Search

Electrolytes and phospholipids of cartilage fractions were partitioned by extraction with organic and aqueous solvents into six solubility groups: Electrolytes I, II and III, and Lipids I, II and III. Of the total Ca, only 4% was water soluble (Electrolytes I); 4-12% was complexed with lipids (Electrolytes II); while the majority (84-92%) was insoluble (Electrolytes III). In contrast, nearly half of the Mg and Pi were water soluble. Of the neutral phospholipid, 95% was not complexed with mineral ions (Lipids I), but 30-45% of the acidic phospholipid was (Lipids II). Ca/Pi ratios were extremely low in the water-soluble phase, but were in the range of amorphous calcium phosphate (ACP) in the insoluble. Molar ratios of the lipid-mineral complex were: Ca:Mg:Pi:acidic phospholipid, 4:3:2:2. Mg/Ca ratios in the soluble fraction were high (5.5-8.9), sufficient to stabilize ACP. Kinetic studies revealed rapid turnover of soluble Ca, insoluble turning over much more slowly. Labeling of lipid-complexed Ca was rapid in cells, but occurred later in matrix vesicles, suggesting transfer. While lipid-Ca-Pi complexes can nucleate apatite in vitro, those present in vivo inside matrix vesicles apparently do not because of the excess Mg. We conclude therefore, that in vesicle-mediated calcification, lysis of the membrane may be essential to allow release of internal Mg.

1. Intact cells, cell fragments (membranes) and matrix vesicles were isolated from the proliferating and calcifying layers of epiphyseal cartilage by sequential hyaluronidase and collagenase digestion and differential centrifugation. Lipids were extracted and analyzed for various lipid classes and their fatty acid composition by column, thin-layer, paper and gas-liquid chromatography. 2. On a protein basis the isolated matrix vesicles had more total lipid than either the membrane or cell fractions, the vesicles and membranes being richer in non-polar lipids and containing smaller quantities of phospholipids than whole cells. Expressed as a percentage of the total lipid, the cells were richer in triacylglycerols and lower in free fatty acids than in the membrane or vesicle fractions. The proportion of free cholesterol and the cholesterol/phospholipid ratio were nearly twice as high in the matrix vesicles as in the other tissue fractions. Choline and ethanolamine phosphoglycerides progressively declined in the membrane and matrix vesicle fractions, whereas serine phosphoglycerides and sphinogomyelin increased. Non-phosphorus-containing polar lipids were present in all fractions, the vesicles being richer in polyhexosyl ceramides, cerebrosides, glycosyldiacylglycerols and certain uncharacterized acidic polar lipids. 3. Fatty acid patterns of the matrix vesicles were distinctive from those of isolated cells, being generally richer in 18 : 0 and 18 : 2, and lower in 16 : 1 and 18 : 1 fatty acids. Monoacyl forms were similarly increased in 16 : 0 and/or 18 : 0, and reduced in 16 : 1, 18 : 1 or 20 : 2 fatty acids, depending on the lipid class. The fatty acid composition of diphosphatidylglycerol from cells and matrix vesicles was markedly different, providing evidence that the cardiolipin in the vesicles was not from mitochondrial components. 4. Based on the fact that the matrix vesicles were significantly enriched in free cholesterol, sphingomyelin, glycolipids and serine-phosphoglycerides, it is concluded that they are derived from the plasma membrane of the cell, supporting earlier conclusions based upon morphological and enzymological evidence.

Fractions composed primarily of cells (Fraction I), membrane fragments (Fraction II) and matrix vesicles (Fraction III) were isolated from chick epiphyseal cartilage. The characteristics of the alkaline phosphatase (EC 3.1.3.1), pyrophosphatase (EC 3.6.1.1) and ATPase (EC 3.6.1.3) activities in the matrix vesicle fraction were studied in detail. Mg-2-+ was not absolutely essential to any of the activities, but at low levels was stimulatory in all cases. Higher concentrations inhibited both pyrophosphatase and ATPase activities. Both the stimulatory and inhibitory effects were pH-dependent. Ca-2-+ stimulated all activities weakly in the absence of Mg-2-+. However, when Mg-2-+ was present, Ca-2-+ was slightly inhibitory. Thus, none of the activities appear to have a requirement for Ca-2-+, and hence would not seem to be involved with active Ca-2-+ transport in the typical manner. The distribution of alkaline phosphatase, pyrophosphatase, and Mg-2-+ ATPase activities among the various cartilage fractions was identical, and concentrated primarily in the matrix vesicles. Conversely, the highest level of (Na-+ + K-+)-ATPase activity was found in the cell fraction. All activites showed nearly identical sensitivities to levamisole (4 - 10-3 M) which caused nearly complete inhibition of alkaline phosphatase and pyrophosphatase. About 10-15% of the ATPase activity was levamisole-insensitive. The data are consistent with the concept that the Mg-2-+-ATPase and pyrophosphatase activities of matrix vesicles stem from one enzyme, namely, alkaline phosphatase.

Chondrocytes in epiphyseal cartilage were examined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) using freeze-fracture techniques. Freeze-fracture replicas showed large numbers of fingerlike, 0.11-0.15 micrometer diameter, projections from the chondrocyte surface, with numerous 95-180 A diameter intramembranous particles associated with both the cell membrane surface and these projections. With SEM, these cytoplasmic projections were also obvious, but appeared collapsed into clusters of globular-shaped projections on the surface of the chondrocytes. With freeze-fracture techniques, in which shrinkage artifacts were essentially eliminated, the cytoplasmic projections were often seen in intimate contact with the extracapsular matrix. However, with chondrocytes prepared by both SEM and conventional TEM, there was evidence of shrinkage, the cytoplasmic projections having little contact with the extracapsular matrix. These findings show that the cytoplasmic processes are not artifacts of tissue processing and provide morphological evidence in support of the hypothesis that matrix vesicles are of cellular origin.

Because lipids had been consistently detected histologically at sites of new calcification, the lipids of epiphyseal cartilage and bone in various stages of mineralization were examined. Lipids were extracted before and after demineralization and analyzed. Lipid content increased during proliferation and calcification of epiphyseal cartilage. Much less was seen in the adjacent cancellous bone; this corroborates histochemical findings. Similar phospholipid compositions were seen in the total lipids of cartilage and bone. Neutral (dipolar) phospholipids accounted for nearly 90% of the total lipid P and were almost completely extracted before demineralization. Serine- and inositol-containing phospholipids and two other, unidentified, acidic lipids could not be effectively extracted from calcifying tissues until after demineralization. Since the extraction of the acidic lipids was closely related to the degree of mineralization, it is possible that they form part of a lipoprotein-mineral complex in the calcifying matrix. Lysophospholipids were detected in all extracts, but primarily in those made after decalcification. It is concluded that acidic lipids are mainly responsible for the sudanophilia detected histologically at sites of new calcification.

A simple, rapid column procedure for the removal of nonlipid contaminants from lipid extracts is described. Lipid P is completely recovered and nonlipids completely removed.

Two-dimensional chromatography on commercially available silica gel-loaded paper for the microanalysis of polar lipids from various tissues is described. All common phospholipids and their lyso derivatives can be reproducibly separated. As many as 22 lipid components were separated on a single chromatogram. Improved methods for staining lipids and for determining phosphorus in the chromatographic spots are reported.

In the presence of acidic phospholipids, inorganic phosphate greatly enhances the net migration of calcium ions from the aqueous phase to the organic phase, an effect that does not occur at less than the physiological pH. The calcium complex in the organic phase is shown by electron microscopy to consist of spherules, composed of stoichiometric amount of calcium, inorganic phosphate, and phospholipid. The demonstration of complex formation between calcium phosphate and acidic phospholipids adds Support to the concept that phospholipids are involved in biological mineralization.

The ion-binding capacity of highly purified reconstituted calf-skin collagen, and the effects of these ions on the precipitation and solubility of the collagen, were studied with a variety of salt solutions at ionic strength 0.16 and pH7.4. Only a small percentage of the total theoretically available anionic and cationic groups was available for ion-binding. In view of this, it appears that most of the ionizable groups of collagen are involved in intramolecular or intermolecular linkages, or both. Nevertheless, marked differences in the binding of the various ions by collagen were observed. Bivalent cations were bound in extremely small but remarkably similar quantities. In contrast, sodium was bound both in much higher and more variable quantities. Of the anions, pyrophosphate and sulphate were bound in the largest quantities, followed by phosphate, fluoride and chloride, in that order. Despite the minimal uptake by collagen of bivalent cations, they prevented the aggregation of tropocollagen into fibrils, and disaggregated fibrillar collagen. In the presence of multivalent anions, tropocollagen aggregated readily and its fibrillar stability was maintained. On the basis of the imbalance in the binding of ion pairs by the sodium pyrophosphate- and sodium phosphate-treated collagens, it was apparent that a reduced number of side-chain carboxyl groups were dissociated in the presence of these salts.