Extracellular calcium regulates distribution and transport of heparan sulfate proteoglycans in a rat parathyroid cell line.

The regulation of the cellular distribution of proteoglycans in a clonal rat parathyroid cell line by extracellular Ca2+ concentrations ([Ca2+]e) was studied. Proteoglycans synthesized by the cells metabolically labeled with [35S]sulfate have been shown to be almost exclusively heparan sulfate (HS) proteoglycans (Yanagishita, M., Brandi, M.L., and Sakaguchi, K. (1989) J. Biol. Chem. 264, 15714-15720), which are generally associated with the plasma membrane. The proportion of HS proteoglycans on the cell surface was approximately 20% in 2.1 mM (high) [Ca2+]e, whereas it increased to 50-60% in 0.05 mM (low) [Ca2+]e. Cell-associated HS proteoglycans redistribute in response to changing [Ca2+]e with a t 1/2 less than 4 min; HS proteoglycans appear on the cell surface as [Ca2+]e decreases and disappear from the cell surface as [Ca2+]e increases. Further, HS proteoglycans on the cell surface recycle to and from an intracellular compartment approximately 10 times before their degradation in low [Ca2+]e but do not recycle in high [Ca2+]e. The distribution of newly synthesized HS proteoglycans is regulated by [Ca2+]e but is independent of [Ca2+]e during biosynthesis. In low [Ca2+]e, at least 50% of the HS proteoglycans pulse-labeled for 10 min are transported from the Golgi complex to the cell surface or to the recycling compartment with a t 1/2 of approximately 20 min. Another approximately 10% appear on the cell surface in either low or high [Ca2+]e in a compartment with a long half-life. Addition of Mg2+ or Ba2+ to the low [Ca2+]e cultures had little effect on the distribution of HS proteoglycans. These observations suggest that [Ca2+]e specifically regulates the distribution and recycling of cell-associated HS proteoglycans in the parathyroid cells.

usually required for the optimal response to stimulators (4,5). Although several mechanisms are involved in controlling parathyroid hormone secretion (6-lo), precise mechanisms coupling the signal ([Ca"+],) and the cell secretory machinery are still far from being understood.
A rat parathyroid cell line established by Sakaguchi et al. (11) retains certain characteristics of normal parathyroid cells. Secretory processes and growth of the cells are suppressed by an increase of [Ca*+], (11,12). We have reported that the proteoglycans synthesized by this cell line are almost exclusively (>95%) heparan sulfate-proteoglycans (HS-PGs) of two distinct types; HS-PGl has a mass of -160 kDa with a single HS chain (-12 kDa) and a core protein of -150 kDa including oligosaccharides, and HS-PG2 has a mass of -170 kDa with three to four HS chains (-30 kDa) and a core protein of 70-80 kDa including oligosaccharides (13

RESULTS
Trypsin Accessibility of Heparan Sulfate Proteoglycarw-Trypsin treatment of metabolically radiolabeled cell cultures has been shown to be a convenient method to classify operationally cell-associated HS-PGs into cell surface (trypsinaccessible) and intracellular (trypsin-inaccessible) forms (14). For the critical assessment of the amounts and the temporal profiles of HS-PGs in these two forms, it is necessary to determine concentrations of trypsin and incubation times required for the maximum removal of accessible HS-PGs from the cell cultures. 3 Rat parathyroid cells labeled with [35S]sulfate for 20 h in low [Ca"], were treated with 200 pg/ml trypsin at 37 "C for 2 min, and trypsin-accessible and -inaccessible materials were extracted separately. In other cell culture systems, such as rat ovarian granulosa cell cultures (14), 100 pg/ml trypsin is sufficient to cleave all trypsin-accessible proteoglycans in 2 min at 37 "C. Macromolecules metabolically labeled in parathyroid cells with [35S]sulfate contained 35S-glycoproteins/ glycopeptides and 35S-HS-PGs, which were clearly separated into two peaks by Q-Sepharose chromatography ( Fig. 1). In subsequent experiments, the two peaks were separately pooled by step elution with 0.3 M NaCl, 8 M urea, and 4 M guanidine HCl buffers and quantitated.
35S-HS-PGs comprised two species, HS-PGl and HS-PG2, and the latter contained more than 90% of the radioactivity incorporated into proteoglycans (13). Thus, the results of this study reflect primarily, if not exclusively, the metabolism of the HS-PG2 molecules. Amounts of trypsin-accessible 35S-HS-PGs and 35S-glycoproteins in parathyroid cells labeled for 20 h both in low and high [Ca'+], were determined after treatment with 20 rg/ml, 200 pg/ml, or 10 mg/ml trypsin for 2 or 15 min as indicated ( Table I) Trypsin-accessible and -inaccessible ""S-macromolecules excluded from Sephadex G-50 columns were applied to Q-Sepharose chromatography and eluted with a NaCl gradient.
A cell extract without trypsin shows three peaks (WI, GP2, and PGs) in the 0.15-1.0 M NaCl gradient (A). The two early peaks contained 3"S-labeled glycoproteins, and the third peak contained proteoglycans (13). The same sample as for A eluted with a 0.3-1.2 M NaCl gradient shows base-line separation into the two peaks, unbound and bound to the column (GPs and PCs), containing W glycoproteins and proteoglycans, respectively (B). Both trypsin-inaccessible (C) and -accessible (D) molecules were separated into two fractions in the same manner.
ml for 2 min, and 10 mg/ml for 2 min) gave essentially the same values for trypsin-accessible 35S-HS-PGs (26-29%) and of 35S-glycoproteins (28-34%), values which are significantly less than those in low [Caz+Je (Table I). Since treatment with 200 pg/ml trypsin for 2 min or 20 pg/ml for 15 min at 37 "C was sufficient to release all accessible 35S-HS-PGs and 35Sglycoproteins in both low and high [Ca2+Je, these conditions were used for determining the localization of the cell-associated HS-PGs in subsequent experiments.
The amounts of HS-PGs removed by increasing trypsin digestion times for cultures in low [Ca2+le were quantitated in more detail (Fig. 2). The extrapolated value for time 0 of digestion (48 + 3%, mean f S.E.) indicates the actual proportion of HS-PGs present on the cell surface at any instant in time, and the maximum value (82 k 3%, mean k S.E.) indicates that some HS-PGs (-34%) that are initially inaccessible become accessible to the enzyme during the incubation. The translocation of HS-PGs from the trypsin-inaccessible to the trypsin-accessible compartment can be approximated by an exponential process with a tl,> of 3.8 min (S.E. of 27%) (Fig. 2,   The time courses for the redistribution of HS-PGs were significantly different for the two changes, from high to low or from low to high [Ca"],. This difference may reflect the time required for intracellular Ca*+ to re-equilibrate when medium is changed from high to low [Ca2+le. For this reason, a second protocol was tested in which cells labeled in high [Ca"'], were chased in Ca'+-free medium containing 0.5 mM EGTA to chelate Ca2+ rapidly; as a control the same medium was added to cells labeled in low [Ca"], as well (Fig. 3B). Addition of Ca'+-free medium with EGTA induced a rapid (t, -1-2 min) increase of trypsin-accessible HS-PGs, a time course of redistribution close to that observed in the first experiment when cultures were changed from low to high [Ca"], (Fig. 3A) In low [Ca*+],, the amount decreased from -100 to -60% with the same kinetics as for the appearance of 35S-HS-PGs in the trypsin-accessible (cell surface) compartment, reflecting the transport of HS-PGs from the site of sulfation in the Golgi complex to the cell surface. Similar kinetics have also been observed in the rat ovarian granulosa cell system (14). There was an additional, slower decrease between 30-120 min of chase which coincided with the secretion of 3"S-HS-PGs into the medium. Until 120 min of chase there is little net loss of YS-HS-PGs from the cultures (inset, Fig. 4) whereas between 120-240 min chase -65% are lost, reflecting depolymerization in lysosomes. Both the trypsin-accessible and -inaccessible compartments contributed to this loss. In high [Ca"],, the trypsin-inaccessible compartment showed only a small decrease (-10%) between 5-60 min, reflecting the appearance of only -10% of the 35S-HS-PGs in the cell surface compartment.
A further decrease (-10%) was observed between 60-120 min reflecting primarily secretion into the medium. This was followed by a large decrease between 120-240 min reflecting lysosomal depolymerization.
Net loss of YS-HS-PGs from the high and low [Ca2+lr cultures between 60-240 min was essentially the same.
The translocation of HS-PGs from the trypsin-inaccessible to the trypsin-accessible compartment and their simultaneous internalization in a steady-state experiment in low [Ca2+le was demonstrated in a previous section. This observation was further studied by following the movement of 35S-HS-PGs pulse labeled for 10 min in low and high [Ca2+le. Cultures were treated with trypsin for either 2 or 15 min after the indicated chase periods. In low [Ca*+],, the 15-min trypsin digestion released more ?S-HS-PGs than the 2-min digestion at all times after 30 min of chase (Fig. 5A). The difference reached a maximum (-24% of the total) at 60-min chase and gradually decreased thereafter (Fig. 5, inset). By 120 min, about 20% of the total was secreted into the medium. In high [Ca*'],, in contrast, 15-min trypsin digestions failed to release more ?S-HS-PGs than the 2-min digestions, and -80% remained in the trypsin-inaccessible compartment throughout the 120-min chase (Fig. 5B). This indicates that most HS-PGs labeled during the lo-min pulse did not reach the cell surface in high [Ca"],. These results from the low [Ca"], experiments provide additional support for the translocation of HS-PGs from the intracellular compartment to the cell surface and also provide evidence for translocation from the cell surface to the intracellular compartment.  [Ca2+], Regulates Proteoglycan Distribution in Rat Parathyroid Cells [Ca"], in media in a dose-dependent manner with -0.3 mM [Ca"], for the EDn, (Fig. 6).
We next examined at which times [Ca"+], affects the transport and distribution of HS-PGs. Cultures were incubated in low or high [Ca"'], for 20 h and pulse labeled for 10 min in the same [Ca2+le medium. The cultures were then washed and chased either with the medium containing the same [Ca*+], or with the medium-containing reciprocal [Ca2+le. The trypsin accessibilities of 35S-HS-PGs observed after changing [Ca2+le from low to high are virtually identical to those for cultures maintained in high [Ca"], throughout (Fig. 7A). Changing [Ca'+], from high to low gave nearly identical results as for cultures maintained in low [Ca"'], throughout (Fig. 7B) This observation was further verified by switching [Ca*+], at the indicated times during the chase following a lo-min pulse (Fig. 8). The parathyroid cells responded rapidly (tl& < 5 min) to a change of [Ca"], throughout the 120-min chase (Fig. 8, dashed lines). As calcium ions in high Ca'+ medium were chelated by EGTA, 2-min trypsin-accessible 35S-HS-PGs rapidly increased to 40-45% at any time between 30-and 90-min chase (Fig. 8B). Amounts of secreted 35S-HS-PGs were also stimulated to -20% in response to lowering [Ca'+],. Changing from low to high [Ca'+], by adding an aliquot of CaC12 at a 30-min chase rapidly decreased the trypsin-accessible 35S-HS-PGs to the level obtained in the constant high [Ca"], treatment (Fig. 8A) (17)(18)(19). Although precise functions of cell surface HS-PGs are uncertain, they appear to be involved in cell-matrix interactions (20)(21)(22), cytoskeletal organization (23), and as receptors for molecules, such as transforming growth factor-p (24)(25)(26). The rat parathyroid cell line used in this study synthesizes proteoglycans which are almost exclusively HS-PGs (13). The predominant form, HS-PG2, has a structure similar to that described for many other cell surface HS-PGs; a core glycoprotein of -70 kDa with three to four HS chains attached4 (15). HS-PG2 contains more than 90% of the radioactivity incorporated into proteoglycans in all cellular compartments studied in this paper when [35S]sulfate is used as a metabolic precursor (13 value it is evident that -57% of the labeled HS-PGs in this experiment were in population 2, of which -23% were on the cell surface and -34% in the intracellular compartment. The time of digestion required to reach halfway between the initial value (time 0) and the plateau value is -3.6 min and corresponds to the tl,+ for exocytosis from the intracellular compartment to the cell surface. Conversely, the tXA for endocytosis from the cell surface to the intracellular compartment is -2.6 min5 Assuming that the functional life time of population 2 HS-PGs is 90 min (30-120 min of chase) (Fig. 4A), these molecules would recycle -10 times6 between these two compartments. The kinetics exhibited by these HS-PGs are similar to those observed for lysosomal acid phosphatase, most of which migrates from the Golgi complex to the cell surface where it recycles with the endosome compartment for -5 h before being transferred to lysosomes (27). Parameters for recycling of the HS-PGs are also similar to those reported for cell surface receptors which recycle with the endosome compartment (28), such as receptors for transferrin, low density lipoprotein, and mannose 6-phosphate/insulin-like growth factor II (29). Thus, it is likely that population 2 HS-PGs are intercalated in plasma membrane and cycle between selective sites on the cell surface, such as in coated pits, and endosomes (30). The sequestration of the HS-PGs in high [Ca'+],, which suppresses secretory processes in these cells (11,12), is similar to the sequestration of glucose transporters in adipocytes which can be stimulated to redistribute in response to increased insulin levels (31,32). Thus, the HS-PGs in population 2 may be closely associated with the specialized molecules in the cell membranes which sense [Ca"'], and initiate the biological response(s) to [Ca"'], changes that are characteristic of parathyroid cell functions.
We have noted differences among biological responses of parathyroid cells to the level of [Ca2+le. For example, parathyroid hormone secretion is normally at the maximum level in 0.5 mM [Ca2+le (6,9). This concentration, however, is not sufficiently low to redistribute the maximum amount of HS-PGs to the surface of the cells examined in this study. Further, neither Mg2+ nor Ba2+ substituted for Ca2+ in terms of distribution of HS-PGs, although divalent cations, such as MgZ (3), have been shown to suppress the parathyroid hormone secretion. Thus, hormonal secretion and distribution of HS-PGs may be regulated by [Ca2+le at least partially independently. These differences may reflect diverged cellular responses initiated by a common [Ca"'], sensing mechanism which is yet to be identified. Further studies with the system ' The tH for endocytosis of population 2 HS-PGs is based upon the assumption that the cycling process is in steady state and that all aonulation 2 HS-PGs molecules have the same probabilities for (a) exocytosis from the intracellular pool and (5) endocytosis from the cell surface.