Volume 384, Issue 3

One of the essential maturation steps to yield functional tRNA molecules is the removal of 3'-trailer sequences by RNase Z. After RNase Z cleavage the tRNA nucleotidyl transferase adds the CCA sequence to the tRNA 3terminus, thereby generating the mature tRNA. Here we investigated whether a terminal CCA triplet as 3'-trailer or embedded in a longer 3'- trailer influences cleavage site selection by RNase Z using three activities: a recombinant plant RNase Z, a recombinant archaeal RNase Z and an RNase Z active wheat extract. A trailer of only the CCA trinucleotide is left intact by the wheat extract RNase Z but is removed by the recombinant plant and archaeal enzymes. Thus the CCA triplet is not recognized by the RNase Z enzyme itself, but rather requires cofactors still present in the extract. In addition, we investigated the influence of acceptor stem length on cleavage by RNase Z using variants of wildtype tRNATyr. While the wild type and the variant with 8 base pairs in the acceptor stem were processed efficiently by all three activities, variants with shorter and longer acceptor stems were poor substrates or were not cleaved at all.

Ribozymes have a great potential for developing specific gene silencing molecules. One of the main limitations to ensure the efficient application of ribozymes is to achieve effective binding to the target. Stem-loop domains support efficient formation of the kissing complex between natural antisense molecules and their target sequence. We have characterized catalytic antisense RNA hybrid molecules composed of a hammerhead ribozyme and a stemloop antisense domain. A series of artificial RNA substrates containing the TAR-RNA stem-loop and a target for the hammerhead ribozyme were constructed and challenged with a catalytic antisense RNA carrying the TAR complementary stemloop. The catalytic antisense RNA cleaves each of these substrates significantly more efficiently than the parental hammerhead ribozyme. Deletion of the TAR domain in the substrate abolishes the positive effect. These results suggest that the enhancement is due to the interaction of both complementary stemloop motifs. A similar improvement was corroborated when targeting the LTR region of HIV-1 with either hammerhead and hairpinbased catalytic antisense RNAs. Our results indicate that the TAR domain can be used as an anchoring site to facilitate the access of ribozymes to their specific target sequences within TAR-containing RNAs. Finally, we propose the addition of stable stem-loop motifs to the ribozyme domain as a rational way for constructing catalytic antisense RNAs.

An extensive screening of the human neuronal nitric oxide synthase (nNOS) mRNAs in various human tissues and cell lines unraveled an extreme complexity in the transcription of this gene. Using 5' rapid amplification of cDNA ends (5'-RACE), ten different exons 1 named 1a-1l) were identified. They were spliced in a cell-specific manner to a common exon 2, which bears the translational start site. Three first exons (1d, 1g and 1f) were used predominantly for the transcription of the nNOS gene (146 out of 197 5'-RACE clones contained these exons). Exon 1k was found alone, but in many instances was interposed between exons 1b, 1d, 1g, 1i or 1j and the common exon 2. In addition to the cellspecific heterogeneity of human nNOS transcripts, nNOS is highly regulated at the transcriptional level. In resting A673 neuroepithelioma cells, the prevalent nNOS transcript was the exon 1g mRNA with minor expression of exons 1d+1k and exon 1f mRNAs). When the cells were treated with dibutyrylcAMP, nNOS mRNA was markedly upregulated. This upregulation was solely due to an increase in exon 1f mRNA, while the expression of the other mRNA species remained unchanged. Human HaCat keratinocyte like cells expressed the exon 1i+1k and 1i nNOS transcripts under basal conditions. When stimulated with epidermal growth factor, only the exon 1i+1k transcript was upregulated. Although these nNOS transcripts do not differ in their translated region, the various mRNAs may trigger post-transcriptional effects such as changes in mRNA stability and translation efficiency.

The core antigen (HBcAg) of hepatitis B Virus (HBV) can be expressed in Escherichia coli where it assembles into icosahedral particles containing 240 or 180 subunits. Analysis of the two kinds of particles by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) showed that a substantial proportion of their subunits were smaller than the full-length HBcAg monomer and of variable size, but all had the same N-terminal sequence showing that the smaller species were heterogeneous in their arginine-rich C-terminal regions. Around 50% of these arginine residues are encoded by the triplet AGA which is rare in E. coli. Supplementation of the level of AGA tRNA in the cell by transformation with plasmids expressing the T4 AGA tRNA gene significantly enhanced the yield of HBcAg. Fusion phage carrying a ligand specific for HBcAg showed no significant difference in the affinity for the two sizes of HBcAg particles, but in similar reactions in solution HBV surface antigen exhibited differential affinities for the same two HBcAg preparations.

The crystal structure of cystathionine γ-lyase (CGL) from yeast has been solved by molecular replacement at a resolution of 2.6 å. The molecule consists of 393 amino acid residues and one PLP moiety and is arranged in the crystal as a tetramer with D2 symmetry as in other related enzymes of the CysMetmetabolism PLP-dependent family like cystathionine β-lyase (CBL). A structure comparison with other family members revealed surprising insights into the tuning of enzymatic specificity between the different family members. CGLs from yeast or human are virtually identical at their active sites to cystathionine γ-synthase (CGS) from E. coli. Both CGLs and bacterial CGSs exhibit γ-synthase and γ-lyase activities depending on their position in the metabolic pathway and the available substrates. This group of enzymes has a glutamate (E333 in yeast CGL) which binds to the distal group of cystathionine (CTT) or the amino group of cysteine. Plant CGSs use homoserine phosphate instead of O-succinyl-homoserine as one substrate. This is reflected by a partially different active site structure in plant CGSs. In CGL and CBL the pseudosymmetric substrate must dock at the active site in different orientations, with S in γ-position (CBL) or in δ-position (CGL). The conserved glutamate steers the substrate as seen in other CGLs. In CBLs this position is occupied by either tyrosine or hydrophobic residues directing binding of CTT such that S is in the in γ-position. In methionine γ-lyase a hydrophic patch operates as recognition site for the methyl group of the methionine substrate.

A key step in the development of new hydrophilic pharmaceuticals is to get them through biological barriers. Cell-penetrating peptides, CPPs, have been shown previously to enter cells both in vitro and in vivo by a non-endocytotic mechanism and to be able to carry large cargo molecules with them. Recently, we showed that a small peptide, pVEC, from murine vascular endothelial cadherin, has the characteristics to be classified as a protein derived CPP. Here we have further investigated pVEC together with its all-D analog for cellular uptake, intra- and extracellular stability, and their enzymatic degradation. The two peptides, pVEC and all-D pVEC, translocate into aortic endothelial cells and murine fibroblasts by a nonendocytotic mechanism. In phosphate buffer, pVEC remains intact while the C-terminal lysine is quickly removed in human serum and serumcontaining media. Both pVEC and pVEC without the Cterminal Lys were detected by mass spectrometry inside the two cell types tested. The pVEC half-life is 10.5 min in phosphate buffer containing 10 units of trypsin and 44.6 min in phosphate buffer containing 4.2 units of carboxypeptidase A and 18 units of carboxypeptidase B. In contrast to pVEC, the all-D analog remains intact in serum and resists enzymatic degradation.

The ubiquitous calpains, u- and m-calpain, have been implicated in essential physiological processes and various pathologies. Cell-permeable specific inhibitors are important tools to elucidate the roles of calpains in cultivated cells and animal models. The synthetic Nacetylated 27-mer peptide derived from exon B of the inhibitory domain 1 of human calpastatin (CP1B) is unique as a potent and highly selective reversible calpain inhibitor, but is poorly cell-permeant. By addition of N-terminal cysteine residues we have generated a disulfide-conjugated CP1B with the cell-penetrating 16-mer peptide penetratin derived from the third helix of the Antennapedia homeodomain protein. The inhibitory potency and selectivity of CP1B for calpain versus cathepsin B and L, caspase 3 and the proteasome was not affected by the conjugation with penetratin. The conjugate was shown to efficiently penetrate into living LCLC 103H cells, since it prevents ionomycin-induced calpain activation at 200-fold lower concentration than the non-conjugated inhibitor and is able to reduce calpain triggered apoptosis of these cells. Penetratin-conjugated CP1B seems to be a promising alternative to the widely used cell-permeable peptide aldehydes (e.g. calpain inhibitor I) which inhibit the lysosomal cathepsins and partially the proteasome as well or even better than the calpains.

Six GIP1-30NH2 analogs were synthesized with modifications (de-protonation, N-methylation, reversed chirality, and substitution) at positions 1, 3, and 4 of the N-terminus, and additionally, a cyclized GIP derivative was synthesized. The relationship between altered structure to biological activity was assessed by measuring receptor binding affinity and ability to stimulate adenylyl cyclase in CHO-K1 cells transfected with the wild-type GIP receptor (wtGIPR). These structureactivity relationship studies demonstrate the importance of the GIP N-terminus and highlight structural constraints that can be introduced in GIP analogs. These analogs may be useful starting points for design of peptides with enhanced in vivo bioactivity.

Interleukin-6 (IL-6) is a helical cytokine exerting pleiotropic activities including the regulation of hematopoiesis, B cell activation and acute-phase reaction. The structure-function relationship of the molecule is the subject of intensive investigation using point and deletion mutants. Our objective was to analyse the role of the N-terminal 18–46 region in IL-6-mediated expression of junB protooncogene and fibrinogen production, reflecting the acute phase response, with synthetic overlapping peptides. mRNA expression of junB was monitored by competitive RT-PCR, while sandwich ELISA was used for the detection of fibrinogen in the supernatant of HepG2 human hepatoma cells. We found that even short synthetic octapeptides can be stimulatory (in the absence of IL-6) or inhibitory (in the presence of IL-6) in both assays. To establish the molecular mechanism by which synthetic peptides exert their biological effects electromobility shift assay was carried out using HepG2 nuclear extracts. Peptides inducing junB expression initiate gel shifts of STAT3/DNA complexes, which may indicate the involvement of this signal transduction pathway. Circular dicroism spectroscopy data suggest that 8-11-mer peptides representing different parts of the 18–46 region have a marked tendency to adopt ordered conformations in a water/trifluoroethanol (1:1 v/v) mixture. Competition studies with rhIL-6 and selected fluorophorelabelled peptides indicate the presence of more than one binding site on soluble IL-6 receptor. Considering the possible multiple etiologic role of IL-6 in the pathogenesis of various diseases, these peptides could be useful for dissection of IL-6 related biological effects.

Plasminogen activator inhibitor-1 (PAI-1) and twochain high molecular weight kininogen (HKa) exert antiadhesive properties in vitronectindependent cell adhesion. Here, the hypothesis was tested that these anti-adhesive components promote apoptosis in vascular cells. PAI-1 or HKa induced a 2- to 3-fold increase in apoptosis of human umbilicalvein endothelial cells (HUVEC) and vascular smooth muscle cells (VSMC) adherent to vitronectin, as determined by annexin VFACS assay, similar to αvintegrin inhibitor cyclo-(Arg-Gly-Asp-D-Phe-Val)-peptide (cRGDfV). Apoptosis occurred after 12 h incubation and was attributable to caspase 3 activation that in turn induced DNA fragmentation. Induction of apoptosis strongly correlated with the antiadhesive effect of PAI-1 and HKa on these cells. In contrast, PAI-1 and HKa did not affect fibronectin-dependent adhesion or cell survival. uPA did not influence apoptosis in vitronectin- or fibronectin-adherent cells. In atherosclerotic vessel sections, congruent distribution of vitronectin, PAI-1, HK, and of components of the urokinase plasminogen activator/receptor system with apoptotic cells lining foam cell lesions was demonstrated by immunostaining. These results indicate that inhibition of vitronectindependent cell adhesion through PAI-1 and HKa correlates with apoptosis induction in vascular cells mediated through the caspase 3 pathway. Co-distribution of apoptosis with plasminogen activation system components in atherosclerosis exemplifies the significance of antiadhesive mechanisms and apoptosis for tissue remodeling, such as in neointima development.

Oxylipins of the jasmonate pathway and synthetic functional analogs have been analyzed for their elicitor-like activities in an assay based on the induced accumulation of glyceollins, the phytoalexins of soybean (Glycine max L.), in cell suspension cultures of this plant. Jasmonic acid (JA) and its methyl ester showed weak phytoalexin-inducing activity when compared to an early jasmonate biosynthetic precursor, 12-oxo-phytodienoic acid (OPDA), as well as to the bacterial phytotoxin coronatine and certain 6-substituted indanoyl-L-isoleucine methyl esters, which all were highly active. Interestingly, different octadecanoids and indanoyl conjugates induced the accumulation of transcripts of various defense-related genes to different degrees, indicating distinct induction competencies. Therefore, these signaling compounds and mimics were further analyzed for their effects on signal transduction elements, such as the transient enhancement of the cytosolic Ca2+ concentration and MAP kinase activation, which are known to be initiated by a soybean pathogenderived βglucan elicitor. In contrast to the β-glucan elicitor, none of the other compounds tested triggered these early signaling elements. Moreover, endogenous levels of OPDA and JA in soybean cells were shown to be unaffected after treatment with β-glucans. Thus, OPDA and JA, which are functionally mimicked by coronatine and a variety of 6-substituted derivatives of indanoyl-L-isoleucine methyl ester, represent highly efficient signaling compounds of a lipidbased pathway not deployed in the β-glucan elicitor-initiated signal transduction.

The activities of the lysosomal cysteine proteinases cathepsin B and L are regulated by their endogenous inhibitors, stefins A and B, and cystatin C, and their imbalance may be associated with increased invasiveness and development of the malignant cell phenotype. The aim of this study was to investigate mRNA, protein and activity levels of the above proteins in relation to in vitro invasiveness and to the reported in vivo tumorigenicity of four human breast tumor cell lines: the spontaneously immortalized cell line MCF10A, its cHaras transfectant MCF10AT, and two tumorigenic derivative cell lines, MCF10AT-Ca1a and MCF10AT-Ca1d. Invasiveness did not correlate with tumorigenicity, since the MCF10AT cell was the most invasive and the remaining three were at about half of its level. Cathepsin B expression paralleled the in vitro invasiveness through matrigel at all levels of expression, but cathepsin L did not. Stefin levels were elevated severalfold in the tumorigenic cell lines, but not in MCF10AT. The hypothesis that cathepsin B plays an active role in the invasion of breast cancer cell lines was confirmed by the fact that synthetic cysteine proteinase inhibitors, particularly those selective for cathepsin B, significantly reduced the invasion of the MCF10AT cells.

Ribonuclease P (RNase P) is a ubiquitous and essential enzyme that endonucleolytically cleaves all tRNA precursors to produce the mature 5'-end. We have investigated the effect of synthetic rertinoids (all-trans retinoic acid, acitretin) and arotinoids (Ro 13-7410, Ro 15-0778, Ro, 13-6298 and Ro 15-1570) on RNase P activity isolated for the first time from normal human epidermal keratinocytes (NHEK). All tested compounds but one (Ro 15-1570) revealed a dose-dependent inhibition of RNase P activity, indicating that they may have a direct effect on tRNA biogenesis. Detailed kinetic analysis showed that all retinoids behave as classic competitive inhibitors. On the basis of the Ki values Ro 13-7410 was found to be the strongest inhibitor among all compounds tested.

Cellular defense systems against reactive oxygen species (ROS) include thioredoxin reductase (TrxR) and glutathione reductase (GR). They generate sulfhydryl-reducing systems which are coupled to antioxidant enzymes, the thioredoxin and glutathione peroxidases (TPx and GPx). The fruit fly Drosophila lacks a functional GR, suggesting that the thioredoxin system is the major source for recycling glutathione. Whole genome in silico analysis identified two non-selenium containing putative GPx genes. We examined the biochemical characteristics of one of these gene products and found that it lacks GPx activity and functions as a TPx. Transgene-dependent overexpression of the newly identified Glutathione peroxidase homolog with thioredoxin peroxidase activity (Gtpx-1) gene increases resistance to experimentally induced oxidative stress, but does not compensate for the loss of catalase, an enzyme which, like GTPx-1, functions to eliminate hydrogen peroxide. The results suggest that GTPx-1 is part of the Drosophila Trx antioxidant defense system but acts in a genetically distinct pathway or in a different cellular compartment than catalase.

The estrogen receptorrelated receptor γ (ERγ/ERR3/NR3B3), a member of the nuclear receptor superfamily, activates transcription in the absence of ligands. In order to identify ligand-independent mechanisms of activation, we tested whether calmodulin (CaM), a key regulator of numerous cellular processes and a predominant intracellular receptor for Ca2+-signals, interacts with ERRγ. In vitro pulldown experiments with calmodulin-Sepharose demonstrated a Ca2+-dependent interaction with cellularly expressed ERRγ. As shown by truncation analysis, the CaM binding site is highly unusual in that it is composed of two discontinuous elements. Moreover, by surface plasmon resonance (SPR) biosensor technology, we detected a direct interaction of immobilized bacterially expressed ERRγ fusion protein with Ca2+-calmodulin. This is best described by a model which assumes a conformational change of the initially formed complex to a more stable form. Whereas in vitro DNA binding was calmodulinindependent, transient transfection analysis revealed a Ca2+-influxdependent ERRγmediated transcriptional activation of a luciferase reporter gene. Thus, we propose that CaM acts as a mediator in the Ca2+-dependent modulation of ERRγ.

In the human B cell line P493-6 two mitogenic signals, the Epstein-Barr virus nuclear antigen 2 (EBNA2) and myc, can be independently regulated by means of an estrogen receptor fusion construct or an inducible expression vector, respectively. Shut off of EBNA2, either in the presence or absence of myc, leads to a significant increase in enzymatic activity and surface expression of ecto-5'nucleotidase (CD73) as well as an increased adenosine receptor response in cyclic AMP formation. Shut off of myc expression has a small additional positive effect on CD73 activity. Among the four different subtypes of adenosine receptors, the A2a receptor exclusively is subject to regulation in this system, which is substantiated by pharmacologic data (specific agonists and inhibitors), as well as on the mRNA level. With upregulated CD73 and A2a, cells also respond to 5'-AMP with increased cyclic AMP formation. Turn on of EBNA2 has the reverse effect of repression of CD73 and A2a expression. The time course of both induction and repression of CD73 and A2a is rather slow.

Bauhinia ungulata factor Xa inhibitor (BuXI) inactivates factor Xa and LOPAP, a prothrombin activator proteinase isolated from the venom of Lonomia obliqua caterpillar bristles. The reactive site of the enzyme -inhibitor interaction was explored to design specific substrates for both enzymes. Methionine is crucial for LOPAP and factor Xa substrate interaction, since the change of both Met residues in the substrates abolished the hydrolysis. Synthetic substrates containing the sequence around the reactive site of BbKI, a plasma kallikrein inhibitor, were shown to be specific for trypsin hydrolysis. Therefore, these substrates may be an alternative in studies aiming at a characterization of trypsin-like enzyme activities, especially non-mammalian enzymes.

EMPO [5-(ethoxycarbonyl)-5-methyl-1-pyrroline N-oxide] is a highly hydrophilic cyclic nitrone spin trap, whose superoxide adduct is considerably more stable (t1/2=8.6 min) than DMPO (5,5-dimethyl-1-pyrroline Noxide, t1/2=45 s). EPR spectra of spin adducts of EMPO and its derivatives are very similar to those of the respective DMPO spin adducts, in contrast to the rather complex spectra obtained using DEPMPO [5- (diethoxyphosphoryl)-5-methyl-1-pyrroline Noxide]. Several EMPO derivatives, with both the ethoxycarbonyl group and the methyl group at position 5 of the pyrroline ring being replaced by other substituents, were synthesized and characterized by 1H and 13C NMR spectroscopy. Thus, a series of derivatives was obtained that exhibit large differences in the stability of their superoxide adducts, ranging from less than one to more than 25 min. The stability of the superoxide adducts was mainly determined by the steric environment of the nitroxyl group: in compounds with less bulky 5-alkoxycarbonyl substituents the nitroxyl group is sterically less shielded, which resulted in a lower stability of the superoxide adducts. The spin density distribution, as obtained from DFT computations, was found to be nearly identical for all compounds, so that in contrast to the steric influences the spin density did not seem to be a crucial factor for the stability of the superoxide adducts.