Sequestration of adenoviral vector by Kupffer cells leads to a nonlinear dose response of transduction in liver.

Systemic administration of a recombinant adenovirus encoding the human interferon-beta gene (H5.110CMVhIFN-beta) results in transduction of hepatocytes and detectable circulating levels of IFN-beta protein. In preclinical studies in mice, we noticed a distinctly nonlinear dose response, with low levels of virus (1-3 x 10(10) viral particles) yielding barely detectable levels of IFN-beta but with a higher viral dose (1 x 10(11) particles) resulting in disproportionately high IFN-beta levels. Further studies showed that transgene expression levels from low viral doses could be dramatically enhanced by coadministering an unrelated recombinant adenovirus (H5.110CMVlacZ), suggesting that there was a viral dose threshold effect for efficient viral transduction and/or IFN-beta expression. This enhancement of reporter expression by a nonreporter adenovirus, effective upon coadministration, was further enhanced by preadministration of H5.110CMVlacZ (up to 8 h), but was ineffective if the helper virus was administered as little as 5 min after the H5.110CMVhIFN-beta reporter virus. Our data suggest that the reticuloendothelial system plays a role in this threshold effect, such that low doses of virus are efficiently taken up by the RES/Kupffer cells without leading to appreciable transgene expression, whereas high doses saturate these cells and are able to productively transduce hepatocytes. A better understanding of this phenomenon could have an impact on gene therapy clinical trial safety and efficacy.

Fibronectin type III repeats mediate RGD-independent adhesion and signaling through activated beta1 integrins.

Many cell-surface and extracellular matrix proteins contain multiple modular domains known as fibronectin type III (FNIII) repeats. Cells adhere to the extracellular matrix proteins fibronectin and tenascin in part by the interaction of certain integrins with the Arg-Gly-Asp (RGD) sequence, displayed on specific FNIII repeats. We have found that, after experimental activation of beta1 integrins, a number of cell types adhere and spread on FNIII repeats lacking RGD, derived from extracellular matrix proteins and cytokine receptors. Interaction between individual FNIII domains and beta1 integrins mediates focal adhesion kinase phosphorylation and subsequent stress fiber and focal contact formation. These data suggest that many FNIII-containing proteins may bind and signal through activated beta1 integrins, dramatically expanding the potential for integrin-dependent intercellular and cell-matrix communication.

The alpha1beta1 integrin is expressed during neointima formation in rat arteries and mediates collagen matrix reorganization.

Remodeling of the extracellular matrix by activated mesenchymal cells (myofibroblasts) is a critical aspect of wound repair in all adult organs. Collagen-dependent gel contraction, a process requiring integrin function, is an established in vitro assay thought to mimic in vivo matrix remodeling. Numerous data have implicated the alpha2beta1 integrin in various cell types as the primary collagen receptor responsible for collagen gel contraction. However, evidence from the literature suggests that the major collagen binding integrin expressed on mesenchymally derived cells in situ is the alpha1beta1 integrin, not the alpha2beta1 integrin. In this report, we use a rat vascular injury model to illustrate that the alpha1beta1 integrin is the major collagen receptor expressed on vascular smooth muscle cells after injury. Using two smooth muscle cell lines, expressing either the alpha1beta1 integrin alone or both the alpha1beta1 and alpha2beta1 integrins, along with Chinese hamster ovary cells transfected with the alpha1 integrin, we demonstrate that alpha1beta1 supports not only collagen-dependent adhesion and migration, but also gel contraction. These data suggest that in vivo the alpha1beta1 integrin is a critical collagen receptor on mesenchymally derived cells potentially involved in matrix remodeling after injury.

Identification of residues in the estrogen receptor that confer differential sensitivity to estrogen and hydroxytamoxifen.

We have generated mutant mouse estrogen receptors which differ in their sensitivity to estrogen and the antiestrogen 4-hydroxytamoxifen. Mutation of the glycine at position 525 and the methionine and/or serine at positions 521/522 virtually abolishes the ability of the receptor to bind estradiol and stimulate transcription. In contrast, the mutant receptors retain the partial agonist activity exhibited by the wild-type receptor in the presence of 4-hydroxytamoxifen. The mutations do not affect the expression and DNA-binding activity of the receptor, but do abolish the estrogen-induced increase in the mobility of the receptor-DNA complex observed with the wild-type receptor. Other mutant receptors that were able to bind and stimulate transcription in the presence of estradiol also failed to show the agonist-induced increase in the mobility of the receptor-DNA complex, suggesting that it is unlikely to reflect the formation of a hormone-dependent transcriptional activation function.

Effect of ligand binding and DNA binding on the structure of the mouse oestrogen receptor.

We have investigated the effects of ligand and DNA binding on the structure of the oestrogen receptor by performing limited proteolysis and analysing DNA binding activity by gel shift analysis. The effects of oestradiol, 4-hydroxytamoxifen and ICI 164,384 have been examined and we have found that despite differences in the DNA binding activity or relative mobility of the receptor-DNA complex we were unable to detect differences in the cleavage pattern produced by trypsin, chymotrypsin, Staphylococcus aureus V8, papain or elastase. Inhibition of DNA binding by ICI 164,384 was lost in receptor fragments that lacked the hormone binding domain. In contrast to the full-length receptor, proteolytic fragments produced by chymotrypsin differed in their ability to bind to an oestrogen response element (ERE) vs a thyroid response element (TRE). Evidence is presented that this difference can be accounted for by the inability of fragments lacking the hormone binding domain to dimerise on a TRE.

Analysis of oestrogen receptor dimerisation using chimeric proteins.

Sequences essential for dimerisation have been identified in the hormone binding domain of the mouse oestrogen receptor by insertional and point mutagenesis and sequence comparisons reveal that equivalent residues may be conserved in other members of the nuclear hormone receptor superfamily. To assess functional compatibility of this region between members of the receptor superfamily, peptide sequences corresponding to the equivalent regions of the human androgen receptor and retinoic acid receptor have been substituted for the dimerisation domain of the mouse oestrogen receptor. The resulting chimeric proteins were analysed for high affinity DNA binding using a gel retardation assay and shown to bind with reduced affinity compared to the wild type oestrogen receptor. The reduction in DNA binding observed may result from the intramolecular incompatibility of functional elements within the hormone binding domain of nuclear hormone receptors.

Mechanism of estrogen receptor-dependent transcription in a cell-free system.

RNA synthesis was stimulated directly in a cell-free expression system by crude preparations of recombinant mouse estrogen receptor (ER). Receptor-stimulated transcription required the presence of estrogen response elements (EREs) in the test template and could be specifically inhibited by addition of competitor oligonucleotides containing EREs. Moreover, polyclonal antibodies directed against the DNA-binding region of ER inhibited ER-dependent transcription. In our cell-free expression system, hormone-free ER induced transcription in a hormone-independent manner. Evidence is presented suggesting that ER acts by facilitating the formation of a stable preinitiation complex at the target gene promoter and thus augments the initiation of transcription by RNA polymerase II. These observations lend support to our current understanding of the mechanism of steroid receptor-regulated gene expression and suggest strong conservation of function among members of the steroid receptor superfamily.

A 22-amino-acid peptide restores DNA-binding activity to dimerization-defective mutants of the estrogen receptor.

We have identified residues within the estrogen receptor that are required for dimerization and high-affinity DNA binding. A 22-amino-acid peptide encompassing these residues was sufficient to restore DNA-binding activity to a mutant receptor lacking most of the hormone-binding domain. Point mutagenesis of the fusion protein confirmed that this sequence continued to mediate dimerization in a manner similar to that within the native receptor, although its position relative to the DNA-binding domain was appreciably altered.

Inhibition of estrogen receptor-DNA binding by the "pure" antiestrogen ICI 164,384 appears to be mediated by impaired receptor dimerization.

Many estrogen-antagonist and -agonist ligands have been synthesized, some of which have proved clinically important in the treatment of hormone-dependent breast tumors and endocrine disorders. Here we show that the "pure" antiestrogen ICI 164,384 inhibits mouse estrogen receptor-DNA binding in vitro. The effects of this steroid on DNA binding can be overcome by addition of anti-receptor antibody whose epitope lies N-terminal to the receptor DNA-binding domain. Since this antibody is also capable of restoring DNA-binding activity to receptor mutants that either lack the dimerization domain or bear deleterious mutations within it, we propose that ICI 164,384 reduces DNA binding by interfering with receptor dimerization. In contrast, when complexed with the antagonist/partial agonist tamoxifen, the estrogen receptor is capable of binding to DNA in vitro, but tamoxifen does not promote the agonist-induced conformational change obtained with estradiol. The implications of these data are discussed in relation to the in vivo properties of these drugs.

Characterization and colocalization of steroid binding and dimerization activities in the mouse estrogen receptor.

We have identified a region within the steroid binding domain of the mouse estrogen receptor that is required for both receptor dimerization and high affinity DNA binding. Analysis of sequences in this region revealed that a heptad repeat of hydrophobic residues was conserved in all members of the nuclear receptor superfamily. Single amino acid substitutions of residues in the N-terminal half, but not the C-terminal half, of the repeat prevented receptor dimerization. Steroid binding was abolished by point mutations in the center of the conserved region, implying that the steroid binding and dimerization domains overlap. The role of this region in steroid receptor function is discussed in relation to other models of protein dimerization and DNA binding.

Identification of two transactivation domains in the mouse oestrogen receptor.

We have identified two discrete transactivation domains within the mouse oestrogen receptor whose relative activities vary according to the target promoter. One domain lies within the N-terminal region and is active in the absence of oestradiol. The second domain is contained within the C-terminal portion of the protein and depends upon oestrogen binding for its activity. The location and oestrogen dependence of this domain has been confirmed using chimaeric receptors containing the Lex A DNA binding domain. Although transactivation by the C-terminal domain is dependent upon ligand binding the analysis of receptor deletion mutants has demonstrated that these two functions are not entirely coincident.

A proposed consensus steroid-binding sequence--a reply.

A consensus sequence has been proposed as necessary for steroid binding on the basis of sequence homology in steroidogenic enzymes, steroid binding proteins and steroid receptors. We have mapped the limit of sequences important for steroid binding in the mouse estrogen receptor using Scatchard analysis of deletion mutants. Since a mutant lacking the entire proposed consensus is still able to bind estradiol with essentially wild type affinity, the functional significance of this motif is unclear.

Identification of constitutive and steroid-dependent transactivation domains in the mouse oestrogen receptor.

We have identified two transactivation domains in the mouse oestrogen receptor whose activities depend on the target promoter. The major domain is contained within the C-terminal portion of the protein and depends upon oestrogen binding for its activity. The location and oestrogen dependence of this domain has been confirmed using chimaeric receptors containing the Lex A DNA binding domain. Although transactivation by the C-terminal domain is dependent upon ligand binding the analysis of receptor deletion mutants has demonstrated that these two functions are not entirely coincident. The second transactivation domain lies within the N-terminal region and is active in the absence of oestradiol. The differences in oestrogen requirement for the activity of the two transactivation domains may account for the partial agonist activity of certain antihormones.

A specific insulin receptor and tyrosine kinase activity in the membranes of Neurospora crassa.

Cells of the wall-less ("slime") strain of Neurospora crassa possess specific high affinity insulin binding sites on their cell surface. 125I-labeled bound insulin was not displaced from these cells by insulin-like growth factor II (IGF-II), and was only weakly displaced by IGF-I and proinsulin. Cross-linking of 125I-labeled insulin with N. crassa cells using disuccinimidyl suberate resulted in the labeling of a single band of ca. 67 kDa m.w. on a polyacrylamide gel. Two proteins of ca. 66 and 59 kDa m.w. were purified from detergent solubilized plasma membrane preparations by passage over an insulin-agarose affinity matrix. Antibodies against an autophosphorylation site on the human and Drosophila insulin receptors (anti P2) immunoprecipitated a single phosphoprotein of ca. 50 kDa m.w. from detergent solubilized plasma membranes, which possessed protein tyrosine kinase activity when histone H2 was used as substrate.

Effects of mammalian insulin on metabolism, growth, and morphology of a wall-less strain of Neurospora crassa.

Addition of mammalian insulin to a nutritionally rich, chemically defined culture medium affects Neurospora crassa "slime" (wall-less) cells, as indicated by enhancement of growth, extension of viability at the stationary phase of growth, alteration of morphology, and stimulation of glucose oxidation. Bovine, porcine, and recombinant human insulin had similar effects on growth and morphology, while proinsulin, reduced insulin, and several other proteins were inactive. Insulin added in the presence of excess antiinsulin antibody was without activity. Intact cells possessed high affinity insulin-binding sites, represented by a curvilinear Scatchard plot, suggesting that effects are mediated through insulin receptors on the cell surface. These findings establish a role for insulin or insulin-like molecules in regulating growth and metabolism in this fungal cell and demonstrate a close similarity to insulin effects on certain mammalian cells.

Stimulation by mammalian insulin of glycogen metabolism in a wall-less strain of Neurospora crassa.

Addition of bovine insulin to cells of the wall-less variant FGSC4761 of Neurospora crassa ("slime") produced several significant effects on glycogen metabolism. 1) Intracellular levels of the glycogen precursor UDP-glucose decreased 17-18% (P less than 0.01) within 30 min of insulin addition. 2) Cells grown with insulin possessed 40% more glycogen than did control cells. 3) The incorporation of 14C-labeled glucose into glycogen increased 41% after 30-min treatment with 100 nM bovine insulin (P less than 0.01). 4) Insulin treatment of the cells caused activation of the enzyme glycogen synthase from a glucose-6-phosphate-dependent form to an independent form. Half-maximum activation occurred with 2 nM insulin. These are similar to insulin-induced effects in some mammalian cells. In contrast, no insulin-induced effect on glucose transport could be demonstrated in these cells.

Formation of rat copulatory plug: purified seminal vesicle secretory proteins serve as transglutaminase substrates.

An in vitro system has been used to study the role of purified rat seminal vesicle proteins in the formation of the copulatory vaginal plug. Proteins II, IV (or S) and V (or F) were each separately coagulated using the transglutaminase in coagulating gland extracts. In each case the coagulum required Ca2+ ions for its formation and was insoluble in denaturing solvents. In experiments with [3H]lysine, proteins II and S incorporated [3H]lysine into glu-lys dipeptide with similar kinetics. Both the N-terminal and C-terminal glutamine residues of protein S participated in the reaction.

Comparison of seminal vesicle secretory proteins of rodents using antibody and nucleotide probes.

The copulatory vaginal plug is a conspicuous feature of rodent reproduction. The five major seminal vesicle secretory proteins of Rattus norvegicus (proteins I-V), which form the copulatory plug, constitute a closely related androgen-regulated family that appears to share a common evolutionary origin. The relationships between these rat proteins and the major seminal vesicle proteins of other rodents were explored using antibodies specific for the individual rat proteins. Immunoblotting of proteins separated by SDS-PAGE showed that the vesicular proteins of R. rattus are identical to those of R. norvegicus except for an additional protein related to protein III. No differences were seen in inbred and outbred strains of R. norvegicus. Of the major proteins of Mus musculus, one showed strong homology with rat protein II and three others were weakly homologous to proteins I, IV (or S) and V (or F); none showed homology to rat protein III. The only homology between the vesicular proteins of Mesocricetus auratus (Syrian hamster) and Meriones ungulatus (Mongolian gerbil) was with rat protein II while those of Cavia porcellus (guinea pig) showed no homology at all with the rat proteins. In addition, cDNA probes for rat genes IV and V both detected weak homologues in seminal vesicle RNA from mice but not guinea pigs.(ABSTRACT TRUNCATED AT 250 WORDS)

Tissue distribution, developmental profile and hormonal regulation of androgen-responsive secretory proteins of rat seminal vesicles studied by immunocytochemistry.

The seminal vesicles of the rat synthesise large amounts of androgen-regulated secretory proteins. Indirect immunofluorescence cytochemistry and immunoblotting with monospecific polyclonal antibodies against three of the major secretory proteins (II, S and F) have been used to investigate the tissue distribution, subcellular localisation, androgen-regulation and developmental profile of secretory protein synthesis. There was no evidence for regional specialisation of the seminal vesicle epithelium; every epithelial cell synthesizes all three proteins via a classical secretory involving storage in secretory vesicles. Proteins S and II are contained within the same secretory vesicles. The time course of deinduction of proteins S and F after castration and their reinduction by testosterone closely followed that for their specific mRNAs described previously. During development, proteins S and F first appear between 10 and 15 days after birth. A protein immunologically related to seminal vesicle protein II is present in the lateral and dorsal lobes of the prostatic complex.