Phosphorylation of Parathyroid Hormone by Human and Bovine Parathyroid Glands"

Human and bovine parathyroid gland slices were incubated in vitro for varying time periods with inorganic szP and [35S]meth10nine or ['HIserine. Tissue was then extracted with aqueous medium, and parathyroid hormone (PTH) purified. Incorporated "P was found to coelute with immunoreactive PTH in multiple chromatographic systems, and a peak of phosphorylated material could be resolved from the nonphosphorylated hormone by reversed phase high pressure liquid chro- matography. The amino acid composition of both the phosphorylated and the nonphosphorylated entities conformed to that of the major glandular species of PTH, and phosphorylated hormone accounted for 10- 20% of the total. A time course revealed slow incorporation of "P into hormone, and after a 4-h preincu- bation with inorganic "P, co-elution of "P with both PTH and its precursor was observed. Phosphoserine was identified in purified PTH labeled with ISH]serine. Additionally dilute acid hydrolysis of PTH, labeled with [S5S]methionine and containing 32P, generated an 36S-labeled fragment with which 32P co-chromato- graphed. The results are consistent with in vitro phosphorylation of PTH on serine residues within the NH2- terminal region of the hormone by both human and bovine glands and suggest that phosphorylation of the prohormone occurs as well.

graphed. The results are consistent with in vitro phosphorylation of PTH on serine residues within the NH2terminal region of the hormone by both human and bovine glands and suggest that phosphorylation of the prohormone occurs as well.
The major glandular form of PTH' is known to be an 84 amino acid straight chain peptide hormone in all species so far examined (1-4). The hormone is synthesized on the ribo-*This work was supported by Grant MT-5775 of the Medical Research Council of Canada. A preliminary report of this work was presented at the Fifth Annual Meeting of the American Society for Bone and Mineral Research, San Antonio, Texas. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The abbreviations used are: PTH, parathyroid hormone; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl; HPLC, high pressure liquid chromatography; ProPTH, proparathyroid hormone. The numbering of PTH residues is as follows: the designation -6 refers to the NH,terminal Lys and -1 to the COOH-terminal Arg of the prohormone hexapeptide sequence; for the hormone sequence +l refers to the NH2-terminal residue. When the native hormone or hormonal fragments are referred to, sequence positions are denoted by numbers not preceded by plus or minus. somes of the rough endoplasmic reticulum as a precursor extended by 31 residues at the NH2-terminus and termed PreProPTH (5,6); however, the predominant intraglandular precursor is a shorter moiety extended at the NH2 terminus by only a basic hexapeptide and termed ProPTH (7)(8)(9).
In view of the important role of phosphorylation in the function of many peptides (10, 11) and in view of the recent reports of phosphorylation of peptide hormones (12)(13)(14)(15), we have examined the capacity of parathyroid cells to phosphorylate PTH. Our studies indicate that in vitro phosphorylation of PTH occurs in two species, bovine and human; explore the in vitro time course of this reaction and localize the phosphorylation sites to serine residues within the NH*-terminal region of the molecule.

EXPERIMENTAL PROCEDURES
Biosynthetic Labeling-Human parathyroid tissue (either adenomas or hyperplastic glands) obtained at surgery, or bovine parathyroid glands, freshly dissected at a local abattoir, were placed immediately into ice-cold calcium-and magnesium-free Hank's balanced salt sohtion and processed in the laboratory within 1-2 h. Glands were cut into l-mm slices, and approximately 150-to 330-mg aliquots were equilibrated with 95% O2 and 5% Con. These were incubated for varying time periods at 37 "C in 2 ml of RPMI 1640 medium (Grand Island Biological Co.), supplemented with 5% fetal bovine serum and with 3zPi by addition of (22P)orthophosphoric acid (carrier-free, New England Nuclear) to a final specific activity of 89 pCi/pmol. Incubations were performed with added [36S]methionine (0.25 mCi/ml, 800 Ci/mmol, New England Nuclear) or added (3H]serine (0.1 mCi/ml, 5-25 Ci/mmol, New England Nuclear) with the corresponding unlabeled amino acid omitted. At the end of the incubation, solid urea and concentrated HCI were added to a final concentration of 8 M urea, 0.1 N HC1. The mixture was then homogenized, and the homogenate centrifuged for 15 min at 2200 X g. The supernatant containing extracted hormone was then purified sequentially by gel filtration, ion exchange chromatography, and reversed phase HPLC (16).
Alternatively, at the end of the incubation period, tissue and medium were homogenized at 0 "C in acetone (30 ml/g); the insoluble precipitate was then homogenized in n-hexane (30 ml/g), and the insoluble material from this procedure then homogenized in acetone (30 ml/g) (16). The dried and defatted residue was then extracted in a homogenization medium consisting of 1 M HCI containing 5% (v/ v) formic acid, 1% (w/v) NaC1, and 1% (v/v) trifluoroacetic acid (F&COOH), as previously described (16). The supernatant from this procedure was then passed through cartridges of octadecylsilyl-silica (Cia SepPak, Waters Associates) which were washed with 0.1% F3CCOOH and eluted with 80% acetonitrile (16). The eluate was then purified by reversed phase HPLC.
Gel Filtration-Gel filtration was performed on columns (1.2 X 50 cm) of Bio-Gel P-100 (Bio-Rad) at 22 "C eluting with a buffer of 0.1 M ammonium acetate, pH 5.0. Column elution was monitored by UV absorbance at 280 nm using a Pharmacia UV-1 continuous monitor (Pharmacia, Uppsala, Sweden). Additionally, aliquots of eluted 650gl fractions were taken for determination of 32P and 36S radioactivity by scintillation counting and for radioimmunoassay. Elution of hormone was determined by radioimmunoassay, and the appropriate tubes were pooled and lyophilized.
Ion Exchange Chromatography-Ion exchange chromatography was performed at 22 "C using a column (0.7 X 15 cm) of CM-cellulose (Whatman CM-52, Reeve Angel Co.). After applying the lyophilized hormone from the P-100 chromatography step in the starting buffer, a linear conductivity gradient was developed, using 70 ml of 0.01 M ammonium acetate, conductivity 1.9 mS, pH 5.0, and 70 ml of 0.10 M ammonium acetate, conductivity 16.5 mS, pH 6.5. Fractions of 1 ml were collected, and aliquots were taken for determination of 32P and 35S radioactivity by scintillation counting, and for radioimmunoassay.
ter (Copenhagen) model CDM 83 conductivity meter with a PP1042 All conductivity measurements were made at 22 "C using a Radiomeelectrode. In some cases eluted fractions containing hormone, as determined by radioimmunoassay, were pooled and lyophilized, and then taken up in 0.1% F3CCOOH and pumped onto a Cls pBondapak column for reversed phase HPLC.
Reuersed Phase HPLC-Reversed phase HPLC was performed as previously described (16,17) on a Waters Associates (Milford, MA) HPLC system consisting of two 6000-A pumps controlled by a M720 system controller and using C18 p Bondapak columns. Samples for loading were diluted with 0.1% F3CCOOH or 0.13% heptafluorobutyric acid (FTC3COOH) and were generally pumped directly onto the HPLC column using one of the unused ports on the "aqueous" pump. Columns were then developed over 60-90 min at a flow rate of 1.5 ml/min, with linear gradients of acetonitrile containing 0.1% F3CCOOH or with linear gradients of acetonitrile containing 0.13% F7C3COOH. Column eluates were monitored for UV absorbance at 210 nm using a variable wavelength flow-through spectrophotometer (model "480, Waters Associates) and a M730 data module (Waters). Radioactivity was determined in aliquots of eluted fractions either by scintillation counting or by using Cerenkov radiation (18). Other aliquots were employed for radioimmunoassay.
For chromatography of dansyl amino acids, the mixture of derivatized residues with standards was dissolved in 0.1% F3CCOOH, injected directly onto a CIS pBondapak column using a U6K (Waters) injector and then eluted with an isocratic system of 7.5% acetonitrile in 0.1% F3CCOOH (19).
Radioactiue Counting-Aqueous samples (0.05 to 0.1 ml) were counted in 10 ml of fluor (New England Nuclear, Formula 947) on a Packard Tri-Carb liquid scintillation spectrometer (Packard Instrument Co.) with window settings such that no correction for spillover of 3H or 35S into the 32P window was needed for these samples. Alternatively 32P in 1.5-ml fractions from HPLC was counted in minivials, without fluor, via Cerenkov radiation (18). All 32P data from a given incubation were corrected for decay and are reported as counts per min per total chromatographic fraction on the date of the incubation; analyses were completed within 15 days of the incubation.
Radioimmunoassay-Radioimmunoassays for PTH were carried out as previously described (16), employing as tracer 1251-labeled bovine PTH-(1-84) and, as standard, purified bovine PTH-(l-84) (either kindly supplied by Dr. H. T. Keutmann, Endocrine Unit, Massachusetts General Hospital, Boston, MA, or purchased from Bachem Corp., Torrance, CA). Dextran-charcoal was used for phase separation. Dried portions of column fractions were dissolved in plasma known to have no detectable PTH and assayed in appropriate dilutions in duplicate. Two different antisera were employed. With antiserum AS1127/23 (purchased from Burroughs-Wellcome, London, U.K.), synthetic PTH-(53-84) (Bachem Corp.) inhibited tracer binding by over 70%, whereas minimal inhibition was seen with PTH-(1-34) (Bachem Corp.); this antiserum, therefore, recognized antigenic determinants in the COOH-terminal region of PTH-(1-84), and assays employing this COOH-terminal antiserum were, therefore, termed "C assays." Antiserum R-4 raised in rabbits to synthetic bovine PTH-(1-34) was kindly supplied by Dr. G. Hurst, Oklahoma State University, Stillwater, Oklahoma. Assays employing this antiserum, therefore, recognized the NH1-terminal region of bovine PTH-(1-84) and were termed "N assays." Limited Acid Hydrolysis and Dansylation-Mild acid hydrolysis of an aliquot of HPLC-purified PTH-(1-84) labeled with [3H]serine was carried out in 0.2-0.5 ml of constant boiling HCl (Pierce Chemical Co.) under nitrogen at 110 "C for 2 h. Hydrolysis was stopped by addition of 0.2 M Na2C03 and drying the mixture. Derivatization of the freed residues was then carried out by addition of 0.2 ml of dansyl chloride, 2.5 mg/ml, and incubation at 37 "C for 1 h (20). Dansyl derivatives of standard serine and phosphoserine (Sigma) were similarly prepared. The amino acid mixture and 240 nmol of each standard were then dried, taken up in 0.1% F3CCOOH, and injected together onto a CI8 pBondapak column.
Dilute Acid Hydrolysis-HPLC-purified PTH-(l-84) containing 32P and internally labeled with [%]methionine was lyophilized, then dissolved in 0.2 ml of 0.03 N HCl, and heated to 110 "C for 4 h in sealed ampoules that had been flushed with nitrogen (21). The mixture of peptides generated by the hydrolysis was then dried, taken up in 0.1% FSCCOOH, and injected onto a Cl8 pBondapak column.
Amino Acid Analysis-Approximately 10 pg of isolated bovine PTH was hydrolyzed in an evacuated glass tube in 0.25 ml of 6 M HCl (with a crystal of phenol to prevent loss of tyrosine) for 18 h at 110 "C. Amino acid analysis was performed on a Beckman high performance system 6300 analyzer.

RESULTS
Identification of Phosphorylated PTH-In initial experiments after 4 h of incubation of human parathyroid tissue with 32Pi and [%]methionine, tissue and medium were extracted with 8 M urea, 0.1 N HC1 and the extract applied to a column of Bio-Gel P-100 (Fig. lA). Fractions eluting in the position of immunoreactive PTH were pooled and subjected to ion exchange chromatography on CM-cellulose (Fig. 1B). A major peak of concordant 32P and 35S and of NH2 and COOH immunoreactivity was seen eluting at a conductivity of 6 mS (in the position of standard PTH-(1-84)). When the material in the peak was then subjected to reversed phase HPLC, again a major peak of concordant 32P, 35S, NHz, and COOH immunoreactivity eluted from the column (Fig. IC). The ratio of 32P to PTH immunoreactivity in this peak (C) remained constant when the material in this peak was rechromatographed by reversed-phase HPLC. Similar results were obtained in experiments with bovine tissue.
Separation of Phosphorylated from Unphosphorylated PTH-Separation of phosphorylated from unphosphorylated PTH was achieved by sequential purification by HPLC (Fig.  2). Phosphorylated peptides were extracted from bovine parathyroid tissue after a 4-h incubation with 32Pi, partially purified on cartridges of octadecylsilyl-silica and then subjected to reversed phase HPLC on a CIS pBondapak column employing a linear gradient of acetonitrile in 0.1% F3CCOOH (Fig.  2, upper left). The major immunoreactive peak (Fig. 2, peak  A ) was then rechromatographed on the same column employing a linear gradient of acetonitrile in 0.13% F3C3COOH as a counterion (Fig. 2, upper right). Two peaks (Fig. 2, B and C ) of co-eluting immunoreactivity and UV absorbance were now seen; 32P co-eluted only with the earlier peak. Each peak was then separately re-chromatographed on the same column eluting with a shallow linear gradient of acetonitrile in 0.1% F3CCOOH. The earlier eluting peak ( B ) , when re-chromatographed (Fig. 2, lower left), retained the same approximate ratio of 32P, immunoreactivity, and UV absorbance and appeared symmetrical (peak B'). The later eluting peak (C) when re-chromatographed (Fig. 2, lower right) also appeared symmetrical (peak C') but contained no 32P. The amino acid composition of both peptides, after complete acid hydrolysis, conformed to the published composition of bovine PTH-(1-84) ( In four experiments, as assessed by immunoreactivity, UV absorbance, and amino acid analysis, the approximate ratio of phosphorylated to unphosphorylated hormone was 1:6 (range 1:9 to 1:4).
Time Course of Phosphorylation-To assess the time course of phosphorylation, human parathyroid slices were incubated in vitro with 32Pi and [35S]methionine for increasing time periods, following which hormone was extracted in 8 M urea, 0.1 N HCl and purified by gel filtration followed by ion exchange chromatography. At early incubation times 35S was incorporated into a peak which eluted on ion exchange chromatography as a basic peptide in the elution position of ProPTH (Fig. 3, A and B). No co-elution of 32P with 35S was seen at the earliest time period (Fig. 3A). With increasing "P and 35S were determined in all fractions in the experiment shown in C; however, no radioactivity was detected other than in the peaks shown. (1-84) (Fig. 3, B and C; see also Fig. 1). Increasing 32P was observed co-eluting with this material, with increasing time of incubation ( Fig. 3; see also Fig. 1). When the total 35S and 32P radioactivity was determined in peaks eluting in the position of PTH-(1-84) and ProPTH, the overall time course suggested the early appearance and then disappearance of labeled amino acid in the peptide eluting in the position of ProPTH, the later appearance but then reduction of labeled amino acid in the peptide eluting in the position of PTH, and the progressive increase in the incorporation of 32P into a peptide eluting in the position of PTH (Fig. 4).

2.
The effect of preincubation of glandular tissue with 32Pi was then determined. After a 4-h preincubation of human parathyroid tissue in RPMI 1640 medium with 32Pi, tissue slices were washed with medium and then reincubated with fresh RPMI 1640 medium containing [35S]methionine but no unlabeled methionine. After 15 min of incubation, hormone was extracted in 8 M urea, 0.1 N HCl and purified by gel filtration followed by ion exchange chromatography as before. Two peaks of 35S were now seen, one smaller one, eluting in the Time ( min position of PTH-(1-84) and a more basic peak eluting later, in the position of ProPTH (Fig. 5). Additionally, two major peaks of 32P were now seen as well, co-eluting with the two 35S peaks (Fig. 5).
Site of Phosphorylation-To provide further evidence of the existence of phosphorylated hormone and to ascertain the nature of the phosphorylated residue, bovine parathyroid gland slices were incubated with [3H]serine, and then labeled hormone was extracted and purified by HPLC as described above. An aliquot of the purified 3H-labeled phosphorylated hormone was then subjected to mild acid hydrolysis, and the hydrolysate derivatized with dansyl chloride. The dansyl amino acids, with added standard dansyl serine and dansyl phosphoserine, were then injected onto a C,a IBondapak column and eluted isocratically (Fig. 6). Two major peaks of 3H radioactivity were observed, co-eluting with the derivatized standard phosphoserine and serine (Fig. 6).
Region of Phosphorylation-To determine the region of the molecule phosphorylated, HPLC-purified PTH, labeled with [35S]methionine and containing 32P, was subjected to dilute acid hydrolysis. This procedure is known to cleave PTH at aspartic acid residues (22) generating only one methionine-  32P (c".), immunoreactivity, and UV absorbance was then rerun on the same CIS pBondapak column eluting with a linear (36-48%) gradient of acetonitrile in 0. 13% FvC3COOH over 60 min (upper right). The peak ( B ) of co-eluting 32P, immunoreactivity, and UV absorbance was then rerun, eluting with a linear gradient of 28-40% acetonitrile in 0.1% F3CCOOH over 60 min (lower left), resulting in an apparently homogeneous peak ( B ' ) which was then subjected to amino acid analysis. Similarly the peak (C) of coeluting immunoreactivity and UV absorbance (upper right) was rerun eluting with a linear gradient of 28-40% acetonitrile in 0.1% F3CCOOH over 60 min (lower right) resulting in an apparently homogeneous peak (C') which was also subjected to amino acid analysis.
containing (35S-containing) fragment, PTH-(1-29), and several non-methionine-containing fragments. When the mixture of PTH fragments obtained by this procedure was resolved by HPLC, only a single peak of 32P was seen, comigrating with the single peak of 35S (Fig. 7); consequently phosphorylated material co-chromatographed with an NHZterminal fragment generated from purified PTH.

DISCUSSION
Our studies have demonstrated co-elution of incorporated 32P and PTH immunoreactivity from extracts of parathyroid tissue in multiple chromatographic systems. Thus phosphorylated material could not be distinguished from PTH-  in systems resolving on the basis of size (gel filtration), charge (ion exchange chromatography), or hydrophobicity (HPLC). Consequently the phosphorylated material appeared similar to native PTH-( 1-84) in many of its chemical characteristics. Only when heptafluorobutyric acid, a stronger ion-pairing agent than trifluoroacetic acid, was employed in the solvent system was it possible to readily resolve the phosphorylated from the unphosphorylated entities (Fig. 2). NH2-and COOHterminal immunoreactivity was retained in the phosphorylated as well as the nonphosphorylated peaks; amino acid analysis of each peak revealed, in each case, the composition of PTH-(1-84) and failed to disclose any contaminating entity in the phosphorylated peak. Finally, after dilute acid hydrolysis, the 32P co-migrated with an NHz-terminal fragment generated from phosphorylated PTH-(1-84) (Fig. 7). These results, therefore, provide strong evidence for the phosphorylation of PTH by parathyroid glands incubated in uitro.
The failure of progressive incorporation of radiolabeled amino acids into newly synthesized hormone during prolonged in uitro incubation (Fig. 4) suggests that with continued incubation either new biosynthesis was decreasing, or, more likely, hormonal degradation was increasing, or both events were occurring. Such an increase in proteolytic activity could include accelerated conversion of ProPTH to PTH, as suggested by the presence of radiolabeled ProPTH after a 20 min  incubation ( Fig. 3) but not after 60 min. However, when tissues were rinsed after a 4-h incubation and then reincubated with fresh medium and radiolabeled amino acids for 15 min, proteolysis appeared to be slowed since both newly synthesized ProPTH and PTH were seen (Fig. 5). Of interest is the identification of 32P in a peak co-eluting with PTH ( Fig. 3B) when only minimal 32P was seen in a ProPTH peak.
This finding is consistent with phosphorylation of the hormone only after cleavage from the prohormone. However, complex in vitro kinetics involving accelerated conversion of ProPTH to PTH with increasing time of incubation, at which time phosphorylation with exogenous 32P may only have begun, could have obscured the capacity to demonstrate modest quantities of phosphorylated ProPTH prior to conversion. When a 4-h preincubation of parathyroid tissue with inorganic 32P was performed, the tissue washed and reincubated for 15 min, phosphorylation of both ProPTH and PTH was now clearly seen (Fig. 5). Whether this was due to reduced proteolysis, consequent to modification of the initial experimental protocol, requires further investigation. Nevertheless, phosphorylation of the hormone prior to cleavage from the prohormone was demonstrable. This, of course, does not exclude the possibility of phosphorylation of additional hormonal loci after cleavage. Phosphorylation of serine, threonine, and tyrosine, and less frequently of histidine and lysine residues has been reported (10). Inasmuch as evidence for phosphorylation of both the bovine and human PTH molecules was obtained, but the bovine molecule lacks a threonine and the human is deficient in tyrosine, it was decided to explore the possibility that the site of phosphorylation would be serine residues. The results (Fig. 6) demonstrated clearly that a major site of hormonal phosphorylation appears to be on serine residues.
Multiple serine residues are present within the sequence of the PTH- (1-84) molecule (1, 2, 4). Our studies employing limited chemical cleavage (via dilute acid hydrolysis) of the molecule localize the phosphorylation site to residues within the NH2-terminal 1 to 29 sequence of the molecule. Previous reports have suggested the possibility of co-translational phosphorylation of nascent peptides with phosphoseryl-tRNA (23). Additionally, post-translational modification (10) could account for the phosphorylation of PTH. In both bovine and human species the serine residue at position 17 is located within the sequence Ser-X-Glu which is a recognition site for the action of noncyclic AMP-dependent kinases ("physiological casein kinases") (24, 25), and, therefore, this position could be a site of post-translational phosphorylation. As well, in both species of hormone the serine residue at position 3 could be a site for cyclic AMP-dependent phosphorylation when located within the prohormone prior to cleavage inasmuch as the prohormone sequence (26) Lys-2-Arg"-X+'-Y+2-SeP3 constitutes a substrate-specific region for the cyclic AMP-dependent kinases. In view of the fact that the bioactivity of PTH is known to reside within the NH2-terminal region of the molecule (27, 28), phosphorylation of serine residues within this region could be of particular significance in modulating hormonal action. Further studies are, therefore, required to identify the precise site, mechanism, and extent of hormonal phosphorylation  Fig. 1. Chromatography was performed on a CM-cellulose column using an ammonium acetate buffer gradient, as described under "Experimental Procedures." Two major peaks, each of 32P and %S were seen, eluting in the position of PTH-(1-84) and of ProPTH. PTH immunoreactivity, determined in eluted fractions with COOH-directed assays, coincided with the major peaks of 32P and 35S as shown in Fig. 1.
under various conditions of ambient calcium and the effect of hormonal phosphorylation on transport through the parathyroid cell and on bio-and immunoreactivity. Development of specific immunoassays for phosphorylated versus nonphosphorylated hormone could help clarify the in vivo significance of these findings which have been demonstrated by in vitro techniques. The identification of phosphorylated PTH may, therefore, be of major importance in further elucidating multiple aspects of the physiology of parathyroid hormone and of the parathyroid cell.