Charles B. Underhill

Extracellular matrix protein 1 (ECM1) is over-expressed in malignant epithelial tumors

The extracellular matrix protein 1 (ECM1) is a secreted protein that has been implicated with cell proliferation, angiogenesis and differentiation. In the present study, we used immunohistochemical staining to examine the expression of ECM1 in a panel of human tumors and found that it was closely correlated with some types of tumors including: invasive breast ductal carcinoma (83%), esophageal squamous carcinoma (73%), gastric cancer (88%) and colorectal cancer (78%). Significantly, ECM1expression was correlated with the metastatic properties of the tumors. Primary breast cancers that had formed metastases were 76% positive while those that had not metastasized were only 33% positive. ECM1 expression was also correlated with PCNA a marker for proliferation, but not with CD34, a marker for endothelial cells. These results indicate that ECM1 tends to be preferentially expressed by metastatic epithelial tumors.

A Dominant Interference Collagen X Mutation Disrupts Hypertrophic Chondrocyte Pericellular Matrix and Glycosaminoglycan and Proteoglycan Distribution in Transgenic Mice

Collagen X transgenic (Tg) mice displayed skeleto-hematopoietic defects in tissues derived by endochondral skeletogenesis.(1) Here we demonstrate that co-expression of the transgene product containing truncated chicken collagen X with full-length mouse collagen X in a cell-free translation system yielded chicken-mouse hybrid trimers and truncated chicken homotrimers; this indicated that the mutant could assemble with endogenous collagen X and thus had potential for dominant interference. Moreover, species-specific collagen X antibodies co-localized the transgene product with endogenous collagen X to hypertrophic cartilage in growth plates and ossification centers; proliferative chondrocytes also stained diffusely. Electron microscopy revealed a disrupted hexagonal lattice network in the hypertrophic chondrocyte pericellular matrix in Tg growth plates, as well as altered mineral deposition. Ruthenium hexamine trichloride-positive aggregates, likely glycosaminoglycans (GAGs)/proteoglycans (PGs), were also dispersed throughout the chondro-osseous junction. These defects likely resulted from transgene co-localization and dominant interference with endogenous collagen X. Moreover, altered GAG/PG distribution in growth plates of both collagen X Tg and null mice was confirmed by a paucity of staining for hyaluronan and heparan sulfate PG. A provocative hypothesis links the disruption of the collagen X pericellular network and GAG/PG decompartmentalization to the potential locus for hematopoietic failure in the collagen X mice.

Hyaluronan synthesis by epiphysial chondrocytes is regulated by growth hormone, insulin-like growth factor-1, parathyroid hormone and transforming growth factor-β1

In a previous study, we presented evidence that the synthesis of hyaluronan by hypertrophic chondrocytes is one of the principal factors driving the interstitial expansion of the growth plate (Pavasant et al., J. Cell Sci. 109: 327-334, 1996). To test this possibility further, we used two different approaches to examine the effects of hormones on the production of hyaluronan in the growth plate. In the first approach, we examined the growth plate of the lit/lit mouse that lacks growth hormone and found that its hypertrophic lacunae were smaller and contained less hyaluronan than those of wild type mice. Moreover, the ratios of hyaluronan staining density to total area of the lacunae were similar for the lit/lit and the wt/wt mice, indicating that the amount of hyaluronan is directly related to lacuna size. In the second approach, we examined the effects of hormones on segments of the epiphysial growth plate placed in organ culture. Under normal culture conditions, a band of hyaluronan staining progressed across the length of the growth plate, reflecting the maturation of chondrocytes into the hypertrophic stage. When insulin-like growth factor-1, a factor known to promote chondrocyte maturation, was added to the culture medium, the production of hyaluronan and the enlargement of the lacunae were stimulated. In contrast, when either parathyroid hormone or transforming growth factor-beta 1, both of which inhibit chondrocyte differentiation, was added to the medium of cultured segments, new pericellular hyaluronan was not detected and the lacunae did not enlarge. Taken together, these results indicate that factors that either stimulate or inhibit the maturation of epiphysial chondrocytes have a corresponding effect on the production of hyaluronan. This, in turn, further supports the importance of hyaluronan in the process of lacuna enlargement.

Naturally-occurring antibodies which bind hyaluronate.

Sera from a number of different species were found to contain a hyaluronate binding activity. This binding activity was detected by mixing the serum with [ 3 H] hyaluronate and then precipitating the proteins and any bound [ 3 H] hyaluronate by the addition of an equal volume of saturated (NH 4 ) 2 SO 4 . The binding activity associated with horse serum was examined further and found to be saturable, of relatively high affinity and specific for hyaluronate. Competition experiments showed that the binding was not influenced by the molecular weight of the hyaluronate, suggesting that the binding protein interacts with an intrachain region of the hyaluronate rather than the terminals. The binding activity co-purified with the IgG and IgM fractions of serum after (NH 4 ) 2 SO 4 precipitation, ion-exchange and molecular-sieve chromatography as well as after affinity chromatography.

Hyaluronan can be non-enzymatically linked to protein through an alkali sensitive bond

To test for the presence of hyaluronan to protein linkages, both purified and unpurified preparations of hyaluronan were blotted onto nitrocellulose, and the adsorbed hyaluronan was detected using a biotinylated probe derived from cartilage. While purified hyaluronan did not adsorb to nitrocellulose, the unpurified hyaluronan from rat fibrosarcoma (RFS) cells and chick embryos did. This adsorption appeared to require the presence of protein, since it was inhibited by treatment with proteases. Furthermore, when hyaluronan from RFS cells was subjected to conditions which break most non-covalent bonds, it retained its ability to adsorb to nitrocellulose, suggesting that this hyaluronan was covalently bound to protein which mediated its adsorption to nitrocellulose. While this bond was resistant to acidic buffers, it was readily broken by alkaline buffers. Additional experiments demonstrated that when either purified [3H] hyaluronan or oligosaccharide fragments of hyaluronan were incubated with a variety of proteins, they slowly gained the ability to adsorb to nitrocellulose. However, this process could be blocked by the addition of low molecular weight amines or by reducing the [3H] hyaluronan with NaBH4. These results are consistent with the possibility that the reducing terminal of the hyaluronan reacts with amine groups of protein to form a Schiff base which then rearranges to a stable bond. Such a bond could account for the association of hyaluronan with the surfaces of RFS cells.

A protein binding assay for hyaluronate

A relatively quick and simple assay for hyaluronate was developed using the specific binding protein, hyaluronectin. The hyaluronection was obtained by homogenizing the brains of Sprague-Dawley rats, and then centrifuging the homogenate. The resulting supernatant was used as a source of crude hyaluronectin. In the binding assay, the hyaluronectin was mixed with [3H]hyaluronate, followed by an equal volume of saturated (NH4)2SO4, which precipitated the hyaluronectin and any [3H]hyaluronate associated with it, but left free [3H]hyaluronate in solution. The mixture was then centrifuged, and the amount of bound [3H]hyaluronate in the precipitate was determined. Using this assay, we found that hyaluronectin specifically bound hyaluronate, since other glycosaminoglycans failed to compete for the binding protein. In addition, the interaction between hyaluronectin and hyaluronate was of relatively high affinity (Kd = 5.7 X 10(-10) M), and the size of the hyaluronate did not appear to substantially alter the amount of binding. To determine the amount of hyaluronate in an unknown sample, we used a competition assay in which the binding of a set amount of [3H]hyaluronate was blocked by the addition of unlabeled hyaluronate. By comparing the degree of competition of the unknown samples with that of known amounts of hyaluronate, it was possible to determine the amount of hyaluronate in the unknowns. We have found that this method is sensitive to 1 microgram or less of hyaluronate, and is unaffected by the presence of proteins.

The hyaluronate-binding site from the plasma membrane is distinct from the binding protein present in brain.

To determine whether the hyaluronate-binding protein from brain is similar or identical to the hyaluronate-binding site from the cell surface, the two molecules were compared with respect to their physical and binding properties. The hyaluronate-binding protein was purified from mouse brains by lectin-affinity chromatography, and then analyzed by molecular-sieve chromatography and rate-zonal centrifugation, which showed that it has a Stokes radius of 6.3 nm, and a sedimentation coefficient of 4.8 S. These values are remarkably close to those obtained previously for the membrane-associated binding site (a = 6.5 nm, s20,w = 4.8 S), indicating that the two molecules have similar shapes and sizes. Binding studies of the semi-purified proteins showed that the dissociation constant for the brain derived binding protein (Kd = 270 ng/0.5 ml) was similar to that of the cell-surface binding site (Kd = 350 ng/0.5 ml). However, when the two molecules were compared with respect to oligosaccharide inhibition of binding, significant differences were observed. The hexasaccharide significantly inhibited the binding of [3H] hyaluronate to the cell-surface binding site but had only a small effect on the binding to the brain derived protein. Differences were also found between the two molecules with respect to the effects of a monoclonal antibody (K-3). This antibody blocked most of the binding activity of the membrane-associated binding site, but had no effect on the protein from brain. Taken together, these results indicate that although the hyaluronate-binding protein derived from brain and the cell-surface binding site are physically similar, they are distinct proteins.