The Journal of Nutrition Vol. 128 No. 5 May 1998, pp. 917-920

Calcium AbsorptionA Paradigm for Mineral Absorption¹

Felix Bronner

Department of BioStructure and Function, The University of Connecticut Health Center, Farmington, CT 06030-3705

	ABSTRACT

Abstract Introduction References

Intestinal calcium absorption proceeds by two mechanisms, an active transcellular process that takes place in the duodenum and a passive paracellular process throughout the small intestine. This article characterizes the three steps of transcellular calcium movemententry, intracellular diffusion and extrusionand identifies conditions that must be satisfied for other mineral ions to move transcellularly as part of a transepithelial transport process. Passive calcium movement is down a chemical gradient with the amount absorbed by this pathway determined in large measure by the sojourn time, most of which is spent in the ileum. Because transcellular movement of most mineral ions other than calcium, where measured, is either small or negligible, passive transport is likely to be the major route of intestinal absorption, the nature of which, however, has not been well established experimentally.

	INTRODUCTION

Abstract Introduction References

Intestinal calcium absorption involves two processes: a transcellular, metabolically driven transport and a paracellular, passive process. This article will describe these processes in detail and raise the question whether and how intestinal absorption of other minerals can be analyzed in comparable terms. It should be stated at the outset that whereas much is known about the steps involved in each of these transport processes for calcium, much less is known with respect to most other minerals. Characterizing each of the two processes for calcium therefore may help identify what still needs to be learned about the transport process for a given mineral. Some generalizations about the active transcellular process in particular may provide insight into boundary conditions that seem to apply to all active mineral transport.

	ACTIVE (SATURABLE) CALCIUM TRANSPORT

Transcellular movement involves three steps: entry across the cell wall, diffusion through the cytoplasm and exit at the basolateral cell pole. Calcium movement across the plasma membrane of the cell is extremely restricted yet is not the rate limiting step for active calcium transport (Bronner et al. 1986). This is shown by work with brush border membrane vesicles, the calcium uptake of which in vitro depends only to a limited degree on the vitamin D status of the animals from which the vesicles have been isolated (Miller and Bronner 1981) even though transcellular calcium transport, as evaluated in everted duodenal sacs or duodenal loops, is totally vitamin D dependent (Pansu et al. 1983). Calcium crosses the plasma membrane via calcium channels that are not voltage gated but respond to verapamil (Miller and Bronner 1981), a compound that also has been shown to lower calcium transport by duodenal loops (Fox and Green 1986). Calcium uptake by brush border membrane vesicles is not energy dependent.

In contrast, calcium extrusion from the cell requires energy in the form of ATP. The extrusion process can be studied with the aid of basolateral membrane vesicles, about half of which are inside out. The enzyme that mediates calcium extrusion is the Ca ATPase (Carafoli 1994), which is located at the inner aspect of the basolateral membrane. ATP addition to a basolateral membrane preparation increases calcium uptake by inside-out vesicles only because ATP cannot enter right-side out vesicles (Bronner 1996, Ghijsen et al. 1982). Therefore vesicle uptake of calcium is equivalent to half the calcium extrusion capacity of an enterocyte population.

Extrusion occurs through the channel that is formed by the transmembrane elements of the CaATPase, the calcium having become bound to a site at the inner aspect of the cell membrane. Extrusion follows a phosphorylation-induced conformational change of the enzyme. The extrusion process is also vitamin D dependent, presumably because vitamin action increases the number of enzymes, i.e., the number of pumps in the endothelial cell (Wasserman et al. 1992). However, calcium extrusion, even in the absence of vitamin D, is not rate limiting (Bronner et al. 1986).

The rate-limiting step in transcellular calcium movement is the diffusion of the calcium ion across the cytoplasm. This may appear surprising, yet the self-diffusion rate of the calcium ion in an aqueous medium at 37°C amounts to only ~ of the experimentally determined transport rate (Bronner et al. 1986). The actual rate of self-diffusion in the cytoplasm is likely to be even slower but has not been determined. The effective intracellular path, whether straight or crooked, the presence of obstruction (organelles) and the density of the cytoplasmic medium are factors that would tend to slow diffusion, whereas a steep calcium concentration gradient between the luminal and basolateral poles of the cell would tend to accelerate diffusion. On the basis of available data (Bronner et al. 1986), self-diffusion of the calcium ion in the cytoplasm would likely be slower than the experimentally determined transport rate by close to two orders of magnitude if the duodenal cell did not contain calbindin D_9K, the vitamin D-dependent enterocytic calcium-binding protein (Bronner et al. 1986, Wasserman and Fullmer 1995, Wasserman et al. 1968). Rat calbindin D_9K is an acidic protein with a pI of 4.8, a K_D for calcium binding of 0.3 × 10⁶ mol/L (Ueng et al. 1979), the biosynthesis of which is totally dependent on vitamin D (Desplan et al. 1989). Calbindin D_9K is found in duodenal cells and is absent from ileal cells and from all intestinal cells in vitamin D-deficient animals (Pansu et al. 1983).

Calbindin D_9K functions by raising the calcium ion concentration, amplifying the calcium flux in bucket brigade fashion in a manner analogous to that by which hemoglobin functions as an oxygen carrier (Stein and Hoshen 1996, Wyman 1966). In a vitamin D-deficient intestinal cell, calcium ion moves across the plasma membrane down its chemical gradient and accumulates along the inner aspect of the plasma membrane with very little calcium found in the cytoplasm. In an intestinal cell from a vitamin D-replete animal, on the other hand, calcium ions occur throughout the cytoplasm (Wasserman and Fullmer 1995). Thus calbindin D_9K acts to augment the intracellular diffusion rate of the calcium ion. Stein (1992) has shown how Fick's diffusion equation can be modified to express the augmentation of calcium movement due to the presence of calbindin. Bronner et al. (1986) have provided experimental evidence that shows calbindin D_9K enhances calcium transport positively and linearly in vivo, whereas Feher et al. (1992) have provided similar evidence for in vitro transport.

How do these three processesentry, diffusion, exitfunction for minerals other than calcium? In a situation where the free intracellular concentration of a mineral ion is higher than its luminal concentrations, metabolic energy would be required to overcome the adverse chemical gradient. If that were true, it is also likely that the intracellular concentration of the mineral ion would be high enough to permit exit out of the cell. Such a situationcell entry mediated by a metabolic pump mechanism, followed by an exit down a chemical gradientdescribes phosphate absorption in the jejunum (Peterlik et al. 1981).

If, however, the free intracellular concentration of a mineral ion is lower than its luminal concentration, entry would be down a chemical gradient. Intracellular diffusion then would be a function of the chemical gradient between the mineral ion concentrations on the luminal and basolateral poles of the intestinal cell and the diffusion constant of the mineral. The free mineral ion concentration of the cytoplasm is of course the average of the concentrations at the two cell poles.

In other words, if one knows the free intracellular concentration of a mineral ion and its diffusion constant, one can calculate the self-diffusion rate of the mineral ion in the intestinal cytoplasm and compare the calculated with the experimental value of transcellular movement. As stated above, for calcium, actual transport exceeds the theoretical value of self-diffusion by >70-fold. If the experimental transport value appreciably and reliably exceeds the theoretical value, it is likely there exists a molecular mechanism for transport enhancement, similar to how calbindin D_9k augments calcium diffusion through the cytoplasm. A first search then could be targeted at a soluble binding molecule in the cytoplasm.

It might be argued that the presence of an active extrusion process, as exemplified by the CaATPase, is sufficient for an active transport process. This would be so if there were no or virtually no barrier to the entry of the mineral ion across the brush border, so that extrusion would become the limiting rate. But in that case, the cellular content of the free mineral ion would equal that of the lumen and could threaten cell survival. A high intracellular concentration of the free calcium ion is toxic. Therefore calcium entry into cells is restricted, and extrusion by the CaATPase of the free calcium, for all practical purposes, is not rate limiting. Instead calcium-transporting cells have become equipped with a calcium-binding molecule, calbindin D_9kor, in the kidney, calbindin D_28k (Taylor et al. 1982)that augments the otherwise rate-limiting self-diffusion of the calcium ion.

Most mineral ions cannot freely enter intestinal cells and, if they are transported transcellularly at a rate that exceeds their self-diffusion rate, are likely to be moved via a binding molecule. The concentration of such a ligand must be high enough to raise the intracellular concentration to the point where transcellular movement can occur at the experimentally determined rate.

How specific must such a ligand be? A number of mineral ionsBa, Sr, Mn, the lanthanides, Pbcompete with calcium for calbindin, some binding more tightly than calcium (Fullmer and Wasserman 1977). Induction of calbindin can explain why vitamin D administration enhances lead absorption (Fullmer 1992). Blumsohn et al. (1994) found that administration of 1,25-dihydroxycholecalciferol in patients with osteoporosis and chronic renal failure stimulated Sr absorption more than calcium absorption. A possible explanation is that Sr binds more tightly than Ca²⁺ to the newly synthesized calbindin D_9k (Fullmer and Wasserman 1977).

On the other hand, I know of no cytosolic-binding protein other than calbindin that is associated with the transcellular transport of a mineral ion or of a ligand that has evolved in response to the need to transport such an ion.

Extrusion of mineral ions from the intestinal cell can be expected to use existing extrusion mechanisms, i.e., one of the ATPases or ion exchangers found in all cells. Many of the divalent cations may well compete with calcium for binding and perhaps transchannel extrusion, although no studies concerning this step are available. ATPase efflux systems for toxic ions like cadmium or copper are known in bacteria and are located on chromosomal genes (Silver 1996). Genes for such extrusion pumps may therefore also exist in mammalian cells.

On the whole, transcellular movement of most mineral ions has not been demonstrated, often because the relevant techniques (everted intestinal sacs or in situ intestinal loops) have not been used. Magnesium does not appear to move transcellularly (Hardwick et al. 1990) and the saturable component of copper transport is very small (Bronner and Yost, 1985).

	PARACELLULAR (NONSATURABLE) TRANSPORT

The second major mechanism by which calcium and, presumably, many other mineral ions move from the intestinal lumen to the circulation is by a paracellular process, i.e., down a chemical gradient through the tight and intermediate junctions and then through the much wider basolateral region. In the case of calcium, it can be shown that the rate at which the divalent cation moves out from an ileal loop exceeds the ion's self-diffusion rate by one to two orders of magnitude (Bronner et al. 1986). Because ileal cells contain no calbindin D_9K, which otherwise might act to augment the self-diffusion rate, calcium must be moving paracellularly.

In rats fed a 1.5 g Ca/100 g diet, true calcium absorption is 60-70 mg/day (Pansu et al. 1993, Sammon et al. 1970), with some 50 mg absorbed by the passive, paracellular route (Pansu et al. 1993). Duflos et al. (1995) have shown by direct experimentation that in rats on such a diet, the total solubilized calcium in the small intestine at any time of day or night amounts to 0.9 mg. With 50 mg Ca absorbed by the paracellular route in a 24-h period, 0.035 mg Ca/min would be moving transepithelially or 3.85% per minute. Duflos et al. (1995) have calculated that if there were no barrier between the ionic calcium in the intestinal lumen and the ionic calcium in the blood plasma, the amount of calcium that is absorbed in 24 h could have been cleared by the blood circulation in ~30 min. This is another way of pointing to the substantial restriction to free ion movement that the tight junctions represent. For this reason, differences in the radii of the hydrated mineral ions probably play a minor role in the rate at which mineral ions travel paracellularly.

The amount of mineral absorbed by the paracellular route is determined by how much is solubilized in the intestinal lumen, i.e., mineral solubility, the paracellular permeability to the ion and the time spent by the chyle in a given intestinal region, i.e., the intestinal sojourn time. In the case of calcium, solubility is a function of the chemical form of the calcium salt and of the pH in a given intestinal region. Thus when, in a rat study, the major form of dietary calcium was in the form of gluconate, the amount of calcium absorbed was directly proportional to the intake over a wide range (Pansu et al. 1993). On the other hand, when calcium was in the form of carbonate and phosphate, there was no further calcium absorption at intakes >500 mg Ca/day (Pansu et al. 1993). Presumably most calcium precipitated or reprecipitated at the alkaline pH of the distal intestine in the carbonate/phosphate diet group, whereas sufficient calcium remained in solution in the lumen of the gluconate diet group for absorption to proceed. It should be pointed, however, that so long as some ions remain in solution, absorption will proceed if there is a concentration gradient between the lumen and the ion concentration in the body fluid. In the case of calcium, the lower limit of the gradient is the calcium concentration of the plasma or of the lymph (~1.25 mmol/L), but there is no such lower limit for ions that do not naturally occur in body fluids.

The time chyle spends in the various sections of the intestine has been measured in the rat (Duflos et al. 1995). Of the total transit time of ~3 h, chyle passes through the duodenum in 2-3 min, spends ~45 min in the jejunum and the remainder of the time, somewhat >2 h, in the ileum. Inasmuch as the amount of solubilized calcium found in the intestine at any time is fairly much the same in all regions (Duflos et al. 1995), and because permeability is the same in all three major regions of the intestine (Bronner and Spence 1988), sojourn time becomes the differentiating factor for how much is absorbed in the three parts of the intestine by the paracellular route. Thus of the 50 mg Ca absorbed by rats fed a 1.5% Ca diet via the paracellular pathway in the course of 24 h, <2% is absorbed in the duodenum, 25% in the jejunum and the remainder in the ileum.

According to Pansu et al. (1993), the maximum rate of calcium efflux from duodenal loops of male rats fed a 1.5% Ca diet was found to be 11.7 µmol Ca·h¹·g¹, equivalent to 11.2 mg Ca in a 24-h period. In rats fed a 3.0% Ca diet, the amount of calcium transported by the active, duodenal route was halved, but the amount absorbed via the paracellular route went up with the increase in the calcium content of the diet. In other words, active calcium transport was down-regulated as calcium intake was raised, but the amount transported by the passive transport route was a positive function of the calcium intake.

For most minerals, the paracellular pathway appears to be the predominant route of entry from lumen to blood. Factors that determine paracellular calcium movement undoubtedly also play a role in mineral absorption, i.e., solubility, intestinal permeability and sojourn time. Solubility in the chyle is specific for each mineral, is influenced by the pH at a given locus and by the nature and quantity of anions. For example, calcium sulfate is quite insoluble, whereas cadmium sulfate is very soluble (Lide 1992). Magnesium citrate is far more soluble than calcium citrate (Lide 1992), even though the latter is one of the more soluble calcium salts that can be used for food supplementation.

The structure and composition of tight junctions are primary determinants of cellular permeability. Moreover, as already pointed out, differences in the hydration radius of mineral ions in all likelihood play a minor role in the rate at which mineral ions move along the paracellular route. This inference is supported by the observation (Bronner and Spence 1988) that Ca²⁺ and phenol red move transepithelially at the same rate in all three intestinal segments, even though the molecular weight of phenol red is nine times that of calcium and its charge would be anionic. Intestinal sojourn time is determined largely by peristalsis and therefore will be the same for essentially all components of the luminal fluid. In other words, absolute absorption of two different minerals will be the same if their ionic concentration in the luminal fluid is the same. But this is unlikely given the vast differences in recommended intakes, ranging in the rat from 0.15 mg I/kg diet to 5 g Ca/kg, a range of more than four orders of magnitude (National Research Council 1978).

In conclusion, active transcellular transport is relatively unimportant for the absorption of most minerals, calcium excepted. The principal route of absorption is the paracellular pathway where mineral solubility, permeability of the intestinal epithelium, and luminal sojourn time are the principal determinants of absorption.

	FOOTNOTES

Manuscript received 4 November 1997. Initial reviews completed 17 December 1997. Revision accepted 24 December 1997.

	LITERATURE CITED

Abstract Introduction References

• Blumsohn A., Morris B., Eastell R. Stable strontium absorption as a measure of intestinal calcium absorption: comparison with the double-radiotracer calcium absorption test. Clin. Sci. 1994; 87:363-368 ^[Medline]
• Bronner F. Cytoplasmic transport of calcium and other inorganic ions. Comp. Biochem. Physiol. B Comp. Biochem. 1996; 115:313-317
• Bronner, F., Pansu, D. & Stein, W. D. (1986) An analysis of intestinal calcium transport across the rat intestine. Am. J. Physiol. 250 (Gastrointest. Liver Physiol. 13): G561-G569.
• Bronner, F. & Spence, K. (1988) Non-saturable Ca transport in the rat intestine is via the paracellular pathway. In: Cellular Calcium and Phosphate Transport in Health and Disease (Bronner, F. & Peterlik, M. eds.), pp. 277-284. Alan R. Liss, New York, NY.
• Bronner, F. & Yost, J. H. (1985) Saturable and nonsaturable copper and calcium transport in mouse duodenum. Am. J. Physiol. 249 (Gastrointest. Liver Physiol. 12): G108-G112.
• Carafoli E. Biogenesis: plasma membrane calcium ATPase: 15 years of work on the purified enzyme. FASEB J. 1994; 8:993-1002 [Abstract]
• Desplan C., Brehier A., Perret C., Thomasset M. Rat duodenal calcium-binding protein messenger RNA: induction by 1,25-dihydroxyvitamin-D₃. J. Steroid Biochem. 1983; 19:1577-1582 [Medline]
• Duflos C., Bellaton C., Pansu D., Bronner F. Calcium solubility, intestinal sojourn time and paracellular permeability codetermine passive calcium absorption in rats. J. Nutr. 1995; 125:2348-2355
• Feher, J. J., Fullmer, C. S. & Wasserman, R. H. (1992) Role of facilitated diffusion of calcium by calbindin in intestinal calcium absorption. Am. J. Physiol. 262 (Cell Physiol. 31): C517-C526.
• Fox J., Green D. T. Direct effects of calcium channel blockers on duodenal calcium transport in vivo. Eur. J. Pharmacol. 1986; 129:159-164 [Medline]
• Fullmer C. S. Intestinal interactions of lead and calcium. Neurotoxicol. 1992; 13:799-808 [Medline]
• Fullmer, C. S. & Wasserman, R. H. (1977) Bovine intestinal calcium-binding protein: cation-binding properties, chemistry and trypsin resistance. In: Calcium Binding Proteins and Calcium Function (Wasserman, R. A., Corradino, E., Carafoli, R. H., Kretsinger, R. H., MacLennan, D. H. & Siegel, F. L. eds.), pp. 303-312. North Holland, New York, NY.
• Ghijsen W.E.J.M., DeJong M. D., Van Os C. H. ATP-dependent calcium transport and its correlation with Ca²⁺-ATPase activity in basolateral plasma membranes of rat duodenum. Biochim. Biophys. Acta. 1982; 689:327-336 ^[Medline]
• Hardwick, L. L., Jones, M. R., Buddington, R. K., Clemens, R. A. & Lee, D.B.N. (1990) Comparison of calcium and magnesium absorption: in vivo and in vitro studies. Am. J. Physiol. 259 (Gastrointest. Liver Physiol. 22): G720-G726.
• Lide, D. R., ed. (1992) Handbook of Chemistry and Physics, 3rd ed, pp. 8-14. CRC Press, Boca Raton, FL.
• Miller III, A., Bronner F. Calcium uptake in isolated brush-border vesicles from rat small intestine. Biochem. J. 1981; 196:391-401 ^[Medline]
• National Research Council (1978) Nutrient Requirements of Laboratory Animals. National Academy of Sciences, Washington, D.C.
• Pansu, D., Bellaton, C., Roche, C. & Bronner, F. (1983) Duodenal and ileal calcium absorption in the rat and effects of vitamin D. Am. J. Physiol. 244 (Gastrointest. Liver Physiol. 7): G695-G700.
• Pansu D., Duflos C., Bellaton C., Bronner F. Solubility and intestinal transit time limit calcium absorption in the rat. J. Nutr. 1993; 123:1396-1404
• Peterlik, M., Fuchs, R. & Cross, H. S. (1981) Phosphate transport in the intestine: cellular pathways and hormonal regulation. In: Calcium and Phosphate Transport Across Biomembranes (Bronner, F. & Peterlik, M. eds.), pp. 173-179. Academic Press, New York, NY.
• Planas-Bohne, F. Cadium metabolism. In: Handbook of Metal-Ligand Interactions. Bioinorganic Medicine (Berthon, G. ed.), vol. 1. pp. 636-639. Marcel Dekker, New York, NY.
• Sammon P. J., Stacey R. E., Bronner F. Role of parathyroid hormone in calcium homeostasis and metabolism. Am. J. Physiol. 1970; 218:479-485 [Free Full Text]
• Silver S. Bacterial heavy metal resistance: new surprises. Annu. Rev. Microbiol. 1996; 50:753-789 [Medline]
• Stein, W. D. (1992) Facilitated diffusion of calcium across the intestinal epithelial cell. J. Nutr. 122 (suppl. 3): 651-656.
• Stein W. D., Hoshen M. B. The role of ligand-binding proteins in increasing the transcellular flux of metabolites, in particular that of calcium, bilirubin and fatty acids. Comp. Biochem. Physiol. B Comp. Biochem. 1996; 115:307-312
• Taylor A. N., McIntosh J. E., Boudreau J. E. Immunocytochemical localization of vitamin D-dependent calcium-binding protein in renal tubules of rabbit, rat and chick. Kidney Int. 1982; 21:765-773 [Medline]
• Ueng T.-H., Golub E. E., Bronner F. The effect of age and 1,25-dihydroxyvitamin D₃ treatment of the intestinal calcium-binding protein of suckling rats. Arch. Biochem. Biophys. 1979; 196:624-630 [Medline]
• Wasserman R. H., Corradino R. A., Taylor, A. N., Vitamin D-dependent calcium binding protein: purification and some properties. J. Biol. Chem. 1968; 243:3978-3986 [Abstract/Free Full Text]
• Wasserman, R. H., & Fullmer C. S. (1995) Vitamin D and calcium transport: facts, speculations and hypotheses. J. Nutr. 125 (suppl.): 1971S-1979S.
• Wasserman R. H., Smith C. S., Brindak M. E., de Talamoni N., Fullmer C. S., Penniston J. T., Kumar R. Vitamin D and mineral deficiencies increase the plasma membrane calcium pump of chicken intestine. Gastroenterology 1992; 102:886-894 [Medline]
• Wyman J. Facilitated diffusion and the possible role of myoglobin as a transport molecule. J. Biol. Chem. 1966; 241:115-121 [Abstract/Free Full Text]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences