L-amino acids regulate parathyroid hormone secretion.

Parathyroid hormone (PTH) secretion is acutely regulated by the extracellular Ca(2+)-sensing receptor (CaR). Thus, Ca(2+) ions, and to a lesser extent Mg(2+) ions, have been viewed as the principal physiological regulators of PTH secretion. Herein we show that in physiological concentrations, l-amino acids acutely and reversibly activated the extracellular Ca(2+)-sensing receptor in normal human parathyroid cells and inhibited parathyroid hormone secretion. Individual l-amino acids, especially of the aromatic and aliphatic classes, as well as plasma-like amino acid mixtures, stereoselectively mobilized Ca(2+) ions in normal human parathyroid cells in the presence but not the absence of the CaR agonists, extracellular Ca(2+) (Ca(2+)(o)), or spermine. The order of potency was l-Trp = l-Phe > l-His > l-Ala > l-Glu > l-Arg = l-Leu. CaR-active amino acids also acutely and reversibly suppressed PTH secretion at physiological ionized Ca(2+) concentrations. At a Ca(2+)(o) of 1.1 mm and an amino acid concentration of 1 mm, CaR-active amino acids (l-Phe = l-Trp > l-His = l-Ala), but not CaR-inactive amino acids (l-Leu and l-Arg), stereoselectively suppressed PTH secretion by up to 40%, similar to the effect of raising Ca(2+)(o) to 1.2 mm. A physiologically relevant increase in the -fold concentration of the plasma-like amino acid mixture (from 1x to 2x) also reversibly suppressed PTH secretion in the Ca(2+)(o) concentration range 1.05-1.25 mm. In conclusion, l-amino acids acutely and reversibly activate endogenous CaRs and suppress PTH secretion at physiological concentrations. The results indicate that l-amino acids are physiological regulators of PTH secretion and thus whole body calcium metabolism.

Extracellular Ca 2ϩ ions are recognized as the principal physiological regulators of parathyroid hormone secretion acting to close a classical endocrine feedback loop, whereby PTH 1 elevates Ca 2ϩ o , and elevated Ca 2ϩ o , in turn, suppresses PTH secretion. A major insight into the mechanism by which Ca 2ϩ o interacts with parathyroid cells was gained with the cloning of the extracellular Ca 2ϩ -sensing receptor (CaR) (1). This receptor, which belongs to subgroup C of the G-protein-coupled receptor superfamily, mediates the Ca 2ϩ o -stimulated activation of intracellular signaling pathways in parathyroid cells leading to the activation of various enzymes including phosphatidyl inositol-specific phospholipase C and attendant intracellular Ca 2ϩ mobilization (for reviews, see Refs. 2 and 3). Targeted deletion of the CaR in the CaR null mouse eliminates feedback control of PTH secretion and results in a severe form of neonatal hyperparathyroidism (4). A similar condition in humans, neonatal severe hyperparathyroidism, has been shown to arise in homozygotes or compound heterozygotes with inactivating mutations of the CaR (5) (for review see Ref. 3).
Molecular analysis of the CaR indicates that it is composed of several functional domains that are conserved across related members of subgroup C. The N-terminal Venus Fly Trap domains of many of these receptors recognize amino acids (especially glutamate) or the amino acid analog, ␥-aminobutyric acid. More recent work indicates that several cloned members of the family are also broad-spectrum amino acid sensors. In mammals, these include the CaR, which is allosterically activated by aromatic, aliphatic, and polar amino acids as well as plasma-like amino acid mixtures (6) and a heterodimeric amino acid taste receptor, which has broad selectivity for aliphatic, polar, and charged amino acids but not aromatic amino acids (7). In human embryonic kidney (HEK)293 cells that stably express the human CaR, L-amino acids markedly enhanced the sensitivity of the receptor to Ca 2ϩ and other cationic agonists including spermine and Gd 3ϩ (6).
Because the CaR mediates the acute control of PTH secretion, the finding that it is allosterically activated by L-amino acids raises the possibility that PTH secretion is acutely regulated not only by adjustments in extracellular Ca 2ϩ o but also by physiological changes in amino acid concentration. We have tested and confirmed this hypothesis in the current study on normal human parathyroid cells. The data indicate that Lamino acids and plasma-like mixtures of L-amino acids allosterically activate endogenous parathyroid CaRs contributing to the control of intracellular signaling pathways and PTH secretion. In the presence of physiological concentrations of extracellular Ca 2ϩ , L-amino acids (also at physiological concentrations) stereoselectively activated Ca 2ϩ mobilization and inhibited PTH secretion. The data support the view that fluctuations in serum amino acid levels acting via the CaR acutely regulate PTH secretion and thus whole body calcium metabolism.

Origin of Tissue and Preparation of Human Parathyroid Cells-
Samples of normal human parathyroid transplants were obtained at neck surgery at the Royal North Shore Hospital, St. Leonards, New South Wales, Australia and the Mater Private Hospital, North Sydney, New South Wales, Australia under guidelines established by the relevant institutional ethics committees. The parathyroid tissue, typically taking the form of a single 1-1.5-mm diameter "chip", was transported to the laboratory in ice-cold Hanks' balanced salt solution (Invitrogen) containing CaCl 2 1.25 mM. It was either used immediately or stored overnight at 4°C in MEM (number 11380-037, Invitrogen). The tissue was then transferred into 10 ml of MEM that contained 1 mg/ml collagenase (type I, Worthington, Scimar, Victoria) and 0.2 mg/ml DNase I (type IV, Sigma). The composition of the MEM formulation was as follows: NaCl 137 mM, KCl 5.4 mM, NaHCO 3 4.2 mM, CaCl 2 1.2 mM; MgSO 4 0.81 mM; Na 2 HPO 4 0.34 mM, KH 2 PO 4 0.44 mM, D-glucose 1 g/liter, Phenol Red 10 mg/liter, MEM amino acids, and vitamins. Collagenase was stored at Ϫ80°C and brought out for weighing on ice or dry ice. The suspension was briefly oxygenated and then incubated at 37°C. After 20 min, the enzyme solution was decanted and the parathyroid tissue was transferred into 5 ml of MEM that contained 1 mg/ml of bovine serum albumin (protease-free, Sigma A-3059). The tissue was then triturated by repeated passage (10 -15 times) through the tip of a disposable 5-ml syringe (no needle attached). The cloudy suspension containing clumps of parathyroid cells was passed through a 200-m pore-size nylon filter and then sedimented (Sorvall H-1000B rotor, 100 ϫ g, 2 min). The cell pellet was gently resuspended and washed twice in 5 ml of bovine serum albumin-containing MEM (Invitrogen, 11575-032). It was finally resuspended in bovine serum albumin-containing MEM. The remaining pieces of undigested parathyroid tissue were returned to medium containing collagenase and DNase I and incubated for a further 20 min at 37°C. The enzyme solution was then decanted, and the parathyroid tissue subjected to trituration and centrifugal isolation as above. In general, the second cell harvest was used for analyses of cell function because of greater cell yields; however, the first and second cell harvests were also frequently combined.
Amino Acid Solutions-Stock amino acid-containing solutions were routinely made up in physiological saline at 100 or 200 mM with the exception of tryptophan (50 mM because of poor solubility). The amino acid composition of the 1ϫ basal amino acid solution that emulated fasting human amino acids (all L-isomers in physiological saline solution, pH 7.4) was as follows (in M): 50 Phe, 50 Trp, 80 His, 60 Tyr, 300  Ala, 200 Thr, 30 Cys, 50 Asn, 600 Gln, 125 Ser, 30 Glu, 250 Gly, 180 Pro,  250 Val, 30 Met, 10 Asp, 200 Lys, 100 Arg, 75 Ile, and 150 Leu. Stocks of this basal amino acid solution (40-fold concentrated in amino acids) were made up in physiological saline solution and stored at Ϫ20°C. They were thawed and diluted in amino acid-free physiological saline as required. The control physiological saline used in all microfluorimetry experiments had the following composition: 125 mM NaCl, 4.0 mM KCl, 0.2 or 0.5 mM CaCl 2 , 0.5 mM MgCl 2 , 20 mM HEPES (NaOH), 0.1% D-glucose, pH 7.4.
Microfluorimetry for Determining Cytoplasmic Ca 2ϩ Concentration-Parathyroid cells were loaded with fura-2/AM (1 M, 20 min, 37°C) in physiological saline solution containing bovine serum albumin 1 mg/ml. The cells were then sedimented (Sorvall H-1000B rotor, 100 ϫ g, 2 min) and resuspended in albumin-free physiological saline solution. Fura-2/ AM-loaded cells were transferred into a superfusion chamber and placed in the light path of a Nikon Diaphot microscope as described previously (8). Excitation was performed at alternating wavelengths (340 and 380 nm). The fluorescent light (F, peak 510 nm) was detected by a photomultiplier, and its digitized recording was achieved using Acknowledge software for Macintosh. The control superfusion solution had the following composition: 125 mM NaCl, 4.0 mM KCl, 0.2 or 0.5 mM CaCl 2 , 0.5 mM MgCl 2 , 20 mM HEPES (NaOH), 0.1% D-glucose, pH 7.4. Data for cytoplasmic free Ca 2ϩ concentrations were expressed either as uncorrected excitation ratios (F340/F380) or converted to ionized Ca 2ϩ concentration using a calibration procedure (8).
Determination of PTH Secretion-Perifusion of normal human parathyroid cells was undertaken in low molecular mass (4 -5 kDa) cut-off gel filtration media so that intact PTH (ϳ9 kDa) would appear in the void volume. Gel filtration media were pre-equilibrated with physiological saline that contained 1ϫ basal amino acid mixture and 1 mg/ml bovine albumin (protease-free, Sigma A-3905). 20,000 -50,000 cells were loaded onto the surface of a 1-ml bed volume of Bio-Gel P-4 (nominal exclusion limit 4 kDa) and then gently covered with a 1-ml bed volume of Sephadex G-25 (medium, nominal exclusion limit 5 kDa) in a small perifusion column. After tubing connections were established downstream to a peristaltic pump and upstream to a reservoir, the column was suspended in a water bath (37°C) and perifused at 1.5 ml/min with 37°C equilibrated control physiological saline, 125 mM NaCl, 4.0 mM KCl, 1.25 mM CaCl 2 , 1.0 mM MgCl 2 , 0.8 mM NaH 2 PO 4 , 20 mM HEPES (NaOH, pH 7.4), 0.1% D-glucose, 1 mg/ml bovine albumin, 1ϫ basal amino acid mixture. Prior to starting all experiments, the columns were perifused for a 20-min control period. Routinely, 2-min (i.e. 3-ml) samples were collected into tubes immersed in an ice bath, and the tubes were transferred to dry ice upon completion of each collection period. As required, solutions were changed to permit variations in extracellular calcium or amino acid concentration. All samples were stored at Ϫ80°C until the analysis of intact human PTH using an Immulite 2000 autoanalyzer (Diagnostic Products Corporation, CA).
Statistical Analysis and Curve Fitting-The Ca 2ϩ mobilization data were expressed as concentration-response curves and fitted to the following form of the Hill equation, R ϭ b ϩ (a Ϫ b)* C n /(e n ϩ C n ), where R ϭ fractional response, a ϭ maximum fractional response, b ϭ basal fractional response, C ϭ extracellular Ca 2ϩ concentration (in mM), e ϭ EC 50 (the concentration of Ca 2ϩ that induced a half-maximal response), and n ϭ Hill coefficient.
The PTH secretion data were fitted to the following equation, o concentration that yielded half-maximal inhibition of secretion), and n ϭ Hill coefficient. The first of three or more data points, corresponding to a short equilibration period, was routinely ignored when calculating the mean secretion rate for any given treatment.
Estimates of the curve-fitting parameters and their standard errors were obtained using MacCurveFit 1.5 for Macintosh. Differences between the parameter values obtained from concentration-response curves were tested for statistical significance using the F-test as described in Ref. 9. Differences in PTH secretion rates were tested for statistical significance using the paired or unpaired t tests. Other data are routinely expressed as means Ϯ S.E. (number of experiments).  (Fig. 1B), or L-Arg (not shown). CaR-active amino acids exhibited similar effects in the presence of the polycationic CaR agonist, spermine (not shown). The effects of amino acids on intracellular Ca 2ϩ mobilization were stereoselective i.e. L-amino acids were much more effective than D-amino acids (Fig. 1). In addition, the effects of L-amino acids were concentration-dependent (Table I, Fig. 2). At a Ca 2ϩ o concentration of 2.0 mM, the apparent EC 50 s for L-Trp, L-Phe, L-His, and L-Ala were 0.04 Ϯ 0.01 mM, 0.06 Ϯ 0.01 mM, 0.1 Ϯ 0.01 mM, and 0.41 Ϯ 0.02 mM, respectively, and the order of potency was L-Trp ϭ L-Phe Ͼ L-His Ͼ L-Ala Ͼ L-Glu Ͼ L-Leu ϭ L-Arg (Table I). Individual L-amino acids were also effective at lower extracellular Ca 2ϩ concentrations (1.5 and 1.0 mM); however, the EC 50 values increased progressively, and the amino acid-induced maximum responses fell progressively as the Ca 2ϩ concentration fell (Fig. 2, C and D, Table I). An appreciation of the physiological significance of these effects can be gained by dividing the fasting plasma concentration of an amino acid by its EC 50 value (conc/EC 50 ). At 2.0 mM Ca 2ϩ o , four of the seven amino acids tested, L-Trp, L-Phe, L-His, and L-Ala all exhibited conc/EC 50 ratios close to 1.0. At 1.5 and 1.0 mM Ca 2ϩ o , the apparent effectiveness of these individual amino acids fell progressively (Table I). Between 1.0 and 1.5 mM Ca 2ϩ o , encompassing the normal physiological range, the data indicate that physiologically significant modulation of the CaR could only arise if multiple CaR-active amino acids acted together in concert.

Effects of Individual Amino Acids on Cytoplasmic Free Ca 2ϩ Concentration and Sensitivity to Ca
Active L-amino acids also left-shifted the concentration response curves for extracellular Ca 2ϩ . In the absence of amino acids, the apparent EC 50 for Ca 2ϩ was 2.6 Ϯ 0.2 mM (n ϭ 4); in the presence of 3 mM L-Phe, the concentration response was markedly left-shifted, and the apparent EC 50 for Ca 2ϩ was 1.5 Ϯ 0.2 mM (Fig. 3, n ϭ 4). F-test analysis confirmed that the difference was statistically significant (F [1,28] ϭ 30.5, p Ͻ 0.001). In addition, L-Phe induced an increase in the maximum response as determined by curve-fitting the cumulative fluorescence ratio data. In the absence of L-Phe, the maximum response was 1.08 Ϯ 0.05, in the presence of 3 mM L-Phe, the maximum response was 1.33 Ϯ 0.08 (F [1,28] ϭ 18.7, p Ͻ 0.01). However, L-Phe had no effect on the basal fluorescence ratio at an extracellular Ca 2ϩ of 0.2 mM (F [1,28] ϭ 1.95, p Ͼ 0.1). Similar findings were obtained using another CaR-active amino acid, L-Ala, and L-Phe also left-shifted the concentration response curve for spermine (not shown).
Effects of Amino Acid Mixtures on Cytoplasmic Free Ca 2ϩ Concentration and Sensitivity to Ca 2ϩ o -To further assess the physiological significance of the above results, the effects of plasma-like amino acid mixtures were also examined. Plasmalike amino acid mixtures activated intracellular Ca 2ϩ mobilization in a -fold concentration-dependent fashion (Fig. 4A) and markedly left-shifted the Ca 2ϩ o sensitivity of the fura-2/AMloaded normal human parathyroid cells (Fig. 4, B and C). In the absence of amino acids, the EC 50 for Ca 2ϩ was 2.6 Ϯ 0.2 mM (n ϭ 8). In the presence of a fasting plasma-like L-amino acid mixture (total amino acid concentration, 2.8 mM), the EC 50 for Ca 2ϩ was 1.6 Ϯ 0.1 mM (n ϭ 4, F [1,28] ϭ 78, p Ͻ 0.001). As the -fold concentration was increased from 0.5 to 2ϫ, encompassing the range normally described between protein restriction and high dietary protein intake in humans, respectively, a small apparent decrease in the EC 50 for Ca 2ϩ o and an increase in the maximum response were observed. Under these conditions, the EC 50 for Ca 2ϩ o dropped from 1.7 Ϯ 0.2 mM (n ϭ 4) to 1.6 Ϯ 0.1 mM (n ϭ 4), and the maximum response, in the form of a cumulative fluorescence ratio measured from baseline, increased from 0.62 Ϯ 0.04 (n ϭ 4) to 0.69 Ϯ 0.04 (n ϭ 4, Table II). To further evaluate whether receptor response is sensitive to physiologically relevant variations in amino acid concentration, the data were replotted as a function of -fold concentration (Fig. 4D). A threshold for the amino acid-dependent responses was observed at a Ca 2ϩ o of ϳ0.5 mM. At both 1.0 and 1.5 mM Ca 2ϩ o , i.e. either side of the normal range (1.1-1.3 mM), the receptor response increased as the -fold concentration rose above 0.5ϫ (Fig. 4D).
Effects of Individual Amino Acids and Plasma-like Amino Acid Mixtures on PTH Secretion-In control experiments, normal human parathyroid cells were perifused with a physiological saline solution that contained a fasting plasma-like Lamino acid mixture (total concentration 2.8 mM) and 1 mg/ml bovine serum albumin. Under these conditions, the apparent IC 50 for the effect of extracellular Ca 2ϩ on the secretion of intact PTH was 1.12 Ϯ 0.02 mM (n ϭ 7). In addition, an increase in extracellular Ca 2ϩ concentration from 0.9 to 1.4 mM induced a maximal suppression of intact PTH secretion of 84 Ϯ 7% (n ϭ 7). The addition of L-Phe (3 mM) or L-Trp (3 mM) acutely and reversibly suppressed PTH secretion rates at Ca 2ϩ concentrations between 0.8 and 1.5 mM (Fig. 5A, L-Phe) and 0.9 and 1.4 mM (Fig. 5B, L-Trp). In the presence of 3 mM L-Phe, the IC 50 value for extracellular Ca 2ϩ concentration fell from 1.13 Ϯ 0.02 to 1.03 Ϯ 0.04 mM. A similar left-shift in the concentrationresponse curve was observed for L-Trp (Fig. 5, B and C, n ϭ 3). Upon return to the control physiological saline solution containing the 1ϫ amino acid mixture, the IC 50 values returned to control levels (Fig. 5, A and B). In addition, 3 mM L-Phe or L-Trp induced small drops in intact PTH secretion rates at high Ca 2ϩ concentrations (Fig. 5C). For example, the PTH secretion rate (in fg min Ϫ1 cell Ϫ1 , n ϭ 4) at the high Ca 2ϩ concentration plateau fell from 0.40 Ϯ 0.08 to 0.30 Ϯ 0.12 in the presence of 3 mM L-Phe and recovered to 0.51 Ϯ 0.08 upon return to the control solution (F [1,32] ϭ 7.2, p ϭ 0.01). However, the CaRactive amino acids appeared to have no effect on the low Ca 2ϩ  concentration plateau. Prior to L-Phe, for example, the low Ca 2ϩ concentration plateau secretory rate (in fg min Ϫ1 cell Ϫ1 , n ϭ 4) was 1.95 Ϯ 0.08, in the presence of 3 mM L-Phe, it was 1.68 Ϯ 0.22, and upon return to the control solution, it was 1.65 Ϯ 0.08. Statistical testing of the data in the L-Phe series using the paired t test confirmed that the observed differences were significant at all Ca 2ϩ o concentrations between 0.9 and 1.5 mM (for 0.9 mM, p ϭ 0.002; for 1.0 mM, p ϭ 0.0001; for 1.1 mM, p ϭ 0.0005; for 1.2 mM, p ϭ 0.002; for 1.3 mM, p ϭ 0.005; for 1.4 mM, p ϭ 0.0001; and for 1.5 mM, p ϭ 0.03). Similar results were obtained upon statistical analysis of the data in the L-Trp series.
The effects of various individual amino acids were evaluated in cells exposed to a physiologically significant drop in extracellular Ca 2ϩ concentration from 1.2 to 1.1 mM in the presence of the 1x amino acid mixture (Fig. 6, A and B, Table III and L-Ala. These effects were stereoselective. Furthermore, the CaR-inactive amino acids L-Arg and L-Leu had no effect (Fig. 6, A and B, Table III).
An analysis of the effects of raising the -fold concentration of the amino acid mixture from 1 to 2ϫ was performed to evaluate the impact of a physiologically relevant variation in total serum amino acid level (Fig. 6C). The 2-fold concentration of the amino acid mixture reversibly suppressed PTH at all three Ca 2ϩ o concentrations tested. With respect to Ca 2ϩ o concentration, the inhibitory effects of the 2-fold amino acid mixture were as follows, 25 Ϯ 4% at 1.05 mM, 36 Ϯ 8% at 1.15 mM, and 34 Ϯ 5% at 1.25 mM (n ϭ 5). Paired t test analyses confirmed that the effects of basal amino acid 2ϫ were statistically significant at all three Ca 2ϩ o concentrations (p ϭ 0.01 for 1.05 mM, p ϭ 0.02 for 1.15 mM, and p ϭ 0.002 for 1.25 mM). These results indicate that the PTH secretion rate is highly sensitive to physiologically relevant fluctuations in serum amino acid concentration and that changes in total amino acid concentration inhibit PTH secretion at all levels of Ca 2ϩ o concentration within the normal range. DISCUSSION The data described in this study support the conclusion that physiologically relevant increases in the concentrations of Lamino acids acutely activate the extracellular Ca 2ϩ -sensing receptor and reversibly inhibit PTH secretion from normal human parathyroid cells. These findings indicate that PTH secretion is under the acute physiological control of serum L-amino acid levels and provide a theoretical basis for a link between protein and calcium metabolism.
L-Amino acids, including L-Phe, L-Trp, and L-Ala but not L-Arg, L-Leu, and L-Ile stereoselectively activated intracellular Ca 2ϩ mobilization from fura-2/AM-loaded normal human parathyroid cells in the presence of the CaR agonist Ca 2ϩ o (Figs. 1 and 2, Table I). In addition, CaR-active amino acids markedly enhanced the sensitivity of the receptor to Ca 2ϩ o (Figs. 3 and 4, Table II). The amino acid selectivity of these responses, L-Trp ϭ L-Phe Ͼ L-His Ͼ L-Ala ϭ L-Glu Ͼ L-Leu ϭ L-Arg, closely resem- L-Amino Acid Sensing by Parathyroid Ca 2ϩ -Sensing Receptors bles that described previously for the cloned human CaR stably expressed in HEK293 cells (6,10). In addition, an L-amino acid mixture that emulated the composition of fasting human plasma also activated intracellular Ca 2ϩ mobilization and enhanced the sensitivity of the receptor to extracellular Ca 2ϩ (Fig. 4). A threshold extracellular Ca 2ϩ concentration for the effects of plasma-like L-amino acid mixtures was detected around 0.5 mM (Fig. 4). The Ca 2ϩ mobilization response to a change in -fold concentration of the plasma-like L-amino acid mixture from 0.5 to 2ϫ (the range normally described between protein restriction and high protein intake) took the form of a small decrease in EC 50 (from 1.7 to 1. CaR-active amino acids including L-Phe, L-Trp, L-His, and L-Ala stereoselectively suppressed PTH secretion from normal human parathyroid cells that were perifused with physiological saline solution that contained a 1-fold mixture of L-amino acids that emulated normal fasting human plasma (11)(12)(13). CaRinactive amino acids including L-Arg and L-Leu had no inhibitory effect (Fig. 6, A and B, Table III Table III). In C the effect of an elevation in the -fold concentration of the amino acid mixture from 1 to 2ϫ and the reversibility of this effect is shown from data obtained in five experiments in which cells were exposed to repetitive stepwise changes in ionized Ca 2ϩ o concentrations 1.25, 1.15, 1.05, and 1.25 mM. Basal amino acid 1ϫ (baa) corresponds to the amino acid composition of fasting plasma.

L-Amino Acid Sensing by Parathyroid Ca 2ϩ -Sensing Receptors 38157
-fold concentration of the plasma-like amino acid mixture from 1 to 2ϫ (Fig. 6) were observed in the physiological Ca 2ϩ o concentration range i.e. from 1.1 to 1.3 mM. The data indicate that amino acids are also effective in the wider pathophysiological range from ϳ0.8 to 1.5 mM and above.
In the current work the observed effects of amino acids are not simply pharmacological, because plasma-like amino acid mixtures, as well as individual amino acids, are effective CaR activators (Figs. 4 and 6, Table II). The effects are also not simply "permissive," because the response of the receptor was clearly submaximal in the presence of the 1ϫ mixture (Figs. 4 and 6, Table II). The conclusion that physiologically relevant fluctuations in the plasma levels of amino acids modulate the Ca 2ϩ o sensitivity of CaR and parathyroid hormone secretion relies on the observed impact of variations in amino acid concentration in the context of a normal fasting amino acid mixture. Most important was the observed impact of an increase in the -fold concentration of the plasma-like amino acid mixture from 1 to 2ϫ, which reversibly suppressed PTH secretion by 25-40% at all ionized Ca 2ϩ o concentrations tested in the range 1.05-1.25 mM (Fig. 6C). A -fold concentration change of this magnitude approximates what is observed in response to a protein-rich meal (11,(13)(14)(15) or between the 24-h trough and peak levels of the normal circadian rhythm (16,17). In addition, the observed inhibitory effects of individual L-amino acids on PTH secretion in the presence of the 1ϫ mixture (Figs. 5 and 6, Table III) indicate that there is a "reserve" amino acidinducible sensitivity even under normal physiological conditions.
Links between protein and calcium metabolism have been previously identified including evidence that elevated dietary protein intake promotes urinary calcium excretion (18,19) and reduced protein intake induces secondary hyperparathyroidism in the absence of changes in plasma-ionized Ca 2ϩ concentration (20). The finding that the CaR in normal human parathyroid cells is modulated by physiologically relevant variations in amino acid levels provides a possible explanation for secondary hyperparathyroidism in human subjects on moderately low protein diets (20). In this respect, the impact of fluctuations in the concentrations of CaR-active amino acids on PTH secretion in vivo requires analysis. Similar modulation of CaRs by amino acids in renal thick ascending limb cells, which are a major site of CaR-regulated calcium transport (for review see Ref. 2), may contribute to the mechanism by which elevated protein intake provokes hypercalciuria.
Although the impact of -fold concentrations of plasma-like amino acid mixtures on PTH secretion has obvious physiological implications, the selectivity of the CaR for aromatic and aliphatic amino acids seems puzzling. It might be wondered whether situations arise physiologically whereby selective elevations or reductions in the serum levels of these amino acids are associated with, respectively, suppressed or elevated serum levels of PTH. The established patterns of the circadian rhythms for extracellular Ca 2ϩ , PTH, and L-amino acids may provide an answer. Normal human subjects exhibit a highly reproducible nocturnal peak in the serum PTH level that occurs independent of changes in plasma ionized Ca 2ϩ concentration (21,22). The origin of this peak, which is reported to be defective in patients with osteoporosis (23) and is not driven by changes in plasma-ionized Ca 2ϩ concentration (24,25), is currently unclear. However, it coincides with a 30% drop in the serum levels of aromatic amino acids (16,26). The data described in the current study indicate that this association may be causative i.e. the nocturnal drop in the serum levels of aromatic and other amino acids may drive the nocturnal peak inPTH.Inturn,thenocturnalPTHpeakmaydrivetheosteoblastdependent synthesis of bone in a manner comparable with that described for daily subcutaneous injections of PTH in patients with osteoporosis (27).
The finding that normal parathyroid cells are regulated by physiologically relevant variations in amino acid levels indicates that fluctuations in amino acid levels may modulate CaR-dependent cell function in other settings. With respect to whole body calcium metabolism, this may include CaR-expressing cells of the renal tubules as discussed above (28). Amino acids may also regulate CaR-expressing cells in the gastrointestinal tract such as gastrin-secreting G cells (29) and acidsecreting gastric parietal cells (30). Exposure of these cells to elevations in amino acid levels following the ingestion, digestion, and absorption of a protein-rich meal results in enhanced gastrin release and gastric acid production, respectively (see Ref. 31).
In conclusion, the evaluation of the Ca 2ϩ o -sensing receptor in normal human parathyroid cells, confirms its sensitivity to L-amino acids (6,10). CaR-active L-amino acids acutely and reversibly inhibit PTH secretion in physiologically relevant concentration ranges for ionized Ca 2ϩ o and amino acids. Amino acids, like Ca 2ϩ ions, should be viewed as physiological regulators of parathyroid hormone secretion.

TABLE III
Effects of amino acids on intact PTH secretion from normal human parathyroid cells The data are means Ϯ S.E. and were derived using the second and third data points from the sets of three data points in experiments such as those shown in Fig. 6, A and B. Normal human parathyroid cells were perifused with physiological saline solutions that contained 1.1 mM Ca 2ϩ , a 1-fold mixture of L-amino acids, and 1 mg/ml bovine albumin in the presence or absence of added L-or D-amino acids (all 1 mM final concentration). The data are expressed as a percentage of the mean control response at 1.1 mM Ca 2ϩ o . The control data for 1.1 mM Ca 2ϩ (Nil) were derived by expressing the first control as a percentage of the second control in all experiments. Expressed as a percentage of the control PTH secretion rate at 1.1 mM Ca 2ϩ , the PTH secretion rate at 1.2 mM Ca 2ϩ was 52.5 Ϯ 3.8 % at the start and 54.4 Ϯ 3.8 % at the end of these experiments. Statistical analysis was performed using the unpaired t test.