Correlation of umbilical cord blood haematopoietic stem and progenitor cell levels with birth weight: implications for a prenatal influence on cancer risk.

We examined the relation with birth weight and umbilical cord blood concentrations of haematopoietic stem and progenitor populations in 288 singleton infants. Across the whole range of birth weight, there was a positive relation between birth weight and CD34+CD38(-) cells, with each 500 g increase in birth weight being associated with a 15.5% higher (95% confidence interval: 1.6-31.3%) cell concentration. CD34+ and CD34+c-kit+ cells had J-shaped relations and CFU-GM cells had a U-shaped relation with birth weight. Among newborns with >or=3000 g birth weights, concentrations of these cells increased with birth weight, while those below 3000 g had higher stem cell concentrations than the reference category of 3000-3499 g. Adjustment for cord blood plasma insulin-like growth factor-1 levels weakened the stem and progenitor cell-birth weight associations. The positive associations between birth weight and stem cell measurements for term newborns with a normal-to-high birth weight support the stem cell burden hypothesis of cancer risk.

Neural precursor cells form rudimentary tissue-like structures in a rotating-wall vessel bioreactor.

We have analyzed the biology of embryonic, epidermal growth factor-responsive murine neural precursor cells cultured in the high-aspect ratio vessel (HARV). Within 2-3 d of rotary-cell culture, such cells formed multiple, macroscopic, three-dimensional structures that were orders of magnitude larger than the cellular clusters ("neurospheres") formed by these cells in conventional stationary-flask cultures. Each HARV structure was composed of a multilayered cellular shell surrounding one or more central cavities that were bordered by pyknotic cell nuclei. Although the cells in the HARV structures were more pleomorphic than those in neurospheres, the structures did not appear to represent primitive neural tumors: the formation of HARV structures by precursor cells was not an irreversible phenotypic change, and the structures did not originate from the clonal expansion of single-progenitor cells; the growth rate and invasiveness of the cells in HARVs were less than those in flasks; and HARV-cultured cells did not form tumors after subcutaneous inoculation into the flanks of NOD-scid/scid mice. Immunohistochemical analysis suggested that HARV structures might be novel "prototissues" characterized by a crude, but organized, architecture, with a surface layer of immature proliferating cells (nestin- and proliferating cell nuclear antigen-positive) that enclosed strata of more differentiated cells (beta-tubulin III- and glial fibrillary acidic protein-positive) within. Rotary-cell culture may have significant implications for the eventual utility of neural precursors for clinical neurotransplantation.

Interleukin-11 receptor expression in primary ovarian carcinomas.

OBJECTIVES: The objectives of this study were to determine (1) the frequency of expression of the interleukin-11 receptor alpha subunit (IL-11Ralpha) and its signal transducing subunit, gp130, among primary ovarian carcinomas; (2) the frequency of expression of IL-11 in ovarian carcinomas; and (3) the potential role IL-11 might have in ovarian cancer cell biology. METHODS: An immunohistochemical assay was used to determine the expression of IL-11Ralpha and the gp130 cofactor among primary ovarian carcinomas; the expression of IL-11 in ovarian malignancies was determined using reverse transcription polymerase chain reaction (RT-PCR). The ability of IL-11 to stimulate [3H]thymidine incorporation in IL-11R-expressing ovarian carcinoma cell lines (OVCAR-3 and SKOV-3) and/or abrogate cell death mediated by apoptosis-inducing agents using an ELISA assay that quantitates DNA fragmentation was also studied. RESULTS: IL-11Ralpha was expressed in the malignant epithelial cells of 45 of 48 (93.8%) primary ovarian carcinoma samples studied. In 45 primary ovarian carcinoma samples where both components of the IL-11 receptor (IL-11Ralpha and gp130) were examined, coexpression was observed in 42 (93.3%). Expression of the IL-11 receptor components was also found in the stromal layer. Coexpression of IL-11Ralpha and gp130 was commonly observed in both benign ovarian tumors and in the epithelial layer of normal ovaries. In contrast, IL-11 mRNA was expressed in only 3 of 21 malignant samples studied (14.3%). Recombinant human IL-11 was unable either to stimulate [3H]thymidine incorporation or to block cell death effected by paclitaxel or Fas-activating antibodies in in vitro assays using OVCAR -3 or SKOV-3 cells. CONCLUSIONS: The IL-11 receptor system is commonly expressed in both malignant and nonmalignant ovarian tissues, although its function in ovarian epithelial cell biology remains unclear.

Increased expression of the interleukin-11 receptor and evidence of STAT3 activation in prostate carcinoma.

Previous investigations have shown that interleukin-6, a member of the JAK-STAT activating family of cytokines, plays an important role in prostate carcinoma. Here we demonstrate the co-expression of another member of this cytokine family, interleukin-11 (IL-11), and components of its receptor (interleukin-11 receptor; IL-11R), ie, IL-11Ralpha (involved in ligand recognition), and gp130 (involved in signal transduction) in cultured normal and malignant prostate-derived epithelial cell lines. In the DU-145 prostate carcinoma cell line, rhIL-11 stimulates a transient and dose-dependent increase in the tyrosine 705-phosphorylated, active form of STAT3 (STAT3 P-Tyr705), involved in the downstream signaling of IL-11R and other members of the gp130-dependent receptors. The ability of IL-11 to activate STAT3 in prostate-derived cells may be mechanistically important, given recent data suggesting that constitutively activated STAT3 may be associated with the malignant phenotype. In 51 human primary tissues derived from normal prostate, benign prostatic hyperplasia, and prostate carcinomas, IL-11Ralpha and gp130 were commonly expressed, with a statistically significant elevation in the expression of IL-11Ralpha in prostate carcinoma. Also, the tyrosine-phosphorylated, activated form of STAT3 was observed more prominently in the nuclei of cells residing in malignant glands compared to those in nonmalignant samples. Thus, the IL-11 receptor system is up-regulated in prostate carcinoma, and may be one part of a cytokine network that maintains STAT3 in its activated form in these tissues.

Coexpression of granulocyte colony stimulating factor and its receptor in primary ovarian carcinomas.

Immunohistochemistry was used to determine the expression of granulocyte colony-stimulating factor (G-CSF) and its receptor (G-CSFR) in primary ovarian carcinomas. The expression of G-CSFR was observed in the malignant cells of each of the 46 primary carcinomas examined; G-CSF was coexpressed in both the malignant epithelial cells and the stroma of 56.5% of the specimens. Thus the majority of ovarian carcinomas harbor both potential autocrine and paracrine G-CSF axes. In 37% of the samples, G-CSF was expressed only within stromal cells, suggesting that only a potential paracrine system is in place. In a preliminary, retrospective, evaluation, the survival of patients whose tumors expressed only the apparent paracrine loop was significantly worse than patients whose tumors expressed both potential autocrine and paracrine G-CSF-based regulatory loops (14.5 vs. 42.5 months, respectively). Studies on the potential function of G-CSF were performed using the G-CSFR-expressing OVCAR-3 ovarian carcinoma line. As a single agent, rhG-CSF failed to stimulate [3H]-thymidine incorporation in these cells, but enhanced the mitogenic action of epidermal growth factor (EGF) in a dose-dependent manner. Thus, potential autocrine and/or paracrine loops involving G-CSF and its receptor occur in over 90% of primary ovarian carcinomas, and may act to modulate the action of growth factors.

Expression of multiple angiogenic cytokines in cultured normal human prostate epithelial cells: predominance of vascular endothelial growth factor.

The cytokines that regulate angiogenesis in normal and malignant prostate tissue are not well studied. Using an RT-PCR-based screen, we observed that cultured, low-passage normal human prostate epithelial cells (PrECs) express a variety of cytokines which have been shown to have angiogenic and/or endothelial cell-activating properties in various systems. These include vascular endothelial growth factor (VEGF), basic fibroblastic growth factor (bFGF), transforming growth factor-alpha (TGF-alpha), transforming growth factor-beta (TGF-beta), interleukin-8 (IL-8), tumor necrosis factor-alpha (TNF-alpha), granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF). Expression of VEGF, bFGF, GM-CSF, G-CSF, TGF-alpha and TNF-alpha in these cells was confirmed by immunohistochemistry. Culture medium conditioned by normal human PrECs for periods of up to 96 hr were found to contain VEGF, GM-CSF, G-CSF, IL-8, TGF-beta1 and TGF-beta2 but not TNF-alpha or bFGF, as determined by ELISA. Of these, VEGF was by far the most prominently expressed angiogenic cytokine (approx. 2,500 pg/ml conditioned medium at 96 hr vs. 30 to 100 pg/ml conditioned medium for the other cytokines). PrEC-conditioned medium induced an approximately 2-fold stimulation of [3H]-thymidine incorporation in cultured human umbilical cord endothelial cells (HUVECs) deprived of the endothelial growth factors VEGF and bFGF; this stimulation was abolished by neutralizing antibodies directed against VEGF but not bFGF, IL-8, GM-CSF or TNF-alpha. VEGF expression by PrECs was not markedly altered by administration or deprivation of other angiogenic cytokines for which these cells have receptors, suggesting that there is not a hierarchy of cytokines controlling its expression; however, retinoic acid, a component of PrEC growth medium, was found to modestly suppress VEGF at physiological concentrations (0.1 ng/ml). These data suggest that normal PrECs express a variety of angiogenic cytokines, most prominently VEGF, to recruit a supporting vasculature, even in culture. Our data also suggest that the ability of malignant PrECs to stimulate angiogenesis may be intrinsic and does not need to be acquired during oncogenesis.

Expression of methylthioadenosine phosphorylase cDNA in p16-, MTAP- malignant cells: restoration of methylthioadenosine phosphorylase-dependent salvage pathways and alterations of sensitivity to inhibitors of purine de novo synthesis.

5'-Deoxy-5'-methylthioadenosine phosphorylase (MTAP) is involved in the salvage of adenine and methylthio moieties of 5'-deoxy-5'-methylthioadenosine, a byproduct of polyamine synthesis, to adenine nucleotides and methionine, respectively. The gene encoding MTAP, MTAP, is frequently codeleted along with the tumor suppressor gene p16 in malignant cells bearing homozygous deletions in the chromosome 9p21 region. p16-, MTAP- malignant cells have been shown to be more susceptible to the purine de novo inhibitory actions of antifolates such as methotrexate than are p16+, MTAP+ cells. To understand the underlying mechanism, we reintroduced MTAP activity into two p16-, MTAP- cell model systems, the MiaPaCa-2 and PANC-1 human pancreatic carcinoma cell lines, by transfection with MTAP cDNA. It was found that transfection with MTAP cDNA (i) restored both the MTAP-dependent adenine and methionine salvage pathways, (ii) decreased the rates of purine de novo synthesis (18-47% lower than the wild-type or sham-transfected counterparts), and (iii) decreased cellular sensitivity to the antipurine-related growth-inhibitory actions of methotrexate and azaserine. These data support the hypothesis that operation of the MTAP-dependent adenine salvage pathway renders MTAP+ cells less dependent on de novo purine synthesis and hence less susceptible than MTAP- malignant cells to the growth-inhibitory actions of agents (e.g. antifolates) whose mechanism of action in part involves the de novo purine pathway. These findings provide a theoretical basis for the relatively selective action certain antifolates may have against MTAP-deficient malignancies.

Gene deletion chemoselectivity: codeletion of the genes for p16(INK4), methylthioadenosine phosphorylase, and the alpha- and beta-interferons in human pancreatic cell carcinoma lines and its implications for chemotherapy.

Pancreatic carcinoma cells lines are known to have a high incidence of homozygous deletion of the candidate tumor suppressor gene p16 (MTS1/CDKN2), which resides in the chromosome 9p21 region. Here we: (a)examined a series of these cell lines for the incidence of codeletion of genes located near p16, in particular, the gene for the enzyme 5'-deoxy-5'-methylthioadenosine phosphorylase (MTAP) and the genes of the IFN-alpha and -beta cluster (IFNs); and (b) investigated whether therapeutic strategies could be developed that target malignant cells that have undergone the codeletion of such genes. Five of the eight pancreatic carcinoma cell lines were p16(-), MTAP was codeleted in all five cases. Because MTAP phosphorolyzes 5'-deoxy-5'-methylthioadenosine (MTA), generated as a byproduct of polyamine synthesis, to the salvageable purine base adenine, loss of this pathway in p16(-), MTAP(-) cells might sensitize these cells to methotrexate (MTX), the mechanism of action of which involves, in part, an inhibition of purine de novo synthesis. MTAP(+) normal keratinocytes and pancreatic carcinoma lines had relatively poor sensitivity, in terms of efficacy, to the purine nucleotide-starving actions of MTX. This may be in part due to the MTAP-dependent salvage of adenine moieties from endogenously generated MTA, because the MTAP inhibitor 5'-chloro-5'-de- oxyformycin A potentiates the antipurine actions of MTX in some of these MTAP(+) lines. Also, exogenous MTA (10 microM) reverses the growth-inhibitory actions of MTX in these lines. In contrast, MTAP(-) cell lines, which cannot recycle purines from endogenous MTA, have a relatively high sensitivity to the antipurine actions of MTX, which is not modulated by 5'-chloro-5'-deoxyformycin A or exogenous MTA. Thus the MTAP loss in malignant cells may be an example of gene deletion chemoselectivity, in which genetic deletions that occur as part of the oncogenic process render these cells more sensitive to particular anticancer agents than normal cells, which have not undergone such deletions. We also examined whether the loss of IFN genes sensitize cells to the growth-inhibitory actions of these cytokines. Three of the five p16(-) cell lines bore homozygous deletions of IFNA1 and IFNB1 genes, representing each end of the IFN-alpha,-beta gene cluster; one cell line bore a codeletion of the IFNA1 gene but retained the IFNB1 locus. Whereas the cell lines that were most sensitive to the growth-inhibitory effects of IFN-beta or IFN-alpha(2b), tended to be those with IFN deletions, there were enough exceptions to this pattern to indicate that the IFN genotype does not reliably predict IFN responsiveness.

Codeletion of the genes for p16INK4, methylthioadenosine phosphorylase, interferon-alpha1, interferon-beta1, and other 9p21 markers in human malignant cell lines.

In this study, 27 malignant cell lines, including leukemias, gliomas, and lung and bladder carcinomas were screened for homozygous deletions of the putative tumor suppressor gene p16 (MTS1/CDK4I/CDKN2) and other markers within the chromosome 9p21 region; these include the genes for interferon-alpha1 (IFNA1), interferon-beta1 (IFNB1), methylthioadenosine phosphorylase (MTAP), and two microsatellite markers, D9S171 and D9S169. The purpose of this study was to determine the incidence of codeletion of these markers. Screening for homozygous deletions was carried out using direct polymerase chain reaction of genomic DNA, or, in the case of MTAP, a functional enzyme assay. Of these cell lines, 14 (52%) were found to have homozygous deletions of the p16 gene. Two of the 14 p16-negative cell lines (14%) were found to have homozygous deletions within the p16 domain but but no other 9p21 marker. MTAP was codeleted in 12 of the 14 p16-negative cell lines (86%), whereas IFNA1 was codeleted with p16 in eight of these lines (57%); IFNB1 was codeleted in five (36%) of the p16-deleted cell lines. The D9S171 marker, which may lie greater than 3 cM centromeric to p16, is codeleted in three cell lines (21%); the D9S169 marker, which maps even further toward the centromere, was codeleted in only one cell line (7%). Loss of any 9p21 marker, e.g., MTAP or IFNA1, were invariable predictors of the loss of the p16 gene. In addition, loss of IFNA1 always predicted a loss of MTAP (eight of eight cell lines), although loss of MTAP did not always predict a loss of IFNA1 (four of 12 MTAP-deleted cell lines did not have homozygous deletions of IFNA1). Thus loss of nearby genes occurs in a high percentage of cell lines that bear homozygous deletions of the p16 locus. Codeletion of MTAP or IFN in p16-negative malignant cells is of interest, as loss of these genes may influence the biologic behavior of these cells and render them susceptible to certain therapeutic approaches.

Site-directed mutagenesis of the rat m1 muscarinic acetylcholine receptor. Role of conserved cysteines in receptor function.

There are 9 cysteine residues in the rat m1 muscarinic acetylcholine receptor (mAChR) that are conserved among all five mammalian mAChR subtypes sequenced to date. To study the role of these cysteines in rat m1 mAChR function, site-directed mutagenesis was used to convert each Cys residue to Ser, and the mutant receptor genes were transfected into mAChR-deficient Chinese hamster ovary (CHO) cells. Substitution of Cys391 (extracellular loop III) or Cys421 and Cys435 (carboxyl terminus) produces receptors with wild type phenotype. Cells transfected with Ser98 or Ser178 (extracellular loops I and II, respectively) receptor genes display no carbachol-mediated hydrolysis of phosphoinositides (PI), and membranes prepared from these cells do not bind the muscarinic antagonist [3H] quinuclidinyl benzilate, even though the cells express transcripts of the m1 mAChR as determined by RNA hybridization analysis. Since biochemical evidence suggests that these cysteines form a disulfide bridge (Curtis, C. A. M., Wheatley, M., Bansal, S., Birdsall, N. J. M., Eveleigh, P., Pedder, E. K., Poyner, D., and Hulme, E. C. (1989) 264, 489-495), our findings imply that this disulfide linkage may be critical for formation of the ligand binding domain or for proper protein folding. The Ser394 mAChR (extracellular loop III) exhibits a 44% decrease in efficacy for carbachol-mediated stimulation of PI hydrolysis relative to the wild type receptor, but displays normal ligand binding affinities. The Ser407 m1 mAChR (transmembrane helix VII) displays a decreased efficacy for eliciting carbachol-mediated PI hydrolysis (39% that of CHO cells transfected with the wild type receptor) and a 4-fold shift to the right in the carbachol dose-response curve, which is consistent with the 4-fold decrease in carbachol affinity at the Ser407 m1 mAChR. In contrast, the Ser417 m1 mAChR (transmembrane helix VII) displays an increase in carbachol affinity and a shift to the left in the carbachol dose-response curve for PI hydrolysis. These findings suggest that cysteine residues in the seventh transmembrane helix of the m1 mAChR may influence agonist binding and the efficiency of receptor activation.

Posttranscriptional regulation of c-myc proto-oncogene expression and growth inhibition by recombinant human interferon-beta ser17 in a human colon carcinoma cell line.

Recombinant human interferon-beta ser17 (IFN-beta ser17), a cytokine that exhibits both antiviral and antiproliferative activity against a wide variety of cell types, causes a time- and dose-dependent inhibition of monolayer growth and of the expression of the c-myc proto-oncogene in DLD-1 Clone A human colon-carcinoma cells. The suppression of c-myc expression mediated by IFN-beta ser17 is due to a posttranscriptional destabilization of c-myc mRNA rather than to an inhibition of c-myc mRNA transcription. There is evidence suggesting that the selective reduction in the half-life of c-myc mRNA in IFN-beta ser17-treated cells occurs through an increase in the activity of the 2',5'-oligoadenylate synthetase/RNase L [2',5'-oligo (A) synthetase] pathway in DLD-1 Clone A cells. Cotreatment of these cells with IFN-beta ser17 and the anticancer agent N-methylformamide leads to the partial abrogation of 2',5'-oligo (A) synthetase activity and the stabilization of c-myc mRNA. These findings suggest that there is a correlation between the IFN-beta ser17-mediated suppression of c-myc expression and the induction of 2',5'-oligo (A) synthetase activity in DLD-1 clone A cells.

5'-deoxy-5'-methylthioadenosine phosphorylase--V. Acycloadenosine derivatives as inhibitors of the enzyme.

Various adenosine acyclonucleoside derivatives were tested as inhibitors of 5'-deoxy-5'-methylthioadenosine (MeSAdo) phosphorylase, an enzyme involved in the salvage of adenine and methionine from MeSAdo. The 2-halogenated derivatives of acyloadenosine [9-(2-hydroxyethoxy-methyl)adenine], including the chloro-, bromo- and iodo-congeners, all inhibited murine Sarcoma 180 (S180) MeSAdo phosphorylase, with Ki values in the range of 10(-6) to 10(-5) M. Halogenated derivatives of 9-(1,3-dihydroxy-2-propoxymethyl)adenine, which more closely resemble the natural substrate, were substantially more potent inhibitors of the enzyme, with Ki values in the range of 2-7 x 10(-7) M. 5'-Methylthio and 5'-halogenated analogs of 2'-deoxy-1',2'-seco-adenosine were weak inhibitors, with Ki values of 10(-4) M or greater. 9-[(1-Hydroxy-3-iodo-2-proxy)methyl]adenine. (HIPA), the derivative with the lowest Ki values among these analogs, was a competitive inhibitor of S180 MeSAdo phosphorylase. In preliminary studies, HIPA inhibited MeSAdo phosphorylase in intact HL-60 human promyelocytic leukemia cells, as it limited the incorporation of [8-14C]MeSAdo into cellular adenine nucleotide pools. In addition, 9-(phosphonoalkyl)adenines, representing potential multisubstrate inhibitors of MeSADo phosphorylase, were synthesized. Of these the heptyl derivative was the most potent inhibitor, with a Ki of 1.5 x 10(-5) M at low (3.5 mM) phosphate concentrations. The inhibitory effects of these analogs could be ablated at high phosphate concentrations (50 mM), suggesting that they interact with the phosphate binding site on the enzyme. Some of these novel MeSAdo phosphorylase inhibitors may have a role in cancer chemotherapy as potentiators of agents that block purine de novo synthesis, e.g. antifolates and 6-methylmercaptopurine ribonucleoside.

Reversible suppression of c-myc expression in a human colon carcinoma line by the anticancer agent N-methylformamide.

The anticancer agent N-methylformamide (NMF), which at high concentrations (170 mM) induces cultured DLD-1 Clone A human colon carcinoma cells to increase their doubling times and lose their tumorigenicity in nude mice (Cordeiro, R.F., and Savarese, T.M. Cancer Res., 46: 1297-1305, 1986), suppresses the expression of the c-myc protooncogene in these cells in a dose- and time-dependent manner. This suppression involves an inhibition of c-myc transcription rather than an increased degradation of c-myc mRNA, and is reversed if NMF is removed from the culture medium. Expression of the glyceraldehyde 3-phosphate dehydrogenase gene, which is thought to be constitutive, is phosphate dehydrogenase gene, which is thought to be constitutive, is relatively unaffected by NMF treatment. The NMF-mediated decrease in c-myc expression may be associated with the ability of this agent to increase the doubling time of these cells, but there is no direct temporal link between the loss of c-myc expression and the NMF-induced loss of tumorigenicity.

Substrate specificities of 5'-deoxy-5'-methylthioadenosine phosphorylase from Trypanosoma brucei brucei and mammalian cells.

The separation by chromatofocusing of two distinct purine nucleoside cleaving activities from crude extracts of Trypanosoma brucei brucei is described. One catalyzes the reversible phosphorolysis of 5'-deoxy-5'-methylthioadenosine (MeSAdo) and adenosine (Ado) and was designated an MeSAdo/Ado phosphorylase, while the other catalyzes the hydrolysis of adenosine, inosine, and guanosine but not MeSAdo. The substrate specificity of trypanosomal MeSAdo/Ado phosphorylase differed from that of a mammalian MeSAdo phosphorylase (derived from murine Sarcoma 180 cells) in that it was able to phosphorolyze 2'-deoxyadenosine, 3'-deoxyadenosine and 2',3'-dideoxyadenosine. In addition, the trypanosomal phosphorylase was able to utilize the nucleoside analog, 6-methylpurine 2'-deoxyribonucleoside, as an alternative substrate, whereas the mammalian enzyme could not. Because of these differences, cytotoxic analogs of MeSAdo may be designed that are selectively activated by the trypanosomal MeSAdo/Ado phosphorylase.

5'-deoxy-5'-methylthioadenosine phosphorylase--IV. Biological activity of 2-fluoroadenine-substituted 5'-deoxy-5'-methylthioadenosine analogs.

5'-Deoxy-5'-methylthioadenosine phosphorylase (MTAPase) phosphorolyzes 5'-deoxy-5'-methylthioadenosine (MTA) generated during polyamine biosynthesis to adenine and 5-methylthioribose-1-phosphate. Two doubly-substituted, 2-fluoroadenine-containing analogs of MTA, 5'-deoxy-2-fluoroadenosine (5'-dFAdo) and 5'-deoxy-5'-iodo-2-fluoroadenosine (5'-IFAdo), were synthesized and studied as substrates of MTAPase: their reaction with this enzyme resulted in the liberation of the cytotoxic base, 2-fluoroadenine, as well as potentially cytotoxic analogs of 5-methylribose-1-phosphate. The activities of these MTA analogs were compared to that of the singly-substituted analog, 5'-deoxy-5'-methylthio-2-fluoroadenosine (5'-MTFAdo). The cytotoxic action of these MTA analogs depended primarily on their conversion to 2-fluoroadenine-containing nucleotides, as a cell line that contains both MTAPase and adenine phosphoribosyltransferase (APRT) activity (HL-60 human promyelocytic leukemia) readily converted these MTA analogs to 2-fluoroadenine-containing nucleotides (especially 2-fluoroadenosine triphosphate) and was highly sensitive to the growth-inhibitory effects of all three compounds (IC50 values in the 10(-8) M range), whereas cell lines lacking MTAPase (CCRF-CEM human T-cell leukemia) or APRT (HL-60/aprt1 cells) did not form analog nucleotides and were relatively insensitive to these compounds (IC50 values in the 10(-5) M range). The doubly-substituted analogs were not more growth inhibitory than 5'-MTFAdo in wild type HL-60 cells as the potent effects of 2-fluoroadenine may mask the activity of the 5-methylthioribose-1-phosphate analogs generated in the reaction of these compounds with MTAPase. 5'-dFAdo and 5'-IFAdo also were irreversible inhibitors of S-adenosylhomocysteine hydrolase, which may explain in part the weak but observable growth inhibitory action of these compounds against MTAPase-deficient cell lines.

Role of glutathione depletion in the mechanism of action of N-methylformamide and N,N-dimethylformamide in a cultured human colon carcinoma cell line.

The mechanism of the antitumor action of N-methylformamide (NMF), an agent currently undergoing clinical trials, and its congener, N,N-dimethylformamide (DMF), was examined in the DLD-1 Clone A human colon carcinoma cell line in vitro. The primary action of NMF and DMF on these cells is a depletion of cellular reduced glutathione levels which results in cytostasis. Evidence to support this hypothesis include (a) the extent of growth inhibition produced by NMF and DMF is directly proportional to the extent of depletion they effect on cellular reduced glutathione levels; (b) removal of NMF (170 mM) or DMF (103 mM) from the culture medium results in a parallel restoration of cell growth and cellular glutathione levels; (c) coaddition of the glutathione precursor, L-cysteine (0.5 mM), or certain precursors for this amino acid, significantly reverses NMF- and DMF-induced glutathione depletion and cytostasis; and (d) the specific glutathione-depleting agent, buthionine sulfoximine (7.5 mM) mimics the ability of NMF and DMF to induce cytostasis. In addition, NMF- and DMF-mediated reduced glutathione depletion accounts for the previously reported ability of these agents at high concentrations (less than 100 mM) to induce a more benign phenotype in DLD-1 Clone A human colon carcinoma cells. This is evidenced by the findings that (a) L-cysteine (0.5 mM) reverses NMF- and DMF-mediated (170 and 103 mM, respectively) increases in doubling time, decreases in saturation density, and decreases in clonogenicity; (b) buthionine sulfoximine (7.5 mM) mimics these actions of NMF and DMF; and (c) the tumorigenicity of DLD-1 Clone A cells in nude mice, which is completely eliminated by the in vitro treatment of these cells for 4 passages with 170 mM NMF prior to inoculation, is fully restored if the cells are passaged in the presence of 170 mM NMF and 0.5 mM L-cysteine. Thus, NMF- and DMF-induced depletion of cellular reduced glutathione is responsible for not only the cytostatic effect of these agents on human colon carcinoma cells but also for their ability to induce more benign characteristics in these cells.

5'-Deoxy-5'-methylthioadenosine phosphorylase--III. Role of the enzyme in the metabolism and action of 5'-halogenated adenosine analogs.

5'-Deoxy-5'-halogenated adenosines are alternative substrates for 5'-deoxy-5'-methylthioadenosine phosphorylase (MTAPase), an enzyme responsible for the metabolism of 5'-deoxy-5'-methylthioadenosine (MTA), a by-product of polyamine biosynthesis. The relative reactivity of these nucleosides with MTAPase from HL-60 human promyelocytic leukemia cells is MTA greater than 5'-deoxy-5'-fluoroadenosine (5'-FlAdo) greater than 5'-chloro-5'-deoxyadenosine (5'-ClAdo) greter than 5'-bromo-5'-deoxyadenosine (5'-BrAdo) greater than 5'-deoxy-5'-iodoadenosine (5'-IAdo). In MTAPase-containing cells, the adenine released from the 5'-halogenated adenosine was incorporated into adenine nucleotide pools; cleavage by (MTAPase appeared to be the rate-limiting step in this process. 5'-BrAdo and 5'-IAdo were growth inhibitors (EC50 values less than 10 microM) of MTAPase-containing cell lines (HL-60 human promyelocytic leukemia and the L5178Y murine lymphoblastic leukemia) but were much less active (EC50 values greater than 65 microM) against MTAPase-deficient cell lines (the CCRF-CEM human T cell leukemia and the L1210 murine leukemia). The full cytotoxicity of these compounds, therefore, appeared to be related to their phosphorolysis by MTAPase. Indirect evidence suggests that 5-halogenated ribose-1-phosphate derivatives of 5'-BrAdo or 5'-IAdo produced by the MTAPase reaction were the active metabolites of these 5'-halogenated adenosines.

Reversal by L-cysteine of the growth inhibitory and glutathione-depleting effects of N-methylformamide and N,N-dimethylformamide.

N-Methylformamide and N,N-dimethylformamide, which can induce differentiation in selected malignant cell lines, are known to increase doubling times, inhibit clonigenicity in agar, and to effect responses against particular human colon carcinomas in vivo. At concentrations which inhibit growth and clonigenicity, N-methylformamide (170 mM) and N,N-dimethylformamide (103 mM) deplete total intracellular glutathione levels of DLD-1 Clone A human colon carcinoma cells in a dose and time dependent manner. In the presence of 0.5 mM 1-cysteine, both the growth and glutathione levels of polar-solvent treated DLD-1 Clone A cells are restored. 1-Cysteine also reverses the inhibition of clonigenicity mediated by NMF. The mechanism of action of N-methylformamide and N,N-dimethylformamide against this cell line, at least in vitro, is therefore related to its effects on cysteine/glutathione metabolism. Furthermore, this evidence suggests that glutathione plays a key role in regulating the growth of these cells.

Characterization of a defect in the pathway for converting 5'-deoxy-5'-methylthioadenosine to methionine in a subline of a cultured heterogeneous human colon carcinoma.

5'-Deoxy-5'-methylthioadenosine (methylthioadenosine) is cleaved to adenine and 5-methylthioribose-1-phosphate (methylthioribose-1-P). Methylthioribose-1-P is converted to 2-keto-4-methylthiobutyrate ( ketomethylthiobutyrate ) which is transaminated to methionine. We report that one subline of a heterogeneous human colon carcinoma, DLD-1 Clone D, only forms methylthioribose-1-P from methylthioadenosine or 5'-deoxy-5'-methylthioinosine (methylthioinosine), a deaminated derivative of methylthioadenosine, whereas Clone A converts methylthioadenosine and methylthioinosine to methionine, as shown by growth studies in culture of Clone A and Clone D cells and radioactive studies utilizing [methyl-14C]methylthioadenosine or [methyl-14C]methylthioinosine in the presence of extracts of these cells lines. To characterize this defect, we utilized three protein fractions isolated from rat liver which together convert methylthioribose-1-P to ketomethylthiobutyrate . Addition of only Fraction A to Clone D sonicates restores its ability to convert methylthioadenosine to methionine. This fraction is responsible for converting methylthioribose-1-P to 5- methylthioribulose -1-phosphate; radioactive studies confirm this observation. Thus, Clone D is deficient in an enzyme contained in Fraction A; this represents a qualitative biochemical difference between the two clones derived from a single human tumor.