Pathophysiologic Changes in Extracellular pH Modulate Parathyroid Calcium-Sensing Receptor Activity and Secretion via a Histidine-Independent Mechanism

Abstract

The calcium-sensing receptor (CaR) modulates renal calcium reabsorption and parathyroid hormone (PTH) secretion and is involved in the etiology of secondary hyperparathyroidism in CKD. Supraphysiologic changes in extracellular pH (pH_o) modulate CaR responsiveness in HEK-293 (CaR-HEK) cells. Therefore, because acidosis and alkalosis are associated with altered PTH secretion in vivo, we examined whether pathophysiologic changes in pH_o can significantly alter CaR responsiveness in both heterologous and endogenous expression systems and whether this affects PTH secretion. In both CaR-HEK and isolated bovine parathyroid cells, decreasing pH_o from 7.4 to 7.2 rapidly inhibited CaR-induced intracellular calcium (Ca²⁺_i) mobilization, whereas raising pH_o to 7.6 potentiated responsiveness to extracellular calcium (Ca²⁺_o). Similar pH_o effects were observed for Ca²⁺_o-induced extracellular signal-regulated kinase phosphorylation and actin polymerization and for L-Phe-induced Ca²⁺_i mobilization. Intracellular pH was unaffected by acute 0.4-unit pH_o changes, and the presence of physiologic albumin concentrations failed to attenuate the pH_o-mediated effects. None of the individual point mutations created at histidine or cysteine residues in the extracellular domain of CaR attenuated pH_o sensitivity. Finally, pathophysiologic pH_o elevation reversibly suppressed PTH secretion from perifused human parathyroid cells, and acidosis transiently increased PTH secretion. Therefore, pathophysiologic pH_o changes can modulate CaR responsiveness in HEK-293 and parathyroid cells independently of extracellular histidine residues. Specifically, pathophysiologic acidification inhibits CaR activity, thus permitting PTH secretion, whereas alkalinization potentiates CaR activity to suppress PTH secretion. These findings suggest that acid-base disturbances may affect the CaR-mediated control of parathyroid function and calcium metabolism in vivo.

Mammalian calcium homeostasis is maintained by the regulated secretion of parathyroid hormone (PTH) and renal calcium reabsorption under the control of the calcium-sensing receptor (CaR).^1,2 Hypercalcemia stimulates the CaR to suppress PTH secretion, while chronic underactivation of CaR can contribute to pathologically elevated PTH secretion, such as in secondary hyperparathyroidism of CKD. In a heterologous human embryonic kidney 293 (HEK-293) cell-based expression system, Quinn et al.³ found that the Ca²⁺_o potency of the human CaR is sensitive to supraphysiologic changes in extracellular pH (pH_o). By varying the pH of the experimental buffer in 0.5-unit steps from 5.5 to 9, the sensitivity of CaR to Ca²⁺_o (and Mg²⁺_o) was altered; external acidification decreased CaR sensitivity and alkalinization increased CaR sensitivity.³ However, blood pH levels (7.35–7.45 approximately 40 nM H⁺ concentration) rarely vary beyond ±0.4 pH units in vivo, even under extreme pathologic conditions. Even smaller changes (±0.2) are more generally seen, such as with the acidosis of CKD; however, even this still represents an approximately 58% increase in H⁺ concentration. Secondary hyperparathyroidism is another complication of CKD. In animal experiments, induction of metabolic acidosis results in increased serum PTH levels and hypercalcemia, whereas induction of alkalosis suppresses serum PTH levels.^4–8

Therefore, because acidosis has been associated with increased serum PTH levels in vivo, we examined whether much smaller, pathophysiologically relevant changes in pH_o (±0.2) can alter the Ca²⁺_o sensitivity of CaR-mediated signaling pathways in HEK-293 cells and whether such effects can be replicated in parathyroid cells. We also examined whether a specific histidine or free cysteine residue could account for CaR pH_o sensitivity and, finally, considered the effect of such pathophysiologic changes in pH_o on PTH secretion.

Results

Fura 2–loaded CaR-HEK cells were stimulated with 2.5mM Ca²⁺_o (pH, 7.4) to elicit CaR-induced Ca²⁺_i mobilization and were then switched to the same buffer at pH 7.2 or 7.6 before being returned to pH 7.4. At pH 7.6, CaR responsiveness (i.e., Δ350/380 fura ratio area under the curve) was significantly increased (Figure 1A), an effect that was immediately reversible upon return to pH 7.4. In contrast, lowering pH_o to 7.2 significantly inhibited CaR responsiveness (Figure 1B), the effect being immediately reversible. The effect of ±0.2-unit pH changes on Ca²⁺_i mobilization was then tested over a range of Ca²⁺_o concentrations, from 0.5 to 10 mM. Again, CaR responsiveness was significantly inhibited with pH 7.2 (half maximal effective concentration for Ca²⁺_o, 3.6±0.2 mM in 7.2 versus 3.1±0.1 in 7.4; P<0.05) and significantly stimulated with pH 7.6 (2.8±0.0 mM in 7.6; P<0.05). In addition, similar effects of changing pH_o (±0.2 unit) were also seen using bicarbonate/CO₂ buffers (Figure 1D). See Supplemental Material for the complete methods.

Figure 1.

CaR-induced Ca²⁺_i mobilization is sensitive to pathophysiologic pH_o changes in CaR-HEK cells. (A) Representative trace showing Ca²⁺_i changes (Fura-2 ratio) in a single cell (upper trace, “cell”) and “global” (lower trace) cluster of (>10) cells in response to elevated [Ca²⁺]_o (2.5 versus 0.5 mM control) when pH_o was changed from 7.4 to 7.6. Changes in [Ca²⁺]_i shown as percentage control of the area under the curve. n=4 coverslips. (B) Identical except testing the effect of decreased pH_o (7.2). n=4 coverslips. *P<0.05 and ***P<0.001 versus first pH 7.4 treatment; ⁺⁺⁺P<0.001 versus pH 7.6 treatment by repeated-measures ANOVA (Tukey post-test) performed on the raw data. (C) Cells were exposed to buffers containing increasing concentrations of Ca²⁺_o (0.5–10 mM) with pH_o 7.2, 7.4, or 7.6 with the resulting concentration-effect curves for Ca²⁺_i mobilization shown (left) together with their half maximal effective concentration (EC₅₀) (right). While the EC₅₀ values were significantly different, the maximal responses were not significantly different, despite the apparent trend. *P<0.05 versus pH 7.4 by repeated-measures ANOVA (Dunnett) performed on log EC₅₀ values from four independent experiments (two to five coverslip replicates per data point, >15 cells per coverslip). (D) Similar results were obtained with use of bicarbonate/CO₂ buffers, with pH 7.6 increasing Ca²⁺_i mobilization and pH 7.2 decreasing mobilization in a single cell (upper trace) and “global” cluster of (>10) cells (lower trace) in response to 2.5 mM Ca²⁺_o as before. Relative changes in [Ca²⁺]_i are shown in the bar graph. The appropriate pH values were obtained by varying the bicarbonate (and NaCl) concentrations in buffers gassed continuously with 5% CO₂/95% O₂ (g). For the complete methods, see the Supplemental Material. *P<0.05 versus pH 7.4 treatment by one-tail paired t test; n=7 coverslips.

To determine whether this apparent pH_o sensitivity of CaR is specific to intracellular Ca²⁺_i mobilization or applies to other effector pathways, two other readouts of CaR activity were used: extracellular signal-regulated kinase (ERK) phosphorylation and actin polymerization.^9,10 Indeed, mild acidosis (pH_o 7.2) significantly inhibited Ca²⁺_o-induced ERK phosphorylation in CaR-HEK cells (3.5 mM Ca²⁺_o) (Figure 2A), while mild alkalosis (pH_o 7.6) increased ERK activation. Similarly, a 0.2-unit reduction in pH_o inhibited 1.5 mM Ca²⁺_o-induced actin polymerization while a 0.2-unit increase in pH_o potentiated 1.8 mM Ca²⁺_o-induced actin polymerization (Figure 2B). Together these results suggest that the effect of altering pH_o affects the Ca²⁺_o sensitivity of the CaR rather than the mechanism of Ca²⁺_i mobilization. Indeed, in the presence of 0.5 mM Ca²⁺_o (which renders the CaR inactive), variations in pH_o between 7.2 and 7.6 had no effect on actin polymerization. This finding indicates that changing pH_o does not otherwise affect baseline signaling.

Figure 2.

Pathophysiologic pH_o changes also affects CaR-induced ERK activation, actin polymerization, and L-Phe responsiveness in CaR-HEK cells. (A) Representative immunoblot showing ERK phosphorylation (pERK) after treatment with 3.5 mM Ca²⁺_o for 10 minutes under different pH_o conditions (7.4±0.2). Mean changes in ERK phosphorylation were determined by densitometry and expressed on a bar graph in arbitrary units. n=9 dishes from two independent experiments. *P<0.05 and **P<0.01 versus pH_o 7.4 by one-way ANOVA (Dunnet). (B) Representative immunofluoresence images (i) showing phalloidin-TRITC staining following CaR-HEK treatment with 1.5 mM Ca²⁺_o for 3 hours in buffers of pH_o ranging from 7.2 to 7.6. Quantification presented on a bar graph as pixels (phalloidin-TRITC) per unit area in five randomly selected regions of the cells is shown below (iii). Also shown are bar graphs from separate experiments where the cells were exposed to 0.5 (ii) or 1.8 (iv) mM Ca²⁺_o. n≥5 coverslips. *P<0.05 and **P<0.01 versus pH_o 7.4 by one-way ANOVA (Dunnet). (C) Representative Ca²⁺_i trace (i) showing CaR-HEK cells stimulated first with 10 mM L-Phe (in 1.8 mM Ca²⁺_o) and then with 3 mM Ca²⁺_o under various pH_o conditions as shown (0.5 Ca²⁺_o, baseline, pH 7.4 unless otherwise stated). The traces include a typical response in a single cell (gray) as well as the total response in the cell cluster (>10 cells; “global”, black). Quantification (ii) is shown as a bar graph (percentage change in L-Phe responsiveness relative to pH_o 7.4 control). n=7 coverslips from two independent experiments. **P<0.01 by repeated-measures ANOVA (Bonferroni).

Next, we examined the effect of 0.2-unit pH_o changes on the response to the CaR-positive allosteric modulator L-Phe (10 mM) to determine whether the pH_o sensitivity is exclusive to orthosteric agonism with Ca²⁺_o. We observed significant potentiation of L-Phe-induced Ca²⁺_i mobilization in pH_o 7.6 and attenuation in pH_o 7.2 (Figure 2C), similar to that observed for Ca²⁺_o.

We then assessed whether pathophysiologic changes in pH_o also affect CaR-induced Ca²⁺_i mobilization in parathyroid cells, in which the CaR is expressed endogenously. Indeed, raising pH_o to 7.6 while stimulating CaR-induced Ca²⁺_i mobilization in bovine parathyroid cells (2.5 mM Ca²⁺_o) significantly potentiated the response, an effect that was fully and immediately reversible upon return to pH_o 7.4 (Figure 3A). In addition, lowering pH_o to 7.2 promptly inhibited the Ca²⁺_i mobilization elicited in parathyroid cells in response to 2.5mM Ca²⁺_o (Figure 3B).

Figure 3.

CaR-induced Ca²⁺_i mobilization is sensitive to pathophysiologic pH_o changes in bovine parathyroid cells. (A) Trace showing Ca²⁺_i changes (quantified as a bar graph, as before) in a single cell (upper trace; “cell”) or the “global” (lower trace) cluster of cells in the field of view, in response to elevated [Ca²⁺]_o (2.5 versus 0.8 mM control) at pH_o 7.4 or 7.6 (resulting changes normalized to pH 7.4 control). (B) Identical experiment except examining the effect of lowering pH_o (7.2) during CaR stimulation. n=6–8 coverslips. *P<0.05 and **P<0.01 versus initial pH 7.4 response; ⁺⁺P<0.01 versus pH 7.6 by Kruskal–Wallis (Dunn post-test).

In considering the potential (patho)physiologic relevance of CaR pH_o-sensitivity, it should be noted that serum albumin can bind calcium in a pH-dependent manner. Thus, albumin releases calcium under acidic conditions, which might tend to counteract the concomitant CaR inhibition, whereas alkaline conditions promote calcium binding to albumin and thus potentially counteract the CaR stimulation. Thus, we tested the effects of pH_o on CaR-induced Ca²⁺_i mobilization in the presence of a physiologic concentration of 5% (w/v) albumin. Exposure of CaR-HEK cells to 3 mM Ca²⁺_o (in 5% [w/v] albumin) at pH 7.4 resulted in an increase in Ca²⁺_i concentration, and as before, lowering pH_o to 7.2 resulted in a reversible attenuation of the CaR response (Figure 4Ai). In addition, increasing pH_o to 7.6 while stimulating CaR with 3 mM Ca²⁺_o (in 5% albumin) potentiated the response (Figure 4Aii). This was further tested in bovine parathyroid cells; again, raising pH_o to 7.6 potentiated CaR activity despite the presence of 5% (w/v) albumin (Figure 4B).

Figure 4.

Physiologic albumin concentration fails to prevent changes in pH_o from modulating CaR activity. (A) Changes in CaR-HEK cell Ca²⁺_i concentration (measured as before) in a single cell or the “global” cluster of cells in response to elevated [Ca²⁺]_o (3 versus 0.5 mM control) in buffer supplemented with 5% (w/v) BSA at pH_o 7.6 (i) or 7.2 (ii). The resulting changes (area under the curve) are normalized to pH 7.4 control and displayed as a bar graph. (B) Identical experiment except using bovine parathyroid cells (2.5 versus 0.8 mM control at pH_o 7.4 or 7.6). n≥5 coverslips. *P<0.05 and **P<0.01 versus initial pH 7.4 response; ⁺P<0.05 versus pH 7.6 (or 7.2) by Kruskal–Wallis (Dunn post-test).

To test whether the changes in pH_o might instead be acting by altering intracellular pH (pH_i), CaR-HEK and bovine parathyroid cells were loaded with the pH-sensitive dye BCECF and exposed to CaR stimulation in the presence or absence of even greater changes in pH_o, namely pH 7.0 for acidosis and pH 7.8 for alkalosis. However, in neither cell type did pH_o 7 or 7.8 (or indeed CaR activation itself) affect pH_i over the timescales used (Figure 5). Therefore, these data suggest that the consequence of altering pH_o on intracellular signaling is via an effect on the CaR per se. Ammonium chloride and sodium acetate were used as positive controls to demonstrate that pH_i changes could be detected in the BCECF-loaded cells. Indeed, ammonium chloride elicited intracellular alkalinization in both cell types, whereas sodium acetate induced a marked intracellular acidification in CaR-HEK cells and had a much smaller effect in bovine parathyroid cells.

Figure 5.

Acute pH_o changes do not affect pH_i levels in CaR-HEK or parathyroid cells. (A, i) Trace showing pH_i changes (BCECF ratio) in a cluster of CaR-HEK cells in response to elevated [Ca²⁺]_o (2.5 versus 0.5 mM control) at pH_o 7.4 and 7.8. Ammonium chloride (20 mM, NH₄Cl₂) was used as a positive control. Quantification comparing control (baseline extrapolated) and treated (actual) values for calculated pH_i. (A, ii) Identical experiment except examining the effect of lowering pH_o (7.0) and with sodium acetate (20 mM, NaAc) used as positive control. (B, i) Trace showing pH_i changes (BCECF ratio) in a cluster of bovine parathyroid cells in response to elevated [Ca²⁺]_o (2.5 versus 0.8 mM control) at pH_o 7.4 and 7.8 (NH₄Cl positive control). (B, ii) Identical experiment except examining the effect of lowering pH_o (7.0). n=3–7 coverslips. No significant change (by paired t test) in pH_i occurred after acute changes in pH_o in either cell type.

Effect of Altered pH_o on Parathyroid Hormone Secretion

Parathyroid hormone secretion was measured in human parathyroid cells. In the presence of physiologic free Ca²⁺_o concentration (1.2 mM), lowering pH_o (7.2 then 7.0) caused an initial increase in PTH secretion whereas raising pH (7.6 then 7.8) suppressed PTH secretion (Figure 6). When pH_o was lowered to normal (7.4), PTH secretion then rose again, suggesting that the suppression was fully reversible. Preliminary experiments investigating the effect of 0.4-unit pH_o changes on PTH secretion from both bovine and human parathyroid cells (and conducted before those shown in Figure 6) showed a similar pH_o sensitivity, with alkalosis suppressing PTH secretion significantly in both cases (not shown).

Figure 6.

Extracellular pH modulates CaR-induced suppression of PTH secretion. (A) Human parathyroid cells were perifused in 1.2 mM calcium-containing buffers at pH_o 7.0–7.8 (at 5-minute intervals) and PTH secretion quantified as fg per minute per cell. See Supplemental Material for the complete methods. (B) Quantification of these changes is shown as percentage control. n=3 independent experiments. *P<0.05 by unpaired t test (one-tail) on PTH values following preliminary experiments investigating the effect of 0.4-unit pH_o changes on both human and bovine parathyroid cells (not shown).

Investigation of Extracellular Histidine/Cysteine Residues as Possible Sites of CaR pH_o Sensitivity

Finally, modeling the homodimeric CaR extracellular domain using sequence alignment with the metabotropic glutamate receptor (mGlu) structure 2e4u^11,12 revealed the proximity of certain extracellular histidine sites (in black) to clusters of aspartate and glutamate residues (in gray) previously proposed to be sites of Ca²⁺_o binding (Figure 7A).^13,14 The amino acids whose side chain pK values are closest to the physiologic pH_o and, thus that might account for pH_o sensitivity in CaR, are histidine (approximately 6.5) and free cysteine (approximately 8.5). Indeed, the human CaR contains 16 extracellular histidines plus 1 free cysteine residue (Cys-482). Therefore, each of these residues was mutated individually, histidines to valines and the cysteine to serine, with the resulting 17 CaR mutants expressed transiently in HEK-293 cells. The effects of ±0.2-unit pH_o changes on CaR-induced Ca²⁺_i mobilization were then tested in the mutant CaR-expressing cells and compared with their effects on wild-type CaR-expressing cells. For this, concentration-effect curves for Ca²⁺_o-induced Ca²⁺_i mobilization were generated to determine whether the mutations elicit substantial changes in Ca²⁺_o potency/CaR responsiveness per se. All but two of the mutants exhibited significant, sigmoidal Ca²⁺_o sensitivity (at concentrations <10 mM), the exceptions being CaR^H41V and CaR^H595V (Table 1). See Supplemental Material for the complete methods.

Figure 7.

CaR extracellular domain model showing histidine and free cysteine residue locations relative to aspartate and glutamate clusters. (A) Histidines and the free cysteine are shown in black with the three aspartate/glutamate clusters in light gray.^21,22 Not shown are residues CaR^H377 and CaR^H766 (located in a transmembrane extracellular loop). (B) Immunoblots showing the mutant CaR proteins lacking each of the 17 extracellular histidine or free cysteine residues CaR immunoblots compared with wild-type (wt) CaR and β-actin loading controls.

Table 1.

Characterization of the Ca²⁺_o-sensitivity, relative maximal responses, and pH_o sensitivity of the CaR histidine and free cysteine mutants versus wild-type

Mutant	EC50 (mM)	Emax (%)	(n)	Inhibition by Acidosis (%)	Stimulation by Alkalosis (%)	(n)
WT-CaR	4.3±0.3	100±19	16	−20±2	14±3	20
H134V	4.3±0.3	93±19	6	−22±6	9±4	9
H192V	7.6±1.0a	118±16	5	−22±9	34±13	7
H254V	3.9±0.6	90±22	6	−10±5	35±9	6
H312V	3.3±0.5	80±8	5	−26±5	32±10	6
H338V	5.7±0.1	162±17	4	−22±7	16±4	6
H344V	6.1±0.5b	82±6	8	−38±10	55±26	6
H359V	5.2±0.5	152±22	8	−24±13	11±5	6
H377V	5.8±0.7	143±11	8	−43±6	19±12	7
H413V	3.9±0.4	104±20	5	−16±9	13±3	9
H429V	5.7±0.7	91±11	3	−13±13	9±3	7
H463V	5.9±1.1	90±13	6	−40±5	23±6	6
H463V/H466V	4.7±0.3	70±9	9	−17±8	21±5	7
H495V	4.8±0.6	180±52	3	−11±7	11±6	8
H766V	3.3±0.2	131±9	9	−22±4	20±8	9
C482S	3.7±0.7	81±39	3	−36±5	26±8	7
H429V/H495V	5.0±0.4	53±7	7	−35±4	19±5	7
H41V	>10	–	9	−43±6	22±11	8
H595V	>10	–	9	−22±7	13±5	9

Two independent series of experiments tested (1) the Ca²⁺_o senstivity and maximal response for each CaR mutant (extracellular histidine or free cysteine residues replaced with valine or serine respectively) versus wild-type and (2) the effect of decreasing or increasing pH_o by 0.2 unit on receptor responsiveness. All but four of the CaR mutants exhibited Ca²⁺_o sensitivity not significantly different from wild-type CaR. For CaR^H192V and CaR^H344V the EC₅₀ values for Ca²⁺_o (mM) were increased significantly. Regarding maximal responsiveness to Ca²⁺_o, none of these CaR mutants responded differently to wild-type (Kruskal–Wallis, Dunn multiple comparison test) with the exception of CaR^H41V and CaR^H595V that required additional cotreatment with a positive allosteric modulator (1 mM R-467) to achieve robust Ca²⁺_i mobilization (not shown). Next, none of the mutants exhibited significantly altered CaR pH_o sensitivity (one-way ANOVA with Dunnett) in response to pathophysiologic alkalosis (pH_o 7.6) or acidosis (pH_o 7.2) in the presence of 3.5 mM Ca²⁺_o. Because CaR^H41V and CaR^H595V lacked responsiveness to 3.5 mM Ca²⁺_o, the percentage change values quoted for them are not directly relative to wild-type CaR, but merely represent the percentage change of their CaR response from the immediately prior response in pH 7.4. EC₅₀ half maximal effective concentration; E_max, maximal responsiveness. n=coverslips, from a minimum of two independent transfections in the case of the pH_o experiments.

^aP<0.05 by one-way ANOVA, Dunnett post hoc test.

^bP<0.001 by one-way ANOVA, Dunnett post hoc test.

For the 15 CaR mutants that exhibited robust Ca²⁺_o sensitivity, the effect of ±0.2-unit pH_o changes on CaR-induced Ca²⁺_i mobilization was then tested after exposure to 3.5 mM Ca²⁺_o. Acidosis-elicited attenuation and alkalosis-mediated potentiation of CaR activity did not significantly decrease in any of the CaR mutants (relative to wild-type CaR), suggesting that none of them contribute to pH_o sensitivity in CaR (Table 1). Having failed to identify a sole histidine residue as being responsible for eliciting CaR pH_o sensitivity, the two histidine mutants that produced the largest trend reductions in pH_o sensitivity, namely CaR^H429V and CaR^H495V, were then comutated to determine whether together they might inhibit CaR pH_o sensitivity. However, the pH_o sensitivity of CaR^H429V/H495V was not significantly different from that of wild-type CaR.

For the two remaining mutants that exhibited reduced Ca²⁺_o sensitivity (and lower protein abundance of the mature 160-kDa CaR protein), more robust stimulation was required, namely 5 mM Ca²⁺_o plus 1 μM R467 (positive allosteric modulator) for CaR^H595V and 30 mM Ca²⁺_o plus 1 μM R467 for CaR^H41V (Figure 7B, Table 1). Because these treatments represent supramaximal stimuli for wild-type CaR and thus are unsuitable for testing the effects of altering pH_o on the wild-type receptor, the resulting data for CaR^H41V and CaR^H595V cannot be compared directly to wild-type CaR but are appropriate for qualitative comparison by examining their signaling before and after the change in pH_o. In the case of both mutations, lowering pH_o significantly inhibited the CaR response while raising pH_o significantly increased activity in the mutant CaRs as before. Therefore, taken together, there was no evidence that any of the 17 histidine residues or single free cysteine (Cys-482) contributed to the pH_o sensitivity of CaR.

Discussion

Ca²⁺_o potency for CaR is sensitive to large changes in ambient pH_o.^3,15 The pH of human plasma is maintained between 7.35 and 7.45, representing a 12.5% deviation above and below the normal H⁺ concentration of 4×10⁻⁸ M (i.e., pH 7.40). Indeed, a 1-unit pH change would represent a lethal 10-fold change in H⁺ concentration,¹⁶ and even a smaller drop to 7.1 and below represents a medical emergency. This led us to investigate whether the CaR can also respond to smaller, pathophysiologically relevant changes in pH_o. Because metabolic acidosis and secondary hyperparathyroidism are both common consequences of CKD, the current data may even provide a mechanistic link between the two.

Here we have demonstrated that pathophysiologically relevant (0.2-unit) changes in pH_o significantly modulate Ca²⁺_o-induced Ca²⁺_i mobilization in CaR-HEK cells and elicit similar effects on two other readouts of CaR-mediated signaling: ERK phosphorylation and actin polymerization. Thus, the effect of the pH_o change is not specific for one particular signaling pathway and thus likely occurs at the level of the CaR. Indeed, no acute changes in pH_i were detected in CaR-HEK or bovine parathyroid cells after exposure to larger (0.4-unit) changes in pH_o over the timescale tested, supporting the idea that the change in pH_o elicits an extracellular, as opposed to nonspecific intracellular, effect on CaR activity. Quinn et al.³ reported slow cytoplasmic acidification/alkalinization in CaR-HEKs in response to much larger 1- to 2-unit pH_o decreases/increases, respectively; however, the slow rate of change failed to account for the much faster pH_o-mediated change in CaR sensitivity.

Regarding the CaR agonist selectivity of the pH_o effect, Quinn et al.³ found that Mg²⁺_o was similarly affected by pH_o as for Ca²⁺_o, both of these cations being orthosteric CaR agonists. Here we found that L-Phe–induced Ca²⁺_i mobilization was also sensitive to pathophysiologic changes in pH_o, suggesting that allosteric CaR modulation is similarly sensitive to pH_o and thus unlikely to counteract the effect in vivo.

Of note, the enhancement of CaR signaling with pH_o 7.6 and inhibition with pH_o 7.2 was also observed in bovine parathyroid cells, in which the CaR is expressed endogenously. Consistent with this observation was the clear suppression of PTH secretion from bovine and human parathyroid cells after perifusion in alkaline buffer (pH_o 7.8; preliminary experiments not shown). Repeating these experiments with use of 0.2-unit pH_o changes, we observed a transient rise in PTH secretion from human parathyroid cells that was not sustained when pH_o was lowered to 7.2; we also noted sustained suppression of PTH secretion when pH_o was increased to 7.6. These in vitro data are consistent with observations that metabolic acidosis is associated with increased serum PTH and calcium levels and that alkalosis suppresses serum PTH in vivo.^4,6–8 The current findings suggest a CaR-based mechanism for these previous data from animal studies. It will be interesting therefore to determine therefore whether acidosis and/or alkalosis elicit chronic changes in PTH secretion in vivo. Interestingly, at age 80 years, human blood H⁺ concentration is 6%–7% higher than at age 20,¹⁷ and there is a concomitant rise in serum PTH levels.¹⁸

With regard to metabolic acidosis and secondary hyperparathyroidism in CKD, while secondary hyperparathyroidism is known to be due partly to hyperphosphatemia and partly to decreased calcitriol levels (resulting in part to lowered plasma free calcium concentration), it is interesting to speculate that acidosis may also contribute to elevated PTH secretion rates by suppressing CaR sensitivity, as observed here. In addition, the current data suggest that raising pH_o promotes CaR-mediated suppression of PTH secretion and thus may provide the basis for a useful adjuvant therapy for secondary hyperparathyroidism in CKD. Clinically, low bicarbonate concentration in predialysis patients predicts subsequent coronary artery calcification.¹⁹ Because oral sodium bicarbonate supplementation slows the rate of decline of renal function in patients with CKD and low plasma bicarbonate concentrations,²⁰ the effect of this co-therapy on the development of secondary hyperparathyroidism and vascular calcification might be considered for investigation. Of note, acidosis also increases fibroblast growth factor-23 expression in osteoblasts, raising the question of whether treating acidosis in CKD may lower fibroblast growth factor 23²¹ and, indeed, PTH secretion.

Low pH displaces bound calcium from serum albumin and could compensate for the acid-mediated decline in CaR responsiveness by increasing the free calcium concentration.²² However, in the current study, inclusion of 5% (w/v) albumin in the physiologic saline solution (i.e., at a physiologically relevant concentration) had little or no effect on the pH_o sensitivity of the CaR. Indeed, albumin was present in both sets of experiments measuring human and bovine PTH secretion, and this did not prevent detection of an alkalosis-induced suppression of PTH secretion. Thus, over the pathophysiologic pH_o range, the effect of changing pH_o on CaR activity appears greater than any potential effect on calcium buffering/displacement. Indeed, one study found that between pH 6.8 and 7.4, the calcium-binding affinity of albumin is not altered significantly, whereas at pH 8 the association constant increases 4-fold.²³

Extracellular histidine residues account for proton sensitivity in several membrane proteins, including the anion exchange protein 2,²⁴ ovarian cancer G protein–coupled receptor-1²⁵ and purinergic P2×4 receptor.^26,27 Despite this, we found no evidence that any of the 16 extracellular histidine residues or the one free-cysteine residue (Cys-482) are individually responsible for CaR pH_o sensitivity, at least over the pathophysiologic pH_o range tested. Among members of the homologous family C GPCRs, mGluR4 is also inhibited by acidity and activated by alkalinity whereas mGluRs 1, 5 and 8 are pH_o insensitive.²⁸ However, no histamines are shared exclusively between CaR and mGluR4. Therefore, we conclude, by exclusion, that the most likely molecular mediators of CaR pH_o sensitivity are extracellular clusters of aspartate and glutamate residues.³ That is, despite the low pKa values of Glu and Asp side chains in their free amino acid forms (around 4), when clustered, their pKa values may lie closer to the physiologic pH range (approximately 7).^29,30 Finally, while we cannot rule out the possible contribution of other pH_o-sensitive membrane proteins to these data, the pH_o changes had no effect in cells lacking the CaR (not shown) or incubated in low Ca²⁺_o concentration (Figure 2Bii).

In conclusion, the human CaR exhibits pH_o sensitivity over the pathophysiologic range of pH_o witnessed in vivo. This pH_o sensitivity of the CaR exerts functionally significant effects on PTH secretion and thus might have wider relevance for whole-body calcium homeostasis and indeed for the secondary hyperparathyroidism of CKD.

Concise Methods

Cell Culture and Calcium-Sensing Receptor Functional Assays

HEK-293 cells, stably transfected with human parathyroid CaR, were cultured in DMEM (supplemented with 10% [v/v] FBS), loaded with Fura-2/AM and assayed for Ca²⁺_i by dual-excitation wavelength microfluorometry in experimental buffer (20 mM HEPES, pH 7.4, 125 mM NaCl, 4 mM KCl, 1.2 mM CaCl₂, 0.5 mM MgCl₂, 5.5 mM glucose) as described previously.¹⁰ Intracellular pH was quantified similarly using the pH-sensitive fluorescent dye BCECF. CaR-induced ERK phosphorylation was detected by semi-quantitative immunoblotting using a phospho-specific polyclonal antibody⁹ against the lysates of cells solubilized in RIPA buffer supplemented with protease and phosphatase inhibitors. To assess actin stress fiber assembly, paraformaldehyde-fixed cells were stained with Phalloidin-TRITC and imaged by fluorescence microscopy.¹⁰

Site-Directed Mutagenesis

CaR mutations were introduced into the human CaR by QuikChange (Stratagene) site-directed mutagenesis, then transiently transfected into HEK-293 cells using FuGENE-6.

Parathyroid Gland Preparation and PTH Secretion Assay

Bovine parathyroid cells (abattoir-sourced) were obtained by collagenase-digestion⁹ and cultured on collagen-coated coverslips. Normal human parathyroid cells were obtained by collagenase-digestion after neck surgery⁹ (procedures performed under institutional ethical guidelines and with patients’ written informed consent). PTH levels were quantified by ELISA after perifusion with buffer containing 125 mM NaCl, 4 mM KCl, 1.25 mM CaCl₂, 1 mM MgCl₂, 0.8 mM Na₂HPO₄, 20 mM HEPES (pH 7.4, NaOH) supplemented with 0.1% D-glucose, 2.8 mM basal amino acid mixture (defined previously⁹), and 1 mg/ml BSA.

Statistical Analyses

Data are presented as means±SEM, and statistical significance was determined using GraphPad Prism software. For the complete methods, see the Supplemental Material.

Disclosures

None.