Volume 381, Issue 11

The GAGA box of the rat serine protease inhibitor 2 ( spi 2 ) genes not only acts as a basal promoter element, but also mediates transcriptional activation by growth hormone and interleukin-6. The GAGA box is separated from the TATA box by only 12 bp, and this close association is required for efficient transcription. Hence, the GAGA box may influence transcription efficiency through interactions between GAGA box binding proteins and some components of the RNA polymerase II complex. Here we report the cloning of two GAGA box-binding proteins termed p38 and p40, that belong to the type A/B heterogeneous nuclear ribonucleoprotein subgroup. GAGA box mutations that diminish the affinity for p38 and p40 decrease basal and GH-induced reporter gene expression. Furthermore, nuclear extracts depleted of p38 and p40 can no longer support GAGA box-dependent in vitro transcription. Therefore, two polypeptides previously assigned to a family of RNA processing proteins also act as DNA-binding, promoter-specific transcription factors.

Reticulocyte lysates contain ribosome-bound and free populations of 5S RNA. The free population is sensitive to nuclease cleavage in the internal loop B, at the phosphodiester bond connecting nucleotides A54 and A55. Similar cleavage sites were detected in 5S rRNA in 60S subunits and 80S ribosomes. However, 5S rRNA in reticulocyte polysomes is insensitive to cleavage unless ribosomes are salt-washed. This suggests that a translational factor protects the backbone surrounding A54 from cleavage in polysomes. Upon nuclease treatment of mouse 60S subunits or reticulocyte lysates a small population of ribosomes released its 5S rRNA together with ribosomal protein L5. Furthermore, rRNA sequences from 5.8S, 28S and 18S rRNA were released. In 18S rRNA the sequences mainly originate from the 630 loop and stem (helix 18) in the 5′ domain, whereas in 28S rRNA a majority of fragments is derived from helices 47 and 81 in domains III and V, respectively. We speculate that this type of rRNA-fragmentation may mimic a ribosome degradation pathway.

Poly(ADP-ribose) is a nuclear polynucleotide involved in the regulation of chromatin functions via covalent and/or noncovalent modification of nuclear proteins. Using a binding assay on protein blots, we searched for poly(ADP-ribose) binding proteins in nuclear matrices from testes of differently aged rats as well as from various adult rat tissues (brain, liver, spleen). We found that nuclear matrix proteins represent a significant subset of the nuclear proteins that can establish noncovalent interactions with poly(ADP-ribose). The profiles of poly(ADP-ribose) binding nuclear matrix proteins appeared to be tissue-specific and changed during postnatal development in the testis. The isolation and analysis of endogenous poly (ADP-ribose) from rat testes showed that the ADP-ribose polymers that bind nuclear matrix proteins in vitro are also present under physiologic conditions in vivo . These results further substantiate the possibility that poly(ADP-ribose) may affect chromatin functions through noncovalent interaction with specific protein targets, including nuclear matrix components.

Human and porcine aminoacylase 1 (Acy1) contain two peculiar sequence motifs located near the interface between the two major domains of the Acy1 subunit. Each motif consists of the sequence PWW preceded by four strongly polar residues. In order to examine the significance of these sequences for Acy1 stability and/or catalysis, we used site-directed mutagenesis of human Acy1 to replace the tryptophan residues in either motif with alanines. Both mutants showed unchanged zinc binding and normal substrate affinity. Modification in PWW motif 1 (residues 192–194) and motif 2 (residues 321–323) resulted in markedly reduced specific activity in the first case and diminished stability in either mutant. Fluorescence quenching measurements showed that all four tryptophans of the PWW motifs are solvent-accessible. We conclude that PWW motif 1 maintains the native conformation of the active site by creating the proper spatial relationship between dimerization domains and catalytic domains, while motif PWW2 is necessary for the stability of the catalytic domain.

The mitochondrial isoenzymes of creatine kinase (MtCK), ubiquitous uMtCK and sarcomeric sMtCK, are key enzymes of oxidative cellular energy metabolism and play an important role in human health and disease. Very little is known about uMtCK in general, or about sMtCK of human origin. Here we have heterologously expressed and purified both human MtCK isoenzymes to perform a biochemical, kinetic and structural characterization. Both isoenzymes occurred as octamers, which can dissociate into dimers. Distinct Stokes' radii of uMtCK and sMtCK in solution were indicative for conformational differences between these equally sized proteins. Both human MtCKs formed 2D-crystals on cardiolipin layers, which revealed further subtle differences in octamer structure and stability. Octameric human sMtCK displayed p 4 symmetry with lattice parameters of 145 Å, indicating a ‘flattening’ of the octamer on the phospholipid layer. pH optima and enzyme kinetic constants of the two human isoenzymes were significantly different. A pronounced substrate binding synergism ( K d > K m ) was observed for all substrates, but was most pronounced in the forward reaction (PCr production) of uMtCK and led to a significantly lower K m for creatine (1.01mM) and ATP (0.11mM) as compared to sMtCK (creatine, 7.31 mM; ATP, 0.68 mM).

We have cloned and sequenced the gene for the glycerol kinase of Trypanosoma brucei (Tb GLK1 ), obtained by RT-PCR. The corresponding mRNA is 2.3 kb in size and contains an ORF encoding a protein with high homology to known glycerol kinases of other organisms. It is 512 amino acids in length with a PTS1-like targeting sequence (AKL) at its C-terminus, suggesting glycosomal compartmentalization of this enzyme. Although Northern blot analysis revealed higher mRNA levels in slender bloodstream forms than in the procyclic insect forms, specific glycerol kinase activities were found to be virtually identical in both life stages. Southern blot analysis suggested a single copy gene, but we were able to clone two alleles utmost similar to each other. Heterologous expression of the trypanosomal glycerol kinase in E. coli enabled us to perform a kinetic analysis of this enzyme. In particular, we have been able to monitor ATP production from glycerol-3-phosphate and ADP, a reaction which, although thermodynamically very unfavorable, is regarded essential for the survival of Trypanosoma brucei under anoxic conditions. Since the unique spatial separation of glycolysis in the kinetoplastida imposes important consequences for the regulation of the energy metabolism in these organisms, we discuss the observed differences between TbGLK1 and glycerol kinases from other organisms in view of its physiological relevance.

The total protein mixture from the 50S subunit (TP-50) of the eubacterium Thermus thermophilus was characterized after blotting onto PVDF membranes from two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) and sequencing. The proteins were numbered according to their primary structure similarity with their counterparts from other species. One of them has been marked with an asterisk, namely L*23, because unlike the other known ribosomal proteins it shows a very low degree of homology. A highly acidic 5S rRNA binding protein, TL5, was characterized and compared with the available primary structure information. Proteins L1 and L4 migrate similarly on 2D-PAGE. Protein L4, essential for protein biosynthesis, is N-terminally blocked and shows a strikingly low homology to other L4 proteins. In addition to L4, two other proteins, namely L10 and L11, were found to be N-terminally blocked. In conclusion, 33 proteins from the large subunit were identified, including TL5. Homologs to rpL25 and rpL26 were not found.

An iron-containing superoxide dismutase (SOD; EC 1.15.1.1) of the hyperthermophilic archaeon Acidianus ambivalens (Aa-SOD) has been purified and characterized and the gene has been cloned and sequenced. The SOD from the facultatively aerobic member of the crenarchaeota could be expressed in E. coli . Both, the native as well as the heterologously overproduced protein turned out to have extraordinarily high melting temperatures of 128 °C and 124.5 °C, respectively. To the best of our knowledge, this is the highest directly measured melting temperature of a native protein. Surprisingly, neither the native nor the recombinant superoxide dismutase displays 100% occupation of the metal coordination sites. Obviously it is not the incorporation of a metal ion that confers the extreme thermostability. Expression of the superoxide dismutase in the presence of different metals such as Fe, Co, Ni, Mn and Cu offered the possibility of studying the hitherto unknown cofactor preference of iron-superoxide dismutase. The recombinant enzyme displayed the highest preference for incorporation of cobalt although iron is used as the natural cofactor. Spectroscopic analysis by EPR, atomic absorption and UV/Vis spectroscopy as well as activity measurements and differential scanning calorimetry of the metal substituted superoxide dismutases were performed. However, the superoxide dismutase of A. ambivalens is active only with iron but may incorporate other metals equally well in the catalytic center without loss of conformational stability or heat tolerance. The co-form of the enzyme could be crystallized.

Angiogenesis is the sprouting of new capillary blood vessels from pre-existing ones. The kinin family of vasoactive peptides, formed by the serine protease tissue kallikrein from its endogenous multifunctional protein substrate kininogen, is believed to regulate the angiogenic process. The aim of this study was to determine the expression of tissue kallikrein and kinin receptors in an in vitro model of angiogenesis. Microvascular endothelial cells from the bovine mature and regressing corpus luteum were used only if they reacted with known endothelial cell markers. At first the cultured endothelial cells began sprouting, and within four weeks formed three-dimensional, capillary-like structures. Immunolabelling for tissue prokallikrein and the mature enzyme was intense in the angiogenic endothelial cells derived from mature corpora lutea. Immunoreactivity was lower in non-angiogenic endothelial cells and least in angiogenic endothelial cultures of the regressing corpus luteum. Additionally, using specific antisense DIG-labelled probes, tissue kallikrein mRNA was demonstrated in cells of the angiogenic phenotype. Immunolabelled kinin B2 receptors, but not kinin B1 receptors, were visualised on angiogenic endothelial cells. Our results suggest an important regulatory role for kinins in the multiple steps of the angiogenic cascade that may occur in wound healing and cancer cell growth.

The interaction of leech-derived tryptase inhibitor (LDTI) with bovine liver capsule tryptase (BLCT) and bovine trypsin has been studied using both thermodynamic and kinetic approaches. Several differences were detected: (i) the equilibrium affinity of LDTI for BLCT ( K a = 8.9 × 10 5 M −1 ) is about 600-fold lower than that for bovine trypsin ( K a = 5.1 × 10 8 M −1 ); (ii) LDTI behaves as a purely non-competitive inhibitor of BLCT, while it is a purely competitive inhibitor of bovine trypsin. These functional data are compared with those previously reported for the LDTI binding to human tryptase, where tight inhibition occurs at two of the four active sites of the tetramer ( K a = 7.1 × 10 8 M −1 ). Amino acid sequence alignment of BLCT, human βIItryptase and bovine trypsin allows us to infer some possible structural basis for the observed functional differences.

N. Bst NBI is a unique restriction endonuclease isolated from Bacillus stearothermophilus . We have characterized the recognition sequence and the cleavage site of N. Bst NBI. Mapping of cleavage sites of N. Bst NBI showed that it recognizes an asymmetric sequence, 5′ GAGTC 3′, and cleaves only on the top strand 4 base pairs away from its recognition sequence. To verify the nicking activity of N. Bst NBI, we have constructed two plasmids containing a single recognition sequence (pNB1) or no recognition site (pNB0). When pNB1 and pNB0 were incubated with the enzyme, N. Bst NBI nicked only the plasmid pNB1, suggesting that N. Bst NBI is a specific nicking endonuclease.

Recently we described a new signal transduction-based tumor therapeutic strategy involving first sensitization of tumor cells by trichostatin A (TSA), an inhibitor of histone deacetylation, and thereafter efficient apoptotic triggering by ribotoxic agents, which activate stress-activated protein kinases. In the present work we investigate the molecular basis of the sensitization step in this therapeutic model system and describe TSA-induced changes in mRNA and protein expression of several candidate genes identified previously by complex hybridization. Furthermore, activities of 15 different protein kinases were followed after TSA application, using a new filter-based technique (PhosphoSpots ® -Assay). The obtained data suggest that TSA induces pro-apoptotic genes like ID1, ID2, ID3, and down-regulates anti-apoptotic genes like Hsp27 and Bcl-xL, thereby shifting the cellular equilibrium from life to death. Furthermore, activities of calcium/calmodulin-dependent kinase II and protein kinase C, which have been assigned pro-apoptotic function in other systems, are induced.

The co-chaperone murine stress-inducible protein 1 (mSTI1), a Hsp70/Hsp90 organizing protein (Hop) homolog, functions as a physical link between Hsp70 and Hsp90 by mediating the formation of the mSTI1/ Hsp70/Hsp90 chaperone heterocomplex. We show here that mSTI1 is an in vitro substrate of cell cycle kinases. Casein kinase II (CKII) phosphorylates mSTI1 at S 189 , and cdc2 kinase (p34 cdc2 ) at T 198 , substantiating a predicted CKII-p34 cdc2 -NLS (CcN) motif. The possible implications of this phosphorylation as a cell cycle checkpoint are discussed.

It has been suggested that downstream signaling from the insulin receptor to the level of the protein kinases and protein phosphatases is accomplished by prostaglandylinositol cyclic phosphate (cyclic PIP), a proposed second messenger of insulin. However, evidence points also to both phosphatidylinositol 3-kinase, which binds to the tyrosine phosphorylated insulin receptor substrate-1, and the Ras complex in insulin's downstream signaling. We have examined whether a correlation exists between these various observations. It was found that wortmannin, a specific inhibitor of phosphatidylinositol 3-kinase, prevented insulin-induced, as well as cyclic PIP-induced activation of glucose transport, indicating that PI 3-kinase action on glucose transport involves downstream signaling of both insulin and cyclic PIP. Wortmannin has no effect on cyclic PIP synthase activity nor on the substrate production for cyclic PIP synthesis either, indicating that the functional role of PI 3-kinase is exclusively downstream of cyclic PIP.

Cystatins are physiological cysteine proteinase inhibitors. Here we report a novel function for some family 2 cystatins that is not related to these activities. The release of interleukin-6 (IL-6) and interleukin-8 (IL-8) by the gingival fibroblasts and that of IL-6 by murine splenocytes were measured using ELISA systems specific for these cytokine molecules. Family 2 cystatins, including cystatins C, SA1, SA2, S, and egg white cystatin, upregulated the IL-6 production by two human gingival fibroblast cell lines or murine splenocytes and also the IL-8 production by gingival fibroblasts at physiological concentrations. After complete saturation with papain, those family 2 cystatins still upregulated IL-6 production, suggesting that the papain- inhibitory site was not involved in the cytokine-inducing activity.