The calcium sensing receptor (CaSR) is a class C G-protein coupled receptor that is crucial for the feedback regulation of calcium homeostasis. While extracellular calcium (Ca2+o) is considered the primary physiological ligand, the CaSR is activated physiologically by a plethora of molecules including polyamines and L-amino acids. Activation of the CaSR by different ligands has the ability to stabilize unique conformations of the receptor which may lead to preferential coupling of different G-proteins, a phenomenon termed 'ligand-biased signalling'. Mutations of the CaSR are currently not linked with any malignancies; however altered CaSR expression and function are associated with cancer progression. Interestingly, the CaSR appears to act both as a tumour suppressor and an oncogene, depending on the pathophysiology involved. Reduced expression of the CaSR occurs in both parathyroid and colon cancers, leading to loss of the growth suppressing effect of high Ca2+o. One the other hand, it seems that activation of the CaSR might facilitate metastasis to bone in breast and prostate cancer. A deeper understanding of the mechanisms driving CaSR signalling in different tissues, aided by a systems biology approach, will be instrumental in developing novel drugs that target the CaSR or its ligands in cancer.
Key words: calcium sensing receptor, parathyroid adenoma, colorectal cancer, breast cancer, prostate cancer, bone metastsis, wnt pathway, RANKL, PTHrP, proliferation, differentiation, beta catenin, calcimimetics, calcilytics, Monod-Wyman-Changeux model, systems biology, ligand-biased signalling
2. Introduction
The calcium ion (Ca2+) is crucial for the control of many important cellular functions such as proliferation, differentiation, fluid secretion and skeletal integrity. Many organisms express cell-surface sensors for extracellular Ca2+ (Ca2+o). Ca2+o is the primary physiological ligand of a G protein-coupled receptor (GPCR), called the calcium sensing receptor (CaSR). The receptor is expressed primarily within the chief cells of the parathyroid glands where, when activated by an excess of Ca2+o, it decreases the release of the calcium-retaining hormone, parathyroid hormone (PTH) [1] to maintain Ca2+o within the physiological range (1.1-1.3 mM). Vice versa, in the case of hypocalcemia, the CaSR is inactive and PTH is released, an event which restores normocalcemia by increasing the reabsorption of calcium from its target tissues, namely the kidneys, intestine and bone [1]. A large number of studies have shown that the CaSR is expressed in many other tissues in the body, which do not play an obvious role in the maintenance of calcium homeostasis, such as the breast, blood vessels, liver, and placenta, to mention but a few [1]. Altered CaSR expression and/or activity is associated not only with disorders of the parathyroid glands, but also with other conditions like osteoporosis, vascular calcification and cancer.
The CaSR is involved in the regulation of a number of diverse processes, such as hormone secretion, gene expression, ion channel activity, modulation of inflammation, proliferation, differentiation, and apoptosis, depending on cell type, and therefore represents a key molecule in functional physiology [1]. The CaSR is also able to respond to a variety of ligands, including polyvalent cations, amino acids, to name but a few. Furthermore, changes to pH and ionic strength affect the activity of the receptor, making the CaSR uniquely capable of integrating several metabolic signals.
Furthermore, activation of the CaSR by different ligands may stabilize the CaSR in unique activation states allowing preferential stimulation of different signalling pathways - termed "ligand-biased signalling" [2]. Altered CaSR signalling contributes to a number of pathophysiological states. In cancer the CaSR-mediated signalling either becomes reduced or even lost, or different pathways will be encompassed. Therefore, there is need for a better understanding of how individual ligands effect CaSR-mediated signalling, and how will this, in turn benefit development of novel pharmaceutical therapies targeting the CaSR.
This review article aims to provide an overview of the signalling mechanism mediated by the CaSR under normal physiological conditions and in cancer.
2.1 CaSR signalling in physiology
The CaSR is a pleiotropic, GPCR that is extremely sensitive to very small deviations in plasma Ca2+o (i.e., less than 10%) within the physiological range. There is a very steep, inverse sigmoid relationship between Ca2+o and PTH release, with the steepest part of the curve being centred around the physiological serum Ca2+o of 1.2 mM, at which concentration PTH secretion is already suppressed by ~25% of its maximal value [1].
Evidence gathered since the late eighties showed that CaSR activation and inhibition of PTH release are linked through Gaq/11 subunits of trimeric G proteins. CaSR activation in the parathyroid glands and many other cellular systems promotes phosphoinositide turnover through activation of membrane-bound phospholipase C (PLC), with production of inositol 1,4,5 tris-phosphate (IP3) and diacyl glycerol. These, in turn, promote release of Ca2+ from intracellular, IP3-sensitive stores [3] and protein kinase C (PKC) activation, respectively. Indeed, PTH levels of mice lacking both Gaq and Ga11 are greatly elevated [4]. Genetic studies in humans demonstrate the importance of CaSR activation in the suppression of PTH secretion and parathyroid cell proliferation [5-7].
Activation of the CaSR can also induce intracellular calcium (Ca2+i) oscillations, which have been linked to inhibition of proliferation in colonic epithelial cells [8]. Such CaSR-mediated Ca2+i oscillations have been observed in a number of CaSR-expressing cells [9], including parathyroid cells [10, 11] and CaSR-expressing HEK293 cells (HEK-CaSR) [12]. Recent work has examined how variations in CaSR-mediated Ca2+i oscillation frequency and amplitude are physiologically significant in a variety of cell contexts. For example, in human colonic epithelial cells the CaSR can induce two separate oscillatory pathways and while CaSR-mediated high frequency (~ 3 - 4 min-1) sinusoidal Ca2+i oscillations induced inhibition of proliferation, low frequency (~ 1.5 min-1) transient Ca2+i oscillations have not [8]. Thus, the mechanisms that control Ca2+i oscillation frequency and amplitude are critical for various CaSR-dependent biological responses.
In HEK-CaSR cells, the sinusoidal oscillations (~ 4 min-1 at 37 °C) [13, 14] arise from the dynamic phosphorylation and dephosphorylation of T888 [15-17], the primary PKC phosphorylation site of the CaSR [18]. Thus, receptor-induced activation of PKC leads to phosphorylation of T888 to uncouple the receptor from Gαq/11-induced PLC activation and Ca2+i mobilisation [18, 19]. T888 is then dephosphorylated by protein phosphatases to restore receptor coupling to PLC and Ca2+i mobilisation [17]. Interestingly, T888 phosphorylation exhibits a bi-phasic, 'bell-shaped' profile in which phosphorylation of T888 peaks at Ca2+o concentrations around 2 - 3 mM at 10 min [17]. Further increases in Ca2+o lead to decreases in T888 phosphorylation, and Ca2+o concentrations ≥ 4.0 mM elicit sustained elevations in Ca2+i rather than Ca2+i oscillations [17]. Disruption of the T888 PKC phosphorylation site produces an increase in Ca2+o sensitivity and recently the mutation T888M was identified in a case of Autosomal Dominant Hypocalaemia (ADH) [20], demonstrating that the T888 residue and its regulation by PKC is critical for physiological CaSR function in vivo.
The CaSR also couples to a number of different signalling pathways. For example, in thyroid C cells activation of the CaSR is linked to calcitonin-secretion through the activation of voltage-gate calcium channels [21], while in kidney cells, CaSR activation is coupled to the metabolism of arachidonic acid by cytochrome P450 (CYP450) and cyclooxygenase (COX) pathways via a Gi-dependent mechanism [22].
The CaSR is also expressed in the colon where it plays an important role in nutrient sensing and intestinal fluid transport. The group of Hebert et al., using isolated crypts from rat [23, 24] and the CaSR-/-::GCM2-/- knock out mouse model demonstrated that one of the mechanisms involved in this function is via CaSR-dependent degradation of cyclic nucleotides by phosphodiesterases [24]. Further, it has been suggested that the presence of the CaSR in nerve ends of smooth muscles along the colon could indicate involvement of the CaSR in intestinal motility [25].
CaSR activation is linked also to pro-proliferative stimuli in many cell systems. Pro-proliferative CaSR signalling involves activation of the mitogen-activated protein (MAP) kinases extracellular-signal-regulated kinases (ERK) and p38MAPK protein kinases [26, 27]. ERK phosphorylation occurs through "triple-pass" signalling in which CaSR activation leads to the release of an epidermal growth factor (EGF)-like peptide by matrix metalloproteinases, which in turn evokes EGF receptor-mediated cell signalling [28].
Another mechanism by which CaSR signalling might play a role in mediating pro/anti-oncogenic signalling is through regulation of cell migration. Indeed, CaSR-dependent ERK phosphorylation requires receptor binding to the cytoskeletal protein filamin A while CaSR activation in breast cancer cells evokes rho-dependent actin filament formation through Ga12/13 [29]. Since Ga12/13 proteins have been implicated in many different processes, including cell migration [30], it has been hypothesized that CaSR-dependent signalling in breast and prostate cancer cells could favour the metastatic spread of tumours (a concept that is discussed further below) [31].
3. CaSR in cancer: tumour suppressor or oncogene?
Intracellular Ca2+-dependent signalling mechanisms are frequently remodelled or deregulated in cancer cells. The current understanding is that the CaSR can either prevent, or promote tumourigenesis depending on the type of cancer [32]. The mechanisms behind its impact on carcinogenesis are multiple and not well understood.
The expression of the CaSR can be decreased or even absent, as it is in parathyroid and colorectal cancer. In these tumours dearth of CaSR expression results in loss of the growth suppressing effects of high levels of Ca2+o. Activation of the receptor inhibits proliferation of these cancer cells, suggesting a tumour suppressor function for CaSR. In contrast, increased expression is observed in highly metastatic primary breast and prostate cancer cells. Furthermore, in breast cancer cells CaSR activates preferentially Gαs proteins and not Gαi, as in normal breast cells [33], resulting in increased production of parathyroid hormone-related peptide (PTHrP), which is a primary cause of hypercalcaemia of malignancy, and a contributor to metastatic processes involving bone. In these settings the CaSR seems to have an oncogenic role.
Another mechanism leading to altered signalling through the CaSR is loss of one or more signalling partners (e.g. Gαq in parathyroid tumours). The impact of changes in ligand composition and/or concentration on the preference of the CaSR for a certain signalling pathway (as it might be the case in the intestinal tract) needs further investigation. There is no evidence as yet that mutations of the CaSR play a role in tumourigenesis.
Pharmacological agonists and antagonists of the receptor might find a much broader therapeutic usage, depending on whether activation or inhibition of the receptor is required, but first we have to understand the mechanisms driving CaSR signalling in the different tissues.
3.1. The role of the CaSR in parathyroid tumourigenesis
Parathyroid adenoma, parathyroid hyperplasia and parathyroid cancer are the main disorders of the parathyroid glands. Parathyroid adenoma and cancer lead to primary hyperparathyroidism (pHPT). pHPT is a relatively common endocrine disorder, with 25 - 30 new cases per 100,000 people each year [34]. Parathyroid adenomas account for 80% of the pHPT cases, whereas parathyroid cancer is an extremely rare disease responsible for less than 1% of all pHPT cases. Parathyroid adenomas are more frequent in 42-59 year old women, while parathyroid carcinoma develops usually much later in life with equal frequency in both sexes. No dietary or lifestyle factors were associated with higher or lower risk to develop parathyroid cancer. [35]. The CaSR seems to be important for the prevention of these benign and malignant parathyroid tumours.
3.1.1. Expression of the CaSR in parathyroid tumours
Several studies have demonstrated that the expression levels of CaSR mRNA and protein are decreased in parathyroid adenomas, as well as in hyperplastic parathyroid glands from patients with uremic secondary hyperparathyroidism, compared with normal glands [36-38]. In addition, the study of Haven et al. suggests a role for the CaSR in prevention of malignant parathyroid tumours as the CaSR expression is decreased or absent in parathyroid carcinomas, compared with parathyroid adenomas and hyperplastic glands [39].
The molecular mechanisms that drive the loss of CaSR expression, as well as the mechanisms underlying abnormal PTH secretion in response to Ca2+o, in parathyroid tumours are still not completely understood. One possible explanation for the loss of CaSR expression was the loss of CaSR allele(s) on chromosome 3q leading to decreased stability of the CaSR mRNA, observed in pathological parathyroid glands, however, this has turned out not to be the case [38, 40]. Furthermore, no mutations in the coding region of the CaSR have been identified in parathyroid adenomas, hyperplasia or carcinomas, suggesting that mutations are not involved either in loss of CaSR expression or abnormal PTH release in these tissues [40].
3.1.2. CaSR-mediated signalling in parathyroid tumours
Uncontrolled parathyroid cell proliferation is a common feature in pathological parathyroid glands. Physiological activation of the CaSR appears to be related with suppression of parathyroid proliferation [5, 39, 41]. Patients with inactivating CaSR mutations and mice homozygous for the CaSR knockout generally exhibit marked parathyroid hyperplasia, indicative of the inhibitory role of the CaSR in parathyroid cell proliferation [5]. Furthermore, Haven et al. showed that the Ki67 proliferation index is significantly higher in parathyroid carcinomas, where it is significantly correlated with down-regulation of CaSR expression, than in parathyroid adenomas and hyperplasia [39].
The inhibitory role of this receptor in parathyroid proliferation was demonstrated by using calcimimetics, allosteric activators of the CaSR. Recent work by Miller et al., using a rodent model of chronic kidney disease characterized by parathyroid hyperplasia and excessive PTH secretion, showed that treatment with the calcimimetic cinacalcet mediated regression of parathyroid hyperplasia and PTH decrease, which could be reversed by discontinuation of the treatment. These data further support the need of an active CaSR for inhibition of parathyroid proliferation [42]. Interestingly, calcimimetics not only activate the CaSR but seem to be able to increase its expression as well [43, 44].
Impaired signalling through the CaSR is considered to be a major pathway promoting parathyroid hyperplasia and abnormal Ca2+o sensing. However, some parathyroid adenomas show a reduced sensitivity to Ca2+o even when CaSR expression levels are normal, suggesting that loss of CaSR expression is not the only factor involved in the abnormal PTH secretion in parathyroid adenomas [36, 45]. A number of possible candidates have been proposed that could interact with the CaSR and modulate its signalling, such as cyclin D1, regulator of G protein signalling 5 (RGS5) and caveolin-1.
Cyclin D1 belongs to the cyclin protein family and is involved in regulation and progression of the cell cycle. In a subset of 20 - 40% of parathyroid adenomas, due to chromosomal rearrangement, cyclin D1 becomes controlled by the 5'-regulatory region of PTH and, as a result, is overexpressed [46, 47]. It has been suggested that cyclin D1 can interfere with the CaSR and support development of parathyroid tumours by increasing parathyroid proliferation [48]. The role of cyclin D1 in parathyroid tumourigenesis is supported by the study of Imanishi et al. [46], who have shown a transgenic mouse model of hyperparathyroidism that mimics the overexpression of cyclin D1 in parathyroid adenomas, presented with decreased CaSR expression levels, increased parathyroid cell proliferation and a right-shift in Ca2+o- dependent PTH response [46, 49]. Furthermore, Corbetta et al. [50] showed that in parathyroid adenomas the CaSR inhibits cyclin D1 expression in the presence of growth factors, such as basic fibroblast growth factor (bFGF) and EGF, preventing the oncogenic actions of cyclin D1 in these tumours. Interestingly, the CaSR was unable to inhibit cyclin D1 activation in the absence of bFGF and EGF, supporting the existence of an interaction between the CaSR and these growth factors [50].
Koh et al. have recently shown that RGS5 is up-regulated in parathyroid tumours when compared with normal parathyroid glands [45]. RGS5 is part of the R4 subtype of the RGS proteins that inhibit signal transduction through regulation of heterotrimeric G-proteins. Class C GPCRs, including the CaSR, are regulated by RGS proteins. In HEK-CaSR cells, transiently expressing RGS5 ERK1/2 phosphorylation was inhibited, indicating that RGS5 is able to inhibit CaSR signal transduction in this heterologous expression system. Furthermore, RGS5-/- knockout mice displayed abnormally low plasma PTH levels with normal Ca2+ serum levels and normal responsiveness to Ca2+o. Therefore, it has been suggested that the RGS5 is a negative regulator of CaSR activity in parathyroid cells. It seems that RGS5 may compete with the CaSR in binding to Gαi and Gαq proteins preventing CaSR activation and thus maintaining the sensitivity of the receptor to deviations from normal Ca2+o levels. Furthermore, overexpression of RGS5 in parathyroid adenomas could inhibit normal CaSR signalling and contribute to the abnormal Ca2+ sensing observed in parathyroid tumours [45].
Caveolin-1 is a major component of caveolae, a plasma membrane organelle where the CaSR is localized in bovine parathyroid cells, and is thought to inhibit signal transduction and proliferation. Approximately 62% of parathyroid adenomas express caveolin-1, and these tumours appear to have a better PTH response to Ca2+o compared with those with low or lost caveolin-1 expression [51]. Studies by Kifor et al. [51] have shown in freshly isolated bovine parathyroid cells that activated ERK 1/2 colocalized with caveolin-1in the plasma membrane, whereas in bovine parathyroid cells cultured for 10 days translocation of activated ERK1/2 and caveolin-1 to the nucleus and cytosol was observed, as well as decreased expression of caveolin-1. The activation of ERK1/2 was increased both at low and high Ca2+o in bovine cells cultured for 10 days, which is in accordance with the possible role of caveolin-1 as a negative regulator in the MAPK cascade. Similarly, in parathyroid adenomas where caveolin-1 expression is decreased or lost, ERK1/2 was localised in the cytosol and nucleus, and a reduced ability for high Ca2+o-mediated suppression of PTH secretion was observed. In the majority of the adenomas ERK1/2 was activated independently of Ca2+o. [51]. Thus, contrary to bovine parathyroid glands, the ERK1/2 signalling pathway appears to be lost in parathyroid adenomas.
Furthermore, Corbetta et al. have reported that Gαq protein levels were lower in pathological parathyroid glands than in normal glands [36]. These results are supported by another study in which increasing Ca2+o concentrations and the CaSR agonist, gadolinium (Gd3+), failed to activate ERK1/2 in parathyroid adenomas [52]. Consequently, in parathyroid adenomas a low expression of caveolin-1 and a low level of signalling molecules could possibly cause altered PTH release in response to Ca2+o [36, 51].
In conclusion, parathyroid tumours are characterized by decreased CaSR expression levels, increased cell proliferation and abnormal PTH secretion. Although the exact mechanisms that lead to these events are still unknown, a number of studies have suggested that the CaSR has a role in this process, including CaSR-mediated modulation of cyclin D1 expression and abnormal CaSR signalling due to RGS5 overexpression in these tumours. In addition, CaSR-induced activation of the ERK1/2 signalling pathway does not seem to be active in parathyroid tumours. However, further studies are needed to fully understand the mechanisms responsible for the development of parathyroid tumours and the possible CaSR signalling mechanisms involved.
3.2. The role of CaSR in colorectal tumourigenesis
Colorectal cancer is another cancer in which the CaSR seems to play an important protective role. Colorectal cancer is the fourth most frequent form of cancer in men and the third in women worldwide [53, 54]. Diet plays a major role in colorectal tumourigenesis and epidemiological studies show an inverse correlation between calcium intake and the risk of tumour development [55]. Calcium exerts its chemopreventive features through a plethora of mechanisms, such as binding toxic secondary bile acids and/or ionized fatty acids and neutralizing them in form of insoluble calcium soaps [56, 57], or by activating several downstream signalling cascades such as stimulating cell differentiation, inducing apoptosis and inhibiting proliferation [58-60]. There is evidence that some of these molecular mechanisms are mediated, at least in part by the CaSR [61-64] .
3.2.1. Expression of the CaSR in the colon
In 1991, Whitfield [65] hypothesized that the Ca2+o concentration in the lumen of colonic crypts increases from sub-physiological levels at the base reaching physiological or even higher levels at the top. Cell proliferation is driven by the combination of several factors in the lower two-thirds of the crypt while proliferation is stopped and differentiation is triggered when the cells reach a critical level in the Ca2+o concentration. In a later review Whitfield hypothesized that it is the CaSR that senses the changes in Ca2+o levels and acts as a molecular switch that turns off proliferation and turns on differentiation in colonocytes [66, 67].
Furthermore, besides calcium, there are numerous other CaSR ligands (like polyamines, amino acids) in the gastrointestinal tract, ingested through the diet or produced by local bacterial microflora that might alter the responsiveness of the receptor to Ca2+o. These ligands activate a wide spectrum of signalling pathways in a cell specific manner.
Although, the CaSR is present along the entire length of the digestive tract [68] the exact localization and expression pattern in the colon is not clear. Published data are discordant. Chattopadhyay and colleagues [25] have shown that the CaSR is expressed in different regions of the rat colon, including the basal region of the crypt, the serosa, sub mucosa and also in the nerve endings around the myentric plexus. Sheinin et al. [69] found that the CaSR was present only in the entero-endocrine cells of the human colonic crypt. Later, Cheng et al. [23] showed cytoplasmic and membrane expression of the CaSR protein in rat and human colonic epithelial cells both at the surface and base of the crypt and in the enteric nervous system [70]. Chakrabarty and his colleagues detected perinuclear CaSR expression in the columnar epithelial cells of the colonic crypt, and showed an increase in CaSR expression along the crypt axis, from the basal region to the top of the crypt [71, 72]. Contrary to that, Ahearn et al. found highest expression of the CaSR at the base of the crypt, decreasing towards the apex [73].
A clear consensus regarding the expression of the CaSR protein in the colon is needed to validate Whitfield's hypothesis. Similarly, the existence of a Ca2+o gradient seems to be a hypothesis based on Ca2+i concentrations in colonocytes lining the crypt [23, 74]. Whether or not this gradient in the crypt lumen exists, also needs to be clarified.
Contrary to the inconsistent data of protein expression pattern in the normal colon, there is a consensus that the CaSR expression is down-regulated during colorectal tumourigenesis [69, 71, 72, 75] just as in parathyroid tumours. As a further similarity, the underlying mechanisms leading to this loss of expression are not completely understood in the colon either.
Epigenetic alterations such as DNA methylation and histone modifications, which are frequent events in tumours [76], might cause silencing of the CaSR gene, as suggested by Hizaki et al. [75]. In normal cells DNA methylation assures regulation of proper gene expression and generally affects cytosines within CpG islands [77], regions of the DNA with clusters of cytosines followed by guanine [78]. The CaSR has two promoters and two 5'-untranslated exons (exon 1A and exon 1B) yielding alternative transcripts, but encoding the same protein [79, 80]. The upstream promoter contains a TATA and a CAAT box and only few sporadic CpGs, whereas the downstream promoter contains a large CpG island that could be susceptible to methylation [79, 80].
Interestingly, it seems that exon 1A might have more impact in regulating the expression level of the CaSR then exon 1B. In parathyroid adenomas the expression of exon 1A was much lower than in the normal gland [79]. Similarly, in colorectal tumours exon 1A mRNA expression correlated inversely with tumour grade [81] while expression of exon 1B did not change significantly.
Several transcription factor binding sites like vitamin D response elements (VDRE), and NFкB, Stat1/3, Sp1/3 binding sites are located in both CaSR promoters [80, 82, 83]. The two VDREs in the CaSR promoter suggest a role for vitamin D in regulating CaSR expression. Indeed, injection of 1,25dihydroxyvitamin D3, the active vitamin D metabolite, to rats resulted in an upregulation of the CaSR expression in the parathyroids, thyroids and kidneys [80]. Similar results were observed also in the human colon cancer cell line CBS [71]. Whether lower levels of the vitamin D receptor in colorectal tumours [84] could be one cause of decreased CaSR expression in these tumours, needs to be proven.
3.2.2. CaSR-mediated signalling in colon cancer
The CaSR belongs to an intricate network of calcium signalling pathways that control normal and cancer cell growth, depending on cell-specific coupling to appropriate G-proteins. High Ca2+o levels reduced the rate of proliferation in numerous colon cancer cell lines [61, 85]. The signalling cascade seems to begin with inhibition of phospholipase A2 [81] leading to reduced levels of arachidonic acid, precursor of the proliferative prostaglandins (e.g. prostaglandin E2). Activation of the CaSR sustains differentiation in several colon cancer cell lines [62, 85, 86]. This has been suggested to be dependent on PKC-regulated c-myc down-regulation [61, 87], and mediated via the MAP kinase pathway [85] .
Chakrabarty et al. have shown that in several colon cancer cell lines Ca2+o increases expression of the tumour suppressor E-cadherin [72]. As an intercellular adhesion protein the major role of E-cadherin is maintenance of intact cell-cell contacts for which E-cadherin complexes with ß-catenin, reducing its availability to translocate into the nucleus. In the nucleus the role of ß-catenin is to activate together with TCF4 the so-called canonical Wnt pathway (for rev. see Moon et al. [88]).
The Wingless (Wnt) pathway is crucial for the maintenance of both foetal and adult intestinal crypt architecture and is one of the main pathways regulated by Ca2+o in the colon [89-92]. The adenomatous polyposis coli (APC) protein, a major component of the Wnt pathway is altered in more than 80% of colorectal tumours leading to accumulation of ß-catenin in the nucleus [89, 93]. Calcium inhibits the canonical Wnt pathway by preventing nuclear translocation of ß-catenin, reducing b-catenin-TCF4 complex formation, downregulating c-myc and cyclin D1 expression and thus inhibiting proliferation in colon cancer cell lines, an effect mediated at least in part by the CaSR [72]. Rey et al., using an intestine specific CaSR-/- mouse model, have shown that colonic crypts lacking the CaSR have increased proliferation rate and higher nuclear b-catenin levels. They were able to show in vitro that absence of the CaSR allows phosphorylation of b-catenin at Ser-552/675 which promotes its nuclear translocation [63].
Activation of CaSR by increased Ca2+o mediates Wnt5a secretion from colonic myofibroblasts. Ca2+o enhanced Wnt5a secretion also in HT-29 colon cancer cells expressing truncated APC, leading to inhibition of the defective Wnt signalling in these cells [92, 94]. In colon cancer cells CaSR activation led to up regulation of the receptor tyrosine kinase-like orphan receptor 2 (Ror2) protein which acts as a receptor for Wnt5a [92]. Taken together, it is proposed that the CaSR mediated over expression of Ror2 leads to the Wnt5 mediated inhibition of β-catenin signalling in these cells [92, 94, 95] (see Figure 1). Although Wnt5a can be considered a tumour suppressor, there is also evidence for its involvement in metastasis, as seen in melanoma cells [88]. Furthermore, interactions between CaSR mediated stromal Wnt5a and epithelial Ror2 in colonic epithelial cells seem to increase CDX-2 production suggesting that activated CaSR stimulated epithelial differentiation through Wnt5a mediated non-canonical Wnt signalling [92]. The homeobox gene CDX-2 stimulates expression of sucrase-isomaltase, liver-intestine-cadherin and mucin2 in intestinal epithelial cells, regulating cell growth and differentiation [92, 96]. A number of studies illustrate that although the CDX-2 expression is retained in the intestinal epithelium throughout the adulthood, a clear reduction can be seen in colon cancer [92, 97, 98], probably as a result of loss of CaSR expression.
Figure 1: Regulation of growth control mediated by calcium sensing receptor in colonocytes
The calcium sensing receptor (CaSR) can be activated by several ligands in the extracellular milieu of the colonocytes which regulate proliferation and differentiation in these cells. In addition to regulation of cell growth by activated colonic CaSR directly, there is a cross talk between myofibroblasts and colonocytes. Upon CaSR activation, Wnt5a is secreted by the myofibroblasts which in turn (i) increases the expression of Ror2, a receptor for Wnt5a. Upon Wnt5a-Ror2 binding, differentiation is promoted by signalling cascades leading to regulation of genes like Cdx2, sucrase isomaltase (SI) and villin (green) (ii) The Wnt5a-Ror2 binding can further inhibit nuclear translocation of b-catenin (red) or degradation of b-catenin by upregulation of Siah-2, a ubiquitination ligase (orange) to regulate proliferation by controlling the expression of genes like cyclin D1, C-myc, PCNA and/or replication licensing genes. (iii) A third mechanism of growth control is by the activation of myofibroblast-CaSR which increases secretion of Dickkopf-related protein 1 (Dkk-1) which then inhibits downstream signalling of its receptor, frizzled viz., proliferation (blue).
The anti-proliferative effect of calcium seems to be lost in transformed cells during colon carcinogenesis [99, 100]. Therefore it is of paramount importance to find molecular markers that would identify whether a subject would benefit from calcium or whether high calcium intake would inhibit only the growth of normal cells while supporting proliferation of malignant cells. The CaSR could be this marker, however this still needs scientific proof.
Our knowledge regarding CaSR signalling is still very basic. The exact, step-by-step signalling in colon cancer cells, including other signalling systems involved, is still unravelled. Considering the diversity of ligands present in the colon and the phenomenon of biased signalling as depicted in Figure 1, we need better models and systems to understand the role of the CaSR in normal intestinal physiology and in the process of colon tumourigenesis.
3.3. The role of the CaSR in breast and prostate tumourigenesis
Calcium and the CaSR are considered to be involved in both breast and prostate cancer development as well as in the formation of bone metastasis. Breast and prostate cancer are the most frequent malignancies amongst both sexes, with one in every 8 women and one in every 6 men born in the United States today diagnosed with breast or prostate cancer during their lifetime [101]. Nutritional factors, amongst others dietary calcium, are considered to modulate the risk of breast and prostate cancer. The anti-tumour effects of calcium involve regulation of cell proliferation, differentiation, and apoptosis, partially mediated through interaction with the vitamin D system (reviewed in [102]). However, in contrast to colorectal cancer [103], there is no conclusive evidence to date for the cancer preventive effects of dietary and/or supplemental calcium intake in breast [104-110] and prostate cancer patients [111]. If calcium does influence the risk of breast or prostate cancer, the underlying mechanisms not yet elucidated.
3.3.1. Expression of the CaSR in the breast and prostate
The CaSR is expressed in the ductal epithelial cells of normal breast tissue [112], where it is involved in regulating calcium concentrations in the breast milk during lactation via modulation of PTHrP secretion [113]. Moreover, both ductal and lobular carcinomas express the CaSR. Interestingly, its highest expression was observed in breast cancer patients with bone metastasis, making the authors hypothesize that CaSR-positive tumours are more likely to metastasize to the skeleton [114]. In accordance with this hypothesis, expression of the CaSR was higher in breast cancer cells with relatively increased bone metastatic potential (MDA-MB-231) compared with cell lines showing lower tendency to metastasize to bone (MCF-7, T47D) [115]. Besides breast cancer cells, CaSR mRNA and protein was detected also in the highly metastatic prostate cancer cell lines PC-3 and C4-2B as well as in the low metastatic LnCaP cells, albeit at a lower level [116, 117]. So far, no reports on expression of the CaSR in normal prostate tissue or prostate cancer patients are available.
3.3.2. CaSR-mediated signalling in breast and prostate cancer
CaSR stimulation by high Ca2+o levels promoted proliferation of MCF-7, PC-3 and C4-2B breast and prostate cancer cells known to metastasize to the skeleton. Whereas in PC-3 cells, Cyclin D1 is associated with the Ca2+o-induced proliferative effect [117], in MCF-7 cells, Ca2+o-induced cell proliferation appears to be linked to ERK1/2 phosphorylation, which in turn stimulates expression of the Transient Receptor Potential Canonical 1 (TRPC1) cation channel and subsequent calcium entry [118, 119]. In contrast, proliferation of LNCaP prostate epithelial cells, which do not form bone metastsis, was not affected by Ca2+o [117]. These effects are likely mediated by the CaSR, as its knockdown by shRNA or siRNA decreased Ca2+o-induced cell proliferation in PC-3 and MCF-7 cells in vitro. It has been suggested that physiological concentrations of Ca2+o (1.4 mM) would reduce breast cancer cell proliferation compared with low concentrations of Ca2+o (between 0.175 - 0.2 mM), proposing the CaSR as a tumour suppressor in MCF-7 and MDA-MB-435 breast cancer cells, similar to BRCA1 [120]. These findings might be due to a dose-dependent biphasic response of the CaSR to Ca2+o. Furthermore, proliferation under higher concentrations of Ca2+o was not tested in absence of vitamins, peptone and amino acids, since some of them might have been responsible for "ligand biased signalling" of the CaSR, as further developed later in the present review.
3.3.3. The CaSR in bone metastasis
In advanced breast and prostate cancer, the CaSR seems to play a key role in the development of bone metastasis. For both types of cancer the preferential site of metastasis is the skeleton. Approximately 70% of patients with advanced breast cancer develop bone metastases with a predominantly osteolytic phenotype, while prostate cancer mostly forms osteoblastic bone metastases [121]. For both types of metastases, it is believed that breast and prostate cancer cells and bone cells interact, thereby triggering a vicious circle that results in progressive bone destruction and/or tumour growth as depicted in Figure 2. In particular, cancer cells are capable of producing factors that stimulate bone resorption. In this regard, PTHrP is considered one of the major factors that promote bone turnover via up-regulation of Receptor Activator of NFkB Ligand (RANKL) on osteoblasts. Binding of RANKL to its receptor RANK, which is expressed by cells of the osteoclastic lineage, increases the development, activity, and survival of mature osteoclasts in both physiological and pathological circumstances. Osteoprotegerin, a soluble decoy receptor of RANKL produced by osteoblastic cells as well as cancer cells [122, 123], attenuates osteoclast-mediated bone resorption and it is increased in the serum of patients with metastatic prostate cancer [124, 125]. During bone turnover, bone-derived growth factors and chemoattractant factors, e.g. transforming growth factor-beta (TGF-β), are released, which "drive" cancer cells from the primary tumour site to bone and help them to survive and grow in the bone environment [126, 127]. Moreover, large amounts of calcium are released during bone resorption. Notably, hypercalcemia occurs in 20 to 30% of patients with metastatic cancer and is considered to be caused by systemic secretion of PTHrP by cancer cells leading to excessive bone resorption (reviewed in [128]). In normal breast epithelial cells, PTHrP secretion is subject to a negative feedback regulation by calcium [129]. However, this negative feedback regulation is lost in both malignant breast and prostate cells (MCF-7, MDA-MB-231, LnCaP and PC-3), where activation of the CaSR stimulates PTHrP secretion [130, 131], which is even more increased in combination with TGF-β [116, 130]. As described previously, CaSR signalling is mediated through different Gα subunits of trimeric G proteins. Interestingly, the interaction of the CaSR with G proteins is cell type-dependent and seems to be variable. Mamillapalli et al. demonstrated recently that a shift between G±i and Gαs binding was responsible for the CaSR-dependent stimulation of PTHrP secretion in Comma-D and MCF-7 cells, instead of the inhibitory effect observed in normal mammary epithelial cells [33]. This phenomenon might explain the role played by the CaSR in metastatic breast cancer.
Moreover, calcium is believed to attract cancer cells to the bone via stimulation of the CaSR. Interestingly, breast cancer cells displaying a high metastatic potential in vivo (MDA-MB-231) express the CaSR at a higher level compared with breast cancer cells with a low metastatic potential (MCF-7, T47D). Accordingly, in vitro, MDA-MB-231 cells showed a stronger migratory response to Ca2+o compared with cells with a less metastatic potential. This migratory response was abrogated after transfection of the cells with siRNA targeting the CaSR and after treatment with ERK1/2 or PLCβ pathway inhibitors [115], strongly supporting the role of CaSR-mediated signalling in the migration of breast cancer cells. This hypothesis is also supported by several animal studies. For example, in an intracardiac injection model, Liao and colleagues demonstrated that the CaSR is essential for prostate cancer cells to develop bone lesions in mice. Moreover, they provided evidence that in vitro, Ca2+o stimulates prostate cancer cell proliferation and enhances the attachment of these cells [117], thereby probably promoting their metastatic properties. Recently, in a prostate cancer xenograft model in mice, Li and colleagues showed that ACE-1 and C4-2B prostate cancer cells develop tumours more aggressively when using a mineralized matrix (either neonatal mouse vertebrae, human de-proteinized bone, or a mineralized collagen matrix) compared with a non-mineralized collagen matrix as a scaffold [132]. In consistence with previous studies [133-135], they found that decreasing bone remodelling follows a decrease in tumour growth. Interestingly, tumour growth of prostate cancer cells which do not express the CaSR was similar in mineralized and non-mineralized matrix scaffolds [132], providing further evidence for the importance of calcium and its receptor in metastasis formation.
In contrast to malignancies of the colon and parathyroid, the CaSR seems to play an oncogenic role in advanced breast and prostate cancer through favouring the development of bone metastases.
Figure 2. The vicious circle of bone metastasis in breast and prostate cancer
Figure 2: The vicious circle of bone metastasis in breast and prostate cancer
Breast and prostate tumours produce parathyroid hormone related peptide (PTHrP) which up-regulates receptor activator for NFkB ligand (RANKL) on osteoblastic and stromal cells. Upon binding of RANKL to its receptor on osteoclast precursor cells, i.e. receptor activator for nuclear factor kB (RANK), maturation and differentiation of osteoclast precursors to activated osteoclasts is initiated and bone resorption stimulated. During bone resorption, large amounts of calcium (Ca2+) as well as diverse growth factors are released. Calcium promotes further PTHrP production in the tumour, thereby supporting the vicious circle. Via the calcium sensing receptor (CaSR), Ca2+ is considered to act as a chemoattractant factor and to facilitate tumour cell migration into the bone. Growth factors enable tumour cell survival and growth in the bone microenvironment and in consequence the manifestation of bone metastases which in turn increases the rate of bone turnover, thus feeding the vicious circle.
4. Systems biology of CaSR signalling
The CaSR exerts its physiological role in a huge network of molecular interactions. Systems biology and network-based approaches will be instrumental for the elucidation of the CaSR functions in the cellular signalling network. Studying the dynamics of signalling networks requires quantitative measurements of signalling network responses, e.g. with phosphoproteomics. Computational methods will be required for data integration and inference of biological knowledge. In such an approach, the focus will shift from the CaSR alone to its interactions with its signalling partner and how its responses propagate through the signalling network.
Early signal processing already occurs at the level of the receptor [136, 137]. The CaSR activity is modulated by a large variety of allosteric modulators [138]. Subsequently, downstream signalling to three different G-proteins adds further complexity to CaSR functioning. These signalling processes are modulated by the intracellular signalling state through phosphorylation of CaSR at T888. Many of the dynamic and spatial aspects of CaSR signalling are undoubtedly underappreciated. For other signalling systems it has been shown that dynamics in protein phosphorylation, complex formation and localization reflect the actual decision-making process rather than the static view of whom talks to whom [139, 140].
4.1. Mathematical modelling of the CaSR
The quantitative study of protein mechanisms is deeply rooted in biochemistry where affinity constants, allostery, and cooperativity are fundamental concepts. Systems biology starts from biochemistry and works its way up to the signalling network to study how the underlying molecular processes - and their perturbation in disease - cause network-level properties - cell biology. This approach naturally supports molecular medicine to elucidate the molecular basis of diseases and the development of diagnostic methods and treatment.
Multi-protein complexes, such as the CaSR can display cooperativity, allostery and conformational transitions, central attributes of GPCRs [141]. Cooperativity and conformational changes bring about signal sensitization; in the limit of strong cooperativity the saturation of a receptor with a signal is described by a sigmoidal Hill-type equation. Allostery is the phenomenon that one signal can influence the GPCR affinity for another signal. The models in biochemistry for cooperative proteins explain the functional limits and versatility of such systems due to kinetic, biological and thermodynamic constraints. They facilitate interpretation of experimental data and molecular understanding of altered behaviours of wild-type and mutated cooperative proteins.
In 1965, Monod, Wyman and Changeux proposed a model of cooperativity of multimeric proteins, the MWC model [142]. Several studies on GPCRs concluded that the MWC model applies to membrane receptors, e.g. GPCRs [141, 143-145]. The CaSR is a membrane receptor, it can form homo- and heterodimers, the monomers interact in a cooperative way, its basal signalling activity can be enhanced by point mutations [146] such as those causing Autosomal Dominant Hypocalcaemia (ADH) [147] suggesting that the MWC model may be a suitable description of the CaSR.
4.2 Ligand biased signalling
The net outcome of a MWC description of the CaSR is showing that its intracellular state - a specific conformation - depends on the extracellular signal status and hereby the extracellular signal is "transduced". The next question is how these conformational changes bring about differential signalling of downsteam G-proteins, which become activated by the guanine exchange factor (GEF) activity of the GPCR. The dynamic balance between the GEF activity and the GTPase then sets the activity of the G-protein that propagates further into the intracellular signalling network.
The CaSR exhibits "ligand biased signalling" or "biased agonism", a phenomenon common to GPCRs [148], indicating that distinct ligands or modulators bind preferentially to a specific state of the GPCR and stabilize it. Thereby, ligands induce a bias in signalling towards a specific pathway at the expense of others. By investigating the influence of several modulators on CaSR-associated signalling pathways, it has been shown that the allosteric modulators cinacalcet, NPS-R568 and the inhibitor NPS-2143 induce a preference towards intracellular Ca2+ mobilization rather than the phosphorylation of ERK1/2 as a response of the activated receptor (Figure 3) [137]. The evidence for biased signalling of the CaSR through different orthosteric agonists such as divalent cations, polyamines and aminoglycosides, is increasing. For example, Thomsen et al. showed that barium (Ba2+) biases towards Gi/o signalling in comparison to its potency to activate Gq/11 signalling. Further, the authors describe a general trend of polyamines and aminoglycosides to influence CaSR signalling towards ERK1/2 phosphorylation compared to Gq/11 signalling [149]. The MWC model allows for a molecular explanation of CaSR ligand-biased signalling.
A changed pattern of information flow through a signalling network is one of the defining features of cancer; as an outcome the balance between apoptosis, differentiation and proliferation is disturbed. Alterations in ligand-biased signalling because of mutations in GPCRs could be one of the associated mechanisms. In addition, ligand biased signalling could be used in chemotherapy to enhance the therapeutic effect of drugs, enabling lower dosage and thus reducing side-effects. Indeed, calcium can enhance sensitivity of colon cancer cell lines to cytotoxic drugs [150]. Development of biased CaSR ligands that are more selective for specific signalling pathways could widen the field of utilization of calcimimetics and calcilytics [137]. These examples illustrate the importance of quantitative approaches to GPCRs. Moreover, mathematical models of proteins can be a useful tool for drug design together with protein structural modelling and the analysis of mutated receptors.
Figure 3: Physiological ligand biased signalling of the CaSR.
The receptor conformation is dependent on the amount and type of ligands bound. Different effector molecules bind preferentially to the different receptor states. This results in differential activation of the downstream signalling pathways. Arrow thickness correlates with the direction of the bias signaling. Adapted from [151].
5. The CaSR as a Potential Cancer Drug Target
The role of CaSR agonists and antagonists should certainly encompass actions on tumour prevention or progression. However, at the moment data on the calcimimetic cinacalcet are missing for cancer, except parathyroid neoplasia.
As described above, treatment with cinacalcet was shown to reverse parathyroid gland hyperplasia in a rodent model of chronic kidney disease [42]. Furthermore, cinacalcet has previously been shown to be active in a murine model of primary hyperparathyroidism, which exhibited reduced CaSR expression in parathyroid glands [152]. These improvements may be, in part, due to calcimimetics promoting CaSR expression, as these disorders are characterised by decreased CaSR expression. In nephrectomised rats, for example, the calcimimetic NPS R-568 reversed reductions in CaSR mRNA and protein expression induced by a high-phosphorous diet [43], while the calcimimetic AMG 641 up-regulated CaSR and vitamin D receptor mRNA expression in uremic rats [44].
In addition, the CaSR seems to have a role in chemotherapy. Recent studies have shown that CaSR signalling regulates the expression of thymidylate synthase and survivin and intensifies the effect of 5-fluorouracil, one of the drugs of choice in colon cancer chemotherapy [153, 154]. Furthermore, in breast cancer, knocking down the tumour suppressor gene BRAC1 leads to a down regulation of CaSR expression and, consequently, an up regulation of survivin which reduced sensitivity to paclitaxel, a mitotic inhibitor used in chemotherapy [155].
Targeting the CaSR could therefore be important, not only for finding new therapeutic avenues to prevent/delay malignant transformation, but also to improve the efficiency of current treatments. However, there a number of important considerations that must be taken into account:
The CaSR is widely expressed in many different cell types and tissues throughout the body, therefore it may be important to develop both drugs and delivery forms that only reach the CaSR in a tissue-specific way. This is an important consideration, as off-target effects of potential CaSR-based therapies could be detrimental to patient health. Currently, the class of drugs used for CaSR-based therapy, the phenylalkyamines, are themselves partially selective for the parathyroid CaSR, over the CaSR expressed in the thyroid C cells [156]. Therefore, selective targeting of the CaSR in other organs, such as the breast, colon and prostate, is important.
While Ca2+o has long been considered the main physiological agonist of the CaSR, it is now becoming apparent that a number of different molecules - including polyamines [157], L-amino acids [158, 159] and γ-glutamyl peptides [160] - are in fact primary ligands in some physiological settings. These changes in primary ligands are important, as the different ligands lead to preferential stimulation of diverse signalling pathways, through 'ligand-biased signalling'. Consequently, identification and targeting of different CaSR ligands for specific tissues may also be an important part of the drug development process.
Finally, the CaSR is a pleiotropic GPCR and readily couples to at least three functionally diverse groups of heterotrimeric G-proteins. Furthermore, as mentioned previously, the CaSR is also reported to switch G-protein coupling from Gαi/0 to Gαs in certain pathophysiological situations, including breast cancer [33]. Therefore, potential novel therapies may will most likely need to target specific G-protein usage.
These above considerations underlie the vast array and complexities of CaSR-mediated signalling and highlight the potential difficulties in developing novel chemotherapeutics targeting the CaSR. A greater understanding of these biological networks and the molecular interaction that take place, aided by a systems biology approach, may be essential in helping to address these issues.
6. Conclusions
Epidemiological and experimental evidence suggests that the CaSR might play a role in tumour progression, however unequivocal evidence for a direct link between the CaSR and tumourgenesis is still lacking. CaSR plays a yin and yang role in cancer/cell physiology because it transduces both anti- and pro-proliferative signals, and inhibits or facilitates cell migration (Figure 4). The methodology of systems biology is an excellent tool to help understand these complex signalling networks. Using the ligand-biased signalling properties of the CaSR could be instrumental in designing specific therapeutics in different pathophysiological conditions.
Figure 4: CaSR in cancer: a yin-yang situation?
In cancer, the CaSR has a multi-faceted role which can be represented by the yin-yang model, either acting as a tumour suppressor (green) or as an oncogene (red). Targeting the CaSR holds significant therapeutic potential depending on whether its expression is lost, as seen in colon and parathyroid cancers, or a gain of function takes place as in breast, prostate, ovary and glioma cancers (not reviewed here).
7. Acknowledgements
