The Central Part of Parathyroid Hormone Stimulates Thymidine Incorporation of Chondrocytes"

The stimulation of DNA synthesis in primary cell cultures of chicken chondrocytes by parathyroid hormone was studied by assaying ['Hlthymidine incorpo- ration into DNA. Optimal assay conditions were determined by varying cell age, plating density, and incu- bation time. Under these conditions DNA synthesis was significantly stimulated by parathyroid hormone (PTH) and some of its fragments: cells treated with human (h)PTH( 1-84), bovine (b)PTH( 1-34) and and displayed 2.6-fold enhanced ['Hlthymidine incorporation in a dose-dependent manner. The fragment hPTH(28-48) led to a similar stimulation, whereas [Tyr4']hPTH(43-68) and had no effect. Using a series of synthetic hPTH peptides covering the central region of the hormone molecule (residues 25-47), we could delimitate further this putative mitogenic functional domain to a core region between amino acid residues 30 and 34. The effect of PTH on ['Hlthymidine incorporation could not be mim- icked by forskolin, indicating that the corresponding signal is not mediated by CAMP. It is, however, inhib- ited by EGTA and cannot be provoked in the absence of calcium ions in the medium. Therefore, the results presented indicate a hitherto mg/l, respectively), Hepes (10 mM, pH 7.2)), and FCS (10% v/v) and counted. The highly homogenous cell population isolated by this procedure revealed polygonal morphology which is typical for chon- drocytes (13,18). Their ornithine decarboxylase activity was inducible by PTH(1-34). They were devoid of alkaline phosphatase activity and dedifferentiated upon senescence or secondary culture (data not shown). These observations strongly indicate the identity of chondro- cytes and distinguish them from fibroblasts (ornithine decarboxylase (19)) or osteoblasts (alkaline phosphatase (20, 21)). Assay for Thymidine Incorporation-The DNA synthesis rate was assayed in monolayer culture by the incorporation of [3H]thymidine into perchloric acid-precipitable material. After variation of the essential parameters, the following conditions were found to be optimal to assay the stimulation of thymidine incorporation by PTH: cells were seeded into microtiter plates with 96 wells (6-mm diameter; 14,000 cells/cm2) and a 200-pl volume of medium containing 10% FCS and were incubated at 37 "C in an atmosphere of 5% COz. After 17 h, the medium was replaced by 200 pl of serum-free medium. After 4 days, chondrocytes were incubated for 4 h with the appropriate effector and 1 pCi of [3H]thymidine. Subsequently, medium was removed and the cells were washed twice with 200 pl of phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH2P04, 0.8 mM Na2HPOI, pH 7.0). Then cells were lysed by 100 pl of 2% (v/ v) Nonidet P-40 and 2% (w/v) sodium dodecyl sulfate treatment and the perchloric acid-insoluble material was precipitated by adding an equal volume of ice-cold 2% (w/v)


The Central Part of Parathyroid Hormone Stimulates Thymidine Incorporation of Chondrocytes"
(Received for publication, January 17,1989)  had no effect. Using a series of synthetic hPTH peptides covering the central region of the hormone molecule (residues 25-47), we could delimitate further this putative mitogenic functional domain to a core region between amino acid residues 30 and 34. The effect of PTH on ['Hlthymidine incorporation could not be mimicked by forskolin, indicating that the corresponding signal is not mediated by CAMP. It is, however, inhibited by EGTA and cannot be provoked in the absence of calcium ions in the medium. Therefore, the results presented indicate a hitherto unidentified functional domain of PTH in the central part of the molecule which exerts its mitogenic effect on chondrocytes in a CAMP-independent manner but seems to involve calcium ions for signal transduction.

Klaus-Dieter Schluter, Heribert Hellstern, Edgar Wingender, and Hubert MayerS
Parathyroid hormone (PTH)' is a peptide hormone of 84 amino acid residues which regulates calcium homeostasis by acting on renal cortex, cells of the skeletal system, and other target tissues. It stimulates the activity of adenylate cyclase (see Ref. 1 for review) or the induction of ornithine decarboxylase (2). Furthermore, PTH has been shown to exert effects on a variety of cells (3)(4)(5)(6)(7). In particular, it stimulates the DNA synthesis in secondary cultures of chicken osteoblasts (8), it enhances the growth of chicken embryonic cartilage in *This work was part of Protein Design Project No. 03 8706 9 supported by the Bundesministerium f i r Forschung und Technologie of the Government of the Federal Republic of Germany. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In contrast to the well characterized effect of native PTH, hormone variants, and fragments on the renal receptor which is linked to adenylate cyclase (reviewed in Ref. l), the mechanism through which it acts on cell proliferation remains to be elucidated. Therefore, we attempted to define-thLhnctional domain of the PTH molecule w l m A i $ T s a r y 3 o r the mitogenic response. For this purpose we established a primary cell culture system of chondrocytes, which were isolated from 16-day-old chicken embryos by separating the sterna as described (13). PTH and PTH-derived peptides were assayed for their capability to enhance incorporation of [3H]thymidine by these cells as a criterion for increase of DNA synthesis (9,14,15). Peptide Synthesis and Purification-All syntheses were performed in the automated mode in a modified, computer-controlled version of the Biosearch 9600 Peptide Synthesizer using fluorenylmethoxycar-bony1 strategy (16) and ap-benzyloxybenzyl alcohol resin. For peptide bond formation Castro's reagent (17) was used in the activation step. Protected peptide resins were cleaved in trifluoroacetic acid/thiophenol/anisol 8l:l for 2 h at 40 "C. The free peptides were precipitated afterwards by addition of ether. All crude peptides were routinely purified by gel filtration on Sephadex G10 and preparative reversed phase high performance liquid chromatography (column YMC ODS-5,20 X 250 mm). The purified peptides were characterized by amino acid analysis (Biotronic LC 5001) and fast atom bombardment mass spectrometry (Kratos MS 50).

Chemicals and Supplies
Isolation and Culture of Embryonic Chick Chondrocytes-Chondrocytes were isolated from sterna of 16-day-old embryonic chicks according to Yasui (13) with some modifications. Collagenase was dissolved in Hanks' balanced salt solution (0.4 mM Na2HP04, 140 mM NaCl, 5.3 mM KCl, 0.4 mM MgS04, 1.3 mM CaC12, 0.4 mM KHzPO4, 0.5 mM MgC12, 5 mM glucose, 0.05% gentamycin) at a concentration of 286 units/ml. Sterna were treated for 30 min with collagenase solution at 37 "C. Cells first released by this procedure were discarded because of their very poor responsiveness towards PTH. The collagenase treatment was subsequently repeated leading to a highly homogenous cell population of chondrocytes. The cells obtained by a second collagenase treatment were collected by centrifugation (1000 rpm, 10 min), washed twice with medium (Dulbecco's modified Eagle's medium, GIBCO, supplemented with NaHC03 (45 mM), glutamine (2 mM), penicillin G/streptomycin (65 mg/l and 100 11087 mg/l, respectively), Hepes (10 mM, pH 7.2)), and FCS (10% v/v) and counted. The highly homogenous cell population isolated by this procedure revealed polygonal morphology which is typical for chondrocytes (13,18). Their ornithine decarboxylase activity was inducible by PTH(1-34). They were devoid of alkaline phosphatase activity and dedifferentiated upon senescence or secondary culture (data not shown). These observations strongly indicate the identity of chondrocytes and distinguish them from fibroblasts (ornithine decarboxylase (19)) or osteoblasts (alkaline phosphatase (20, 21)).
Assay for Thymidine Incorporation-The DNA synthesis rate was assayed in monolayer culture by the incorporation of [3H]thymidine into perchloric acid-precipitable material. After variation of the essential parameters, the following conditions were found to be optimal to assay the stimulation of thymidine incorporation by PTH: cells were seeded into microtiter plates with 96 wells (6-mm diameter; 14,000 cells/cm2) and a 200-pl volume of medium containing 10% FCS and were incubated at 37 "C in an atmosphere of 5% COz. After 17 h, the medium was replaced by 200 pl of serum-free medium. After 4 days, chondrocytes were incubated for 4 h with the appropriate effector and 1 pCi of [3H]thymidine. Subsequently, medium was removed and the cells were washed twice with 200 pl of phosphatebuffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH2P04, 0.8 mM Na2HPOI, pH 7.0). Then cells were lysed by 100 pl of 2% (v/ v) Nonidet P-40 and 2% (w/v) sodium dodecyl sulfate treatment and the perchloric acid-insoluble material was precipitated by adding an equal volume of ice-cold 2% (w/v) perchloric acid in the presence of 1% herring sperm DNA as carrier. After storage for 10 min at -20 "C, the precipitated material of each well was transferred to glass fiber filters with a Scatron-As semiautomatic cell harvester. Filters were dried at 80 "C for 20 min, transferred into scintillation vials, and 2 ml of scintillation mixture (scintillator 299; United Technology Packard) were added. Radioactivity of the samples was determined with a p counter (Minimax; United Technology Packard). Each test consisted of 5 up to 10 identically treated wells and was repeated at least twice with freshly prepared chondrocytes. In separate experiments, we confirmed the correlation between enhancement of [3H]thymidine incorporation after 4 h and subsequent increase in cell number after a 24-h incubation period.
Assay for CAMP Determination-Chicken chondrocytes were seeded into 24-well plates with 500 pl of medium and 10% FCS (8000 cells/cm2). Cells were cultivated as described above. After 4 days, chondrocytes were incubated for 20 min with the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthin followed by an additional 10min incubation with the corresponding PTH fragment. The induction was stopped by removing the medium and washing the cells once with ice-cold PBS. Subsequently, the cells were incubated in PBS at 85 "C for 10 min. After replacement of the buffer by ice-cold PBS, the chondrocytes were harvested with a rubber policeman and sonicated in ice with three strokes for 5 s each. The cell homogenate was centrifuged (1800 X g, 15 s). The CAMP content of the supernatant was determined by radioimmunoassay (Amersham, Braunschweig, West Germany).
Protein Determination-The protein concentration was determined using the Bio-Rad assay with bovine serum albumin as standard (22). 50 p1 of cell extract were diluted to 800 pl and mixed with 200 pl of concentrated reagent; after vigorous mixing, absorption was measured at 595 nm.

RESULTS
Effect of PTH Fragments on PH]Thymidine. Incorporation-In a first approach we tested whether parathyroid hormone exhibits a stimulating influence on [3H]thymidine incorporation into acid-precipitable material by chicken chondrocytes. Extensive optimization of the assay system with respect to cell age, plating density, and induction period has been performed using bPTH( 1-34) since this fragment exerts     Tables I and 11 incorporation in a dose-dependent manner and to an extent similar to fetal calf serum as a positive control (Table I). Above a critical value, the stimulation declined for any hitherto unknown reason ( Fig. 1; cf. also Refs. 23 and 24).
Even more su~risingly, hPTH(28-48) was observed to enhance [3H]thymidine incorporation comparably to that of the complete hormone (Table I). These results, which are summarized in Fig. 2

Mapping of the Amino-terminal Border of the Putative
Mitogenic Functional Domuin-For a more precise delineation of the functional PTH domain stimulating [3HJthymidine incorporation, we synthesized a series of peptides with variable amino termini; to avoid racemization of the first amino acid coupled to the resin (25), we started our synthesis with the glycine residue in position 47 and proceeded down to position 34. Subs~uently, an aliquot of the peptide-loaded resin was removed after each condensation cycle. Thus, we obtained the hPTH fragments from amino acid residue 47 down to either position 34, 33, 32, 31, 30, 29, or 28, respectively. These peptides were applied to the [ 3 H~t h~i d i n e incorporation assay described above. The fragments down to a length of 18 amino acid residues ( i e . hPTH(30-47)) were able to stimulate DNA synthesis, whereas those of 16 and less amino acids (corresponding to hWH(32-47)) were inactive in the concentration range studied (Table 11). The peptides hPTH(31-47) and hPTH(33-47) exhibited a slight stimulus on ['Hlthymidine incorporation in some of our experiments (Table 11), but this effect could not be reproduced in a statistically significant manner and also was not reproducible between experiments with different cell charges. Similar to the peptides 28/30-47, the fragment hPTH(25-39) was slightly active (Table 11) indicating that the region beyond position 39 is not essential for stimulating ['Hlthymidine incorporation. However, the shortest fragment assayed which still covers the presumptive functional domain, hPTH(28-39), was inactive (Table 11). Most likely, this peptide comprising 12 amino acids is too short to adopt the structure which is required for binding to the corresponding receptor.
The Effect of PTH on DNA Synthesis Is CAMP-independent-As has been described for the PTH action via CAMP, a genuine amino terminus is required to stimulate the receptorassociated adenylate cyclase. As the stimulation of [3Hfthymidine inco~oration is exerted by a completely distinct region of the hormone molecule, we suggest that the corresponding signal is not mediated by CAMP. To substantiate this hypothesis, we directly measured the CAMP content of chondrocytes after stimulation with bPTH( 1-34), [Nles~38, Tyr34'] bPTH , and [Tyr6z,Asp '6]hPTH(52-84) at concentrations ensuring stimulation of ['Hlthymidine incorporation in the appropriate assay and compared it to nonstimulated control cells (Table 111). Only the fragment which has an intact amino terminus, bPTH(1-34)) was able   to enhance the cellular cAMP content significantly, whereas the sequence 3-34 revealed only very weak activity, and the midregional and carboxyl-terminal regions of PTH were completely inactive. Furthermore, we attempted to mimic the PTH effect by enhancing the intracellular cAMP level with forskolin, which is known to stimulate adenylate cyclase. This agent has been applied in a concentration range between lo-' and lo-' M.
We were able to detect a distinct stimulation of the activity of ornithine decarboxylase in chondrocytes, which is known to be CAMP-mediated (29), by M forskolin and within TABLE I11 Influence of PTH fragments on the CAMP level in chicken chondrocytes Chicken chondrocytes were grown in primary culture as described under "Materials and Methods." After 4 days, they were treated for 10 min with the indicated PTH peptides. The cells were subsequently homogenized, and the cAMP content of the extracts was determined using a radioimmunoassay. Results are aiven as mean f S.E. for duplicates of either induced (E) or controrcells (C). the same time period of 4 h.' However, this agent did not provoke any enhancement of DNA synthesis (see Fig. 3). Accordingly, attempts to provoke stimulation of DNA synthesis by administration of dibutyryl-CAMP failed also (data not shown).
It has been reported that the influx of calcium ions can be stimulated by PTH by a CAMP-dependent as well as by a CAMP-independent mechanism (26). To test if the PTH signal transduction leading to enhanced DNA synthesis is influenced by influx of calcium ions, we simultaneously added PTH and the calcium-chelating agent EGTA to chondrocyte cultures. Increasing concentrations of the chelator led to a progressive loss of the hormonal stimulation capability (Fig.  4), but did not repress the basal level of DNA synthesis. From this finding we conclude that chelation of (presumably) calcium ions selectively abolishes the PTH-induced stimulation of DNA synthesis in chondrocytes rather than damages the cells unspecifically. Moreover, no enhancement of [3H]thymidine incorporation by hPTH  was observed when repeating the experiment in calcium-free medium. After supplementation with calcium ions, however, this activity was restored (Fig. 4, inset). '

DISCUSSION
The results presented above have shown that PTH stimulates I3H]thymidine incorporation of chondrocytes. Additional experiments have also shown that the increase in DNA synthesis we measured after 4 h is followed by an increase in cell number observed after 24 h.
These results agree with those obtained for cartilage organ cultures (9-11), although other investigators described that PTH had little or no effect on DNA synthesis in chondrocytes (27,28). The discrepancy is most likely due to the different assay conditions applied; the reports cited investigated the PTH response of chondrocytes at confluence after hormone treatment for at least 16 h in the presence of 10% FCS (27, 28). We found that high plating density (756,000 cells/cm2), prolonged hormone treatment (76 h), or the presence of calf serum (FCS 7 1.5%) likewise impedes the PTH stimulus on the DNA synthesis of chondrocytes. Finally, it has been shown that in an organ culture system, the [3H]thymidine incorporation of chondrocytes after PTH stimu~ation depends on their stage of differentiation (9). This could conceivably account for the differences discussed here.
Under the conditions applied in this study, chicken chondrocytes respond to PTH and appropriate hormone fragments by an approximately 2.6-fold increase in the rate of thymidine incorporation. This corresponds to the effect PTH exerts on DNA synthesis of rat osteoblasts (29). Somjen et al. (29) could also show that the increase in DNA synthesis coincides with an induction of creatine kinase in these cells.
Up to now, only very few in uitro or in uiuo agonist effects could be detected for amino-terminally truncated PTH fragments. [N~es.1s,Tyr3*]bPTH(3-34)amide has weak agonist properties (30), and this fragment as well as [Tyr34]bPTH(7-34)amide and [T~r~~]hPTH  are still able to stimulate glucose-6-phosphate dehydrogenase activity in the distal renal tubule (31). No function has hitherto been ascribed to the midregional fragment hPTH . Moreover, Takano et aZ. (32) have shown that even the region bPTH( 1-27) is sufficient to enhance the cAMP level as well as ornithine decarboxylase activity, although carboxyl-terminally adjacent sequences support its effects.
However, in the [3H]thymidine incorporation assay optimized for induction with bPTH(1-341, the peptide hPTH(28-48) revealed a distinct stimulating activity. Hormone fragments covering very different regions of the PTH molecule such as residues 13-34 or 30-47 were also active (Fig. 2). This might either indicate (i) that there exist two independent mitogenic regions within these peptides or (ii) that a "core" of the mitogenic functional domain of hPTH resides in the overlapping region, i.e. residues 30-34. We presently favor the second hypothesis, since the central region, which is the most hydrophobic part of the PTH molecule, is highly conserved up to position 39 between different species such as human, bovine, and porcine PTH (Fig. 2) suggesting some physiological significance.
The sequences flanking the ''core domain" residues 30-34 seem to stabilize some minimal structural requirement for receptor binding, since a peptide as short as residues 28-39 is inactive, although it contains the core region ( Table 11) Up to now, most PTH effects were studied either directly by investigating the stimulation of adenylate cyclase (27, 28, 32, 33) or indirectly by examination of CAMP-mediated effects, e.g. ornithine decarboxylase induction (27, 32). However, two lines of evidence strongly indicate that PTH enhances [3H]thymidine incorporation of chondrocytes by a CAMP-independent mechanism: (i) amino-terminally truncated PTH fragments, which are known not to stimulate adenylate cyclase, are active in the [3H]thymidine incorporation assay. This has been confirmed for hPTH   (Table  111). (ii) The adenylate cyclase stimulating agent forskolin as well as dibutyryl-CAMP did not affect DNA synthesis, although both agents were able to induce ornithine decarboxylase in chondrocytes (data not shown), and the enhancement of the cAMP level was confirmed experimentally.
It has recently been shown that calcium channels of osteoblast-like cells are influenced by PTH stimuli in a CAMPindependent manner (26). On the other hand, experiments of other laboratories (35,36) as well as preliminary results of our own group (data not shown) have shown that administration of low doses of a phorbol ester derivative increased the DNA synthesis of chondrocytes. This suggests the involvement of protein kinase C in this signal transfer process (37).
It is known that inositol trisphosphate after phosphorylation to tetrakisphosphate influences the influx of calcium ions into the cells (38), and it has been suggested that calcium might play an important role in the stimulation of cell proliferation by several growth factors (39). The results reported here also indicate that the mitogenic activity of PTH is dependent on external calcium ions (Fig. 4).
To achieve a more profound understanding of the mechanisms of PTH effects on its target cells, additional cell types (e.g. osteoblast-like and k~dney-derived cells) have to be employed. Furthermore, a much broader spectrum of PTH peptides obtained from chemical peptide synthesis or variants of the complete hormone synthesized in Escherichia coli (40) will provide valuable information about the functional domains of the hormone with respect to its different activities. These experiments are in progress.
The finding that different hormonal effects are exerted by to the attractive speculation that its apparently counteracting Natl. Acad. Sci. U. S