Decreased methionine synthesis in purine nucleoside-treated T and B lymphoblasts and reversal by homocysteine.

Purine nucleosides, which accumulate in adenosine deaminase and purine nucleoside phosphorylase deficiency, are toxic to lymphoid cells. Since adenine nucleosides inhibit S-adenosylhomocysteine hydrolase, they could potentially decrease intracellular methionine synthesis. To test this hypothesis, we measured methionine synthesis by the use of [14C]formate as a radioactive precursor in cultured human T and B lymphoblasts treated with varying concentrations of purine nucleosides; 2'-deoxycoformycin and 8-aminoguanosine were added to inhibit adenosine deaminase and purine nucleoside phosphorylase, respectively. In the T lymphoblasts methionine synthesis was inhibited approximately 50% by 10 microM of 2'-deoxyadenosine, adenine arabinoside, or 2'-deoxyguanosine. By contrast, in the B lymphoblasts methionine synthesis was considerably less affected by these nucleosides, with 50% inhibition occurring at 100 microM of 2'-deoxyadenosine and adenine arabinoside; 100 microM of 2'-deoxyguanosine yielded less than 10% inhibition. Adenosine and guanosine were considerably less potent inhibitors of methionine synthesis in both the T and B lymphoblasts. An adenosine deaminase-deficient and a purine nucleoside phosphorylase-deficient cell line, both of B cell origin, exhibited sensitivities to the nucleosides similar to those of the normal B cell lines. In both the T and B cell lines homocysteine reversed the methionine synthesis inhibition induced by the adenine nucleosides and guanosine and largely reversed that induced by 2'-deoxyguanosine. Methionine synthesis from homocysteine generates free tetrahydrofolate from 5-methyltetrahydrofolate, the main intracellular storage form of folate. We conclude that purine nucleoside toxicity may be partly mediated through (a) decreased intracellular methionine synthesis, and (b) altered folate metabolism.

Induction of HL-60 differentiation by starvation for a single essential amino acid but not by protein synthesis inhibitors.

Starvation for a single essential amino acid induced differentiation of the human promyelocytic leukemia line HL-60 into morphologically and functionally mature granulocytes. Differentiation occurred when protein synthesis was inhibited up to 90% but was not simply secondary to growth arrest or protein synthesis inhibition, because neither glucose starvation nor treatment with protein synthesis inhibitors induced differentiation. Induction of differentiation by an aminoacyl tRNA synthetase inhibitor and the effect of cycloheximide and puromycin on amino acid-starved cells suggested an important regulatory role of tRNA molecules during differentiation.

Regulation of gene expression by cyclic GMP-dependent protein kinase requires nuclear translocation of the kinase: identification of a nuclear localization signal.

We recently demonstrated that cyclic GMP (cGMP)-dependent protein kinase (G-kinase) activates the human fos promoter in a strictly cGMP-dependent manner (T. Gudi et al., J. Biol. Chem. 271:4597-4600, 1996). Here, we demonstrate that G-kinase translocates to the nucleus by an active transport mechanism which requires a nuclear localization signal (NLS) and is regulated by cGMP. Immunofluorescent staining of G-kinase was predominantly cytoplasmic in untreated cells, but intense nuclear staining appeared in 8-bromo (Br)-cGMP-treated cells. We identified a putative NLS in the G-kinase ATP binding domain which resembles the NLS of the interleukin-1alpha precursor. Fusion of the G-kinase NLS to the N terminus of beta-galactosidase produced a chimeric protein which localized to the nucleus. Mutation of a single amino acid residue (K407-->E) within the G-kinase NLS produced an enzyme with normal cGMP-dependent activity in vitro which did not translocate to the nucleus and did not transactivate the fos promoter in the presence of 8-Br-cGMP in vivo. In contrast, N-terminally truncated versions of G-kinase with constitutive, cGMP-independent activity in vitro localized to the nucleus and transactivated the fos promoter in the absence of 8-Br-cGMP. These results indicate that nuclear localization of G-kinase is required for transcriptional activation of the fos promoter and suggest that a conformational change of the kinase, induced by cGMP binding or by removal of the N-terminal autoinhibitory domain, functionally activates an otherwise cryptic NLS.

Cyclic-GMP-dependent protein kinase inhibits the Ras/Mitogen-activated protein kinase pathway.

Agents which increase the intracellular cyclic GMP (cGMP) concentration and cGMP analogs inhibit cell growth in several different cell types, but it is not known which of the intracellular target proteins of cGMP is (are) responsible for the growth-suppressive effects of cGMP. Using baby hamster kidney (BHK) cells, which are deficient in cGMP-dependent protein kinase (G-kinase), we show that 8-(4-chlorophenylthio)guanosine-3', 5'-cyclic monophosphate and 8-bromoguanosine-3',5'-cyclic monophosphate inhibit cell growth in cells stably transfected with a G-kinase Ibeta expression vector but not in untransfected cells or in cells transfected with a catalytically inactive G-kinase. We found that the cGMP analogs inhibited epidermal growth factor (EGF)-induced activation of mitogen-activated protein (MAP) kinase and nuclear translocation of MAP kinase in G-kinase-expressing cells but not in G-kinase-deficient cells. Ras activation by EGF was not impaired in G-kinase-expressing cells treated with cGMP analogs. We show that activation of G-kinase inhibited c-Raf kinase activation and that G-kinase phosphorylated c-Raf kinase on Ser43, both in vitro and in vivo; phosphorylation of c-Raf kinase on Ser43 uncouples the Ras-Raf kinase interaction. A mutant c-Raf kinase with an Ala substitution for Ser43 was insensitive to inhibition by cGMP and G-kinase, and expression of this mutant kinase protected cells from inhibition of EGF-induced MAP kinase activity by cGMP and G-kinase, suggesting that Ser43 in c-Raf is the major target for regulation by G-kinase. Similarly, B-Raf kinase was not inhibited by G-kinase; the Ser43 phosphorylation site of c-Raf is not conserved in B-Raf. Activation of G-kinase induced MAP kinase phosphatase 1 expression, but this occurred later than the inhibition of MAP kinase activation. Thus, in BHK cells, inhibition of cell growth by cGMP analogs is strictly dependent on G-kinase and G-kinase activation inhibits the Ras/MAP kinase pathway (i) by phosphorylating c-Raf kinase on Ser43 and thereby inhibiting its activation and (ii) by inducing MAP kinase phosphatase 1 expression.

Transcriptional elongation of c-myb is regulated by NF-kappaB (p50/RelB).

High levels of c-myb expression are necessary for the proliferation of hematopoietic precursor cells whereas down-regulation of c-myb is required for terminal differentiation; this down-regulation occurs through a conditional block to transcriptional elongation in intron I. We previously observed that cAMP analogs prevented the late down-regulation of c-myb during hexamethylene bisacetamide (HMBA)-induced differentiation of murine erythroleukemia (MEL) cells and blocked differentiation; this correlated with the induction of NF-kappaB (p50/RelB) complexes which were shown to bind to NF-kappaB recognition sites flanking the transcriptional pause site of c-myb. We now selected stably-transfected MEL cells which overexpressed p50, RelB or both at levels similar to those induced by cAMP to determine whether these NF-kappaB proteins regulate c-myb expression in intact cells. We demonstrate that transcriptionally active NF-kappaB (p50/RelB) complexes, but not p50 or RelB alone, prevented the early and late down-regulation of c-myb mRNA and increased c-myb transcriptional elongation in HMBA-induced MEL cells. The increase in c-myb expression was sufficient to block erythroid differentiation and allow continuous proliferation of cells in the presence of HMBA. Steady-state c-myb mRNA levels in untreated cells were not affected by overexpression of NF-kappaB, suggesting that p50/RelB specifically modulated the efficiency of transcriptional attenuation during MEL cell differentiation.

Quantitative determination of Rap 1 activation in cyclic nucleotide-treated HL-60 leukemic cells: lack of Rap 1 activation in variant cells.

We have previously isolated variant HL-60 cells that are resistant to cGMP-induced differentiation and showed that they are deficient in proteolytic cleavage and/or carboxyl methylation of Rap 1A (J. Biol. Chem. 269, 32155 - 32161, 1994 and Oncogene 17, 2211 - 2233, 1998). We have now developed an enzyme-based method for assessing Rap 1 activation which is quantitative and provides a measurement of the per cent of Rap molecules in the active GTP-bound state. Using this method, we show that cAMP and cGMP analogs activate Rap 1 in parental HL-60 cells but not in the variant cells and that H-89, a cAMP-dependent protein kinase inhibitor, has no effect on cAMP-induced Rap 1 activation in parental cells. Thus, cAMP activation of Rap 1 in HL-60 cells is likely through a cAMP-regulated guanine nucleotide exchange factor (cAMP-GEF) and since cAMP does not activate Rap 1 in the variant cells, the data suggest that full post-translational processing of Rap 1 is necessary for cAMP-GEF activation of Rap 1. Activation of Rap 1 by cGMP analogs has not been previously found and suggests possible cross-talk between the NO/cGMP signal transduction pathway and Rap 1 signaling. Oncogene (2000) 19, 4029 - 4034.

NO activation of fos promoter elements requires nuclear translocation of G-kinase I and CREB phosphorylation but is independent of MAP kinase activation.

We have shown that nitric oxide (NO) regulates c-fos gene expression via cGMP-dependent protein kinase (G-kinase), but NO's precise mechanism of action is unclear. We now demonstrate that: (1) NO targets two transcriptional elements in the fos promoter, i.e., the fos AP-1 binding site and the cAMP-response element (CRE); (2) NO activation of these two enhancer elements requires the CRE binding protein CREB because a dominant negative CREB fully inhibits NO transactivation of reporter genes whereas dominant negative Fos or CCAAT enhancer binding proteins have no effect; (3) CREB is phosphorylated by G-kinase in vitro and its phosphorylation increases in vivo when G-kinase is activated either directly by cGMP or indirectly by NO via soluble guanylate cyclase; (4) NO activation of fos promoter elements requires nuclear translocation of G-kinase but not activation of mitogen-activated protein kinases.

cAMP-induced NF-kappaB (p50/relB) binding to a c-myb intronic enhancer correlates with c-myb up-regulation and inhibition of erythroleukemia cell differentiation.

During hexamethylene bisactamide (HMBA)-induced differentiation of murine erythroleukemia (MEL) cells erythroid genes are transcriptionally activated while c-myb and several other nuclear proto-oncogenes are down-regulated. Differentiation is inhibited by cAMP analogues and the adenyl cyclase-stimulating agent forskolin. We found that these drugs prevented the late down-regulation of c-myb which is known to be critical for MEL cell differentiation. Since the increase in c-myb mRNA levels was due to increased mRNA transcription, we examined the transcription factors NF-kappaB and AP-1 which have been implicated in the regulation of c-myb expression. Binding of MEL cell nuclear proteins to a NF-kappaB recognition sequence in c-myb intron I was strongly induced by 8-Br-cAMP or forskolin; induction was delayed and correlated with the up-regulation of c-myb. The cAMP-induced NF-kappaB complex contained p50 and RelB; in co-transfection assays, p50 and RelB transactivated a reporter construct containing the c-myb intronic NF-kappaB site upstream of a minimal promoter. 8-Br-cAMP and forskolin also increased the DNA binding activity of AP-1 complexes containing JunB, JunD and c-Fos in MEL cells which could cooperate with NF-kappaB. We conclude that inhibition of MEL cell differentiation by pharmacological doses of cAMP can be explained by the up-regulation of c-myb which is mediated, at least in part, by NF-kappaB p50/RelB heterodimers.

Deficient post-translational processing of Rap 1A in variant HL-60 cells.

Variant HL-60 cells resistant to differentiation induced by nitroprusside and cGMP analogs have normal guanylate cyclase and cGMP-dependent protein kinase (G-kinase) activity (J. Biol. Chem. 269, 32155-32161, 1994). We found decreased phosphorylation of a low molecular weight protein (pp23) in the variant cells and by co-migration on two-dimensional polyacrylamide gels, phosphopeptide mapping, immunoprecipitation and immunoblotting, we showed that pp23 was one of three post-translationally modified forms of Rap 1A expressed in HL-60 cells. Using an in vitro transcription/translation system, we studied each of the posttranslational processing steps of Rap 1A and we showed that pp23 represented fully processed Rap 1A. By immunoprecipitation, immunoblotting and 35S-methionine/cysteine incorporation, we showed that the variant cells were deficient in pp23, and thus in fully processed Rap 1A, but that these cells did express normal amounts of completely unprocessed Rap 1A and geranylgeranylated Rap 1A; the lack of Rap 1A processing beyond geranylgeranylation in the variant cells was not secondary to a change in Rap 1A's amino acid sequence. The variant cells had normal carboxyl methyltransferase activity suggesting they are deficient in proteolytic cleavage of Rap 1A. The deficient post-translational processing of Rap 1A had no effect on Rap 1A's subcellular distribution and we found no evidence for altered post-translational processing of H-Ras.

Impaired erythroid-specific gene expression in cAMP-dependent protein kinase-deficient murine erythroleukemia cells.

Murine erythroleukemia cells rendered deficient in cAMP-dependent protein kinase (A-kinase) activity by gene transfection are severely impaired in hexamethylene bisacetamide (HMBA)-induced differentiation (Pilz, R. B., Eigenthaler, M., and Boss, G. R. (1992) J. Biol. Chem. 267, 16161-16167). We now demonstrate that the A-kinase-deficient cells produce hemoglobin normally in response to exogenous hemin and that the heme precursor delta-aminolevulinate (delta-ALA) significantly increases HMBA-induced synthesis of heme and globin chains in these cells; these data suggest that impaired heme synthesis is at least partially responsible for the cells' deficient hemoglobin synthesis. HMBA-induced expression of the erythroid-specific delta-ALA synthetase, porphobilinogen deaminase, and beta-globin mRNAs was less in A-kinase-deficient cells than in parental cells and was reduced in proportion to the cells' residual A-kinase activity; relative transcription rates of these genes were reduced concordantly. Impaired expression of these three erythroid-specific genes was a feature of many independently-derived A-kinase-deficient clones, and normal expression was found in transfectants with normal A-kinase activity. The A-kinase-deficient cells did not exhibit a generalized defect in gene regulation since mRNA expression and transcription rates of H- and L-ferritin, c-myc, c-myb, and several housekeeping enzymes were similar in HMBA-treated parental and A-kinase-deficient cells. Our data suggest that A-kinase may be involved in regulating genes with erythroid-specific promoters and provide further evidence for heme as a regulator of globin chain synthesis.

Phosphoribosylpyrophosphate synthesis from glucose decreases during amino acid starvation of human lymphoblasts.

When cultured human lymphoblasts are starved 3 h for an essential amino acid, rates of purine nucleotide synthesis decrease markedly because of a decrease in the intracellular phosphoribosylpyrophosphate concentration (Boss, G.R., and Erbe, R.W. (1982) J. Biol. Chem. 257, 4242-4247; Boss, G. R. (1984) J. Biol. Chem. 259, 2936-2941). In amino acid-starved cells, glucose transport was not changed, whereas total glucose consumption and lactate production decreased by approximately 25 and 10%, respectively. Carbon flow through the oxidative pentose phosphate pathway, measured by 14CO2 release from [1-14C]glucose, decreased by 18% during amino acid starvation. However, kinetic studies of ribulose-5-phosphate 3-epimerase and phosphoriboisomerase suggested that the ribulose 5-phosphate produced by this pathway is converted mostly to xylulose 5-phosphate instead of to ribose 5-phosphate so that this pathway produces little phosphoribosylpyrophosphate. The activity of the nonoxidative pentose phosphate pathway, measured by high performance liquid chromatography following the incorporation of [1-14C]glucose into phosphoribosylpyrophosphate, ATP, and GTP, decreased by approximately 55% during amino acid starvation. None of the enzymes of either pathway changed in specific activity during amino acid starvation. We conclude that the nonoxidative pentose phosphate pathway is the major source of phosphoribosylpyrophosphate for purine nucleotide synthesis and that this pathway is regulated by a metabolite which changes in concentration during amino acid starvation.

Chemically induced murine erythroleukemia cell differentiation is severely impaired when cAMP-dependent protein kinase activity is repressed by transfected genes.

During chemically induced differentiation of murine erythroleukemia (MEL) cells, cAMP-dependent protein kinase activity increases, and the enzyme's isozyme pattern changes. To examine the enzyme's role during MEL cell differentiation, we stably transfected MEL cells with recombinant plasmids in which the mouse metallothionein I promoter controlled expression of either a mutant form of the type I regulatory subunit of cAMP-dependent protein kinase (RI) or the enzyme's specific peptide inhibitor (PKI); expressing either sequence rendered cells cAMP-dependent protein kinase-deficient. Chemically induced differentiation of MEL cells as assessed by beta-globin mRNA and hemoglobin accumulation was inhibited in RI mutant and PKI transfectants; adding zinc further inhibited differentiation in the transfectants but had no effect on parental MEL cells. The inhibition of differentiation correlated with the amount of RI mutant mRNA and protein in the RI mutant transfectants and with the cells' degree of cAMP-dependent protein kinase deficiency in both the RI mutant and PKI transfectants. Overexpression of wild type RI did not interfere with differentiation or enzyme activity. We conclude that cAMP-dependent protein kinase activity is important for chemically induced differentiation of MEL cells and that the down-regulation of RI protein which occurs during MEL cell differentiation is not essential for differentiation to proceed.

Molecular and genetic characterization of an ornithine decarboxylase-deficient Chinese hamster cell line.

The ornithine decarboxylase (ODC)-deficient Chinese hamster ovary (CHO) cell line C55.7 has normal amounts of ODC mRNA with very low amounts of immunologically detectable ODC protein, suggesting a structural mutation; however, 5-azacytidine treatment leads to phenotypical reversion (Steglich, C., and Scheffler, I. E. (1985) Somat. Cell Mol. Genet. 11, 11-23). We have demonstrated by chemical cleavage a single base mismatch in DNA heteroduplexes composed of wild-type and mutant cDNA strands. DNA sequencing showed that the mutant phenotype results from an aspartate-glycine substitution at amino acid 381 of the protein. When 5-azacytidine-revertant cell lines were selected for resistance to alpha-difluoromethylornithine, the resulting amplified ODC gene was structurally indistinguishable from the wild type gene. These results suggested the existence of a single active ODC locus in CHO cells. Using the methylation-sensitive restriction endonucleases AvaI and HpaII, we found evidence for two differentially methylated alleles in wild type, ODC-deficient and alpha-difluoromethylornithine-resistant cells. One of the alleles appeared completely inactivated by hypermethylation but could be reactivated by demethylation in spontaneous or 5-azacytidine-induced revertants.

The influence of ribose 5-phosphate availability on purine synthesis of cultured human lymphoblasts and mitogen-stimulated lymphocytes.

The intracellular ribose 5-phosphate concentration was found to be an important determinant of rates of de novo purine synthesis. When ribose 5-phosphate production was reduced in cultured human lymphoblasts by glucose starvation, the intracellular phosphoribosylpyrophosphate concentration and rates of de novo purine synthesis decreased. Inosinate-guanylate:pyrophosphate phosphoribosyltransferase (HPR transferase)-deficient cells were relatively more resistant to glucose starvation. To minimize the effect of purine nucleotide feedback inhibition on the de novo pathway, cells were treated with inhibitors of IMP dehydrogenase and adenylosuccinate synthetase. In normal lymphoblasts, purine synthesis was stimulated only at glucose concentrations greater than 100 microM while in HPR transferase-deficient lymphoblasts, stimulation occurred even in the absence of glucose. The differences between the normal and HPR transferase-deficient cells were lost when ribose reutilization from endogenous nucleotide breakdown was impaired in the HPR transferase-deficient cells by incubation with 2'-deoxyinosine. Endogenous ribose reutilization for purine synthesis is, therefore, important when either glucose availability is limited or synthesis is stimulated. In the absence of glucose, exogenous purine nucleotides restored the intracellular concentrations of ribose 5-phosphate, phosphoribosylpyrophosphate, and purine nucleotides to almost 100% and rates of purine synthesis to 50-75% of those at 10 mM glucose. When ribose 5-phosphate production was increased in peripheral blood lymphocytes by phytohemagglutinin activation, the intracellular phosphoribosylpyrophosphate concentration and rates of de novo purine synthesis increased.

Identification of a ras-related protein in murine erythroleukemia cells that is a cAMP-dependent protein kinase substrate and is phosphorylated during chemically induced differentiation.

Murine erythroleukemia (MEL) cells deficient in cAMP-dependent protein kinase (A-kinase) activity are impaired in chemically induced differentiation (Pilz, R. B., Eigenthaler, M., and Boss, G. R. (1992) J. Biol. Chem. 267, 16161-16167). We identified by two-dimensional polyacrylamide gel electrophoresis two low molecular weight proteins (referred to as pp 21-1 and 21-2) that were phosphorylated when parental MEL cells, but not A-kinase-deficient MEL cells, were treated with the membrane-permeable cAMP analog 8-bromo-cAMP. We showed that pp 21-1 and 21-2: (a) were direct A-kinase substrates; (b) bound GTP; and (c) belonged to the ras superfamily of proteins. The only ras-related proteins that are clearly A-kinase substrates both in vitro and in vivo are Rap 1A and 1B while H- and K-Ras can be A-kinase substrates in vitro; we showed by immunological methods, phosphopeptide mapping, and migration on two-dimensional gels that pp 21-1 and 21-2 were not identical to one of these four proteins. We found a 3-fold increase of 32PO4 incorporation into pp 21-2 in hexamethylene bisacetamide-treated parental MEL cells which was not secondary to an increase in pp 21-2 protein but appeared secondary to increased phosphorylation of pp 21-2 by A-kinase. Thus, pp 21-1 and 21-2 are either new ras-related proteins or are previously identified ras-related proteins not known to be A-kinase substrates, and increased phosphorylation of pp 21-2 occurs during differentiation of MEL cells.

Adenosine dialdehyde and nitrous oxide induce HL-60 differentiation.

Adenosine dialdehyde and nitrous oxide, specific S-adeno-sylhomocysteine hydrolase and methionine synthetase inhibitors, respectively, induced differentiation of the human promyelocytic cell line HL-60. Their effect did not appear to be mediated through changes in transmethylation or decreased S-adenosylmethionine synthesis because (1) there was little correlation between the concentrations of adenosine dialdehyde that induced differentiation and those that changed the ratio of the intracellular concentrations of S-adenosylmethionine to S-adenosylhomocysteine, and (2) inhibition of methionine adenosyltransferase by cycloleucine did not induce differentiation. The differentiation induced by adenosine dialdehyde was prevented by homocysteine and that by nitrous oxide was inversely related to the medium methionine concentration. This suggested that differentiation was secondary to decreased methionine synthesis.

Cell type-specific regulation of B-Raf kinase by cAMP and 14-3-3 proteins.

Cyclic AMP can either activate or inhibit the mitogen-activated protein kinase (MAPK) pathway in different cell types; MAPK activation has been observed in B-Raf-expressing cells and has been attributed to Rap1 activation with subsequent B-Raf activation, whereas MAPK inhibition has been observed in cells lacking B-Raf and has been attributed to cAMP-dependent protein kinase (protein kinase A)-mediated phosphorylation and inhibition of Raf-1 kinase. We found that cAMP stimulated MAPK activity in CHO-K1 and PC12 cells but inhibited MAPK activity in C6 and NB2A cells. In all four cell types, cAMP activated Rap1, and the 95- and 68-kDa isoforms of B-Raf were expressed. cAMP activation or inhibition of MAPK correlated with activation or inhibition of endogenous and transfected B-Raf kinase. Although all cell types expressed similar amounts of 14-3-3 proteins, approximately 5-fold less 14-3-3 was associated with B-Raf in cells in which cAMP was inhibitory than in cells in which cAMP was stimulatory. We found that the cell type-specific inhibition of B-Raf could be completely prevented by overexpression of 14-3-3 isoforms, whereas expression of a dominant negative 14-3-3 mutant resulted in partial loss of B-Raf activity. Our data suggest that 14-3-3 bound to B-Raf protects the enzyme from protein kinase A-mediated inhibition; the amount of 14-3-3 associated with B-Raf may explain the tissue-specific effects of cAMP on B-Raf kinase activity.