A previous report in this laboratory demonstrated that unlike the control experiments in which abundant blood vessels with bone marrow were present in Saos-2 osteosarcoma extract-induced ectopic bones in nude mouse skeletal muscles, the lower zone of plaques (LZP) in atherosclerotic rabbit aortas lacked the evidence of vessel intrusions and bone formation . These observations suggest that the induction of LZP calcification could be caused by the restriction in blood supply to the LZP due to intimal thickening without a prerequisite osteogenesis [10, 11] or remote bone resorption . This contention was further supported by the present data obtained from the experimentation with cultured rabbit aortic smooth muscle cells exposed to serum depletion. In spite of the present in vitro model, factors that underlying the mechanism of the serum depletion-induced LZP calcification remains to be established. A reduced amount or a lack of some serum proteins in the LZP may play a role in LZP calcification. For example, anhydrase and hemoglobin are essential proteins in the regulation of blood pH and their absence could potentially influence pH dependent-mineralization process. The lack of hemoglobin may affect oxygen tension and could in theory alter acid-base chemistry through the isohydric shift, although the role of oxygen in LZP calcification has yet to be established. Moreover, potent mineralization inhibitors such as OPN  and MGP  in blood could be prevented from reaching LZP by intimal thickening. Other factors such as alterations in osmolarity, which can be affected by serum protein deletion, remain unknown. However, the present observation shows that much less amounts of whole serum proteins than serum albumin were needed to inhibit calcification, suggesting that change in osmolarity by serum depletion may not play a significant role in the calcification. An attempt to reproduce intimal thickening in vivo by the addition of cholesterol to SMC culture increased neither cell proliferation nor calcification (not shown). This is expected since aortic calcification in rabbits fed cholesterol diets did not occur until intima was extensively thickened [7, 18, 19].
The minimal requirement of Ca × P ion products of 5.0 mM2 in excess of the serum level of 3.5 mM2 for serum depletion-induced calcification in vitro suggests that additional mechanisms are needed to increase Ca and/or P in the LZP. The current experiments using a sensitive measurement of Ca with Arsenazo III dye  indicated that 5.02 ± 0.25 mM Ca (6 serum samples) was present in rabbit serum vs. 3.23 mM using an electrode assay procedure . A protein-free fraction of serum obtained from 10% trichloroacetic acid (TCA) precipitation followed by the removal of TCA through lyophilization yielded the same result (not shown). The precision and accuracy of the procedure was verified by the use of known quantities of calcium carbonate and calcium mono- or dibasic phosphate dissolved in 0.1 N HCl. Although the difference in quantities between the two procedures could be due to the technical issue, the data suggest a large pool of serum Ca available for calcification through the diffusion process. High ion products can be further achieved via the hydrolysis of phosphoesters in cells. High concentrations of β-glycerophosphate ranging from 5–10 mM were routinely used in numerous studies of calcification in cell cultures including bone cells  and bovine SMCs . However, the current data indicated that rabbit SMCs cultured for up to 14 day in the presence of 10% FBS and/or 10 mM β-glycerophosphate with a biweekly medium changes did not display calcification (data not shown). This is expected since rabbit SMCs and aortas neither have alkaline phosphatase activity nor the ability to hydrolyze β-glycerophosphate . The paradoxical difference in the effect of serum on calcification between rabbit and bovine SMCs remains to be established. Moreover, the Ca × P ion products needed to initiate mineralization with varying Ca and constant P levels appeared to be lower than those with constant Ca and varying phosphate levels (p < 0.05). The changes in FT-IR of amide peak after mineralization also suggest that the perturbation of protein absorption in the infrared spectrum could be attributable to Ca binding or by overlapping mineral deposits.
We previously reported that serum depletion in culture media increased the production of calcifying vesicles, which contain enzymes capable of hydrolyzing nucleotides including Ca-Mg ATPase, AMPase, ADPase, and nucleoside triphosphate pyrophosphohydrolase [21, 22]. The elevated enzyme activities could increase either Ca or P accumulation in favor of mineralization, . The present mass spectral data as shown in Table 1 demonstrate the presence of various calcification-related proteins in calcifying vesicles. These observations, therefore, further support the role of the release of these vesicles in vascular calcification [7, 19, 21, 22]. Furthermore, calcifying vesicles could be trapped and concentrated within LZP by intimal thickening, thereby promoting focal calcification. The ability of serum depletion to induce calcifying vesicles in SMCs is consistent with the observation that calcifying vesicles can be released from human osteosarcoma cells (Saos-2) within 30 min after exposure to serum-free media . Thus, the increased vesicle activity resulting from serum depletion in LZP and the availability of a large pool of serum Ca and/or P diffusible through LZP could provide a mechanism to surpass the threshold ion product needed for calcification.
Apoptotic bodies have been suggested to play a role in physiological and pathological calcification. The present observations of the association between serum-depletion-induced membrane translocation (MT), which is an early event of apoptosis and calcification suggest that the change in membrane topology may play a pivotal role in vascular calcification. Such contention is consistent with our observation that calcifying vesicles were accumulated on the surface of SMCs upon incubation with serum-free culture. The lack of association between peroxide-induced MT and cell-mediated calcification is likely due to the presence of serum in the MT assay, since the MT induced by serum depletion or peroxide can be fully reversed within 30 min by the serum replenishment (data not shown). The less degree of the dye uptake shown in Fig. 8A than in 8B was probably due to a partial inhibition of MT by the inclusion of serum in the assay for peroxide-induced MT. Thus, these data indicate that serum can inhibit both membrane translocation and calcification.
The marked inhibitory effect of extensively diluted serum indicates that minute quantities of some serum protein components play a pivotal role in LZP calcification. It appears from C1/2 data that serum albumin may need to synchronize with other serum proteins to produce maximal effect. Despite these experiments, the mechanisms underlying the inhibitory effect of the minute amounts of serum on calcification are unclear. In contrast to the preventive effect of serum on calcification in rabbit SMC culture reported herein, a rat model in which animals were injected with high dosage of vitamin D, the increase in fetuin-mineral complex level in rat serum was shown to correlate with arterial calcification in the media . The paradoxical differences in the effects of the serum on vascular calcification at the two distinct parts of the arteries between these two species are unclear. It is also plausible that some minor proteins in the serum with the ability to inhibit mineralization such as osteopontin (OPN) [29, 37], matrix γ-carboxyglutamate (Gla) protein (MGP) , and osteoprotegerin (OPG)  may play a major role in the serum effect. Giachelli and colleagues  demonstrated that OPN at a level of 50 ng/ml inhibited bovine SMCs in vitro calcification, which occurred after a 10-day exposure to the media containing 10% serum and 10 mM β-glycerophosphate. Since the effective concentration of OPN in these experiments was 5-fold higher than the serum OPN level , whether the serum level of OPN can inhibit serum depletion-induced calcification remains to be established. A direct test to see if OPN or MGP at the serum levels can inhibit calcification in the rabbit aortic SMC short-term cell cultures awaits the availability of purified proteins for testing. The purified recombinant OPN from EMP Genetech did not inhibit SMC calcification (not shown), probably due to insufficient phosphorylation in the recombinant protein. Since a range of 5–11 ng/ml of OPN was present in human sera , a comparison of C1/2 values among the total serum proteins and regulatory proteins could rule in or out whether one of these proteins is essential for the prevention of unwanted calcification. In addition, whether there is a need of synchronization with other essential proteins can be addressed by the comparison of C1/2 values among these proteins.
Altogether, the data from the in vitro cell culture experiments herein support the hypothesis that serum depletion in LZP imposed by lesion thickening could trigger calcification through the accumulation of calcifying vesicles on the cell surface  and/or the elimination of the potent serum mineralization inhibitors in the LZP. The short-term nature of the effect of serum depletion on calcification in cultures also suggests that the initial events of calcification in rabbit thoracic aortas during dietary cholesterol intervention are independent of osteogenesis and could be related to membrane translocation, an early event in apoptosis [7–9, 23].