TCR gamma/delta; (2) expression of 2.5, 2.2, and 2.6 in the SEZ 627 line, which derives from an HTLV-1+ T-helper leukemia; (3) transcription of 2.3 and 2.6 in both the T-ALL CEM line and four B- ALL lines (interestingly, CALLA- B-ALL lines are constantly 2.3/2.6 RNA+); (4) no HOX-2 gene expression was detected in one T- and two B- ALL lines. Primary blasts from five T- and five pre-B-ALL showed selective expression of one or more HOX-2 genes, namely 2.5, 2.2, 2.6, and 2.7. Our data are compatible with the hypothesis that selected HOX- 2 genes play a role in the IL-2/IL-1 beta-induced activation and/or proliferation of normal NK lymphocytes and possibly in the oncogenetic process of some T- and B-ALL.

The anti‐neoplastic activity of ethylamine‐carboxyborane and triphenylphosphine‐carboxyborane in L‐1210 lymphoid leukemia cells

AbstractStudies on the mode of action of two boroncontaining anti‐neoplastic agents, ethylamine‐carboxyborane and triphenylphosphine‐carboxyborane, are reported. The major site of inhibition was in the pyrimidine de nove synthetic pathway at orotidine monophosphate decarboxylase activity. Additional sites which may facilitate the inhibition of cell growth were IMP dehydrogenase, thymidine kinase, TMP kinase and TDP kinase, m‐RNA, r‐RNA and t‐RNA polymerase activities as well as topoisomerase II activity. The reduction in enzyme activities led to sufficient reduction of d(NTP) levels to suppress DNA synthesis and cell growth. DNA strand scission was evident in the presence of drug. Multiple modes of action are common with amine‐carboxyboranes. Acute toxicity studies in mice showed that both agents were safe in their therapeutic range based on organ weights, histological tissue sections, clinical chemistry values and hematopoietic parameters.

High expression of beta-adrenergic receptor kinase in human peripheral blood leukocytes. Isoproterenol and platelet activating factor can induce kinase translocation.

Receptor phosphorylation is a key step in the process of desensitization of the beta-adrenergic and other related receptors. A selective kinase (called beta-adrenergic receptor kinase, beta ARK) has been identified which phosphorylates the agonist-occupied form of the receptor. Recently the bovine beta ARK cDNA has been cloned and the highest levels of specific mRNA were found in highly innervated tissues. It was proposed that beta ARK may be primarily active on synaptic receptors. In the present study, the cDNA of human beta ARK was cloned and sequenced. The sequence was very similar to that of the bovine beta ARK (the overall amino acid homology was 98%). Very high levels of beta ARK mRNA and kinase activity were found in peripheral blood leukocytes and in several myeloid and lymphoid leukemia cell lines. Since agonist-induced beta ARK translocation is considered the first step involved in beta ARK-mediated homologous desensitization, we screened a number of G-protein-coupled receptor agonists for their ability to induce beta ARK translocation. In human mononuclear leukocytes, beta-AR agonist isoproterenol and platelet-activating factor were able to induce translocation of beta ARK from cytosol to membrane. After 20 min of exposure to isoproterenol (10 microM), the cytosolic beta ARK activity decreased to 61% of control, while membrane-associated beta ARK activity increased to 170%. 20-min exposure to platelet-activating factor (1 microM) reduced the cytosolic beta ARK activity to 42% of control with concomitant increase in membrane beta ARK activity to 214% of control. The high levels of beta ARK expression in human peripheral blood leukocytes together with the ability of isoproterenol and platelet-activating factor to induce beta ARK translocation, suggest a role for beta ARK in modulating some receptor-mediated immune functions.

Stimulation of growth of human malignant lymphoma and lymphoid leukemia cells in vitro.

We have analyzed cultures of malignant lymphoma cells and cells from patients with acute lymphoid leukemia in methylcellulose for their ability to from colonies. Clonogenic growth was examined in the presence or absence of fetal calf serum (FCS), platelet-poor plasma (PPP), medium conditioned by phytohemagglutinin-stimulated leukocytes (PHA-LCM), or irradiated allogeneic bone marrow stroma cells. Cells from 25 lymphoma patients--17 with non-Hodgkin's lymphoma (NHL), eight with Hodgkin's disease (HD)--and from 19 patients with acute lymphocytic leukemia (ALL) were investigated. We show that colony growth can be obtained in a minority of cases (in 3 NHL, 5 HD, and 2 ALL) and that the use of FCS and allogeneic irradiated stroma cells may be required for optimal colony formation.

A gene coding for a zinc finger protein is induced during 12-O-tetradecanoylphorbol-13-acetate-stimulated HL-60 cell differentiation.

ETR103 cDNA was cloned as an immediate early gene in the course of macrophagic differentiation of HL-60 cells stimulated by TPA (12-O-tetradecanoylphorbol-13-acetate). The induction by TPA was immediate-early (within 30 min) and transient. This gene was not induced by vitamin D3 or by retinoic acid, which stimulates differentiation of HL-60 cells to the monocytic or granulocytic lineage, respectively. The ETR103 mRNA was induced by TPA in lymphoid or myeloid leukemia cell lines of several maturation stages. The induction by TPA seems to proceed by a protein kinase C-mediated mechanism, on the basis of the results obtained by using protein kinase C inhibitor (H-7), protein kinase C activator (diC8), and an activator of protein kinase A (dibutyryl cAMP). Okadaic acid, an inhibitor of protein phosphatases, also induced the ETR103 mRNA expression. The nucleotide sequence of the ETR103 cDNA reveals that ETR103 encodes a human zinc finger-containing transcription factor identical to Egr-1 and 225, which is homologous to mouse Egr-1, Zif/268, Krox-24, and TIS8, or to rat NGFI-A.

The anti-neoplastic activity of 2,3-dihydrophthalazine-1,4-dione and N-butyl-2,3-dihydrophthalazine-1,4-dione in human and murine tumor cells.

2,3-Dihydrophthalazine-1,4-dione derivatives demonstrated potent cytotoxicity against the growth of murine leukemia cells and human single cell suspension, i.e. Tmolt3 leukemia and HeLa-S3, as well as colon adenocarcinoma and KB nasopharynx. However, only select compounds demonstrated activity against bronchogenic lung, osteosarcoma and glioma growth. 2,3-Dihydrophthalazine-1,4-dione was active in vivo against L1210 leukemia, Lewis lung and Ehrlich ascites carcinoma growth. In L1210 cells the agents inhibited both DNA and RNA synthesis, and a few of the compounds were capable of inhibiting protein synthesis at 3 times their ED50 values. When 2,3-dihydrophthalazine-1,4-dione and N-butyl-2,3-dihydrophthalazine-1,4-dione were examined for their mode of action in the L1210 lymphoid leukemia cells, the sites of inhibition by the agents appear to be the de novo purine pathway at the enzymes IMP dehydrogenase and PRPP amido transferase. IMP dehydrogenase activity was inhibited at least 45% by 45 min at 100 microM concentration of drugs whereas the remaining enzymes that were affected by the drugs were not inhibited as early. Secondary sites were dihydrofolate reductase and thymidylate synthetase. The d(NTP) levels were also reduced specifically dATP and dCTP levels.

Retinyl esters in chylomicron remnants inhibit growth of myeloid and lymphoid leukaemic cells

We have studied the effects of retinyl esters in chylomicron remnants on cell growth and differentiation of myeloid and lymphoid leukaemic cells. Ten mumol l-1 retinyl ester in chylomicron remnants effectively reduced proliferation of the myeloid leukaemic cell lines HL60, U937 and KG-1, and induced differentiation of 68% and 53% of the HL60 and U937 cells, respectively, in 5 days. While no effect on cell growth of the lymphoid cell lines Daudi, Raji and SOS was observed, 10 mumol 1-1 retinyl esters in chylomicron remnants reduced the growth of the B lymphoid cell line Reh by more than 50%. Primary cell cultures from six patients with acute leukaemia (four non-lymphocytic and two lymphocytic) were incubated with chylomicron remnant retinyl esters and proliferation was measured by means of thymidine incorporation. Among the myeloid leukaemic cells, the monomyelocytic, the two promyelocytic and the monoblastic leukaemic cells were growth inhibited. Chylomicron remnants had no effect on the growth of the c-ALL primary culture, but reduced proliferation of the T-ALL primary culture by approximately 20% after 48 h. These data suggest that high doses of retinol may be used in the treatment of some forms of acute leukaemia.

Differentiation in mature T lymphoid leukemia cells is unstable and reversible to myeloid cells, without the involvement of a common stem cell.

Mechanisms for the unstable and reversible differentiation of a mature CD3+ CD7+TCR-alpha/beta+ T lymphoid leukemia into the myeloid lineage were investigated. Inasmuch as productive rearrangement of the TCR-alpha is a late determinant of T cell differentiation, the TCR-alpha rearrangement was sequenced to determine the state of differentiation of the leukemic multipotent cell. An identical productive rearrangement of J alpha C to a novel V alpha region was found in the myeloid and T lymphoid leukemic cells. Thus, a "terminally" differentiated T lymphoid leukemic cell after productively rearranging TCR-alpha and -beta continues to display potential for multilineage differentiation. Therefore, multilineage potential is due to an unstable and reversible differentiation in a mature T lymphocyte as opposed to differentiation of an uncommitted common T and myeloid precursor cell.

Increased expression of laminin receptors during myeloid differentiation

Variation of laminin receptor levels (LNR) during myeloid-cell differentiation and in acute leukemia were investigated by 125I-laminin-binding determination during HL60 cell differentiation and in cells of patients with different types of leukemia, characterized according to the FAB classification. LNR levels in HL60 cells increased during differentiation, being significantly higher in cells exposed to phorbol myristate acetate (PMA) and ethanol (55,391 +/- 27,845 and 29,314 +/- 6,435 sites/cell respectively) as compared with HL60 controls (8,549 +/- 4,000 sites/cell). The control cells do not adhere to laminin-coated surfaces, but differentiation with PMA results in their rapid adherence on this substratum. Short treatment with PMA does not increase the number of adherent cells or the receptor expression. Granulocytes also presented equally high LNR concentration (29,739 +/- 13,516 sites/cell). The lymphoid cells (lymphocyte, acute lymphoid leukemia and chronic lymphocytic leukemia) shared low LNR numbers (less than 6,500 sites/cell). Myeloid cells displayed a wide range of LN receptors with higher levels being associated with the more differentiated FAB subgroups. 125I-laminin binding to lymphoid or myeloid leukemic cells was mainly inhibited by P1 fragments, whereas granulocytes and differentiated HL60 cells displayed a dual binding pattern for laminin fragments P1 and E8. These results were confirmed by assays using 125I-labelled P1 and E8 fragments. We conclude that magnitude of LNR levels and variation in expression of P1 and E8 receptors appear to be linked to lineage and maturation status in hematopoietic cells.

Significance of aberrant immunophenotypes in childhood acute lymphoid leukemia

Leukemic cells from 51 pediatric patients (younger than 18 years) diagnosed with acute lymphoid leukemia by standard morphologic and cytochemical methods were subjected to flow cytometric studies using a panel of monoclonal antibodies against T-cell (CD1, 2, 3, 4, 5, 7, 8), B-cell (CD10, 19, 20, 21), myeloid (CD13, 14, 15, 33), and HLA-DR antigens. Cases of "conventional" acute lymphoid leukemia (leukemic cells with a normal configuration of B-cell or T-cell differentiation antigens) were observed in 26 of 51 (51%) cases, whereas cases of "aberrant" acute lymphoid leukemia (cells with abnormal patterns of B-cell or T-cell antigens or with concomitant myeloid antigens) were noticed in 25 (49%) cases. Myeloid antigen-positive acute lymphoid leukemia was observed in the leukemic cells of eight (16%) individuals. No significant differences were observed between conventional and aberrant ALL in the distribution of sex, age, leukocyte count, hemoglobin concentration, platelet count, blast count, French-American-British (FAB) type, lymphadenopathy, organomegaly, rate or duration of remission, or survival. When only myeloid antigen-positive cases were compared with myeloid antigen negative-cases, no significant correlations were observed except for duration of first remission (myeloid antigen positive, 26+ +/- 22 months; myeloid antigen negative, 40+ +/- 18 months; P less than 0.001), and duration of survival (myeloid antigen positive, 27+ +/- 24 months; myeloid antigen negative, 62+ +/- 17 months; P = 0.001). These data suggest that pediatric patients with ALL blasts possessing myeloid antigens may represent a high-risk group for length of remission and survival.

Comparison of antineoplastic activity of 2',2'-difluorodeoxycytidine and cytosine arabinoside against human myeloid and lymphoid leukemic cells

2',2'-difluorodeoxycytidine (known as dFdC, Gemcitabine and LY188011) is a new analog of deoxycytidine which has demonstrated excellent antineoplastic activity against many kinds of solid tumors and leukemic cell lines. We were interested in the comparison of the antineoplastic activity of this new antimetabolite with cytosine arabinoside (ARA-C) against HL-60 myeloid, RPMI-8392 B-lymphoid and Molt-3 T-lymphoid leukemic cell lines. Our in vitro experiments showed that dFdC was a more potent cytostatic drug than ARA-C against all the leukemic lines with IC50 ranging from 3 to 10 nM for dFdC and from 26 to 52 nM for ARA-C for a 48 h exposure. The cytotoxicity of both drugs was evaluated by clonogenic assay and dFdC was found to be 100 times more potent than ARA-C against all the leukemic cell lines for both a 2 h and a 24 h exposure. The recovery of DNA synthesis after drug removal was much slower for dFdC than for ARA-C. However, in contrast to cytostatic and cytotoxicity results ARA-C was a more potent inhibitor of DNA synthesis than dFdC for all the leukemic cell lines for short exposure. Uptake and elimination of the drugs showed that dFdC accumulated to a higher degree in the leukemic cells than ARA-C and that elimination of this difluoro analog was slower than that of ARA-C. These results indicate that dFdC has more potent in vitro antileukemic activity than ARA-C.

Elevated Expression of the C-fes Proto-oncogene in Adult Human Myeloid Leukemia Cells in the Absence of Gene Amplification

Expression of the 93-kd tyrosine kinase encoded by the human c-fes proto-oncogene (also known as FES) is restricted to mature hematopoietic cells of the granulocytic and monocytic lineages, suggestive of a function essential to normal myeloid differentiation. However, recent studies have shown that c-fes can transform fibroblasts if sufficient levels of gene expression are achieved. These findings indicate that strict regulation of the c-fes gene is critical to normal myeloid development, whereas elevated c-fes expression may contribute to malignant transformation. In the present study, we compared the c-fes messenger RNA (mRNA) levels in leukemia blasts from patients with myeloid or lymphoid leukemia with those of peripheral monocytes from a normal donor with the use of a quantitative ribonuclease protection assay. The presence of c-fes mRNA was readily detected in both acute myeloid leukemia (AML) and chronic myeloid leukemia (CML) cells, but c-fes mRNA was present in low levels or was absent in lymphoid leukemia cells. The leukemia cells of two of five AML patients and four of four CML patients expressed more c-fes mRNA than monocytes from a normal donor, with more than a threefold elevation in the cells of one CML patient. No evidence of amplification or rearrangement of the c-fes gene was detectable by Southern blot analysis of myeloid leukemia DNA, suggesting that the variation in c-fes mRNA levels are related to differences in transcriptional activity and/or message stability. These results indicate that elevated c-fes expression is a common feature of myeloid leukemia cells that could potentially contribute to the leukemia phenotype.

Detection of Minimal Residual Disease in Adult Acute Lymphoblastic Leukemia by Analysis of Gene Rearrangements and Correlation with Early Relapses

While more than 75% of the adult patients with acute lymphoblastic leukemia (ALL) achieve a complete remission after treatment with intensive chemotherapy, about 40% of them relapse within five years. These relapses are probably due to residual leukemic cells. Gene rearrangements are used as markers of clonality and thereby monoclonal leukemic lymphoid cells can be detected with high sensitivity. In this study, we have applied the analysis of gene rearrangements to detect minimal residual disease in patients considered to be in complete remission. Serial bone marrow samples were studied in 35 patients before and four weeks after initiation of a standardized induction chemotherapy. Gene probes for the joining regions of the human immunoglobulin heavy chain and the constant regions of the human T-cell receptor β-chain were used. In five of the 35 patients, the same gene rearrangements found before therapy persisted and indicated residual disease. Four of them relapsed within a median time of 10 weeks. Six of the 30 other patients without detectable gene rearrangements after induction therapy also relapsed, but median time to relapse was 30 weeks. Two of them had a relapse in the central nervous system without detectable bone marrow infiltration. Our data suggest that minimal residual disease, detected by analysis of gene rearrangements, is associated with a high early relapse rate. Analysis of gene rearrangements at the time of assessing the response to primary therapy seems to be of prognostic value in ALL and may contribute to a stratification of further therapy.

Effect of ( dl)-5-methyltetrahydrofolate on lymphoid leukemia cell lines

We studied the effect of ( dl)-5-methyltetrahydrofolate (mTHF) on the tymphoid cell lines BALM 3, CCRF-SB, CEM, Daudi, MOLT 4 and P3HR1, employing doses in the mM range. The growth of all the lines studied was inhibited by mTHF in a dose-dependent fashion. mTHF demonstrated a substantial cytocidal effect on leukemic lymphoid cells of up to 3 log, as measured by limiting dilution analysis, at a concentration of 10 −3 M. At this dose normal human lymphocyte viability was not affected, and their mitogen-induced proliferation was only slightly impaired. We tested the effect of high doses of mTHF on a clone of CEM cells (CEM-MTXr) infected with the pSDHT retrovirus able to transduce a dominant-acting gene encoding a mutant, less efficient, dihydrofolate reductase. CEM-MTXr cells were inhibited by high doses of mTHF to the same degree as the parental line, thus suggesting that the enzymatic reactions leading to folate reduction are not involved in the cytocidal effect of mTHF.

Identification of RB and p53 mutations in mouse lymphoma cell lines

We analyzed the genomic structure and mRNA of the RB and p53 genes in four mouse lymphoid leukemia cell lines (DL-1, DL-5, DL-8, and DL-12). Although no gross structural alteration of the RB gene was observed in any cell line, abnormalities of RB mRNA were detected in at least two cell lines. RB mRNA expression was greatly reduced in DL-12. In addition, cloning and sequencing analysis of the RB cDNA revealed that the RB mRNA in DL-8 had a 276-nucleotide deletion presumably consisting of exons 10, 11, and 12, suggesting that altered splicing resulted in the loss of these exons. Analysis of the p53 gene indicated that DL-5 had a deletion in both alleles and expressed a smaller mRNA. These results suggest that mutations of the RB or p53 genes, or both, are associated with lymphoid leukemogenesis in mice.