Parathyroid hormone induction of rat interstitial collagenase mRNA in osteosarcoma cells is mediated through an AP-1-binding site.

Interstitial collagenase is secreted by the osteoblast in response to bone-resorbing agents. Previously, we cloned the rat interstitial collagenase cDNA from UMR 106-01 rat osteoblastic osteosarcoma cells. We demonstrated that induction of collagenase by PTH, a powerful resorbing agent, in UMR 106-01 cells is in part transcriptional. In the present study we isolate and characterize the rat interstitial collagenase gene. The gene consists of 10 exons and spans approximately 12 kbp. The major transcriptional start site, determined by primer extension analysis and confirmed by RNase protection assay, is 25 nucleotides upstream of the translational start site. The previously isolated cDNA was missing the 5'-untranslated sequence in addition to 17 nucleotides of the signal sequence of the preproenzyme; therefore, we also present these data. Chloramphenicol acetyl transferase (CAT) analyses were performed on the 5'-upstream region of the gene. These data indicate that PTH appears to mediate its effect through an AP-1 consensus-binding sequence (-51). Footprint analysis demonstrates protein binding to this site. Site-specific mutagenesis markedly decreased protein binding, which correlated directly with a decrease in CAT activation by PTH. Supershift data indicate that cAMP response element binding protein (CREB) is binding to this AP-1 consensus sequence. In addition we demonstrate that PTH induces phosphorylation of CREB.

Parathyroid hormone induces transcription of collagenase in rat osteoblastic cells by a mechanism using cyclic adenosine 3',5'-monophosphate and requiring protein synthesis.

Collagenase is synthesized and secreted by rat osteoblastic cells in response to PTH. We have previously demonstrated that this effect involves a substantial increase in collagenase mRNA via transcription. Northern blots and nuclear run-on assays were performed to further investigate the induction of collagenase by PTH in the rat osteoblastic cell line UMR 106-01. Detectable amounts of collagenase mRNA were not apparent until 2 h of PTH treatment, showed the greatest abundance at 4 h, and declined to approximately 30% of maximum by 8 h. The changes in the rate of transcription of the collagenase gene in response to PTH paralleled and preceded the changes in the steady state mRNA levels. After an initial lag period of about 1 h, collagenase transcription rates increased from very low levels to a maximal response at 2 h, returning to about 50% of maximum by 10 h. The increased transcriptional rate of the collagenase gene was found to be dependent on the concentration of PTH, with a half-maximal response at approximately 7 x 10(-10) M rat PTH-(1-34) and a maximal effect with a dose of 10(-8) M. The PTH-mediated induction of collagenase transcriptional activity was completely abolished by cycloheximide, while transcription of the beta-actin gene was unaffected by the translation inhibitor. These data suggest that a protein factor(s) is required for PTH-mediated transcriptional induction of collagenase. Since PTH increases intracellular levels of several potential second messengers, agents that mimic these substances were employed to determine which signal transduction pathway is predominant in the PTH-mediated stimulation of collagenase transcription.(ABSTRACT TRUNCATED AT 250 WORDS)

Parathyroid hormone induces c-fos and c-jun messenger RNA in rat osteoblastic cells.

PTH is a potent regulator of osteoblast gene expression, yet the nuclear events that mediate PTH action are poorly understood. We were interested in identifying immediate early genes which may regulate PTH-altered gene expression in the osteoblast. Therefore, we examined the effects of PTH on c-fos and c-jun gene expression in a rat osteoblastic cell line (UMR 106-01). Under control conditions, c-fos and c-jun mRNAs were present at low basal levels. After PTH treatment, c-fos mRNA abundance dramatically increased, with a maximal and transient response at 30 min. PTH also stimulated an increase in c-jun mRNA, but in a biphasic manner, with maximal levels at 30 min and 2 h. These responses were dose dependent, not altered by cotreatment with the protein synthesis inhibitor cycloheximide, and preceded PTH-induced expression of matrix metallo-proteinase-1 mRNA. Nuclear run-on assays demonstrated an increased rate of c-fos and c-jun transcription after PTH exposure. To determine the signal transduction pathways involved, second messenger analogs were tested for their ability to mimic the effects of PTH. 8-Bromo-cAMP and phorbol 12-myristate 13-acetate (PMA) caused increases in the abundance of c-fos and c-jun transcripts. Ionomycin had no effect on the expression of these genes. Pretreatment of the cells with PMA resulted in a decrease in basal c-jun expression, but did not alter the PTH-mediated increase in c-fos, c-jun, or matrix metalloproteinase-1 mRNAs.(ABSTRACT TRUNCATED AT 250 WORDS)

Prostanoid-induced expression of matrix metalloproteinase-1 messenger ribonucleic acid in rat osteosarcoma cells.

Individual prostanoids have distinct potencies in activating intracellular signaling pathways and regulating gene expression in osteoblastic cells. The E-series prostaglandins (PGs) are known to stimulate matrix metalloproteinase-1 (MMP-1) synthesis and secretion in certain rodent and human osteoblastic cells, yet the intracellular events involved remain unclear. To further characterize this response and its signal transduction pathway(s), we examined prostanoid-induced expression of the MMP-1 gene in the rat osteoblastic osteosarcoma cell line UMR 106-01. Northern blot analysis demonstrated that prostaglandin E2 (PGE2) and PGE1 were very potent stimulators (40-fold) of MMP-1 transcript abundance, PGF2 alpha and prostacyclin were weak stimulators (4-fold), and thromboxane-B2 had no effect. The marked increase in MMP-1 transcript abundance after PGE2 treatment was first detected at 2 h, became maximal at 4 h, and persisted beyond 24 h. This response was dose dependent and elicited maximal and half-maximal effects with concentrations of 10(-6) and 0.6 x 10(-7) M, respectively. Cycloheximide, a protein synthesis inhibitor, completely blocked this effect of PGE2, suggesting that the expression of other genes is required. Nuclear run-on experiments demonstrated that PGE2 rapidly activates MMP-1 gene transcription, with a maximal increase at 2-4 h. The second messenger analog, 8-bromo-cAMP, mimicked the effects of PGE2 by stimulating a dose-dependent increase in MMP-1 messenger RNA (mRNA) levels, with a maximal effect quantitatively similar to that observed with PGE2. Thus, in UMR 106-01 cells, different prostanoids have distinct potencies in stimulating MMP-1 mRNA abundance. Our data suggest that PGE2 stimulation of MMP-1 synthesis is due to activation of MMP-1 gene transcription and a subsequent marked increase in MMP-1 mRNA abundance. This effect is dependent on de novo protein synthesis and is mimicked by protein kinase-A activation.

Retinoic acid stimulates interstitial collagenase messenger ribonucleic acid in osteosarcoma cells.

The rat osteoblastic osteosarcoma cell line UMR 106-01 secretes interstitial collagenase in response to retinoic acid (RA). The present study demonstrates by Northern blot analysis that RA causes an increase in collagenase messenger RNA (mRNA) at 6 h, which is maximal at 24 h (20.5 times basal) and declines toward basal level by 72 h. This stimulation is dose dependent, with a maximal response at 5 x 10(-7) M RA. Nuclear run-on assays show a greater than 20-fold increase in the rate of collagenase mRNA transcription between 12-24 h after RA treatment. Cycloheximide blocks RA stimulation of collagenase mRNA, demonstrating the need for de novo protein synthesis. RA not only causes an increase in collagenase secretion, but is known to decrease collagen synthesis in UMR 106-01 cells. In this study, the increase in collagenase mRNA is accompanied by a concomitant decrease in the level of alpha 1(I) procollagen mRNA, which is maximal at 24 h (70% decrease), with a return to near-control levels by 72 h. Nuclear run-on assays demonstrated that the decrease in alpha 1 (I) procollagen expression does not have a statistically significant transcriptional component. RA did not statistically decrease the stability of alpha 1 (I) procollagen mRNA (calculated t1/2 = 8.06 +/- 0.30 and 9.01 +/- 0.62 h in the presence and absence of RA, respectively). However, transcription and stability together probably contribute to the major decrease in stable alpha 1 (I) procollagen mRNA observed. Cycloheximide treatment inhibits basal level alpha 1 (I) procollagen mRNA accumulation, demonstrating the need for on-going protein synthesis to maintain basal expression of this gene.

Parathyroid hormone induces rat interstitial collagenase mRNA through Ets-1 facilitated by cyclic AMP response element-binding protein and Ca(2+)/calmodulin-dependent protein kinase II in osteoblastic cells.

Parathyroid hormone (PTH), a powerful bone-resorbing agent, is capable of stimulating interstitial collagenase (MMP-13) mRNA production in osteoblastic cells. In this study, a PEA3 consensus binding sequence (-80; AGGAAGT) in addition to a 'TRE-like' sequence (-89; CGACTCA) in the 5' upstream regulatory region of the rat MMP-13 gene were examined. In response to PTH, there was a time-dependent increase in binding of nuclear factors to an oligonucleotide containing the PEA3 region (-95 to -71). This increase in binding was first observed at 0.5 h, peaked at 4 h (7. 6-fold) then returned to basal levels by 24 h. Mutagenesis of the PEA3 site in a chloramphenicol acetyl transferase (CAT) construct containing 5' upstream regulatory sequence of the rat MMP-13 gene significantly decreased activation by PTH. PTH-mediated binding of nuclear factors to an oligonucleotide containing the mutant PEA3 sequence was decreased as compared with the wild type. Mutation or deletion of the TRE-like sequence affected basal as well as PTH-mediated induction of corresponding CAT constructs. Treatment with KN93, a Ca(2+)/calmodulin-dependent protein kinase II specific inhibitor, greatly reduced the amount of protein binding to the PEA3 region in response to PTH which correlated to a notable decrease in the amount of MMP-13 mRNA produced in response to PTH. Antibodies against Ets-1, cyclic AMP response element (CREB)-binding protein (CBP) and CREB were capable of supershifting proteins binding to the oligonucleotide containing the PEA3 region. These data suggest a possible co-operative interaction of factors binding to the PEA3 and TRE-like sequences and provide the first indication of a role for a calcium-mediated pathway in the PTH induction of MMP-13 mRNA in osteoblastic cells.

Functional analysis of the adenovirus type 5 DNA-binding protein: site-directed mutants which are defective for adeno-associated virus helper activity.

We generated four point mutations in the DNA-binding protein (DBP) gene of adenovirus type 5 by oligonucleotide-directed site-specific mutagenesis. The sites mutated were in the three conserved regions (CR; amino acids 178-186 [CR1], 322-330 [CR2], and 464-475 [CR3]) identified previously by comparative sequence analysis (G. R. Kitchingman, Virology 146:90-101, 1985). The mutations resulted in changes in amino acids 181 (Trp to Leu), 323 (Arg to Leu), 324 (Trp to Leu), and 469 (Phe to Ile). The mutated DBP genes were put under the control of the simian virus 40 early promoter and analyzed by transfection for their ability to help adeno-associated virus replicate its DNA in COS-1 monkey cells. Mutations in the aromatic amino acids 324 and 469 reduced the amount of AAV DNA replication approximately 10-fold, while the mutation in Arg 323 produced a reduction of approximately fourfold. The Trp-to-Leu mutation in amino acid 181 had no effect on AAV DNA replication. The decreased helper activity of the 323, 324, and 469 mutations was not caused by any effect of the mutation on the stability of the DBP. These results suggest that CR2 and CR3 are involved in AAV helper activity, specifically in AAV DNA replication. The relevance of these findings to the identification of residues important for the functions of DBP in adenovirus infection is discussed.

Sequence of the DNA-binding protein gene of a human subgroup B adenovirus (type 7). Comparisons with subgroup C (type 5) and subgroup A (type 12).

The adenovirus type 7 (Ad7) single-stranded DNA-binding protein (DBP) structural gene has been sequenced and located between 66.7 and 62.3 map units. This region codes for a protein that contains 517 amino acid residues with a calculated molecular mass of 58,240 daltons. We compared the Ad7 amino acid sequence with those reported for the Ad5 (Kruijer, W., van Schaik, F.M.A., and Sussenbach, J.S. (1981) Nucleic Acids Res. 9, 4439-4457) and Ad12 (Kruijer, W., van Schaik, F.M.A., Speijer, J.G., and Sussenbach, J.S. (1983) Virology 128, 140-153) DNA-binding proteins. A greater amount of amino acid sequence homology was found in the carboxyl-terminal DNA-binding domain of the molecule. This homology is 61% between Ad7 and Ad5 and 49% when Ad12 was included in the comparison. The NH2-terminal domain of DBP retained a 49% homology between Ad7 and Ad5 and a 23% homology for all three serotypes. The greatest difference between the Ad7 and Ad5 DBPs is the absence, in the Ad7 protein, of 12 amino acids located between the two functional domains in the Ad5 protein (amino acids 151-162). In addition, three regions of high amino acid conservation between Ad5, Ad7, and Ad12 consisting of 9 (178-186), 9 (322-330), and 12 (464-475) consecutive amino acids (numbers refer to Ad5) in the DNA-binding portion of the molecule were revealed. These three regions contain a centrally located basic amino acid (183, 326, and 470) as well as an aromatic amino acid residue (181, 324, and 469). Since basic and aromatic amino acids have been implicated in other single-stranded DNA-binding protein/DNA interactions (Anderson, R.A., Nakashima, V., and Coleman, J.E. (1975) Biochemistry 14,907-917; Kowalczykowski, S.C., Lonberg, N., Newport, J.W., and von Hippel, P.H. (1981). J. Mol. Biol. 145, 75-104), these three conserved regions may represent DBP/DNA contact points.

The effect of increased distraction frequency on soft tissues during limb lengthening in an animal model.

The purpose of this study was to compare the effect of distraction frequency on soft tissues. Ten rabbits underwent high-frequency, automated distraction at 1,440 increments per day, whereas five rabbits were manually distracted at a low frequency of three increments per day. The tibias were lengthened approximately 20% at a rate of 1.05 mm/day. The results indicate that all animals had decreased range of motion of the ankle; however, the high-frequency group fared better compared with the low-frequency group (p<0.01). In addition, the P1 latency of the somatosensory evoked potentials did not change in the high-frequency group, but was statistically higher for the low-frequency group (p<0.05). There was no significant difference found in the histologic studies of muscle of either group. Thus increasing the distraction frequency resulted in better range of motion of the ankle joint and maintained normal somatosensory evoked potentials.

Induction of rat interstitial collagenase (MMP-13) mRNA in a development-dependent manner by parathyroid hormone in osteoblastic cells.

The purpose of this study was to determine whether the production of interstitial collagenase mRNA in response to parathyroid hormone (PTH) changes with osteoblast phenotypic development. To accomplish this, cells derived from fetal rat calvaria were examined. The calvarial osteoblasts, which proliferate when placed in culture, can be made to differentiate after confluence. Studies were performed on cells while they were proliferating, at confluence, and during the differentiation process. The cells were treated with PTH for various times, and interstitial collagenase mRNA was quantified by RNase protection assay. We concluded that the ability of PTH to induce interstitial collagenase mRNA in these cells increased with osteoblast phenotypic development. We also determined that the response could be mimicked by combining the effect of 8-bromo-cAMP and 12-O-tetradecanoyl-phorbol-13-acetate, stimulators of the protein kinase A and protein kinase C pathways, respectively, both known to be activated by PTH. The binding of nuclear factors to two regions previously reported to be important for PTH induction of the gene in UMR 106-01 cells was also examined. These data indicated that the binding of nuclear factors to oligonucleotides encompassing the TRE (-51) or the PEA3 (-80) elements changed with development of the osteoblast phenotype. The latter was also shown to be PTH responsive.

Cloning and regulation of rat tissue inhibitor of metalloproteinases-2 in osteoblastic cells.

Rat tissue inhibitor of metalloproteinases-2 (TIMP-2) was cloned from a UMR 106-01 rat osteoblastic osteosarcoma cDNA library. The 969-bp full-length clone demonstrates 98 and 86% sequence identity to human TIMP-2 at the amino acid and nucleic acid levels, respectively. Parathyroid hormone (PTH), at 10(-8) M, stimulates an approximately twofold increase in both the 4.2- and 1.0-kb transcripts over basal levels in UMR cells after 24 h of exposure. The PTH stimulation of TIMP-2 transcripts was not affected by the inhibitor of protein synthesis, cycloheximide (10(-5) M), suggesting a primary effect of the hormone. This is in contradistinction to regulation of interstitial collagenase (matrix metalloproteinase-1) by PTH in these same cells. Nuclear run-on assays demonstrate that PTH causes an increase in TIMP-2 transcription that parallels the increase in message levels. Parathyroid hormone, in its stimulation of TIMP-2 mRNA, appears to act through a signal transduction pathway involving protein kinase A (PKA) since the increase in TIMP-2 mRNA is reproduced by treatment with the cAMP analogue, 8-bromo-cAMP (5 x 10(-3) M). The protein kinase C and calcium pathways do not appear to be involved due to the lack of effect of phorbol 12-myristate 13-acetate (2.6 x 10(-6) M) and the calcium ionophore, ionomycin (10(-7) M), on TIMP-2 transcript abundance. In this respect, regulation of TIMP-2 and collagenase in osteoblastic cells by PTH are similar. However, we conclude that since stimulation of TIMP-2 transcription is a primary event, the PKA pathway must be responsible for a direct increase in transcription of this gene.

Antibodies to distal carboxyl terminal epitopes in the v-fms-coded glycoprotein do not cross-react with the c-fms gene product.

The product of the v-fms oncogene is an integral transmembrane glycoprotein that is closely related to the cell surface receptor for the macrophage colony stimulating factor, CSF-1. A fragment of the v-fms gene encoding a major portion of the extracellular amino terminal domain, the membrane-spanning segment, and the entire carboxyl terminal tyrosine kinase domain of the glycoprotein was molecularly cloned into an inducible prokaryotic expression plasmid. Polypeptide products consisting only of v-fms-coded amino acids were produced in bacteria and were used to prepare immune reagents that precipitated the v-fms-coded glycoproteins expressed in transformed cells. Whereas rabbit antisera to recombinant polypeptides detected antigenic determinants of the c-fms proto-oncogene product, seven mouse monoclonal antibodies to these same antigens reacted only with v-fms-specific epitopes. Proteolytic mapping experiments and studies with a mutant v-fms-coded glycoprotein lacking the 37 carboxyl terminal amino acids of the wild-type product showed that the monoclonal antibodies were restricted in their reactivity to epitopes at the extreme carboxyl terminus of the glycoprotein. The v-fms and c-fms gene products must differ significantly in this region.

Multilineage differentiation activity by cells isolated from umbilical cord blood: expression of bone, fat, and neural markers.

The stromal cell population in bone marrow has been the focus of much attention since it has been shown that this cell population can be expanded and differentiated into cells with the phenotype of bone, cartilage, muscle, stroma, neural, and fat cells. We evaluated umbilical cord blood (UCB) for the presence of these cells. From the mononuclear fraction of UCB, we demonstrated the presence of a subset of cells that have been maintained in continuous culture for more than 6 months (>10 passages). These adherent cell populations express adhesion molecules CD13+, CD29+, and CD44+, but not antigens of hematopoietic differentiation. Exposure of these cells to osteogenic agents resulted in an increase in expression of alkaline phosphatase and the appearance of hydroxyapatite nodules by Von Kossa staining. Incubation with adipogenic agents resulted in morphological change and staining with Oil Red O. In addition, when exposed to basic fibroblast growth factor and human epidermal growth factor the cells underwent changes consistent with cells of neural origin. These changes were demonstrated by a combination of immunofluorescent labeling and Western immunoblots for neural-specific markers. Thus, similar to what has been previously reported with bone marrow, cord blood contains a population of cells that can be expanded in culture and are able to express the phenotype of multiple lineages. Cord blood multilineage cells are slower to establish in culture, have a lower precursor frequency and a lower level of bone antigen expression, and lack constitutive expression of neural antigens when compared to bone marrow, suggesting a more primitive population. Cord blood may prove to be a new source of cells for cellular therapeutics for stromal, bone, and, potentially, neural repair.

Transmembrane orientation of glycoproteins encoded by the v-fms oncogene.

The retroviral oncogene v-fms encodes a glycoprotein whose transport to the plasma membrane is required for transformation. Tryptic digestion of microsomes from transformed cells yielded membrane-protected amino-terminal fragments 40 kd smaller than intact molecules. These fragments were glycosylated, and they included v-fms-coded epitopes expressed at the cell surface. Deletion of the predicted membrane-spanning peptide generated polypeptides that were completely sequestered within microsomes. The mutant glycoproteins acquired more asparagine-linked oligosaccharide chains than did wild-type molecules, lacked kinase activity in vitro, were not transported to the cell surface, and had no transforming activity. Thus, the membrane-spanning segment in the middle of the glycoprotein interrupts translocation of nascent chains into the endoplasmic reticulum, ultimately orienting the amino-terminal domain outside the cell and the carboxy-terminal kinase domain in the cytoplasm. These topological features are similar to those of several growth factor receptors, suggesting that v-fms transforms cells through modified receptor-mediated signals.

Rat collagenase. Cloning, amino acid sequence comparison, and parathyroid hormone regulation in osteoblastic cells.

We have isolated clones for rat collagenase from a rat osteoblastic cell cDNA library. These clones have been sequenced and the amino acids deduced. The calculated molecular weight is 51,352 for the proenzyme and 42,229 for the active enzyme. The deduced amino acid sequence was compared to those previously reported for: 1) human collagenase, 2) rat transin 1 (stromelysin), 3) human stromelysin, and 4) rabbit collagenase. The number of amino acids conserved was 47, 47, 50, and 47%, respectively. We also compared the collagenase mRNA and protein in different rat cells (osteoblast, uterine smooth muscle, synovial fibroblast) and determined that in rat uterine cells the message is slightly larger, although collagenase protein in all three cell types was identical in size. Parathyroid hormone dramatically induces the 2.9-kilobase collagenase mRNA in the rat osteoblastic cells, UMR 106-01. Nuclear run-on studies in UMR 106-01 cells demonstrated a 4-8-fold induction in the rate of synthesis of collagenase mRNA at 2 and 4 after parathyroid hormone treatment, with steady state levels of mRNA increased 100-fold at 4 h. Thus, parathyroid hormone regulation of the collagenase gene in UMR 106-01 cells is in part transcriptional.