Molecular interactions between breast cancer cells and the bone microenvironment drive skeletal metastases

Breast cancer cells preferentially spread to bone. Bone metastases are currently incurable and therefore better treatments need to be developed. Metastasis is an inefficient, multi-step process. Specific aspects of both breast cancer cells and the bone microenvironment contribute to the development of bone metastases. Breast cancers express chemokine receptors, integrins, cadherins, and bone-resorbing and bone-forming factors that contribute to the successful and preferential spread of tumor to bone. Bone is rich in growth factors and cell types that make it a hospitable environment for breast cancer growth. Once breast cancer cells enter the bone, a highly complex vicious cycle develops, in which breast cancer cells secrete factors that act on bone cells and other cells within the bone (stem cells, T cells, platelets, adipocytes, fibroblasts, and endothelial cells), causing them to secrete factors that act on adjacent cancer cells. The steps in the metastatic cascade and the vicious cycle within bone offer unique targets for adjuvant treatments to treat and cure bone metastases.

Tumor-bone cellular interactions in skeletal metastases

Human tumor cells inoculated into the arterial circulation of immunocompromised mice can reliably cause bone metastases, reproducing many of the clinical features seen in patients. Animal models permit the identification of tumor-produced factors, which act on bone cells, and of bone-derived factors. Local interactions stimulated by these factors drive a vicious cycle between tumor and bone that perpetuates skeletal metastases. Bone metastases can be osteolytic, osteoblastic, or mixed. Parathyroid hormone-related protein, PTHrP, is a common osteolytic factor, while vascular endothelial growth factor and interleukins 8 and 11 also contribute. Osteoblastic metastases can be caused by tumor-secreted endothelin-1, ET-1. Other potential osteoblastic factors include bone morphogenetic proteins, platelet-derived growth factor, connective tissue growth factor, stanniocalcin, N-terminal fragments of PTHrP, and adrenomedullin. Osteoblasts are the main regulators of osteoclasts, and stimulation of osteoblast proliferation can increase osteoclast formation and activity. Thus, combined expression of osteoblastic and osteolytic factors can lead to mixed metastases or to increased osteolysis. Prostate-specific antigen is a protease, which can cleave PTHrP and thus change the balance of osteolytic versus osteoblastic responses to metastatic tumor cells. Bone itself stimulates tumor by releasing insulin-like growth factors and transforming growth factor-beta. Secreted factors transmit the interactions between tumor and bone. They provide novel targets for therapeutic interactions to break the vicious cycle of bone metastases. Clinically approved bisphosphonate anti-resorptive drugs reduce the release of active factors stored in bone, and PTHrP-neutralizing antibody, inhibitors of the RANK ligand pathway, and ET-1 receptor antagonist are in clinical trials. These adjuvant therapies act on bone cells, rather than the tumor cells. Recent gene array experiments identify additional factors, which may in the future prove to be clinically important targets.

Molecular mechanisms of tumor-bone interactions in osteolytic metastases

In patients with advanced disease, several cancer types frequently metastasize to the skeleton, where they cause bone destruction. Osteolytic metastases are incurable and cause pain, hypercalcemia, fracture, and nerve compression syndromes. It was proposed over a century ago that certain cancers, such as that of the breast, preferentially metastasize to the favorable microenvironment provided by bone. Bone matrix is a rich store of immobilized growth factors that are released during bone resorption. Histological analysis of osteolytic bone metastases indicates that the bone destruction is mediated by the osteoclast rather than directly by the tumor cells. These observations suggest a vicious cycle driving the formation of osteolytic metastases: tumor cells secrete factors stimulating osteoclasts through adjacent bone marrow stromal cells; osteoclastic resorption in turn releases growth factors from the bone matrix; finally, locally released growth factors activate the tumor cells. This vicious cycle model has now been confirmed at the molecular level. In particular, transforming growth factor beta (TGF3beta) is abundant in bone matrix and released as a consequence of osteoclastic bone resorption. Bone-derived TGFbeta plays an integral role in promoting the development and progression of osteolytic bone metastases by inducing tumor production of parathyroid hormone-related protein (PTHrP), a known stimulator of osteoclastic bone resorption. In breast cancer cells TGFbeta appears to stimulate PTHrP secretion by a posttranscriptional mechanism through both Smad and p38 mitogen activated protein (MAP) kinase signaling pathways. Osteolytic metastases can be suppressed in vivo by inhibition of bone resorption, blockade of TGFbeta signaling in tumor cells, and by neutralization of PTHrP. Other factors released from bone matrix may also act on tumor cells in bone, which in turn may produce other factors that stimulate bone resorption, following the vicious cycle paradigm established for TGFbeta and PTHrP. An understanding at the molecular level of the mechanisms of osteolytic metastasis will result in more effective therapies for this devastating complication of cancer.

Reproductive tissue renin gene expression in preeclampsia

OBJECTIVE

Using an analogy with renin gene overexpression, low-renin hypertension animal models, we wished to test the hypothesis that renin gene expression is increased in decidua basalis in human gestation with preeclampsia.

METHODS

Human placentas were obtained immediately after delivery from 11 control (C) and 11 preeclamptics (PE). Tissue samples were microdissected and renin gene expression in decidua basalis (DB), chorionic villi (CV), and decidua vera (DV) was measured using dot-blot hybridization.

RESULTS

Overall renin gene expression is highest in decidua basalis (mean +/- SEM, 2.66 +/- 0.69 densitometry area units) compared to chorionic villi (mean +/- SEM, 1.85 +/- 0.5) or compared to decidua vera (mean +/- SEM, 1.63 +/- 0.9) (both t-tests p = 0.001 two-tailed and analysis of variance p = 0.0001). Renin gene expression in DB and in CV was similar in both preeclamptic and normal pregnancies (DB mean +/- SEM C 2.79 +/- 0.96 versus PE 2.54 +/- 1.04, and CV mean +/- SEM C 2.11 +/- 0.91 versus PE, 1.59 +/- 0. 44). Renin gene expression in DV was approximately threefold higher in tissues from preeclamptics compared to control (mean +/- SEM PE 2. 44 +/- 1.76 versus C 0.82 +/- 0.42). Using the median value of 0.5 units for DV as cutoff, the preeclamptics displayed higher renin gene expression (chi square p = 0.033, two tailed).

CONCLUSION

Our data suggest that renin gene expression is increased in preeclampsia in decidua vera. This may explain previously reported increased renin secretion in uterine circulation in preeclampsia.

Rabbit phosphoglucose isomerase/neuroleukin/autocrine motility factor: cloning via interspecies identity

Phosphoglucose isomerase is the first committed enzyme of glycolysis. The protein also has a variety of biological activities on mammalian cells. The molecular basis of these extracellular functions is unclear, and the high resolution three-dimensional structure of a mammalian enzyme has not been described. We report here the cDNA and protein sequence for phosphoglucose isomerase from rabbit muscle. The sequence was obtained directly by PCR without the need to screen clones from a cDNA library and encoded active enzyme when expressed in bacterial cells. The 558 amino acid rabbit coding sequence is the same length as and highly similar (92% residue identity) to the sequences from human and pig and less so (88%) to the mouse enzyme. Non-conservative amino acid changes between the four mammalian sequences are concentrated in the first 35 and last five residues. The rabbit protein has an additional Cys residue and amino acid changes at five positions otherwise invariant in the mammalian enzymes.

NH2-terminal sequence truncation decreases the stability of bovine rhodanese, minimally perturbs its crystal structure, and enhances interaction with GroEL under native conditions

The NH2-terminal sequence of rhodanese influences many of its properties, ranging from mitochondrial import to folding. Rhodanese truncated by >9 residues is degraded in Escherichia coli. Mutant enzymes with lesser truncations are recoverable and active, but they show altered active site reactivities (Trevino, R. J., Tsalkova, T., Dramer, G., Hardesty, B., Chirgwin, J. M., and Horowitz, P. M. (1998) J. Biol. Chem. 273, 27841-27847), suggesting that the NH2-terminal sequence stabilizes the overall structure. We tested aspects of the conformations of these shortened species. Intrinsic and probe fluorescence showed that truncation decreased stability and increased hydrophobic exposure, while near UV CD suggested altered tertiary structure. Under native conditions, truncated rhodanese bound to GroEL and was released and reactivated by adding ATP and GroES, suggesting equilibrium between native and non-native conformers. Furthermore, GroEL assisted folding of denatured mutants to the same extent as wild type, although at a reduced rate. X-ray crystallography showed that Delta1-7 crystallized isomorphously with wild type in polyethyleneglycol, and the structure was highly conserved. Thus, the missing NH2-terminal residues that contribute to global stability of the native structure in solution do not significantly alter contacts at the atomic level of the crystallized protein. The two-domain structure of rhodanese was not significantly altered by drastically different crystallization conditions or crystal packing suggesting rigidity of the native rhodanese domains and the stabilization of the interdomain interactions by the crystal environment. The results support a model in which loss of interactions near the rhodanese NH2 terminus does not distort the folded native structure but does facilitate the transition in solution to a molten globule state, which among other things, can interact with molecular chaperones.

TGF-beta signaling blockade inhibits PTHrP secretion by breast cancer cells and bone metastases development

Breast cancer frequently metastasizes to the skeleton, and the associated bone destruction is mediated by the osteoclast. Growth factors, including transforming growth factor-beta (TGF-beta), released from bone matrix by the action of osteoclasts, may foster metastatic growth. Because TGF-beta inhibits growth of epithelial cells, and carcinoma cells are often defective in TGF-beta responses, any role of TGF-beta in metastasis is likely to be mediated by effects on the surrounding normal tissue. However, we present evidence that TGF-beta promotes breast cancer metastasis by acting directly on the tumor cells. Expression of a dominant-negative mutant (TbetaRIIDeltacyt) of the TGF-beta type II receptor rendered the human breast cancer cell line MDA-MB-231 unresponsive to TGF-beta. In a murine model of bone metastases, expression of TbetaRIIDeltacyt by MDA-MB-231 resulted in less bone destruction, less tumor with fewer associated osteoclasts, and prolonged survival compared with controls. Reversal of the dominant-negative signaling blockade by expression of a constitutively active TGF-beta type I receptor in the breast cancer cells increased tumor production of parathyroid hormone-related protein (PTHrP), enhanced osteolytic bone metastasis, and decreased survival. Transfection of MDA-MB-231 cells that expressed the dominant-negative TbetaRIIDeltacyt with the cDNA for PTHrP resulted in constitutive tumor PTHrP production and accelerated bone metastases. These data demonstrate an important role for TGF-beta in the development of breast cancer metastasis to bone, via the TGF-beta receptor-mediated signaling pathway in tumor cells, and suggest that the bone destruction is mediated by PTHrP.

Properties of soluble fusions between mammalian aspartic proteinases and bacterial maltose-binding protein

The mammalian aspartic proteinases procathepsin D and pepsinogen form insoluble inclusion bodies when expressed in bacteria. They become soluble but nonnative when synthesized as fusions to the carboxy terminus of E. coli maltose-binding protein (MBP). Since these nonnative states of the two aspartic proteinases showed no tendency to form insoluble aggregates, their biophysical properties were analyzed. The MBP portions were properly folded as shown by binding to amylose, but the aspartic proteinase moieties failed to bind pepstatin and lacked enzymatic activity, indicating that they were not correctly folded. When treated with proteinase K, only the MBP portion of the fusions was resistant to proteolysis. The fusion between MBP and cathepsin D had increased hydrophobic surface exposure compared to the two unfused partners, as determined by bis-ANS binding. Ultracentrifugal sedimentation analysis of MBP-procathepsin D and MBP-pepsinogen revealed species with very large and heterogeneous sedimentation values. Refolding of the fusions from 8 M urea generated proteins no larger than dimers. Refolded MBP-pepsinogen was proteolytically active, while only a few percent of renatured MBP-procathepsin D was obtained. The results suggest that MBP-aspartic proteinase fusions can provide a source of soluble but nonnative folding states of the mammalian polypeptides in the absence of aggregation.

Truncations at the NH2 terminus of rhodanese destabilize the enzyme and decrease its heterologous expression

Rhodanese mutants containing sequential NH2-terminal deletions were constructed to test the distinct contributions of this region of the protein to expression, folding, and stability. The results indicate that the first 11 residues are nonessential for folding to the active conformation, but they are necessary for attaining an active, stable structure when expressed in Escherichia coli. Rhodanese species with up to 9 residues deleted were expressed and purified. Kinetic parameters for the mutants were similar to those of the full-length enzyme. Compared with shorter truncations, mutants missing 7 or 9 residues were (a) increasingly inactivated by urea denaturation, (b) more susceptible to inactivation by dithiothreitol, (c) less able to be reactivated, and (d) less rapidly inactivated by incubation at 37 degreesC. Immunoprecipitation showed that mutants lacking 10-23 NH2-terminal amino acids were expressed as inactive species of the expected size but were rapidly eliminated. Cell-free transcription/translation at 37 degreesC showed mutants deleted through residue 9 were enzymatically active, but they were inactive when deleted further, just as in vivo. However, at 30 degreesC in vitro, both Delta1-10 and Delta1-11 showed considerable activity. Truncations in the NH2 terminus affect the chemical stability of the distantly located active site. Residues Ser-11 through Gly-22, which form the NH2-proximal alpha-helix, contribute to folding to an active conformation, to resisting degradation during heterologous expression, and to chemical stability in vitro.

Purification, amino acid sequence, and cDNA sequence of a novel calcium-precipitating proteolipid involved in calcification of corynebacterium matruchotii

Corynebacterium matruchotii is a microbial inhabitant of the oral cavity associated with dental calculus formation. It produces membrane-associated proteolipid capable of inducing hydroxyapatite formation in vitro. This proteolipid was purified from chloroform:methanol extracts by chromatography on Sephadex LH-20 and migrated on SDS-polyacrylamide gel electrophoresis at 6-9 kDa. Removal of covalently attached acyl moieties by methanolic KOH decreased its molecular mass to approximately 5.5 kDa. The amino acid sequence of the apoproteolipid indicated a peptide of 50 amino acids, a calculated molecular weight of 5354 Da, and an isoelectric point of 4.28. Sequence analysis revealed an 8 amino acid sequence with homology to human phosphoprotein phosphatase 2A as well as several potential acylation sites and one phosphorylation site. The purified proteolipid induced calcium precipitation in vitro. Deacylation of the proteolipid by hydroxylamine treatment resulted in >50% loss of calcium-precipitating activity, suggesting that covalently attached lipids are required. Degenerate oligonucleotide primers, based on the amino acid sequence, were used to amplify the gene for the 5.5 kDa proteolipid from total chromosomal DNA of C. matruchotii by PCR. A 166 bp cDNA was isolated and sequenced, confirming the amino acid sequence of the proteolipid. Thus, we have sequenced a unique bacterial proteolipid that is involved in the formation of dental calculus by precipitating Ca2+ and possibly in transport of inorganic phosphate, necessary for hydroxyapatite formation.

Order of fusions between bacterial and mammalian proteins can determine solubility in Escherichia coli

We made fusions between Escherichia coli maltose-binding protein (MBP) and the mammalian aspartic proteinases pepsinogen or procathepsin D. When MBP was at the N-terminus, the fusions were soluble in E. coli. When the order was reversed, the chimeric proteins formed inclusion bodies. The data suggest that the solubility of fusion proteins is controlled by whether the protein domains emerging first from the ribosome normally fold into soluble or insoluble states. The soluble MBP-aspartic proteinase fusions were stable but proteolytically inactive. MBP-pepsinogen, however, was efficiently renatured from 8 M urea in vitro, suggesting that the E. coli cytoplasm does not support folding of the mammalian partner protein to the native state. Thus, inclusion body formation may be the consequence, rather than the cause, of non-native folding in vivo, and in E. coli soluble proteins may fold into states different from those reached in vitro.

Solubility of proteins isolated from inclusion bodies is enhanced by fusion to maltose-binding protein or thioredoxin

When the mammalian aspartic proteinases, procathepsin D or pepsinogen, are expressed in Escherichia coli both accumulate in inclusion bodies. While pepsinogen is efficiently refolded in vitro, recovery of procathepsin D is limited by insolubility. We expressed procathepsin D and pepsinogen in E. coli, with E. coli maltose-binding protein (MBP) or thioredoxin (trx) fused to their C-termini (aspartic proteinase-MBP or aspartic proteinase-trx). The fusion proteins were still found in inclusion bodies. However, the recovery of soluble procathepsin D-MBP and procathepsin D-trx after refolding was facilitated by the bacterial fusion partners. Maltose-binding protein was more efficient than thioredoxin in increasing the recovery of soluble protein. The vector, pET23bMBPH6, can be used for general expression of heterologous proteins in E. coli. The vector includes a histidine tag at the C-terminus of MBP to allow one-step purification of the fusion proteins under denaturing conditions. After purification, the protein of interest can be cleaved from MBP with factor Xa protease and separated from the MBP partner. Refolded pepsinogen-MBP and pepsinogen-trx were enzymatically active, but procathepsin D-MBP and procathepsin D-trx were soluble but largely inactive. The results show that the limited recovery of activity upon refolding of procathepsin D is not the consequence of competing aggregation. Thus, the fusions do not necessarily facilitate native refolding, but they do enhance the recovery of soluble protein. Such fusions could provide a system to study, in soluble form, folding states which are otherwise inaccessible because of aggregation and precipitation.

Preformed GroES oligomers are not required as functional cochaperonins

We have previously shown that the C-terminal sequence of GroES is required for oligomerization [Seale and Horowitz (1995), J. Biol. Chem. 270, 30268-30270]. In this report, we have generated a C-terminal deletion mutant of GroES with a significantly destabilized oligomer and have investigated its function in the chaperonin-assisted protein folding cycle. Removal of the two C-terminal residues of GroES results in a cochaperonin [GroESD(96-97)] that is monomeric at concentrations where GroES function is assessed. Using equilibrium ultracentrifugation, we measured the dissociation constant for the oligomer-monomer equilibrium to be 7.3 x 10(-34)M6. The GroESD(96-97) is fully active as a cochaperonin. This mutant is able to inhibit the ATPase activity of GroEL to levels comparable to wild-type GroES. It is also able to assist the refolding of urea-denatured rhodanese by GroEL. While GroESD(96-97) can function at levels comparable to wild-type GroES, higher concentrations of mutant are required to produce the same effect. These results support the idea that the performed GroES heptamer is not required for function, but they suggest that the oligomeric cochaperonin is most efficient.

Chinese hamster rhodanese cDNA: activity of the expressed protein is not blocked by a C-terminal extension

We describe a cDNA from Chinese hamster ovary cells which encodes a protein 91 and 96% identical to bovine and rat mitochondrial rhodaneses, respectively. Recombinant protein was expressed from the cDNA in Escherichia coli, purified to homogeneity, and found to have kinetic properties nearly indistinguishable from those of the bovine enzyme, the only cloned rhodanese previously verified by characterization of the recombinant protein. The carboxyl-terminus of the enzyme is characterized by a duplicated tripeptide, which can be proteolytically processed in vivo. We constructed a mutant in which the last 5 amino acids were replaced by 28 residues of unrelated sequence. This protein was expressed, purified, and found to have kinetic constants similar to those of the wild-type enzyme. The functionally verified Chinese hamster rhodanese cDNA encodes a protein of 297 residues and differs from the rat enzyme at 13 positions. None of these substitutions occurs at residues suggested to play essential roles in catalysis or structural stabilization.

Lysine residues in the C-terminal lobe and lysosomal targeting of procathepsin D

A major pathway to the lysosome for soluble hydrolases involves the 6-phosphorylation of mannose residues. The initial step in this reaction is catalyzed by a phosphotransferase which recognizes lysosomal precursors. We constructed mutants of human procathepsin D whose targeting to the lysosome could be assayed directly in intact cells. Eight lysine residues were individually converted to glutamic acid on the surface of the carboxyl terminal lobe of the protein. Mutants with as many as four Lys to Glu mutations were normally targeted to the lysosome and processed to the mature form of the enzyme in transfected cells. We conclude that the C-terminal lobe of procathepsin D may not carry a determinant essential for lysosomal targeting in intact fibroblasts.

Role of glycosylation in the expression of human procathepsin D

Human procathepsin D carries two N-linked glycosylation sites at asparagine residues 70 and 199, widely separated on the surface of the folded protein. We created monoglycosylated procathepsin D molecules by site-directed mutagenesis in vitro of the individual glycosylation sites. With only two exceptions, all 12 mutants of this type were expressed efficiently in mammalian cells. The expressed proteins were stable, targeted to the lysosome, and partially secreted into the medium. When both glycosylation sites were eliminated, however, the expressed proteins (9 different mutants) were stable but most were not secreted and targeted poorly to the lysosome. Mammalian fibroblasts appear to sort nascent procathepsin D efficiently only if it is N-glycosylated. Procathepsin D monoglycosylated at N70 is readily distinguished from the endogenous protein in transfected human cells and thus provides an excellent substrate for studying lysosomal targeting in an homologous system.

The propeptide is nonessential for the expression of human cathepsin D

When the 44-amino acid propeptide of human procathepsin D was deleted by mutagenesis in vitro, the mature protein was stably expressed and secreted from transfected mammalian cells. The secreted protein was correctly folded as judged by its binding to pepstatinylagarose. We were unable to detect lysosomal targeting of the propeptide-deleted protein, and targeting was not restored by the substitution of the propeptides from pepsin or renin. We conclude that its propeptide is not essential for the folding of nascent cathepsin D. Efficient lysosomal targeting in mammalian cells appears to require the precursor form of the molecule.

Cloning and identification of annexin II as an autocrine/paracrine factor that increases osteoclast formation and bone resorption

Autocrine products of osteoclasts such as interleukin-6 may play an important role in normal osteoclast formation and activity. To identify novel stimulatory factors for osteoclasts, we have prepared a mammalian cDNA expression library generated from highly purified human osteoclast-like multinucleated cells (MNC) formed in long term bone marrow cultures and screened this library for autocrine factors that enhance MNC formation. A candidate clone which stimulated MNC formation was isolated. Sequence analysis showed that this cDNA encoded annexin II (AXII). Purified recombinant AXII significantly increased MNC formation in human bone marrow cultures in the absence of 1,25-(OH)2 vitamin D3 and enhanced MNC formation in mouse bone marrow cultures treated with 10(-9) M 1,25-(OH)2 vitamin D3. The enhanced MNC formation in murine marrow cultures resulted in increased bone resorption. Treatment of fetal rat long bones with AXII and 1,25-(OH)2 vitamin D3 significantly increased bone resorption compared to 1,25-(OH)2 vitamin D3 alone. Reverse transcriptase polymerase chain reaction analysis demonstrated that AXII mRNA was expressed at high levels in RNA isolated from highly purified giant cells from osteoclastomas, human osteoclast-like MNC, and pagetic bone. Western blot analysis of conditioned media collected from human marrow cultures showed that AXII was present in the media. Furthermore, approximately 50% of total AXII produced by cells transfected with AXII cDNA was present in the conditioned media. These data suggest that the AXII is an autocrine factor that enhances osteoclast formation and bone resorption and demonstrate a previous unknown function for AXII.

Rapid secretion by a nonclassical pathway of overexpressed mammalian mitochondrial rhodanese

Rhodanese is a small (33 kDa) monomeric sulfurtransferase which is synthesized on cytoplasmic ribosomes and imported into the mitochondrial matrix without cleavage of its amino terminus. When we transfected mammalian cell lines with rhodanese cDNA under the control of an efficient viral promoter, up to 40% of the overexpressed protein was secreted into the medium. This secretion was not the result of cell lysis, did not occur via the endoplasmic reticulum, and did not require the amino-terminal mitochondrial import signal. Addition of a carboxyl-terminal peptide extension did not block secretion, nor did a number of inhibitors of cellular sorting processes. Rhodanese polypeptide is known to associate with chaperonin proteins. In the absence of available mitochondrial import sites, such a complex in the cytoplasm of transfected cells could deliver unfolded rhodanese to export pores on the inner surface of the plasma membrane. This mechanism could contribute to the nonclassical secretion of cytoplasmically synthesized interleukins, growth factors, and lectins.

Mutations of noncatalytic sulfhydryl groups influence the stability, folding, and oxidative susceptibility of rhodanese

Mutants of rhodanese (EC 2.8.1.1) which substitute serine residues for each of the 4 cysteine residues have been studied to determine the roles of cysteines in the structure and function of the enzyme. The proteins compared in these studies were: the wild-type, C63S, C247S, C254S, C263S, C254S/C263S, and C63S/C254S/C263S. These current studies show that cysteine 247 is the only cysteine required for the activity of the enzyme. Although the other sulfhydryl groups do not participate in sulfur transfer, mutations of the noncatalytic cysteines result in the destabilization of the native structure of the enzyme. All the active proteins had similar kinetic parameters. Mutants substituting cysteine 254, compared with the other species, were: (a) more resistant than wild-type to inactivation by dithiothreitol, (b) more readily reactivated following oxidative inactivation, and (c) found to adopt conformations that show increased exposure of hydrophobic surfaces following removal of the transferable sulfur. On the other hand, cysteine to serine substitutions had very little effect on: (a) the rates of oxidative inactivation, (b) the increased fluorescence following the removal of transferable sulfur, or (c) the effectiveness of spontaneous refolding after urea denaturation. Forms of rhodanese that were formerly considered to be irreversibly oxidized can be reactivated if the protein is denatured in urea before reductants are used. It is proposed that these forms differ from reversibly oxidized states due to the inaccessibility of intramolecular disulfides to reductants and not to the formation of higher oxidation states of the protein.

Cloning and characterization of the 5'-flanking region of the mouse tartrate-resistant acid phosphatase gene

Little information is available on the molecular mechanisms controlling osteoclastic bone resorption. We used tartrate-resistant acid phosphatase (TRAP) to begin to investigate the regulation of bone resorption at the molecular level. TRAP is expressed at high levels in osteoclasts and may play an important role in the bone resorptive process. Therefore, we isolated the murine TRAP gene from a mouse spleen genomic library and characterized its promoter. A restriction map was generated for the 17 kb TRAP insert. A 2 kb SmaI fragment, containing the 5'-flanking region, was subcloned and the nucleotide sequence determined. Sequence analysis of the SmaI fragment revealed the presence of numerous candidate transcription factor binding sequences, including those for AP1 and H-APF-1. The H-APF-1 site matches the consensus sequence for the IL-6-regulated transcription factor. An intron was identified at -1 to -393 bp relative to the ATG. The presence of an intron was confirmed by PCR analysis of RNA isolated from murine osteoclasts. Primer extension analysis indicated the presence of a transcription initiation site at -552 bp from the ATG. The region from -1846 to 2bp relative to the ATG initiation codon drove the transient expression of a luciferase reporter gene when transfected into HRE H9 rabbit endometrial cells. PMA treatment of HRE H9 cells enhanced luciferase transcription approximately threefold. These data suggest that the TRAP promoter is complex and contains multiple regulatory elements. The availability of the TRAP promoter may also permit production of transgenic mice, which can be used to develop previously unavailable osteoclast cell lines.

Tartrate-resistant acid phosphatase gene expression as a facile reporter gene for screening transfection efficiency in mammalian cell cultures

The efficiency of DNA transfection into mammalian cell cultures has been monitored using a variety of reporter assays. However, the common procedures are expensive, time-consuming and usually cannot identify the transfected cell population directly. In the present communication we describe a simple, inexpensive and efficient method to directly identify DNA transfection in mammalian cells using tartrate-resistant acid phosphatase (TRAP) gene expression. The method involves the transfection of a plasmid (pCT3), which contains TRAP cDNA driven by a CMV promoter, into mammalian cells. The cells can then be stained for TRAP activity, and the transfection efficiency can be determined by simply counting the positively transfected cells in a defined area with a microscope. This method permits screening of mammalian cells for transfection efficiency in multi-well plates. After waiting 30-40 minutes to allow the TRAP assay to saturate, wells can be scored in 1-2 minutes with little difficulty in detecting the transfected cells.

Chinese hamster ovarian cells transfected with human parathyroid hormone-related protein cDNA cause hypercalcemia in nude mice

BACKGROUND

Parathyroid hormone-related protein (PTH-rP) has been implicated as a causative factor in the humoral hypercalcemia of malignancy. Animal models of hypercalcemia of malignancy that have traditionally utilized human or animal tumors or injections or infusions of hypercalcemic factors have limitations that may prevent exact delineation of the biologic effects of tumor-produced PTH-rP.

EXPERIMENTAL DESIGN

To assess the effects of tumor-produced PTH-rP in vivo, we have transfected Chinese hamster ovarian (CHO) cells with cDNA encoding human preproPTH-rP-(1-141) which then stably express only PTH-rP. We inoculated these tumor cells into nude mice, and compared tumor-bearing nude mice with similar tumor-bearing nude mice inoculated with non-transfected CHO cells using standard parameters of calcium homeostasis.

RESULTS

The nude mice carrying tumors expressing PTH-rP became hypercalcemic over 12 to 18 days. Blood ionized calcium values correlated positively with plasma PTH-rP concentration and tumor volume. Plasma PTH-rP concentrations in hypercalcemic mice were similar to those reported in humans with hypercalcemia of malignancy. Bone histology and histomorphometry from hypercalcemic nude mice demonstrated increased bone resorption when compared with mice bearing nontransfected CHO tumors.

CONCLUSIONS

We have produced an animal model of tumor-produced PTH-rP by transfecting CHO cells with the cDNA for PTH-rP and inoculating these tumor cells into nude mice. Hypercalcemia in this model is mediated in part by the effects of PTH-rP to increase osteoclastic bone resorption. The model is advantageous because PTH-rP alone is secreted in a prolonged, constitutive fashion with pharmacokinetics similar to naturally occurring tumors. It should prove to be a useful model to study the effects of tumor-produced PTH-rP in vivo.

Recombinant bovine rhodanese: purification and comparison with bovine liver rhodanese

Recombinant bovine rhodanese (thiosulfate: cyanide sulfurtransferase, EC 2.8.1.1) has been purified to homogeneity from Escherichia coli BL21(DE3) by cation-exchange chromatography. Recombinant and bovine liver rhodanese coelectrophorese under denaturing conditions, with an apparent subunit molecular weight of 33,000. The amino terminal seven residues of the recombinant protein are identical to those of the bovine enzyme, indicating that E. coli also removes the N-terminal methionine. The Km for thiosulfate is the same for the two proteins. The specific activity of the recombinant enzyme is 12% higher (816 IU/mg) than that of the bovine enzyme (730 IU/mg). The two proteins are indistinguishable as to their ultraviolet absorbance and their intrinsic fluorescence. The ability of the two proteins to refold from 8 M urea to enzymatically active species was similar both for unassisted refolding, and when folding was assisted either by the detergent, lauryl maltoside or by the E. coli chaperonin system composed of cpn60 and cpn10. Bovine rhodanese is known to have multiple electrophoretic forms under native conditions. In contrast, the recombinant protein has only one form, which comigrates with the least negatively charged of the bovine liver isoforms. This is consistent with the retention of the carboxy terminal residues in the recombinant protein that are frequently removed from the bovine liver protein.

Expression of cloned bovine adrenal rhodanese

A cDNA for the enzyme rhodanese (thiosulfate:cyanide sulfurtransferase, EC 2.8.1.1) has been cloned from a bovine adrenal library. An initiator methionine codon precedes the amino-terminal amino acid found in the isolated protein. Rhodanese is synthesized in the cytoplasm and transferred to the mitochondrial matrix. Thus, any amino-terminal sequence required for organelle import is retained in the mature protein. Furthermore, the DNA sequence shows that there are three additional amino acids, Gly-Lys-Ala, at the carboxyl terminus that are not found by protein sequencing. Additionally, comparison of the published amino acid sequence with that encoded by the open reading frame revealed three differences in the amino acid sequence. Comparison of the bovine and chicken liver sequences shows an overall level of 70% sequence homology, but there is complete identity of all residues that have been implicated in the function of the enzyme. When two mammalian cells, cos-7 and 293 cells, were transiently transfected with a plasmid containing the rhodanese coding region, rhodanese activity in lysates increased approximately 20-fold. Fluorograms of denaturing polyacrylamide gels detected a large increase in a polypeptide that co-migrated with the native protein and reacted with anti-rhodanese antibodies. Nondenaturing gels showed two active species that co-migrated with the two major electrophoretic forms purified by current procedures. Escherichia coli, transformed with a plasmid containing the rhodanese coding region, showed a 15-fold increase in rhodanese activity over baseline values. When the E. coli recombinant protein was analyzed on a nondenaturing gel, only one species was observed that co-electrophoresed with the more electropositive variant seen in purified bovine liver rhodanese. This single variant could be converted by carboxypeptidase B digestion to a form of the enzyme that co-migrated with the more electronegative species isolated from bovine liver. Thus, two major, enzymatically active electrophoretic variants, commonly observed in mammalian cells, can be accounted for by carboxyl-terminal processing without recourse to other post-translational modifications.

Molecular biology for nonmolecular biologists

Molecular biology offers various powerful tools for understanding the etiology of diabetes mellitus. These methods can be understood by nonmolecular biologists. Topics briefly introduced include 1) control of gene expression, 2) cloning of genes, 3) how gene promoters work, 4) transgenic mice, 5) mutagenesis in vitro, 6) human molecular genetics, and 7) the new powerful polymerase chain-reaction technique, which greatly simplifies many of the other methods.

Cathepsin D and prognosis in breast cancer

We investigated the possibility that cathepsin D, an estrogen-induced lysosomal protease, might have value as a prognostic factor in breast cancer by studying frozen tissue specimens from 397 patients. We measured the 34-kd mature form of the enzyme by Western blot assay and densitometry. Among 199 patients with node-negative disease, but not among 198 with node-positive disease, high levels of cathepsin D proved to be a significant predictor of reduced disease-free survival (median follow-up, 64 months), either as a continuous variable (log cathepsin D; P = 0.018) or as a dichotomous variable with an optimized cutoff point (P = 0.0001). Results were similar for overall survival (P = 0.009 and 0.0001, respectively). Relating the level of cathepsin D to other prognostic factors in the patients with node-negative disease, we found an association with aneuploidy but none with estrogen or progesterone receptors, tumor size, or the age of the patient. In multivariate analyses, a high level of cathepsin D was the most important independent factor in predicting shorter disease-free and overall survival in patients with node-negative disease. As compared with the risk in women with low levels of cathepsin D, the relative risk of tumor recurrence was 2.6 (95 percent confidence interval, 1.6 to 4.4) and the relative risk of death was 3.9 (95 percent confidence interval, 2.1 to 7.3) among those with high levels of cathepsin D. For disease-free survival, cathepsin D status was predictive of outcome primarily among those with aneuploid tumors; the actuarial five-year recurrence rates of aneuploid tumors were 60 percent among women with high levels of cathepsin D and 29 percent among those with low levels, as compared with 22 percent for all diploid tumors. We conclude that cathepsin D may be an independent predictor of early recurrence and death in node-negative breast cancer.

Proinsulin I and II gene expression in inbred mouse strains

Mice and rats express two nonidentical insulins from a pair of unlinked genes. We have applied a nuclease protection assay, which can sensitively quantify each of the mouse insulin mRNAs, to the resolution of the following questions concerning their expression. First, it has not been established whether alterations in expression of one or both of these genes cause differing total insulin biosynthetic capacity noted between several inbred mouse strains. These studies showed that the relative abundance of mRNAs encoding mouse insulins I and II was identical in four separate mouse strains. In spontaneously obese, hyperinsulinemic (db/db)C57BL/KsJ mice, both proinsulin I and proinsulin II mRNAs were increased relative to the levels in normal (+/db) C57BL/KsJ mice, but again the ratio of the two mRNAs did not differ. The ratio was nearly identical to that for the orthologous mRNAs in rats, indicating that the mechanisms which regulate insulin mRNAs in rodents are conserved in both genes in several mouse strains and between rodent species. This finding suggests that differences between mouse strains in insulin biosynthetic capacity result from differences in the glucose sensing/signalling mechanism at a point before coordinate gene transcription. Second, low levels of insulin synthesis have been suggested as an explanation for relatively high levels of insulin in several nonpancreatic tissues. We showed that the ribonuclease protection assay, sufficiently sensitive to measure 1/2000th the amount of insulin mRNA present in pancreas, was unable to detect insulin mRNA in salivary gland. This result indicates that the high levels of radioimmunoassayable insulin detected in salivary glands are not the result of insulin synthesis in situ.

Renin, a secretory glycoprotein, acquires phosphomannosyl residues

Renin is an aspartyl protease which is highly homologous to the lysosomal aspartyl protease cathepsin D. During its biosynthesis, cathepsin D acquires phosphomannosyl residues that enable it to bind to the mannose 6-phosphate (Man-6-P) receptor and to be targeted to lysosomes. The phosphorylation of lysosomal enzymes by UDP-GlcNAc:lysosomal enzyme N-acetylglucosaminylphosphotransferase (phosphotransferase) occurs by recognition of a protein domain that is thought to be present only on lysosomal enzymes. In order to determine whether renin, being structurally similar to cathepsin D, also acquires phosphomannosyl residues, human renin was expressed from cloned DNA in Xenopus oocytes and a mouse L cell line and its biosynthesis and posttranslational modifications were characterized. In Xenopus oocytes, the majority of the renin remained intracellular and underwent a proteolytic cleavage which removed the propiece. Most of the renin synthesized by oocytes was able to bind to a Man-6-P receptor affinity column (53%, 57%, and 90%, in different experiments), indicating the presence of phosphomannosyl residues. In the L cells, the majority of the renin was secreted but 5-6% of the renin molecules contained phosphomannosyl residues as demonstrated by binding of [35S]methionine-labeled renin to the Man-6-P receptor as well as direct analysis of [2-3H]mannose-labeled oligosaccharides. Although the level of renin phosphorylation differed greatly between the two cell types examined, these results demonstrate that renin is recognized by the phosphotransferase and suggest that renin contains at least part of the lysosomal protein recognition domain.

46 kd mannose 6-phosphate receptor: cloning, expression, and homology to the 215 kd mannose 6-phosphate receptor

We have isolated cDNA clones encoding the entire sequence of the bovine 46 kd cation-dependent mannose 6-phosphate (CD Man-6-P) receptor. Translation of CD Man-6-P receptor mRNA in Xenopus laevis oocytes results in a protein that binds specifically to phosphomannan-Sepharose, thus demonstrating that our cDNA clones encode a functional receptor. The deduced 279 amino acid sequence reveals a single polypeptide chain that contains a putative signal sequence and a transmembrane domain. Trypsin digestion of microsomal membranes containing the receptor and the location of the five potential N-linked glycosylation sites indicate that the receptor is a transmembrane protein with an extracytoplasmic amino terminus. This extracytoplasmic domain is homologous to the approximately 145 amino acid long repeating domains present in the 215 kd cation-independent Man-6-P receptor.

Cloning of the bovine 215-kDa cation-independent mannose 6-phosphate receptor

Four overlapping cDNA clones encoding a partial sequence of the cation-independent 215-kDa mannose 6-phosphate receptor have been identified by screening a fetal calf liver cDNA library with oligonucleotide probes. RNA hybridization analysis showed that the length of the mRNA is approximately 9.5 kilobases. Sequence analysis demonstrated that the clones consist of 4647 contiguous nucleotides and contain an open reading frame coding for a polypeptide of 1461 amino acids, which we estimate represents greater than 75% of the primary structure of the receptor. The deduced amino acid sequence indicates that the receptor has a carboxyl-terminal cytoplasmic domain of 163 amino acids that is rich in acidic residues, a 23-amino acid transmembrane segment, and an extracellular domain containing at least eight homologous repeats of approximately 145 amino acids. One of the repeats contains an additional 43-residue segment that is similar to the type II repeat of fibronectin. Each repeat contains a highly conserved 13-amino acid unit bordered by cysteine residues that may be functionally important.

Three rat preprotachykinin mRNAs encode the neuropeptides substance P and neurokinin A

Synthetic oligonucleotides were used to screen a rat striatal cDNA library for sequences corresponding to the tachykinin peptides substance P and neurokinin A. The cDNA library was constructed from RNA isolated from the rostral portion of the rat corpus striatum, the site of striatonigral cell bodies. Two types of cDNAs were isolated and defined by restriction enzyme analysis and DNA sequencing to encode both substance P and neurokinin A. The two predicted preprotachykinin protein precursors (130 and 115 amino acids in length) differ from each other by a pentadecapeptide sequence between the two tachykinin sequences, and both precursors possess appropriate processing signals for substance P and neurokinin A production. The presence of a third preprotachykinin mRNA of minor abundance in rat striatum was established by S1 nuclease protection experiments. This mRNA encodes a preprotachykinin of 112 amino acids containing substance P but not neurokinin A. These three mRNAs are derived from one rat gene as a result of differential RNA processing; thus, this RNA processing pattern further increases the diversity of products that can be generated from the preprotachykinin gene.

Chromosome localization of the human renin gene (REN) by in situ hybridization

Previous studies by Southern blot analysis of human X mouse somatic cell hybrids localized the renin gene to region p21----qter of human chromosome 1. Using a DNA insert encoding exons 2-5, the renin gene was mapped to human chromosome bands 1q25----q32 by in situ hybridization. The sublocalization of the renin gene will facilitate subsequent detailed linkage analysis of human chromosome 1.

Characterization of the two nonallelic genes encoding mouse preproinsulin

We have cloned and sequenced the two mouse preproinsulin genes. The deduced amino acid sequences of the mature mouse insulins are identical to the published protein sequences. However, the nucleotide sequence indicates that the mouse I C-peptide has a deletion of two amino acids compared with the mouse II C-peptide. We used an S1 nuclease assay to confirm the presence of the deletion and to measure the ratio of transcripts from gene I to transcripts from gene II. The mouse preproinsulin I gene, like the rat gene I, is missing the second intervening sequence that normally interrupts the C-peptide region in other insulin genes. Comparison of the 5' flanking sequences of the mouse and rat genes II indicates that they are homologous for at least 1000 base pairs. The preproinsulin I genes also share homology in their 5' flanking DNAs; however, their homology to the preproinsulin II genes extends for only about 500 base pairs.

Human insulin-related DNA sequences map to chromosomes 2 and 11

The insulin-related hormone family consists of four known peptides: insulin, the insulin-like growth factors I and II, and relaxin. To investigate the hypothesis that the family contains additional members, we have isolated a series of human genomic clones using the insulin gene as a hybridization probe. Two of these single copy DNA sequences, hIr1 and hIr2, have been localized to chromosome 2 and 11p11----q13, respectively, by Southern blot analysis of mouse-human somatic cell hybrids, a result consistent with other evidence supporting the dispersal of the insulin gene family throughout the genome. Although a biological function for these DNA sequences has not yet been established, hIr1 and hIr2 are potentially useful as molecular probes for mapping, by analysis of restriction fragment length polymorphisms (RFLP), genetic disorders linked to chromosomes 2 or 11.

Genetic analysis of the hypervariable region flanking the human insulin gene

In order to study the mechanisms for the generation of length diversity within the 5' flanking region of the human insulin gene, we have isolated and sequenced a previously uncharacterized allele. This allele, of a size intermediate between those three already described in the literature, encompasses 1,156 base pairs (bp) and contains 81 reiterated tandem oligonucleotides of 14-15 bp each. Population analysis on 298 independently sampled individuals by Southern blotting of genomic DNA demonstrates that the polymorphic portion of the insulin 5' flanking region varies from 400 to more than 8,000 nucleotides, being encoded by from 30 to over 540 oligomeric repeats. Length variability 5' to the insulin gene is a result primarily of unequal crossing over, which generates an expansion or contraction in the number of tandem repeat units per chromosome. A similar mechanism probably accounts for nondispersed reiterated sequences at other loci in the human genome.

Cloning and sequence analysis of cDNA for human cathepsin D

An 1110-base-pair cDNA clone for human cathepsin D was obtained by screening a lambda gt10 human hepatoma G2 cDNA library with a human renin exon 3 genomic fragment. Poly(A)+ RNA blot analysis with this cathepsin D clone demonstrated a message length of about 2.2 kilobases. The partial clone was used to screen a size-selected human kidney cDNA library, from which two cathepsin D recombinant plasmids with inserts of about 2200 and 2150 base pairs were obtained. The nucleotide sequences of these clones and of the lambda gt10 clone were determined. The amino acid sequence predicted from the cDNA sequence shows that human cathepsin D consists of 412 amino acids with 20 and 44 amino acids in a pre- and a prosegment, respectively. The mature protein region shows 87% amino acid identity with porcine cathepsin D but differs in having nine additional amino acids. Two of these are at the COOH terminus; the other seven are positioned between the previously determined junction for the light and heavy chains of porcine cathepsin D. A high degree of sequence homology was observed between human cathepsin D and other aspartyl proteases, suggesting a conservation of three-dimensional structure in this family of proteins.

Glucose regulated insulin biosynthesis in isolated rat pancreatic islets is accompanied by changes in proinsulin mRNA

Isolated rat pancreatic islets were incubated in high (28 mM) or low (2.8 mM) glucose for 0, 1 or 4 hr and insulin biosynthesis and messenger RNA quantified. During the 4 hr course of the experiments the fraction of protein synthesised specifically into proinsulin increased in the presence of high vs low glucose (14.3 +/- 3.7 vs 0.8 +/- 0.5%, p less than 0.01). The relative amount of proinsulin mRNA was found to be about 5-fold higher in islets incubated in high glucose at 4 hr as determined by RNA blot hybridization. Total translatable islet mRNA was unaffected by glucose in the medium. These data extend observations of changes in proinsulin mRNA in rats in vivo during fasting and glucose injection, and suggest a direct effect of glucose on islets. These data further demonstrate that glucose modulates selectively the level of proinsulin mRNA during incubation of isolated pancreatic islets and that changes in the level of proinsulin mRNA play an important role in regulation of insulin biosynthesis.

Mouse kidney renin gene is on chromosome one

The structural gene for mouse kidney renin (Ren-1) was localized to chromosome 1 by Southern blot analysis of mouse-hamster somatic cell hybrids with a cloned mouse submandibular renin cDNA probe. The submandibular renin gene (Ren-2) also lies on chromosome 1; so it may be, in those mouse lines which carry it, a tandem duplication of the kidney-type Ren-1 gene.

Computer graphics modelling of human renin. Specificity, catalytic activity and intron-exon junctions

A model has been constructed using computer graphics for human renin based on the sequence derived from that of the gene and the 3-dimensional structure defined at high resolution for other homologous aspartic proteinases. Human renin can adopt a 3-dimensional structure close to that of other aspartic proteinases, in which amino acids corresponding to intron-exon junctions in the gene are at surface regions in the 3-dimensional structure. As expected, the essential catalytic residues are retained and the nearby residue 304 is alanine as in the mouse sequence, supporting the idea that Asp 304 of other aspartic proteinases may contribute to the low pH of their optimal activity. There are interesting differences at subsite S3' which may contribute to the specificity of human renin. Certain residues at the surface of the enzyme adjacent to the active site cleft are unique to renins and may play a role in recognition and binding of angiotensinogen.

Human renin gene: structure and sequence analysis

The complete protein precursor of human kidney renin has been determined from the sequence of cloned genomic DNA. The gene spans 12 kilobases of DNA and is interrupted by eight intervening sequences. The nine regions (exons) encoding the protein were mapped with a mouse renin cDNA probe, synthetic oligonucleotide probes, and by hybridization of genomic restriction fragments to a 1600-nucleotide human kidney mRNA. The predicted 403-amino acid preprorenin consists of mature renin and a 66-residue amino-terminal prepropeptide. The DNA sequence 5' to the first exon indicates the location of a transcriptional promoter (T-A-T-A-A-A) for a mRNA encoding preprorenin. An additional transcriptional promoter site is located within the first intron, which, if used, would express a shortened nonsecreted prorenin. The structure of the human renin gene is similar to that of human pepsinogen, a closely related aspartyl protease enzyme. This observation suggests that renin and pepsinogen have a common evolutionary origin.