First Published in R&D Systems' 2007 Catalog
Elementary physiology and biochemistry texts generally do not have sections dedicated to inorganic elements or metals. This is not to say that there is no interest in bioinorganic chemistry. It is simply to say that detailed discussions related to trace elements and other metals are usually confined to specialized texts addressing bioinorganic chemistry.1 Aside from the alkaline earth cations sodium and potassium, and the halogen anions chlorine and iodine, there is certainly no shortage of candidates for targeted discussion: iron, copper, zinc and magnesium, cobalt and nickel, manganese, vanadium and chromium, and the ever-more popular selenium.1 Notably, two of the most important inorganic elements, calcium (Ca2+) and phosphorus, rarely rate their own section, even though they occur in considerable abundance. Ca2+ represents anywhere from 1-2% of total body weight (or 1.0-1.2 Kg), while phosphorus constitutes approximately half that at 1% (~0.5 Kg) of total body weight (based on a 73 Kg human).2 Although almost all Ca2+ (99%) and phosphorus (85%) exist as a complex within bone, each has a number of critical, often unrelated, nonstructural functions related to cell homeostasis. The importance of both Ca2+ and phosphorus in the body is reflected in the exquisite hormonal regulation associated with each element. An exciting number of advances have recently been made that detail the regulation of these two elements. As a result, this review will discuss new information related to metabolism (absorption, transportation, storage), rather than the function of these two elements. A number of excellent reviews on the topic have recently been published. 3, 4, 5, 6, 7, 8, 9, 10
Both Ca2+ and phosphorus (as "phosphate" or Pi/HPO4-2) utilize passive diffusion and saturable/facilitated absorption during intestinal uptake.2, 3, 9, 11 In general, Ca2+ uptake is passive (or paracellular) while Pi uptake is saturable (between cells or transcellular).2 Although facilitated (active/transcellular) Ca2+ absorption will increase with reduced Ca2+ intake, facilitated uptake would always seem to predominate in the Pi system.2 It should be noted, however, that contradictory opinions exist on the extent of passive diffusion that occurs during intestinal Pi uptake.9
In the case of Ca2+, a typical daily diet will contain 1000 mg of elemental Ca2+. Approximately 200 mg will be absorbed and 800 mg excreted.3, 9 Ca2+ is absorbed throughout the intestine, but the low luminal pH (5 to 6) associated with the duodenum and jejunum promotes the ionization of Ca2+ and its efficient absorption.2 When the paracellular route is utilized (generally in jejunum/ileum), Ca2+ exit from the intestinal lumen is likely driven by a transepithelial electrochemical gradient. Although most studies on Ca2+ diffusion have been done on renal epithelium, it is not unrealistic to assume a similar process also occurs in the intestine.3, 12, 13 As such, the gradient will be generated by the action of brush border NKCC2 (Na-K-Cl-2 transporter) and ROMK (ATP-sensitive K+ channel), and basolateral CLCNKB (Cl- channel) and the Na/K ATPase. 14, 15, 16, 17 Na+, K+, and 2 Cl- ions first leave the intestinal lumen (entering the cell) via NKCC2. This is offset by a flux of K+ via ROMK. On the basal side Na+ leaves the cell, and K+ enters the cell via Na/K ATPase, while Cl- leaves the cell via CLCNKB channels. On balance, this creates an electrochemical gradient that drives Ca2+ out of the lumen and into the extracellular fluid via a paracellular route. Although paracellular diffusion can be pictured as simple diffusion between cells, this is likely not the case. As would be expected for the epithelium, tight junctions exist between cells, linking them physically and creating a barrier to free intercellular diffusion. In kidney, a unique junctional complex molecule has been found that regulates both Mg2+ and Ca2+ transit through tight junctions. Termed paracellin-1, it is a member of the claudin superfamily of junctional molecules. 3, 18, 19, 20 It forms a trimeric complex with occludin and JAM (junctional adhesion molecule). Although this claudin is restricted to kidney, other claudins, such as claudin-2, -15 and -20, are found in small intestine and likely serve as regulators of divalent cation transit.20, 21
View Larger Image
|Figure 1. A: Transcellular Ca2+ transport in the intestine (duodenum/jejunum). Ca2+ enters the cells via TRPV6 channels, and Ca2+-binding proteins such as calbindin D9K facilitate Ca2+ transport to the basolateral membrane. Ca2+ can then be extruded by mechanisms that include the NCX1 Na+/Ca2+ exchanger and/or the Ca2+ ATPase termed PMCA1b. VitD may upregulate the expression of both TRPV6 and PMCA1b. B: Pi transport in the intestine (primarily in jejunum). Some passive paracellular Pi transport may occur, although a transcellular mechanism is likely the primary means of Pi transport. Pi enters the cell via the luminal Na+/Pi cotransporter, NPT2b, and leaves the cell by an as yet undiscovered mechanism. VitD may stimulate NPT2b transcription and/or affect transporter activity.|
Facilitated, or transcellular Ca2+ transport (in duodenum/jejunum) actually utilizes a different set of ion transporters/channels (Figure 1A). Again, and principally based on renal epithelial studies, there are both luminal and basolateral transporters/channels.3, 11, 12 The hallmark of transcellular transport is its sensitivity to levels of active 1,25(OH)2 vitamin D3 (abbreviated in this article as VitD). On the luminal side there is a transient receptor potential-vanilloid 6 channel (TRPV6). This is a six-transmembrane domain channel protein that exists as either a homotetramer, or heterotetramer with TRPV5. It is upregulated by low cytosolic Ca2+, opens in the presence of low luminal Ca2+, is inactivated by both phosphorylation and a Ca2+-calmodulin complex, and its levels are increased in response to VitD. 3, 22, 23, 24 Ca2+ enters the cell due to an electrochemical gradient. Once inside, its transit through the cell is mediated by Ca2+-binding proteins (calbindins) whose synthesis is upregulated by VitD. In the intestine, calbindin-D9K is used, which has two Ca2+-binding sites. Calbindin-D9K delivers Ca2+ to one of two basolateral transporters/pumps, 120 kDa NCX1 (a Ca2+-Na+ exchanger) or 138 kDa PMCA1b (plasma membrane Ca2+-dependent ATPase). 9, 12, 25, 26, 27 The NCX1 exchanger internalizes three Na+ for one Ca2+, while PMCA1b pumps Ca2+ out at the expense of ATP. In the intestine, PMCA1b represents the dominant pathway for Ca2+ extrusion.11 As with TRPV6, PMCA1b gene expression is positively regulated by VitD. PMCA1b is also positively regulated by estrogen, which may be critically important during pregnancy and lactation.28, 29
Three things should be noted about the TRPV system. First, in the intestine, both TRPV6 and TRPV5 exist. It is suggested however, that TRPV6 predominates in intestine, while TRPV5 predominates in kidney.3, 12 Second, in kidney, TRPV5 expression is regulated by klotho, a transmembrane, multifunctional protein that exhibits beta-glucuronidase activity. Renal TRPV5 is glycosylated and constitutively active. De-glycosylation by membrane-associated klotho blocks TRPV5 turnover, retaining TRPV5 in the membrane and prolonging its activity. It is not known if the same phenomenon occurs in intestine, but this would seem a possibility.30 Finally, in kidney, parathyroid hormone (PTH) is known to positively regulate all of the molecules involved in cellular Ca2+ transport. The exact effects of PTH on intestinal molecules is uncertain, but presumably parallel those in kidney.9, 31
Phosphorus is remarkably abundant and derived from natural sources such as dairy products, cereals and meat, and unnatural sources such as carbonated beverages. Total daily intake varies, depending on the study, but a representative range is 1000 mg (in women) to 1500 mg (in men). 2, 9, 32, 33, 34, 35, 36 Approximately 70% of dietary phosphorus is absorbed, principally in the jejunum. Again this occurs through one of two ways; a passive intercellular route and a facilitated transport intracellular route (Figure 1B).2, 5, 9 Phosphorus absorption is described as being minimally regulated.34, 37, 38 At issue is whether most of the absorption is passive or facilitated.
Paracellular/intercellular transport is favored electrochemically because the phosphorus concentration of the intestinal lumen exceeds that of the extracellular fluid underlying the epithelium, and the extracellular fluid is electropositive relative to the intestinal lumen. In counterpoint, intercellular junctions are highly impermeable to phosphate ions, and this seems to be the overriding consideration (Figure 1B).2 Thus, on balance, paracellular transport appears to contribute only modestly to phosphate absorption.2
Transcellular/facilitated diffusion is a function of at least three components; a luminal Na+/Pi co-transporter, a basolateral Na+/K+ ATPase, and an as yet unidentified, but hypothesized, basolateral Pi transporter (Figure 1B). The Na+/Pi transporter (or NPT2b) is an 80 kDa, 8-transmembrane domain protein that simultaneously transports one Na+ and one Pi ion into the cell.39 40 Once internalized, Pi exits the cell on the basolateral side. NPT2b is reported to be positively regulated by VitD, and NPT2b likely responds to differences in intracellular Na+ concentration.38 The exact role VitD plays is not clear. While it seems to increase NPT2b expression or "activity," it must act at a posttranscriptional level given the NPT2b gene has no VitD response element.2, 38 It may indirectly impact NPT2b by promoting the activity of the Na+/K+ ATPase, reducing intracellular Na+ and pulling in luminal Na+ accompanied by Pi.2 Another molecule reported to increase Pi uptake is STC-1/stanniocalcin-1. STC-1 is a secreted, dimeric phosphoglycoprotein that is made in kidney. The mechanism of uptake is unclear, although it undoubtedly involves recently discovered STC receptor(s). 41, 42, 43 STC-2, a related protein, has been shown to regulate NPT2a expression in the kidney.44
Non-hormonal factors that impact both Ca2+ and Pi absorption include high luminal levels of both Ca2+ and Pi that generate insoluble CaHPO4 complexes, and the use of antacids that contain aluminum, which renders Pi unabsorbable.2, 9, 45
Once absorbed, both Ca2+ and phosphorus circulate in multiple forms. Approximately 50% of serum Ca2+ is freely ionized, while 45% is bound to protein and 5% exists in poorly-defined complexes.2 Phosphorus is either inorganic (30%) or organic (70%). Of the 30% inorganic phosphorus, 10% is ionically bound to protein (and thus is not filtered by the kidney), while 90% is ionic and freely filtered by the kidney. Within the "ionic 90%," approximately 5% exists as a divalent phosphate salt (Mg or Ca), 30% exists as a Na+ salt, and 65% is free phosphate ion. Approximately 80% of free phosphate ion is HPO4-2, while 20% is H2PO4-1; thus, the designation "Pi" is generally taken to mean HPO4-2.2, 3, 46 Organic phosphorus may take many forms including phospholipids, phosphate esters, phosphoproteins, phosphonucleotides, etc.46
View Larger Image
|Figure 2. Two to four hours of direct sun per week on either the arms or face is enough to ensure the UV-B exposure required for the production of inactive VitD. Once generated in the skin, inactive VitD associates with VitD-binding protein and this complex is taken up by hepatocytes where it is hydroxylated at position #25. The complex is further transported to the kidney where it is filtered and becomes activated by hydroxylation at position #1.|
A number of hormones circulate that impact Ca2+ and Pi metabolism. The first is 1,25(OH)2 vitamin D3 (VitD). The precursor for VitD, 7-dehydrocholesterol, occurs naturally in basal keratinocytes. It is the last step in the synthesis of cholesterol (Figure 2). 7-dehydrocholesterol (7DHC; also known as provitamin D), in the presence of sterol D7-reductase, forms cholesterol. Following exposure to UV-B radiation (290-319 nm), 7DHC is cleaved in its B ring and undergoes spontaneous isomerization to form vitamin D3. This is bio-inactive, but will bind to endothelia-produced, 53 kDa vitamin D-binding protein (VDBP). 7, 47, 48, 49 Each VDBP-vitamin D3 complex transits first to the liver, where -OH is added at position #25, and then to the kidney, where a second -OH is added at position #1. While 1,25 (OH)2 vitamin D3/VitD is considered the active form, 25(OH) vitamin D3 is also suggested to have select bioactivity, particularly in promoting Ca2+ uptake in intestine.50 The second, or position #1 hydroxylation, is performed by 1 alpha-hydroxylase, an enzyme that shows considerable regulation by a number of factors. PTH, IFN-gamma, and IGF-I increase 1 alpha-hydroxylase activity, while Ca2+, Pi and klotho depress 1 alpha-hydroxylase activity.47
The most notable aspect of the second, or kidney-based hydroxylation, is the circuitous route taken by the VDBP-25(OH) vitamin D3 complex. Rather than binding to basolateral VDBP receptor(s) on proximal tubule cells, it is first filtered through the glomerulus, and then binds to luminal, 550 kDa megalin on proximal tubule cells. This induces internalization with apparent complex dissociation.3, 47 Newly freed 25(OH) vitamin D3 now binds to a new intracellular VDBP, termed IDBP-1, which directs it to mitochondrial CYP1/1 alpha-hydroxylase. 3, 47, 51, 52, 53 Following its formation, VitD (1,25(OH)2 vitamin D3) diffuses freely out of the cell to interact with either the 48 kDa vitamin D receptor/VDR or an incompletely characterized 60 kDa membrane receptor that induces rapid, non-transcriptional responses in cells.47, 54 Although VitD has significant effects on bone and renal metabolism (see below), it also likely has the notable effect of down-regulating its own activity. It does so by up-regulating the activity of 24-hydroxylase, an enzyme that replaces the hydroxide at the #1 position with a hydroxide at position #24, inactivating the vitamin.3, 47
What then is the take-home-message for VitD? Simply put, it maintains serum Ca2+ concentrations in the normal range. How? Basically by up-regulating elements associated with the intestinal Ca2+ absorption process such as TRPV6 and calbindin-D9K. It does play a role in immunity, reproduction, and phosphate metabolism, and it does have a complex relationship with other crucial hormones associated with bone metabolism. But, in summary, its target is intestinal.3, 7, 47
It is said that four hours of intense sun exposure per week on either the face or upper extremeties will generate adequate vitamin D3 levels. During winter, or under sunless conditions, either nutritional supplements, fatty fish, or also fortified milk are required to supply needed vitamin D3. Dietary vitamin D3 is absorbed by the gut, transported to the liver by chylomicrons, and either stored in fat or converted to 25(OH) vitamin D3. Nutritional supplements may contain either vitamin D2 or vitamin D3. The difference is only in the source (D2/ergocalciferol from plants; D3/calciferol from animals). Both are convertible into active 1,25(OH)2 vitamin D. Although vitamin D supplementation is often recommended for "healthy bones," some studies strongly recommend a dual supplement composed of vitamin D2/3 and vitamin K. As will be shown later, adequate Ca2+ absorption is only part of the story. It must also be successfully incorporated into bone mineral, a process strongly impacted by vitamin K.47, 55
PTH, or parathyroid hormone, is a 9.4 kDa polypeptide product of the Chief cells of the parathyroid gland (Figure 3). In contrast to VitD, which insures adequate total body Ca2+ stores, PTH regulates the distribution of total body Ca2+.56 Its release results in a rapid mobilization of Ca2+ from bone. As such, it is the principal regulator of minute-to-minute circulating Ca2+ levels, and its secretion is quite sensitive to prevailing Ca2+ concentration. The receptor for circulating Ca2+ is a 140 kDa, 7-transmembrane domain receptor (CaSR) that, when activated, represses PTH release from Chief cells. 56, 57, 58, 59 The CaSR is suggested to easily detect a 200 µM fluctuation in extracellular Ca2+. When circulating levels fall below a threshold, CaSR signaling is reduced and PTH is released.56, 60, 61
View Larger Image
|Figure 3. PTH is produced by the dark Chief cells of the parathyroid. Lowered extracellular Ca2+ removes CaSR-dependent repression of PTH production. PTH is then free to mobilize Ca2+ from stores in bone. PTH production may also be regulated by Pi.|
When released, PTH would seem to exist in a bewildering number of isoforms. It is initially synthesized as a 115 amino acid (aa) prepropeptide that contains a 25 aa signal sequence and a six aa N-terminal pro-segment. The C-terminal 84 aa make up the mature, circulating form of PTH.56, 62 Only the first 34 aa of the mature polypeptide are necessary for bioactivity, and this fact serves as the basis for PTH pharmacological analogs.56 Normally, 20% of circulating PTH is full-length (aa # 1-84), while 80% shows some N-terminal truncation (C-PTH; note: abbreviations vary among authors, and C-PTH here will mean any form not full-length). There are fragments that start at aa position # 4, 7, 8, 10, 15, 34 35, 37, 41, and 43, and perhaps more differing at the C-terminus. 56, 61, 63, 64, 65 All seem to be targets of proteases such as cathepsins.56 Some are generated by Chief cells and some by hepatocytes. Cleavage of the first six aa appears to render the molecules inactive toward the PTH receptor (PTH1R). Remarkably, N-terminally truncated PTH molecules seem to have their own receptor, currently referred to as C-PTHR. It has yet to be characterized. C-PTHs often show activity antagonistic (or anti-calcemic) to that of PTH. The ratio of full-length to truncated forms varies with the ambient level of Ca2+. At low Ca2+ concentrations, additional Ca2+ is needed from mineral stores and full-length PTH represents 30%-40% of total PTH (~18 pM). By contrast, under high Ca2+ conditions, total PTH falls to 5 pM and full-length PTH only represents 5% of this amount.65, 66 In addition to Ca2+-induced variability, PTH shows a circadian rhythm. There is a 30% baseline difference between peak (10 PM-3 AM) and trough (10 AM-Noon) release.67 Notably, osteoporosis patients seem to lose this rhythm.
PTH1R/PTHR1, the receptor for PTH, is a 7-transmembrane domain G protein-coupled receptor (GPCR) found on select cell types, including osteoblasts, osteoclasts, hematopoietic stem cells, and renal tubule cells (proximal and distal epithelium). 56, 65, 68, 69, 70, 71, 72 As noted above, there is also a hypothesized receptor for C-PTH. This receptor is apparently highly expressed by osteoblasts and osteocytes, and when ligated increases intracellular Ca2+ but not cAMP.61, 63 Functional outcomes attributed to C-PTHR include a decrease in circulating Ca2+ and Pi, the promotion of bone formation, and an increase in osteoclast formation and activity.
PTH, by definition, is a normo-calcemic hormone. That is, it exists to maintain blood/extracellular fluid Ca2+ levels within a narrow range. It does so by inducing Ca2+ release from bone, reducing Ca2+ loss through the urine, and promoting VitD production through the upregulation of renal 1a-hydroxylase. It has an indirect, but important effect on phosphorus. By promoting phosphorus excretion and thereby reducing overall phosphorus load, PTH fulfills its principle function, facilitation of Ca2+ release from bone.56
Fibroblast growth factor-23 (FGF-23) is the newest member of a diverse and large FGF family of proteins. As with other members, FGF-23 shows a typical beta-trefoil structure. Unlike most members, however, FGF-23 contains a signal sequence, an atypical intrachain disulfide bond, and an extended C-terminus/pro-segment. This places it in the small FGF-19 subfamily.73, 74 FGF-23 is a 30 kDa, secreted glycoprotein that undergoes post-translational processing.75, 76 Following removal of the signal sequence and pro-segment between Arg179 and Ser180, mature FGF-23 is generated that is 155 aa in length. This mature form, however, is bio-inactive; it appears that the C-terminal pro-segment is essential for bioactivity.77 It is believed to be synthesized by osteoblasts in response to VitD.78, 79 Although its receptor was unknown for some time, it now appears to bind to FGF R1c, 2c, 3c, and FGF R4.80, 81 It is reported that klotho, with its associated carbohydrate moiety, is likely to be the physiological co-receptor for FGF-23.81
Full-length FGF-23 is considered to be a phosphatonin.82 Among other things, phosphatonins decrease plasma phosphate by promoting phosphate excretion. They are considered analogous to calcitonin, which decreases serum Ca2+ levels.83 FGF-23 has two principal actions: promoting phosphate excretion in the urine and suppressing VitD synthesis.73, 74, 77 Its effect on phosphate resorption is mediated by its ability to downregulate phosphorus transporters on the luminal side of renal epithelium. 34, 78, 84, 85, 86 Its effect on VitD synthesis is mediated by blocking 1a-hydroxylase activity in kidney. 37, 73, 78, 87
The relationships between VitD, PTH, and FGF-23 are complex and perhaps not intuitive. It could be said that PTH regulates the Ca2+-VitD axis, while FGF-23 regulates the Pi-VitD axis. One model suggests that under conditions of low circulating/extracellular fluid Ca2+ (a simple Ca2+ deficiency), PTH is released from Chief cells. This withdraws Ca2+ from the bone for a short-term effect (see later section on mineralization). It also induces 1 alpha-hydroxylase activity in kidney to create active VitD. VitD does two important things. First, it promotes Ca2+ absorption, leading to an increase in total body stores. Second, it increases intestinal phosphorus absorption, presumably for the purpose of providing the mineral counterpart to Ca2+ needed during mineralization (hydroxyapatite/Ca2+10(PO4)6(OH)2). Under the conditions of simple Ca2+ deficiency, there is no concomitant phosphorus deficit. With increased VitD activity, however, phosphorus is now in excess. PTH, in the short term, can influence Pi excretion in a manner identical to that of FGF-23; that is, promote excretion rather than reabsorption. On balance, this should eliminate excess phosphorus, but with an increase in circulating Ca2+ due to VitD-mediated absorption, we know PTH release comes to a halt due to the Ca2+-CaSR actions on Chief cells. What is needed longer-term is an additional phosphatonin that will bring phosphorus levels back to normal. That molecule is FGF-23 (and perhaps MEPE and/or sFRP-4).78 In order to "catch up" with continuous Ca2+ and Pi absorption, FGF-23 will downregulate 1a-hydroxylase activity and upregulate 24-hydroxylase activity, thereby removing the stimulus for excess phosphorus uptake.88
Although FGF-23 seems to be made by osteoblasts, what induces its expression? It would appear that VitD induces its expression.88 It has also been suggested that circulating phosphorus drives FGF-23 release, particularly since FGF-23 is supposed to protect against excess phosphorus; i.e.-stimulus-response.89 Indeed, extracellular Pi by itself may independently upregulate a number of genes, including beta5 integrin, STAT5, and osteopontin, and an upregulation of the FGF-23 gene would be consistent with this observation.90, 91 However, in healthy subjects phosphorus intake does not influence FGF-23 levels. High phosphorus diets in renally uncompromised individuals does lead to increased excretion, but without a change in FGF-23.92 Curiously, in rats a high phosphorus diet does lead to increased serum phosphorus and a subsequent increase in FGF-23. In mice, dietary phosphorus is also reported to affect FGF-23 synthesis. Thus, the system may show some species specificity.93
Calcitonin is a 3 kDa, 32 aa peptide that belongs to the calcitonin gene-related peptide family. It is made by C cells of the thyroid gland purportedly in response to elevated blood Ca2+ levels.83, 94 It is synthesized as a 141 aa preproprecursor that is processed into a 32 aa mature peptide. Calcitonin has a potent inhibitory action on osteoclasts mediated by its GPCR, termed the calcitonin R. The molecule seems to have an ontogenic component to it, as it is highly active in the young of species and loses its potency with age. In adult humans, it may act as a stress-related molecule.
sFRP-4 is a member of a small family of secreted proteins that structurally resemble the extracellular domain of the frizzled family of receptors. 4, 95, 96, 97 The mature sFRP-4 molecule is 328 aa in length, contains a 120 aa frizzled/cysteine-rich domain, and a 100 aa netrin-like region. In rat, the molecule is highly spliced, with variants occurring at the C-terminus.98 Similar variants may exist in human. sFRP-4, like other sFRPs, binds to both Wnt ligands and frizzled receptors-1 and -4, and the family is generally acknowledged to be inhibitory to Wnt signaling. 99, 100, 101 There may be additional activities. For instance, sFRP1 is reported to bind to RANK L, an inducer of osteoclast formation.101
Like FGF-23, sFRP-4 is reportedly expressed by osteoblasts and unidentified cells in kidney.101, 102 sFRP-4 is also considered to be a phosphatonin.33 FGF-23 induces internalization of renal phosphate transporters, likely through FGF R signaling. sFRP-4, however, seems to antagonize Wnt signaling, not initiate it. Although highly conjectural, there are at least three possible mechanisms of action. First, FGF-23, when present as a full-length molecule, shows phosphatonin activity. When cleaved into mature N- and C-termini, it loses its activity.103 It has been suggested that a matrix-metalloproteinase (MMP)-like molecule is responsible for cleavage, and MMP-inhibition by a tissue inhibitor of metalloproteinase (TIMP) would ensure continuing FGF-23 activity. sFRP-4 has a TIMP-like domain and has the theoretical potential to neutralize MMPs, thereby, guaranteeing the integrity of FGF-23 and promoting its phosphatonin activity. It should be noted that furin-type convertases have also been reported to cleave FGF-23.104 Second, an absence of Wnt signaling in osteoblasts can lead to apoptosis. Since osteoblasts initiate bone formation, a reduction in osteoblast number would translate into a reduction in mineralization rate that would be accompanied by a reduced need for phosphate. This would translate into a reduction in kidney Pi resorption due to reduced demand. Finally, sFRP-4 inhibits VitD production via 1a-hydroxylase.10, 36, 77 Since VitD is associated with increased Pi uptake, removal of the VitD effect would translate into increased Pi excretion. 36, 105
MEPE is a third phosphatonin. It is a 45-65 kDa, secreted glycoprotein that belongs to the SIBLING (short integrin-binding ligand interacting glycoprotein) family of molecules. 106, 107, 108 The mature molecule is 508 aa in length, serine-rich, contains an RGD motif for cell attachment, and a C-terminal Ser-Asp-Gly-Asp motif associated with glycosaminoglycans. MEPE is synthesized by osteoblasts and osteocytes, particularly during mineralization.109 It is associated with the extracellular matrix (ECM) and with PHEX, a type II transmembrane metalloproteinase on the surface of osteoblasts. 110, 111, 112 PHEX has no proteolytic activity on MEPE, but instead protects MEPE from cleavage by cysteine-proteases such as cathepsin B.113 This is likely due to PHEX acting as a pseudosubstrate for cathepsin(s). When cathepsins have functional access to MEPE, they cleave the molecule between Arg507 and Asp508, generating an 18 aa C-terminal peptide called ASARM.106, 108 This peptide seems to perform two functions; an inhibition of mineralization and the promotion of urinary phosphate loss. The regulation of mineralization may be the principal function for MEPE (or its cleavage product). The phosphatonin effect, while material, may be complementary except under unregulated conditions.
The resorptive mechanism for urinary Ca2+ reflects that found in the intestinal enterocyte. In particular, both paracellular and transcellular processes are found to exist. Of the Ca2+ found in blood, 40% is bound to albumin, 10% is associated with organic ions, and 50% is free ion and has the ability to bind to CaSR.9 Albumin is not filtered by the glomerulus so this Ca2+ fraction is unavailable for resorption. The remaining fractions provide approximately 9000 mg of Ca2+ per day, all of which is resorbed, save for some 200 mg. In the proximal tubule 70% of the glomerular filtrate is resorbed, 20% is resorbed in the thick ascending loop of Henle, and 10% is resorbed in the distal convoluted tubule and collecting duct.9 Almost all paracellular resorption occurs in the proximal tubule and thick ascending limb. Transcellular (hormone-sensitive) resorption occurs in the distal segments. 2, 3, 9, 12, 13, 114
View Larger Image
|Figure 4. A: Paracellular Ca2+ uptake in the proximal tubule is selectively regulated by the junctional protein paracellin-1 and alterations in the transepithelial electrochemical gradient. High basolateral Ca2+ levels trigger the CaSR and suppress ion movement through ROMK K+ channels and NKCC2 transporters. This decreases the driving force for Ca2+ down its electrochemical gradient, reducing paracellular Ca2+ flux. B: Transcellular movement of Ca2+ predominates in the distal tubule. Ca2+ enters through the TRPV5 Ca2+ channel and transport to the basolateral membrane is facilitated by calbindin D28K. Here, Ca2+ may be extruded by the Ca2+ ATPase (PMCA1b) and/or the Na+/Ca2+ exchanger (NCX1). Klotho may enhance the activity of TRPV5 channels by its apparent ability to remove TRPV5 glycosylation, retaining TRPV5 in the membrane and ensuring continuous Ca2+ transport. Both PTH and VitD may also regulate the levels and/or activity of PMCA1b, calbindin D28K, and NCX1.|
Passive uptake in the proximal and straight tubules presumably involves paracellin-1 and the same ion transport mechanisms that exist in enterocytes. These include luminal NKCC2 transporters and ROMK channels, and basolateral Na/K ATPases and CLCNKB (Cl- channels; Figure 4A).2, 3, 114 Paracellular transport in the ascending loop of Henle also involves CaSR. This receptor is found on the basolateral membrane. When extracellular fluid Ca2+ levels are sufficiently high, the receptor is activated and ROMK is inhibited. This reduces the level of K+ in the lumen, causes distortion of the normal electrochemical gradient, and interrupts the paracellular flow of Ca2+.2, 3, 13
In the distal tubule, hormone-sensitive transcellular absorption predominates (Figure 4B). As in the intestine, there is a TRPV Ca2+ channel (TRPV5), a basolateral NCX1 Na/Ca exchanger, and a PMCA1 Ca2+ pump.2, 9, 12 Instead of calbindinD9K, calbindin-D28K is used for intracellular Ca2+ transport. A number of molecules impact Ca2+ uptake. When blood Ca2+ levels drop, the Chief cell CaSR is inactivated and PTH is released. Subsequent binding of PTH to PTH1R on the basolateral surface of distal tubule cells increases three things: the levels of calbindin-D28K and NCX1, and the affinity of PMCA1 for Ca2+.2 Opinions differ on whether TRPV5 is a target of PTH.2, 56 However, TRPV5 is a target for klotho.30 Klotho has both soluble and membrane forms and two glycosyl hydrolase type-1 regions that lack a critical Glu in the "active" site. 115, 116, 117 Nevertheless, the transmembrane form appears to interact with TRPV5 in the cell membrane, and apparently removes some TRPV5 glycosylation. This has the effect of trapping it in the luminal membrane, and forcing it to continue to transport Ca2+.30, 118, 119 The role VitD plays is not clear. It is suggested that calbindin D and the ion transporters/channels involved in Ca2+ flow are not particularly sensitive to VitD action. More study needs to be done on this point.
In the collecting duct, CaSRs closely monitor the level of Ca2+ in the urine. When levels are high due to a reduction in resorption, Ca2+ activates collecting duct CaSRs. This blocks the effects of ADH/vasopressin-mediated water uptake via resident aquaporin channels on collecting duct epithelium. The urine becomes dilute, decreases the mM concentration of Ca2+, and precludes kidney stone formation.3, 120
Circulating and extracellular phosphate levels are essentially a function of renal resorption.2, 34, 37 Circulating Pi is freely filtered by the glomerulus, and 85%-90% of it is resorbed.5, 38 The numbers are impressive. Each day, approximately 7000 mg of Pi enter the glomerular filtrate, while 6000-6300 mg of Pi are recaptured. As with Ca2+, the mechanism for renal Pi resorption parallels that for intestinal Pi absorption.
View Larger Image
|Figure 5. Pi resorption in the proximal tubule is primarily dependent upon the activity of the Na+/Pi transporter, NPT2a. Several mechanisms exist that regulate NPT2a expression and/or activity. VitD upregulates NPT2a expression, and although not yet definitively shown, STC-1 may as well. PTH acting via PTH1R, and sFRP-4 acting via an unclear mechanism may stimulate NPT2a internalization. FGF-23 has also been reported to downregulate NPT2a, possibly via internalization and/or through reduction of transporter expression.|
In the nephron, 60%-75% of Pi resorption occurs in the proximal tubule, 15%-20% occurs in the loop of Henle, and 5%-10% occurs in the distal tubule.46 Resorption is principally dependent upon the activity of luminal NPT2a Na/P transporters (Figure 5). In the proximal tubule, Pi must be actively transported from the urine into the cell against a very strong electrochemical gradient. The concentration of Pi in the cell is at least equal to that of Pi in the urine/lumen and there is an inhibitory electronegative gradient.2 Thus, Pi transport requires luminal Na+ as a "co-transportee." There are three gene families of Na/P co-transporters. Type 1 family members are NPT1, 3, and 4. Their transport function is not limited to Na+ and Pi, and their physiological role is unclear. The type 3 family has two members, Pit1 and 2. Part of their function is to maintain needed intracellular phosphate. In the kidney, they exist on the basolateral side and transport Na+ and Pi into the tubule cell.121, 122 Type 2 family members are three in number, each the product of a separate gene.34, 121, 122 They lie in the luminal border of tubular epithelium. NPT2c is a fetal transporter, while NPT2b is localized to the intestine, and NPT2a to the kidney.34 NPT2a is found exclusively in proximal tubule cells. It is an 8 transmembrane domain protein that exhibits complex regulation.34 Resorbed Pi transits the cytosol uncomplexed, and exits via passive, but facilitated transport. The identity of the basolateral transporter is unknown.2
The number of extracellular regulators reflects the importance of NPT2a. With respect to molecules that promote NPT2a activity, VitD is known to upregulate its expression, likely through a transcriptional mechanism. VitD response elements exist in the NPT2a promoter.38 STC-1 (stanniocalcin-1) is also a molecule that promotes Pi uptake. It is a secreted phosphoprotein that is made by collecting duct epithelium.41, 42 In bone, STC-1 mediates osteoblast Pi uptake via an upregulation of Pit-1 (type 3 family) transporters.123 In kidney, it is posited that STC-1 drives the expression of NTP2a. This has yet to be demonstrated.42, 124, 125
There are also phosphatonins and phosphatonin-like molecules that promote the down-regulation of NPT2a. This reduces phosphate uptake and generates phosphaturia (high urinary phosphate). A number of molecules are involved and include FGF-23, sFRP-4, MEPE, and PTH. 2, 33, 46, 102, 105 FGF-23, PTH and sFRP-4 are all reported to impact NPT2a activity. The effects of sFRP-4 are not well characterized, but have been suggested to involve NPT2a removal from the luminal membrane.102 FGF-23 and PTH both reduce the number of luminal transporters. In one scenario, FGF-23 reduces NPT2a protein levels and mRNA generation, while PTH actually induces transporter internalization. In the absence of PTH, existing NPT2a continues to function until its lifespan is reached.88 Notably, megalin, the vitamin D binding protein that internalizes 25(OH) vitamin D3 from urine, also co-expresses with NPT2a and may be a key target for PTH-mediated NPT2a internalization.126 Other scenarios suggest that FGF-23 actually mediates NPT2a internalization.85 The final word has not been written on the issue. MEPE is perhaps the least well understood phosphatonin.105 Although the precise mechanism is unclear, the C-terminal ASARM motif has been found to be phosphaturic.127 Intriguingly, both FGF-23 and MEPE are associated with PHEX, suggesting an unappreciated dynamic between these three molecules.127
Bone formation is initiated under a variety of circumstances. It occurs during development, disease, and normal bone homeostasis. Three types of bone formation are described: intramembranous, perichondrial, and endochondrial.128 For the purpose of this discussion, aspects of all three processes will be integrated into one generalization that may, or may not, always apply. As a practical matter, the first stage of bone formation involves the recruitment/creation/generation of osteoblasts. It is suggested that they are either pre-existing or migrate to sites of osteogenesis, or they form from resident precursors under the influence of various osteoblast inducers. 129, 130, 131, 132 The following is a potential model for general osteogenesis. First, it is assumed that blood/extracellular fluid Ca2+ levels are elevated, and that there is a need for a repository for excess Ca2+. Such a circumstance might well occur during aggressive hormone-mediated Ca2+ absorption. Elevated circulating Ca2+ in bone may act on macrophages and induce BMP-2 production.133 BMP-2 is known to promote a mesenchymal cell-to-osteoblast transition.132,133 If VitD is the active agent promoting hypercalcemia through increased Ca2+ absorption, VitD may also contribute to the generation of pre-osteoblasts. 3, 7, 132, 133, 134 Once formed, pre-osteoblasts will either align themselves along the surface of previous resorption areas, or arrange themselves in a pattern that facilitates mineralization.128, 129 Once placed, pre-osteoblasts will divide, secrete collagen type I and osteonectin, synthesize alkaline phosphatase, and express PTH1R, as well as alphav, alpha5, beta3, and beta5 integrins.128 Collagen I is the predominant organic fiber in bone. It is suggested by many to be the only collagen in bone.135 Alkaline phosphatase is a GPI-linked membrane protein that generates Pi for internal transport. This ultimately complexes with ionic Ca2+ to form a CaP crystal around which hydroxyapatite may form. 137, 138, 139 Osteonectin is a 40 kDa glycoprotein that may modulate (interfere) with mineralization.135, 139
Following the generation of extracellular Pi and the synthesis of collagen, pre-osteoblasts mature into osteoblasts. These cells are characterized by the expression of collagen I, BSP-1 (bone sialoprotein-1), alkaline phosphatase, PTH1R, E11/gp36, and osteocalcin.128 At this stage, BSP-1 and osteocalcin are of particular interest. BSP-1 is an 85 kDa, sulfated glycoprotein that shows an affinity for hydroxyapatite. 6, 129, 135, 140 It represents approximately 10% of non-collagenous protein in bone and is characterized by the presence of multiple glutamic acid residues. These residues, by virtue of their R-COO- anionic form, are able to bind up to 83 Ca2+ ions. Overall, crystal growth and regulation is not well understood, but in the case of BSP-1, Ca2+ binding is believed to anchor BSP-1 to hydroxyapatite and contribute to organized CaPi formation.140, 141 Osteocalcin (or bone Gla protein), by contrast, is a small 6 kDa, 49 aa residue polypeptide with unclear function. It is one of the many vitamin-dependent proteins associated with bone. Vitamin C is necessary for hydroxylation at Pro9, VitD initiates gene expression, and vitamin K is a cofactor for vitamin K-dependent carboxylase. The latter enzyme adds a gamma-carboxyl residue to the molecule.55, 144, 145 This creates two gamma-carboxyl residues that serve as a binding site for exogenous Ca2+. Osteocalcin has been described as both a promoter and inhibitor of mineralization. 143, 144, 145 PTH/PTH1R activity in the osteoblast downregulates osteocalcin expression. Since PTH is associated with bone demineralization, it may be more important in bone formation. Yet low doses of osteocalcin block the transition of CaHPO4â€¢2H2O to Ca10(PO4)6(OH)2 (i.e.-hydroxyapatite), so the issue is undecided.
Once mineralization commences, that is, osteoid is laid down, osteoblasts begin a two phase process for Ca2+ deposition (Figure 6). It starts with the generation of 100 nm matrix vesicles. These are cytoplasmic membrane-bound vesicles that are created near the osteoid-osteoblast interface. These matrix vesicles contain a variety of substances, including annexin V (on the inner membrane surface), alkaline phosphatase (GPI-linked), calbindin-D9K, pyrophosphatases, carbonic anhydrase, AMPase, BSP-1, osteonectin, and osteocalcin. The matrix vesicle membrane is enriched in cholesterol, phosphatidylserine, and sphingomyelin.138, 139 The matrix vesicles are extruded from the osteoblast in an annexin-dependent process where they attach to the underlying matrix.146 The attachment in bone is not well understood. In cartilage, the matrix vesicle attaches to cartilage matrix collagen types II and X. In bone, there appears to be only type I collagen. Both alkaline phosphatase and annexin are reported to bind to type I collagen, and these molecules, if expressed on the outer surface of the matrix vesicle, may provide the anchor for osteoblast-derived matrix vesicles.147 Once externalized, the matrix vesicle undergoes physical changes. Both Ca2+ and Pi are transported into the vesicle. Annexins form channels that transmit Ca2+. This is accompanied by phosphate uptake, which is tied to external Pi generation via alkaline phosphatase activity.136, 146 ,148 Uptake of Pi occurs via NPT3/Pit1 Na/Pi cotransporters. 91, 136, 146, 147, 148, 149 Once internalized, Ca2+ and Pi form rudimentary (immature?) CaP crystal (amorphous CaP). Early formation is predicated upon initial Ca2+ and Pi interaction with phosphatidylserine. 142, 148, 150, 151, 152 The first crystal(s) formed is referred to as octa-CaPi-like mineral and does not yet qualify as organized hydroxyapatite.150, 151 In the presence of elevated intra-vesicle Ca2+ and Pi, crystal growth occurs internally on the Octa-CaPi-like complex, generating early hydroxyapatite. Undoubtedly, some of the gamma-carboxy-expressing molecules regulate these initial phases of the process. Events to this point constitute the first phase of hydroxyapatite formation.138
|Figure 6. Phase I of the mineralization process takes place in 100 nm matrix vesicles released from adjacent osteoblasts. Ca2+ and Pi are internalized, forming non-crystalline, amorphous CaP. Amorphous CaP is then converted to octa-CaP crystals, and finally into insoluble hydroxyapatite crystals (phase I). During phase II of the mineralization process, phospholipases mediate penetration of the hydroxyapatite through the vesicle membrane. The levels of extra-vesicular Ca2+, Pi, and H+, as well as Ca2+-binding proteins including BSP-1, osteocalcin, and osteopontin regulate the continued nucleation of hydroxyapatite crystals.|
In the second phase, the enlarging CaPi complex actually begins to penetrate the matrix vesicle membrane. Complete matrix vesicle rupture does not seem to occur until later. Rather, phospholipases punch holes in the plasma membrane, exposing the nascent hydroxyapatite to the extracellular fluid. Factors that regulate the continuing growth of the hydroxyapatite include Ca2+ and Pi concentrations, local pH, and the presence of non-collagenous matrix proteins.138 At a minimum, these proteins include BSP-1, osteocalcin, and osteonectin. They also include osteopontin, which would seem to play a major, if not crucial, role in large-scale hydroxyapatite formation. Osteopontin is a secreted phosphoglycoprotein that is induced by VitD. 6, 135, 142, 143 This molecule has an RGD sequence that may provide anchorage for cells. It is also variably phosphorylated, the degree to which apparently impacts its function. When 40% of its phosphorylation sites are filled, it blocks mineralization. When 95% of its sites are filled, it appears to promote hydroxyapatite formation.143 Whether 95% phosphorylated osteopontin occurs regularly in bone is unknown.6, 143 Osteopontin and osteocalcin are suggested to interact and expose Ca2+-binding nucleation sites on osteopontin necessary for hydroxyapatite formation. 143
The process of bone resorption is relegated to the osteoclast, a multi-nucleated, monocyte-derived cell that responds to both low pH and PTH/low extracellular Ca2+ (Figure 7). Although cytokine-induction of osteoclasts has been a favorite topic of experimental biologists over the last decade, acidosis and the response to acidosis by osteoclasts is perhaps just as important. Osteoclasts express a proton-sensing G protein-coupled receptor termed OGR1/GPCR1. In the presence of low pH, this receptor is activated and increases intracellular Ca2+. This begins the process of bone resorption with demineralization. Hydroxyapatite [(Ca10(PO4)6(OH)2] contains a considerable amount of stored OH- ion. During metabolic acidosis, where renal and respiratory homeostatic mechanisms are inadequate, bone is broken down to provide buffering capacity to the blood/extracellular fluid.153
|Figure 7. The activity of bone-resorbing osteoclasts is sensitive to extracellular pH due to their expression of OGR1 H+ receptors. H+ is extruded by transporters and proton pumps, dissolving mineral in the sealed space between osteoclast and bone, while enzymes including Cathepsin K and TRAP digest the organic matrix. Negative feedback mechanisms to bone resorption involve osteoclast apoptosis triggered by Ca2+ influx through TRPV channels. In addition, PDGF-BB and/or VitD enhance osteoblast secretion of OPG, potentially limiting RANK L/RANK-mediated osteoclast development. IFN-beta may also inhibit osteoclast development.|
Osteoclasts are generated from the actions of M-CSF, RANK ligand (RANK L/TRANCE), and likely, IL-1 on osteoclast precursors. M-CSF is secreted by immature osteoblasts and promotes the formation and proliferation of osteoclast precursors from CFU-M.154, 155 In the presence of elevated VitD, IL-11 or a hypocalcemic condition that induces PTH release, osteoblasts will be stimulated to express RANK L on their surface.155, 156 RANK L is a 45 kDa, type II transmembrane glycoprotein that belongs to the TNF superfamily. When expressed, it binds to either constitutively expressed, or M-CSF-induced, RANK on the surface of osteoclast precursors.157 In synergy with M-CSF, RANK L induces differentiation to "pre-fusion" osteoclasts that ultimately fuse under the influence of M-CSF, IL-1, and RANK L. Fusion forms a mature, multi-nucleated osteoclast.155, 156 Finally, RANK L and IL-1 activate the osteoclast, initiating bone resorption. It should be noted that pro-inflammatory cytokines such as IL-1 beta and TNF-alpha will also induce RANK L expression. Although this expression is on T cells rather than osteoblasts, the effect is the same; osteoclast development.158, 159 The osteoclast first generates a sub-osteoclastic space between the osteoclast and underlying bone. A proton pump/transporter (V-ATPase and/or Na+/H+) linked to Cl- transporters/channels (KCC1 and CLC7) facilitates the release of H+ into the sealed space or lacuna.160 This dissolves the underlying mineral and generates high local levels of Ca2+ and Pi. Tartrate-resistant acid phosphatase (TRAP) and cathepsin K are also secreted, and these enzymes digest the accompanying organic matrix.155, 156
There appear to be at least three negative-feedback mechanisms associated with osteoclast activity. The first involves released Ca2+, which is suggested to enter the osteoclast via TRPV channels.157 What opens the channels is not clear. Osteoclasts have CaSRs associated with ryanodine receptor 2 (RyR2), and this complex may initiate Ca2+ transit through the osteoclast into the extracellular fluid. Regardless, Ca2+-influx is reported to induce apoptosis in osteoclasts, potentially limiting the time the bone is exposed to osteolytic agents. Clearly, intracellular Ca2+-binding agents such as calbindin may modulate the effects of any increased cytoplasmic Ca2+ levels. 161, 162, 163, 164, 165, A second potential negative-feedback mechanism involves PDGF-BB. In this scenario, RANK L on osteoblasts will bind to RANK on osteoclast precursors, generating functional osteoclasts. Osteoclasts now secrete a PDGF-BB homodimer that has a down-modulating effect on osteoblast activity. This includes a reduction in type I collagen production and osteocalcin secretion. Perhaps most importantly, PDGF-BB induces osteoblasts to secrete OPG (osteoprotegerin), a natural soluble receptor for RANK L.166 This blocks RANK L activity on osteoclast precursors and limits osteoclast development.161 The effect is amplified by the presence of VitD (a bone anabolic agent), which also promotes OPG secretion by osteoblasts.167 A final mechanism for osteoclast down-regulation involves osteoclast production of IFN-beta, an autocrine inhibitor of osteoclast development. IFN-beta, acting through the type I interferon receptor, reduces c-fos, a transcription factor needed for continued osteoclast activity.155
Although the above provides some insight into the homeostatic processes associated with Ca2+ and Pi metabolism, it does not address the consequences of unregulated or dysregulated metabolism. Recently, the discovery of the phosphatonins has focused attention on phosphorus resorption. When hypophosphatemia (low phosphorus) is present, muscle weakness, rhabdomyolysis (skeletal muscle death), bone pain and rickets (failure to form and maintain bone) may present. These can be accompanied by lethargy, confusion and impaired sphincter control. One particular problem in severe cases is a generalized impairment of energy metabolism. Limiting Pi means limited ATP formation and reduced oxidative phosphorylation. In the blood, this has (or may have) material consequences. In white cells, a reduction in ATP results in an impairment of bactericidal activity and predisposition to infection. In red cells, 2,3-diphosphoglycerate (DPG) is reduced, which shifts the oxygen dissociation curve, making it more difficult to release bound oxygen. Hypophosphatemia occurs during alkalosis, when cells take up circulating Pi and the kidneys do not resorb enough to account for the shortfall. It may also occur in hospitals during cytokine and cancer therapy and when the administration of IV glucose is the only energy source.
It is not uncommon for renal insufficiency to be accompanied by hyperphosphatemia, when the glomerular filtration rate is low and Pi does not appear in urine (for either excretion or resorption). Once present, hyperphosphatemia induces metabolic acidosis. Symptoms of the condition include a prolonged Q-T interval in an EKG, muscle cramps, tetany, and the suppression of VitD formation. One specific danger involves the abnormal deposition of CaP complexes in non-osseous tissue. This is particularly troublesome in blood vessels.2