N-Linked glycan site occupancy impacts the distribution of a potassium channel in the cell body and outgrowths of neuronal-derived cells.

BACKGROUND: Vacancy of occupied N-glycosylation sites of glycoproteins is quite disruptive to a multicellular organism, as underlined by congenital disorders of glycosylation. Since a neuronal component is typically associated with this disease, we evaluated the impact of N-glycosylation processing of a neuronal voltage gated potassium channel, Kv3.1b, expressed in a neuronal-derived cell line, B35 neuroblastoma cells. METHODS: Total internal reflection fluorescence and differential interference contrast microscopy measurements of live B35 cells expressing wild type and glycosylation mutant Kv3.1b proteins were used to evaluate the distribution of the various forms of the Kv3.1b protein in the cell body and outgrowths. Cell adhesion assays were also employed. RESULTS: Microscopy images revealed that occupancy of both N-glycosylation sites of Kv3.1b had relatively similar amounts of Kv3.1b in the outgrowth and cell body while vacancy of one or both sites led to increased accumulation of Kv3.1b in the cell body. Further both the fully glycosylated and partially glycosylated N229Q Kv3.1b proteins formed higher density particles in outgrowths compared to cell body. Cellular assays demonstrated that the distinct spatial arrangements altered cell adhesion properties. CONCLUSIONS: Our findings provide direct evidence that occupancy of the N-glycosylation sites of Kv3.1b contributes significantly to its lateral heterogeneity in membranes of neuronal-derived cells, and in turn alters cellular properties. GENERAL SIGNIFICANCE: Our study demonstrates that N-glycans of Kv3.1b contain information regarding the association, clustering, and distribution of Kv3.1b in the cell membrane, and furthermore that decreased occupancy caused by congenital disorders of glycosylation may alter the biological activity of Kv3.1b.

Nitric oxide-dependent activation of p53 suppresses bleomycin-induced apoptosis in the lung.

Chronic inflammation leading to pulmonary fibrosis develops in response to environmental pollutants, radiotherapy, or certain cancer chemotherapeutic agents. We speculated that lung injury might be mediated by p53, a proapoptotic transcription factor widely implicated in the response of cells to DNA damage. Intratracheal administration of bleomycin led to caspase-mediated DNA fragmentation characteristic of apoptosis. The effects of bleomycin were associated with translocation of p53 from the cytosol to the nucleus only in alveolar macrophages that had been exposed to the drug in vivo, suggesting that the lung microenvironment regulated p53 activation. Experiments with a thiol antioxidant (N-acetylcysteine) in vivo and nitric oxide (NO) donors in vitro confirmed that reactive oxygen species were required for p53 activation. A specific role for NO was demonstrated in experiments with inducible nitric oxide synthase (iNOS)(-/)- macrophages, which failed to demonstrate nuclear p53 localization after in vivo bleomycin exposure. Strikingly, rates of bleomycin-induced apoptosis were at least twofold higher in p53(-/)- C57BL/6 mice compared with heterozygous or wild-type littermates. Similarly, levels of apoptosis were also twofold higher in the lungs of iNOS(-/)- mice than were observed in wild-type controls. Consistent with a role for apoptosis in chronic lung injury, levels of bleomycin-induced inflammation were substantially higher in iNOS(-/)- and p53(-/)- mice compared with wild-type controls. Together, our results demonstrate that iNOS and p53 mediate a novel apoptosis-suppressing pathway in the lung.

Inhibition of beta-globin gene expression by 3'-azido-3'-deoxythymidine in human erythroid progenitor cells.

3'-Azido-3'-deoxythymidine (AZT) treatment in HIV-infected patients is limited by bone marrow suppression including neutropenia and anemia. Previous studies had shown a direct effect of high concentrations of this drug on globin gene expression in K-562 erythroleukemia cells. To better define the mechanism(s) of AZT-induced bone marrow toxicity, the present study evaluates these effects in more relevant human erythroid progenitor liquid cultures, because AZT is 100 times more toxic to human bone marrow cells than K-562 cells. At a clinically relevant concentration of 1 microM, AZT inhibited specifically erythroid cell growth by approximately 58% as compared with untreated cells. The percentage of cells synthesizing hemoglobin was decreased also by 47% in AZT-treated cells with beta-globin mRNA levels accounting for 0.27 pmol in treated cells as compared with 1.44 under control conditions while beta-actin levels remained unchanged. Under the same conditions, AZT inhibited the beta-globin chain synthesis by approximately 60% as compared with the control. Consistent with the data described above was the finding that a concentration as low as 0.1 microM of AZT decreased by almost 40% the binding level of the erythroid-specific transcription factor GATA-1. These findings demonstrate that AZT, at clinical relevant concentrations, specifically inhibits beta-globin gene expression in human erythroid progenitor liquid cell culture.

Effects of 3'-azido-3'-deoxythymidine on erythroid inducible gene expression in human K-562 leukemia cells.

We previously demonstrated that 3'-azido-3'-deoxythymidine (AZT) down regulates hemoglobin (Hb) synthesis and globin gene expression. That inhibition may therefore result either from a direct effect on globin gene transcription or an indirect effect through inhibition of K-562 cell induction, thereby leading to inhibition of other inducible genes of heme biosynthesis. The present results demonstrate that inhibition of globin gene expression by AZT is a direct gene effect rather than a general inhibition of K-562 cell induction as demonstrated by the absence of AZT effects on expression of three other erythroid inducible genes [erythroid-specific aminolevulate synthase (ALAS-E), aminolevulinic acid dehydrogenase (ALAD), and erythroid-specific porphobilinogen deaminase (PBGD-E)].

Phosphorothioate oligonucleotides bind in a non sequence-specific manner to the nucleolar protein C23/nucleolin.

To design optimal strategies for intracellular delivery of antisense phosphorothioate oligonucleotides, it may be useful to understand their interaction with cellular macromolecules. Nuclear extracts from LOX amelanotic myeloma cells were studied for protein binding to phosphorothioate oligonucleotides using a Southwestern protocol. Multiple nuclear proteins bound to the phosphorothioate oligonucleotides but no detectable protein binding was found to phosphodiester oligonucleotides. The protein with the strongest binding signals was shown by immunoprecipitation to be nucleolar C23/nucleolin, a 110 kDa protein. With glutathione S-transferase/nucleolin fusion protein constructs, the region of nucleolin containing the RNA recognition motifs had binding activity to phosphorothioate oligonucleotides.

Antisense phosphorothioate oligonucleotides direct both site-specific and nonspecific RNAse H cleavage of in vitro synthesized p120 mRNA.

To determine the sequence specificity with which various antisense phosphorothioate oligonucleotides to the p120 gene could direct RNAse H-dependent cleavage of p120 mRNA, an in vitro RNAse H assay was done on in vitro synthesized p120 mRNA. Three oligonucleotides tested (ISIS 3462, coding region; ISIS 3466, 3' untranslated region; and ISIS 3782, 3' untranslated region) directed cleavage of p120 mRNA at the target sites. In addition, several additional cleavage products from nontarget sites were detected. A computer sequence homology search revealed that these additional bands probably result from cleavage at regions of complementarity of eight or nine contiguous bases of the oligonucleotide and the p120 mRNA. These findings suggest that such nontarget specific cleavages may account for multiple effects of phosphorothioate oligonucleotides.

Influence of template primary structure on 3'-azido-3'-deoxythymidine triphosphate incorporation into DNA.

In the present study, templates containing a specific segment of the G gamma-globin gene were constructed and incorporation of 3'-azido-3'-deoxythymidine 5'-triphosphate (AZT-TP) or 2',3'-dideoxythymidine 5'-triphosphate (ddTTP) into these templates was compared to that observed in M13 bacteriophage DNA. Investigations on the intrinsic fidelity of T7 DNA polymerase in reactions with AZT-TP and without deoxythymidine 5'-triphosphate, resulted in DNA synthesis beyond the first T site, suggesting that other normal deoxynucleotides misincorporated at these T sites. Modified T7 DNA polymerase incorporated AZT-TP into T sites of elongating DNA strands. Chain termination at noncomplementary sites was also observed with AZT-TP when a genomic DNA template was used and interestingly, this phenomenon was not detected in the presence of a M13 DNA template. These DNA template-dependent effects were not detected with either ddTTP, 2',3'-dideoxycytidine 5'-triphosphate, or 2',3'-didehydro-2',3'-dideoxythymidine 5'-triphosphate (D4T-TP). A variation in the extent of chain termination at T sites was observed with D4T-TP suggesting that each 2',3'-dideoxynucleoside may exhibit unique chain termination patterns along with the template sequence.

Comparative effects of 3'-azido-3'-deoxythymidine and its metabolite 3'-amino-3'-deoxythymidine on hemoglobin synthesis in K-562 human leukemia cells.

We previously demonstrated that 3'-azido-3'-deoxythymidine (AZT) inhibits hemoglobin (Hb) synthesis and globin gene transcription in butyric acid-induced K-562 leukemia cells, suggesting that these effects may play a role in the AZT-induced anemia observed in patients [Mol. Pharmacol. 38:797-804 (1990)]. The recent discovery by our group of a novel metabolite of AZT. 3'-amino-3'-deoxythymidine (AMT), which exhibits a high degree of toxicity toward human hemopoietic cells [Mol. Pharmacol. 39:258-266 (1991); Antimicrob. Agents Chemother. 35:801-807 (1991)], has led us to explore potential effects of this AZT metabolite on Hb production, globin mRNA expression, and heme synthesis in butyric acid-induced K-562 human erythroleukemia cells. AMT inhibited Hb synthesis by approximately 21%, as measured by benzidine staining, at concentrations as low as 25 microM, with slightly increased inhibition at higher AMT concentrations. The inhibition of Hb production by AMT was substantially lower, compared with that of AZT. AMT inhibited globin mRNA steady state levels in a dose-dependent manner to a similar extent as did the parent drug, with approximately 50% inhibition by each compound at a concentration of 100 microM. Nuclear run-on transcription assays demonstrated that inhibition by AMT of globin mRNA synthesis was associated with a decreased rate of globin-specific gene transcription. Globin mRNA stability was not affected by either 100 microM AZT or AMT, as measured after blockage of transcription with actinomycin D. To gain insight into potential mechanism(s) responsible for the different quantitative effects of AZT and AMT on Hb synthesis, the effect of each compound on induction of heme synthesis in K-562 cells was determined. Although heme induction was not affected by AMT, a significant inhibition approximating 20% was observed in the presence of 100 microM AZT. In addition, AZT down-regulated mRNA steady state levels under conditions where heme synthesis was inhibited by succinylacetone. These data suggest that inhibition by AZT of globin gene expression is a direct effect and is not secondary to inhibition of heme synthesis. This study emphasizes the role of AMT in the pharmacodynamic properties of AZT, in relation to its toxicity, and suggest that both AMT and AZT may be involved in the inhibition of erythroid differentiation observed in vivo, through changes in gene expression.

3'-Azido-3'-deoxythymidine inhibits globin gene transcription in butyric acid-induced K-562 human leukemia cells.

We previously demonstrated that 3'-azido-3'-deoxythymidine (AZT) inhibits proliferation of human bone marrow progenitor cells in vitro and that incorporation of AZT into nuclear DNA may be one mechanism responsible for AZT-induced bone marrow toxicity [Antimicrob. Agents Chemother. 31:452-454 (1987); Mol. Pharmacol. 36:9-14 (1989)]. The present study explores possible genetic mechanisms involved in AZT-induced anemia by evaluating the effects of AZT on globin gene expression at both the transcriptional and the translational levels in butyric acid-induced K-562 human erythroleukemia cells. AZT, at concentrations ranging from 10 to 250 microM, was added to cells 25 hr after initiation of induction of hemoglobin (Hb) synthesis with 1.4 mM butyric acid. Hb synthesis, as measured by benzidine staining, was inhibited by AZT in a dose- and time-dependent manner in these cells. AZT inhibition of cell growth was not the major contributing factor in the net inhibition of Hb synthesis in K-562 cells. As assessed by Northern blot analysis, AZT inhibition of Hb synthesis was associated with a decrease in globin mRNA steady state levels without inhibition of total RNA synthesis or actin mRNA steady state levels. In particular, a decrease of globin mRNA levels of 23% by 25 microM AZT was observed, reaching a maximum inhibition of 59% in the presence of 250 microM AZT. In vitro translation experiments demonstrated that essentially all nonglobin translatable mRNAs were not inhibited by AZT concentrations as high as 250 microM, whereas globin mRNAs coding for epsilon, zeta, A gamma, G gamma, and alpha chains were substantially inhibited to similar levels by AZT, in a dose-dependent manner. Transcriptional run-on studies with isolated nuclei from AZT-treated K-562 cells demonstrated a 20 and 50% inhibition of in vitro synthesized globin transcripts from cells exposed to 25 and 100 microM AZT, respectively. 2',3'-Dideoxycytidine also inhibited K-562 cell growth in the same concentration range as AZT but, of importance, had no effects on Hb production. These data suggest that inhibition of globin gene expression may play a role in the cytotoxicity of AZT to the erythroid cell.

Beta globin gene transcripts originating in the promoter region during early hexamethylene bisacetamide and dimethylsulfoxide induction of Friend erythroleukemia cells.

The beta-globin transcripts which are induced by dimethylsulfoxide (DMSO) and hexamethylene bisacetamide (HMBA) have been characterized in order to assess potential differences in their mechanisms of induction. Transcripts which initiate in the 5' flanking promoter region are likely indicators of promoter accessibility and were therefore characterized during the time course of induction with each inducer in Friend Erythroleukemia cells. S1 analysis with probes labeled at - 12 or +82 relative to the (+1) cap site showed no major differences between 5' ends of the upstream initiated transcripts in cells induced by DMSO or HMBA. We detected several upstream bands with each inducer corresponding to beta-globin transcripts with 5' ends between - 190 and -55 relative to the cap site and found that cells induced with DMSO and HMBA show a similar transcription response as measured by initiation in the 5' flanking region of the beta-globin gene. Interestingly, the upstream initiated transcripts reach their peak concentration levels much earlier in the time course of induction than do the mRNA transcripts with 5' ends at the major (+1) cap site. Northern blot analysis detected the upstream initiated transcripts as early as 16 hours after induction with DMSO, primarily in unprocessed large transcripts. We find that the promoter region containing transcripts constitute a higher percent of total beta-globin transcripts at the start of the induction and may therefore have an early function in the multistep induction process.

Accurate in vitro initiation of beta-globin gene transcription in induced Friend-cell nuclei.

The initiation of transcription by RNA polymerase II in isolated murine erythroleukemia cell nuclei was investigated by isolating newly synthesized gamma-thio (gamma-S-)-triphosphate-labeled transcripts by Hg-agarose chromatography. The 5' terminus of transcripts initiated in vitro with [gamma-35S]ATP or [gamma-35S]GTP was identified as the thiotetraphosphate in alkaline hydrolysis products from Hg-agarose-selected RNA. Additional control experiments analyzing the nuclear transcription of two well characterized tRNA genes showed that each gene was initiated with the proper triphosphate, either gamma-S-ATP or gamma-S-GTP, indicating little, if any, exchange of the gamma-S-labeled substrate to the other triphosphates. As determined by S1 mapping, newly synthesized beta-globin gene transcripts initiate only with gamma-S-ATP. Their 5'-terminus is located at the cap site, and their synthesis is inhibited by 1 microgram alpha-amanitin/ml. In reactions containing gamma-S-ATP but not gamma-S-GTP, several additional initiation sites are observed that are located in the 5'-flanking region. We conclude that RNA polymerase II can initiate transcription at the cap site in isolated nuclei.

Nuclear beta-globin RNA transcripts initiated in vitro with gamma-thio nucleoside triphosphate.

We have been able to obtain significant levels of apparently in vitro-initiated beta-globin transcripts from isolated nuclei using mouse Friend cells. The level of hybrid selected beta-globin transcripts was increased 3-9-fold in nuclei after the cells were induced with dimethylsulfoxide. The newly initiated transcripts were detected through incorporation of gamma-thio ATP and isolation by Hg-agarose chromatography. The 5'-terminus was identified by recovery of the thio tetraphosphate in hydrolysis products from in vitro initiated RNA using gamma-35S ATP.