Potential Role of Pancreatic and Enteric Hormones in Regulating Bone Turnover

ADEQUATE NUTRIENT INTAKE and normal gastrointestinal function are essential to bone health. Disorders of nutrient supply caused by either abnormal intake (e.g., anorexia nervosa, starvation) or absorption (e.g., celiac disease, inflammatory bowel disease) may lead to bone disease and an increased fracture risk. (1) The majority of studies examining the effects of nutrients on bone turnover have focused on well-established nutrients (e.g., calcium and phosphate) and homeostatic regulators (e.g., PTH and vitamin D). Three recent observations have suggested the potential involvement of alternative mechanisms regulating the interaction between nutrition and bone turnover. First, bone turnover has a marked circadian rhythm (∼100%), with higher bone resorption at night. The mechanism regulating this rhythm seems to be partially related to the cyclical intake of nutrients. (2-4) Second, there is an acute suppression of bone resorption in vivo (∼50%), which occurs within hours of the ingestion of a mixed meal, glucose, protein, or calcium. (4-7) Third, in a rodent model, providing the normal diet in multiple small fractions decreased bone resorption measured using {3H}tetracycline excretion and increased BMD compared with a matched nutrient load given once a day. (8) These studies suggest acute changes in dietary intake or the pattern of nutrient intake may have an impact on bone homeostasis. This article reviews the evidence supporting an acute nutrient-induced regulation of bone homeostasis, discusses the potential mechanisms involved, and speculates on the possible evolutionary advantage this may confer on animals that have an intermittent supply (or intake) of nutrients. The importance of calcium in maintaining mineral homeostasis is well established. Increased physiological demands result in a reversible demineralization of the skeleton. The maintenance of calcium homeostasis is apparent during acute insufficiency (overnight fast), chronic insufficiency (malnutrition, starvation, anorexia nervosa), or transient increase in demand (e.g., lactation). During these periods, the skeleton acts as a reservoir from which there is rapid mobilization of mineral and buffer. Furthermore, during periods of acute excess (postprandial) or chronic excess (obesity, treatment of anorexia nervosa), mechanisms are in place to restore the balance. Short-term regulators are required to balance acute changes in supply and demand and act to minimize deviations from the set-point. In contrast, longer-term regulators may alter the set-point, resulting in a net increase (or decrease) in bone turnover. The acute nutrient-induced changes in bone turnover may represent the physiological adaptations of the organism to a variable energy and nutrient intake. Bone remodeling is a metabolically demanding process; therefore, mechanisms may exist to ensure that during times of energy (and nutrient) excess bone remodeling is balanced to favor bone formation (in excess of bone resorption), whereas during times of energy insufficiency, bone remodeling is balanced to favor bone resorption (in excess of bone formation) to maintain calcium homeostasis. It is possible that the mechanisms regulating the acute effects of feeding (postprandial) may well be different from (or in addition to) the mechanisms that come in to play during intermediate and long-term changes in nutrient intake (obesity). A similar effect may occur in the regulation of short- and long-term fasting. Similarly, the mechanisms that regulate feeding are likely to be different from those regulating fasting. One intriguing observation is the recent identification of a nutrient responsive signaling system involving SIRT1, which may be involved in energy homeostasis, diabetes, and life span. (9) One or more of the pancreatic or enteric hormones released during a normal physiological response to nutrient ingestion may potentially integrate nutrient intake and bone turnover. Several hormones involved in the postprandial response to nutrients have been shown to regulate bone turnover at pharmacological concentrations. However, the factor(s) regulating the acute postprandial decrease in bone resorption at physiological concentrations remain elusive. Recent studies have shown that a variety of nutrients, including ingestion of a mixed meal, (5) calcium, (10-14) glucose, (4, 6) or protein, (4) result in an acute suppression of bone turnover markers (Figs. 1 and 2). One study examined the effect of feeding a mixed meal to normal subjects and measured the change in serum markers (osteocalcin {OC}, procollagen type I N-terminal propeptide {PINP}, serum bone alkaline phosphatase {sbALP}, serum C-terminal telopeptide of type I collagen {sβCTX}, and serum N-terminal telopeptide of type I collagen {sNTX}) at 1 h and urine markers (urine immunoreactive free deoxypyridinoline {uifDPD}, urine N-terminal telopeptide of type I collagen {uNTX}, and urine C-terminal telopeptide of type I collagen {uCTX}) at 2 h. (5) The magnitude of the decrease with feeding ranged from 3.8 ± 0.9% for PINP (p < 0.0001) to 17.8 ± 2.6% (p < 0.0001) for sβCTX compared with fasting controls. A time-course was not performed. There was a significant decrease in serum markers (range, 3.8-17.8%) and urine markers (range, 7.0-7.9%). (5) The only exception was sbALP, which showed no effect of feeding. However, the long half-life of sbALP (36 h)(15) may obscure an acute response to feeding. The acute effects of oral glucose, protein, and long-chain triglycerides on sβCTX and osteocalcin in healthy individuals compared to fasting controls presented as AUC0–3h (mean ± SE). ANOVA (all) and p values for the individual groups compared with fasting controls are shown: *p < 0.05, ** p < 0.01, and ***p < 0.001. Reproduced with permission of the American Society for Bone and Mineral Research from J Bone Miner Res 2003:18:2180-2189.20 Percentage change from baseline for (A) the bone turnover marker sβCTX and (B) PINP to oral glucose (G+ So) and oral glucose with intravenous octreotide (G+ S+). The horizontal arrows indicate the onset (←; time -30 minutes) and completion (→; time 240 minutes) of the octreotide or saline infusion, and the vertical arrow (↓; time 0 minutes) indicates oral ingestion of either glucose or distilled water. *p < 0.05, **p < 0.01 comparison with fasting baseline (time -45 to -30 minutes) after adjustment for multiple comparisons (n = 8). (6) Copyright 2003, The Endocrine Society.20 In response to oral glucose, there is a rapid suppression in bone resorption within minutes, with a 50% decrease in sβCTX(6, 7) and uNTX. (4, 6) The postprandial decrease in bone resorption markers is similar in magnitude to that observed with many antiresorptive agents. However, unlike the sustained effect seen with antiresorptive agents, the effect of nutrient intake on bone turnover is only transient, with a maximal effect at 2 h, followed by reversal 4–5 h postprandially. (4, 6) The glucose induced suppression in PINP was 8%(6); however, the results for OC were more variable, with one study showing an 16% decrease(6) with a similar time-course for the response to glucose to the bone resorption markers as for OC and PINP. (6) A second study did not show a significant effect of glucose on OC. (7) This may suggest that the factor(s) released in response to a nutrient load modulate bone resorption and bone formation differently or that more than one bone active factor is modulating nutrient-induced changes in bone turnover. In response to glucose, fat, or protein, the preliminary data presented in one study suggested there was no statistically significant difference in the time-course or magnitude of the postprandial effect on sβCTX, a bone resorption marker. (4) When the same data set was analyzed using a fasting control group, only the response to glucose and protein reached statistical significance for sβCTX (Fig. 1). (7) There was no significant effect on osteocalcin after fat or protein ingestion. (7) The pronounced circadian rhythm of bone turnover markers is also attenuated by fasting (Fig. 3), (2, 3) suggesting that this rhythm is mediated, at least in part, by the cyclical intake of nutrients. In circadian studies, the effect of fasting versus feeding on bone turnover has only been observed for sβCTX and uCTX but not for OC, a bone formation marker. (2, 3) There is also a marked circadian rhythm in bone turnover markers in rodents. (16) One study using CTX was not able to show an acute effect of fasting versus feeding on CTX; however, in a food fractionation study, changing the pattern of nutrient intake significantly modified bone resorption determined using urinary {3H}tetracycline excretion and resulted in a net increase in BMD. (8, 17, 18) Effect of fasting (▪) and feeding (▴) on the circadian rhythm of serum ßCTX, a bone resorption marker. (3)20 One study found that the circadian rhythm in bone resorption in vitro was abolished if the serum used was heat treated, suggesting that the factor(s) controlling the nutrient-induced changes in the circadian rhythm were heat labile. (19) In addition, the conventional regulatory mechanisms do not seem to explain the effect of diet on the circadian rhythm of bone turnover. (10, 20-22) The amplitude of the circadian rhythm for bone resorption markers is generally greater than observed for bone formation markers. During the day, feeding results in a net positive balance for bone formation, with a greater suppression in bone resorption compared with bone formation. In contrast, during the night, when energy and nutrient supply is inadequate, there is a net negative balance with bone resorption occurring in excess of bone formation, which may be required to maintain calcium homeostasis. There are multiple hormones released in the postprandial response to nutrient intake. (22-24) Many of these potential mediators have been shown to regulate bone turnover in vitro(25-28) and in vivo, (29) including insulin, glucagon, calcitonin, gastric inhibitory polypeptide (GIP), glucagon-like polypeptide 2 (GLP 2), leptin, and amylin. (7, 28, 30-34) The time-course of change and specificity of response to glucose suggest that adrenaline (Ad), noradrenaline (NA), and growth hormone (GH)(23, 35) are unlikely to mediate the acute nutrient-induced decrease in bone resorption, which is first apparent 20 minutes postprandial. (6) The postprandial increase in GH is first observed at 50 minutes and Ad at 120 minutes, and NA does not change at any time-point. (23, 35) The delayed increase in these hormones suggests that they are unlikely to mediate the acute nutrient-induced changes in bone resorption. In addition, if the morning rise in cortisol is inhibited using metyrapone, there is no effect on the bone resorption markers (DPD, uNTX, and PYD), although the effect on OC is significant. (36) In contrast, in in vitro studies, mifepristone, a glucocorticoid receptor antagonist, partially abolished the amplitude of the circadian rhythm of bone resorption present in serum. (37) Gastrin is released in response to calcium and amino acids but not glucose or fat ingestion. (38) Because a decrease in bone resorption is observed in response to each of these macronutrients, it suggests that gastrin could only have a role in mediating the postprandial decrease in bone resorption if it were acting together with another bone active factor(s). Although exogenous glucagon decreases bone resorption by 10%, (4) glucagon is unlikely to mediate the acute effects of meal ingestion on bone turnover, because there is either no change(22) or a postprandial decrease in glucagon with feeding. (23) In vitro insulin dose-dependently inhibited pit formation by up to 80%, and glucose resulted in a concentration-dependent change in osteoclast activity. (39, 40) In vivo, an intravenous infusion of insulin alone decreased bone resorption by ∼30%, (4) suggesting a role for insulin in the nutrient-induced suppression in bone turnover. However, during a euglycemic hyperinsulinemic clamp study, bone resorption did not change, suggesting that hypoglycemia rather than insulin per se caused the previously observed decrease in bone resorption. (41) Nonetheless, longer-term effects of hyperglycemic hyperinsulinemia may be mediated by indirect effects such as the stimulation of renal calcium excretion(42-45) and intestinal calcium absorption. (46) In another study, the bone turnover response to oral glucose was substantially greater than that after intravenous glucose, (4) suggesting a role for the incretin hormones (GIP and GLP 1). In addition, octreotide abolished the nutrient-induced suppression in bone turnover (Fig. 2), which suggests a potential role for the pancreatic or enteric hormones in the acute suppression of bone turnover. (6) Octreotide is a long-acting analog of somatostatin, which inhibits the basal and postprandial secretion of many pancreatic, enteric, and pituitary hormones. These include insulin, glucagon, gastrin, calcitonin, GLP 1, GLP 2, GIP, GH, and ghrelin. (22, 47-49) The effect did not seem to be mediated by changes in PTH or ionized calcium levels. (6) This raised the possibility of novel endocrine mechanisms mediating an “entero-osseous” interaction between the gastrointestinal tract and bone, which was first proposed by Bollag et al. (26) The evidence suggests that the factor(s) that regulate the acute interaction between nutrient ingestion and bone should be heat labile, (19) respond within minutes of nutrient ingestion, (4, 6) either decrease if the factor stimulates bone resorption or increase if the factor inhibits bone resorption, (4, 6) and must show a reversal in the process within 4–5 h. (4, 6) The majority of studies have focused on the effects of the pancreatic or enteric hormones on and osteoclast in vitro or long-term effects on BMD and bone turnover markers in of GIP, GLP 2, and have been shown to regulate bone turnover in and in (7, in vitro and in vivo studies suggest there is an reversible decrease in bone resorption in response to GLP (7, However, whereas the normal postprandial of GLP 2 after ingestion of protein, fat, or glucose is to a significant decrease in the the for bone resorption, a pharmacological of GLP 2 was which resulted in a of GLP 2, which was higher than the normal physiological postprandial (Fig. (7) In contrast, GLP 1 have not been on (26) and an intravenous of GLP 1 in did not result in a change in sβCTX, although this study was to because of the of an control (7) The effect of GLP 2 on (A) serum GLP 2 and (B) change in sβCTX in The normal postprandial increase in GLP 2 is GLP were given at and and compared with (7) Reproduced with permission of the American Society for Bone and Mineral Research from J Bone Miner Res 2003:18:2180-2189.20 One study showed that are present on and increased and calcium resulting in an increase in the of type I collagen and (26) Furthermore, has been shown to bone in A increased bone formation markers and increased BMD. the receptor a significant decrease in BMD. Although an intravenous of in did not result in a change in sβCTX, this study was to because of the of a control group, the infusion minutes) of is and the minutes) of which may have an effect on bone turnover. (7) is a amino that is with insulin from the in the and after a In stimulates and inhibits osteoclast In addition, pharmacological of result in an effect on bone in resulted in decreased an effect that to be mediated an of with no apparent effect on or There is also evidence suggesting a role for in and regulation of bone which at least in part, explain the effect of on have been on Although may modulate bone turnover, the postprandial increase in in response to nutrient intake is delayed by suggesting that is not the factor mediating the nutrient-induced in bone turnover. There however, evidence that is a hormone involved in the intermediate and long-term balance between energy and bone The role of PTH as a of acute changes in bone turnover is PTH have been on decreases PTH secretion and insulin PTH secretion in PTH secretion is increased by a mixed meal, decreased by glucose, and in the majority of studies, decreased by of serum calcium. were to a difference in the for PTH during fasting compared with glucose ingestion, although the time-course of response did show a small transient decrease in PTH at 20 (6) resulted in either a decrease or no change in the PTH however, it abolished the increase in (2, suggesting that PTH explain the decreased circadian rhythm of bone resorption observed during fasting. (2, 3) When PTH was by a calcium infusion, there was no change in the circadian rhythm in bone resorption, and reversal of the increase in PTH by calcium only partially the in bone resorption. In addition, the effect of PTH using in vitro did not the circadian rhythm of bone resorption present in serum. (37) Several studies have been to show a change in ionized calcium in response to glucose or insulin infusion, acute changes in and it is likely that this is a caused by the of calcium from were to an effect of glucose ingestion compared with fasting on ionized calcium levels. (6) have been in intestinal and the is likely to have an function rather than bone homeostasis the However, to has not been measured in the serum of normal individuals after nutrient ingestion. are present on but not on seems to and osteoclast in Although pharmacological of bone resorption, it is feeding results in a postprandial increase in secretion in either or an increase in does occur after food this is likely related to the calcium of the The a normal basal of bone resorption, increased of the osteoclast to PTH and increased with no bone after In addition, the receptor an increase in bone formation with no apparent changes in bone resorption in GH and modulate bone turnover, and these hormones are by and During there is a marked increase in the secretion of GH, which is by feeding. In addition, octreotide the increase in GH but does not alter the increase in Furthermore, the increase in GH in response to glucose is delayed for at least 50 minutes, (23, 35) which suggests that the changes in GH are unlikely to mediate the effect of feeding on bone turnover. Although nutrient supply is an of serum acute postprandial changes are to an effect on During there is a marked increase in the secretion of a GH that is by feeding. In addition, the and of and may also modulate the effect of on and a long-acting GH bone in the of however, after of this to be to bone in In a model, there was no effect on bone or fat in In secretion is not in response to and it is possible that effects of are during and a bone may only apparent during however, released from the gastric and because to the postprandial decrease in sβCTX, it is unlikely that the acute nutrient-induced decrease in bone resorption. effects may bone resorption, including changes in which may modulate calcium Although studies have suggested that dietary protein may have a potentially effect on the skeleton that could be mediated, at least in part, increased urinary and increased urinary calcium However, nutrient intake in seems to bone resorption of changes in balance. of the which is involved in secretion and the of and to suggesting a role for one or more of these in the of chronic bone disease, if not the acute suppression of bone resorption. Furthermore, there is evidence to an acute effect of fasting versus diet on the at least in although in decreases in hormone hormone and have been it is the possibility of a response mediated by the enteric and similar to that proposed for of bone markers may potentially occur because of postprandial changes in or However, it is unlikely that changes in the of markers have a significant because the of the bone marker response in the majority of studies is the different of which is to represent an of collagen showed a similar time-course of response as found with bone resorption markers measured in the In addition, the magnitude of the decrease in in response to nutrients was similar to urinary and serum bone resorption markers. A of studies have shown an acute change in urinary {3H}tetracycline excretion in response to dietary This is because urinary {3H}tetracycline excretion bone remodeling using a different and results that are with the nutrient-induced changes observed using bone resorption markers. It be to the effect of nutrient-induced changes on bone turnover using bone however, this long-term feeding It is possible that of the between the fasting and postprandial may be caused by such as postprandial However, dietary only a effect on the of (5) In addition, dietary such as have been shown to increase excretion of which is one of the bone resorption but not for urinary A of dietary have been found to be bone including and These at least in to to a rather an effect of the excess In addition, the essential and bone resorption in Furthermore, the and bone resorption in in vivo and in The which are of from the diet and have been shown to and from the which has osteoclast in vitro the of and protein 1 studies have focused on the effect of calcium intake on bone and there is evidence that calcium and There is also evidence that intake of and may BMD and bone turnover in and increase in The impact of dietary protein and the of dietary protein on bone is It is also that decreased intake in the or anorexia in the is to bone health. It however, the and of the meal is or there is a effect of different bone active nutrients. In addition, disease of (e.g., anorexia or (e.g., celiac disease, inflammatory bowel disease, and bowel result in and It is that GLP 2 infusion the bone observed in with bowel GLP 2 therefore, this effect at least in part, to a increase in nutrient absorption rather than a effect on bone turnover. These observations may well be to the of of related to nutrient and chronic inflammatory in and bone turnover and decreases BMD. or a decrease in are with an increased of bone and The mechanism for this is but may include or effects of and The observation that subjects are with a between nutrient ingestion suggests an alternative The increased in dietary intake may result in a more suppression in bone turnover, resulting in an increase in BMD in one study in has showed that an diet given in fractions results in a suppression of bone resorption and a net in bone (8, This suggests that small may the skeleton. the of as in is with significant this is likely caused by multiple of effects of enteric hormones on bone is a possibility that It is that the nutrients and regulators of bone are no to explain the observed between diet and bone (or The evidence a role for the pancreatic or enteric hormones in the acute and chronic regulation of bone 1). the physiological involved in the acute nutrient-induced suppression of bone turnover remain elusive. There are a of that are responsive to nutrients, and evidence suggests that these may represent a of physiological regulators acting on bone The in the showing a or potential effect of the pancreatic and enteric hormones on bone role of pancreatic and enteric hormones in osteoclast and and The the inhibitory and potential effects of the pancreatic and enteric hormones on bone evidence suggests that the effects of pancreatic or enteric hormones may be of and osteoclast In addition, studies effects on and of and These may include regulation of and with indirect effects on and Several and for the and pancreatic hormones or showed a bone in vivo in response to a of (e.g., suggesting a role for the hormone (or in bone However, feeding studies have not been in these and and these data are it not be possible to these hormones in the regulation of nutrient-induced changes in bone turnover. Furthermore, the regulating the of energy and bone are and are likely to significant It is therefore, that as well as studies may be required to the impact of the pancreatic or enteric hormones on bone is is that have a significant to the impact of nutrients on bone and the regulating these