| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Faculty of Medicine-Endocrinology (J.P.W., C.J.S., N.J.F., C.S.K.), Memorial University of Newfoundland, St. John’s, Newfoundland, Canada A1B 3V6; and University of Texas M.D. Anderson Cancer Research Center (A.O.H., R.F.G.), Houston, Texas 77030
Address all correspondence and requests for reprints to: Dr. Christopher S. Kovacs, Faculty of Medicine-Endocrinology, Memorial University of Newfoundland, 300 Prince Philip Drive, St. John’s, Newfoundland, Canada A1B 3V6. E-mail: ckovacs{at}mun.ca.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(Ctcgrp) null mice to determine whether calcitonin also modulates lactational mineral metabolism. During 21 d of lactation, spine bone mineral content dropped 53.6% in Ctcgrp nulls vs. 23.6% in wild-type (WT) siblings (P < 0.0002). After weaning, bone mineral content returned fully to baseline in 18.1 d in Ctcgrp null vs. 13.1 d in WT (P < 0.01) mice. Daily treatment with salmon calcitonin from the onset of lactation normalized the losses in Ctcgrp null mice, whereas calcitonin gene-related peptide- or vehicle was without effect. Compared with WT, Ctcgrp null mice had increased circulating levels of PTH and up-regulation of mammary gland PTH-related protein mRNA. In addition, lactation caused the Ctcgrp null skeleton to undergo more trabecular thinning and increased trabecular separation compared with WT. Our studies confirm that an important physiological role of calcitonin is to protect the maternal skeleton against excessive resorption and attendant fragility during lactation and reveal that the postweaning skeleton has the remarkable ability to rapidly recover even from losses of over 50% of skeletal mineral content. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clinical studies have demonstrated (by dual x-ray absorptiometry) that a woman will lose an average of 5–10% of trabecular mineral content during 6 months of near-exclusive lactation and then fully regain it within several months of weaning (reviewed in Refs. 1, 3 , and 4). Depending upon a woman’s peak bone mass, these losses during lactation are sufficient enough to temporarily reduce her bone density into the osteopenic or osteoporotic range. In contrast, mice and rats will lose approximately 30% of skeletal mineral content (by dual x-ray absorptiometry or by ash weight) during 3 wk of lactation and then fully restore it within 2–3 wk after weaning (1). The rapid and complete recovery from the large losses of skeletal mineral induced by lactation is a clinically important finding, because the adult skeleton normally recovers slowly and incompletely from insults induced by glucocorticoids, prolonged bed rest, menopause, GnRH analog treatment, and weightlessness (1, 5, 6, 7, 8, 9, 10).
The mechanism by which the skeleton demineralizes during lactation is not fully understood (1, 11, 12, 13). Mammary tissue dramatically up-regulates the expression and release of PTH-related protein (PTHrP), which enters the mother’s bloodstream to stimulate bone resorption (12, 14, 15) and, in so doing, mimics the actions of sustained high levels of PTH. At the same time, hyperprolactinemia and other factors suppress the release of GnRH within the hypothalamus, thereby suppressing in sequence the pituitary gonadotropins, ovaries, ovulation, and the sex steroids (especially estradiol) (16). In addition, low estradiol levels of lactation permit enhanced osteoclast-mediated bone resorption and, thereby, potentiate the effect of PTHrP on the skeleton (13).
Calcitonin was originally thought to be an important regulator of calcium and bone metabolism in adults (17), but evidence gathered over the years has largely forced the conclusion that it is a vestigial hormone in mammals (18). The effects of calcitonin to reduce osteoclast-mediated bone resorption (to treat osteoporosis or Paget’s disease) or to enhance urinary calcium excretion (to treat hypercalcemia of malignancy) require pharmacological doses of the hormone and not physiological levels (19, 20). However, several investigators had originally proposed that calcitonin might protect the skeleton against excessive resorption during pregnancy and lactation (21, 22, 23, 24). The experimental approach to test this hypothesis had been to remove the thyroid glands from sheep, goats, or rats, to replace thyroid hormone, and to determine the effect of a full cycle of pregnancy and lactation on the maternal bone mineral content (BMC) or ash weight of these thyroidectomized vs. sham-operated animals (21, 22, 23, 24). Although some investigators determined that these presumably calcitonin-deficient animals had reduced BMC at the time of weaning, other investigators found no such effect (21, 22, 23, 24). It wasn’t appreciated at the time that thyroidectomized animals were not devoid of calcitonin because of extrathyroidal sources of calcitonin in mammary tissue (25, 26), placenta (27, 28), and pituitary (29). Thyroidectomized women experience increases in circulating calcitonin during lactation and express normal levels of calcitonin in milk (26).
We have retested the hypothesis that physiological levels of calcitonin are required to protect the maternal skeleton against excessive resorption during pregnancy and lactation. To do this, we used an established murine model in which exons 3–5 of the calcitonin/calcitonin gene-related peptide- gene (Ctcgrp) have been ablated (30). Calcitonin gene-related peptide- (CGRP) is an alternative splice product of the calcitonin gene, whereas a second gene produces CGRPß. Therefore, Ctcgrp null mice completely lack calcitonin in all tissues but still produce CGRP through the CGRPß gene. Using this model, we determined that ablation of the gene doubled the losses of maternal mineral content during lactation and that the effect was specific to loss of calcitonin and not CGRP.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Reproductive cycles and data collection time points
These studies were completed according to an approximate 70-d reproductive cycle. Baseline scans were collected for 5–10 d before mating and subsequent pregnancy (prepregnancy interval). Pregnancy lasted 19 d, lactation lasted 21 d, and postweaning recovery was 21 d. The four main data collection points included prepregnancy, end of pregnancy (day before expected delivery), end of lactation (forced weaning at d 21 of lactation), and end of postweaning recovery of the skeleton (d 21 after weaning). Additional time points included start of lactation (24–48 h), early lactation (5–7 d), and mid-lactation (10–14 d).
BMC and bone mineral density (BMD)
BMC and BMD were measured using the PIXImus 2 Bone Densitometer (GE Lunar, Madison, WI) and analyzed with PIXImus software version 2.1. A standard phantom (fat 11.9% and BMC 0.063 g) was used to calibrate the PIXImus daily. Anesthesia was induced with isoflurane (Baxter Corp., Toronto, Ontario, Canada) and maintained with a single ip injection of either thiopental sodium (Abbot Laboratories, Toronto, Ontario, Canada) or a 5:1 combination of ketamine hydrochloride (Wyeth Canada, Guelph, Ontario, Canada) and xylazine (Bayer Inc., Toronto, Ontario, Canada). The anesthetized mice were immobilized prone on a plastic tray with the spine straightened, and to maintain reproducibility with less than 1% precision error, the head was excluded in all scans. In additional quality control studies, mice on d 18.5 of pregnancy were scanned immediately before and after a cesarean section in which the pups were removed; this determined that the fetal skeletons contributed 1% or less to the apparent maternal BMC or BMD and were, therefore, negligible. Total body (minus head) and regional (spine and hind limb) BMC and BMD measurements were obtained for each mouse, and the absolute values were normalized to the nonpregnant baseline measurements. In the initial studies to determine the pattern of BMC and BMD change (comprising data in Fig. 1
), each mouse was scanned every 2 d throughout the entire reproductive cycle. In later studies in which tissue would be harvested at specific time points, the mice were scanned at the main time points to ensure that the expected BMC and BMD changes were occurring. There were a few anesthetic-related deaths, but otherwise, the outcomes of the reproductive cycle (live pups and magnitude of BMC loss and recovery) were no different between dams that had been scanned every 2 d vs. those that had been scanned at the main time points. Pups that had received in utero exposure to x-rays were euthanized at weaning, whereas the dams were euthanized at the end of the postweaning recovery phase.
|
Milk collection and analysis
On d 7 of lactation, anesthetized mice were injected with 1.2 U oxytocin (Hospira Healthcare Co., Montreal, Quebec, Canada). After 1–2 min, milk was obtained using a self-designed pump consisting of a 1-ml pipette tip (Mandel Scientific Co. Inc., Guelph, Ontario, Canada) inserted into thick rubber tubing that in turn was connected to vacuum suction. The wide end of the pipette tip was placed over a nipple, and a gentle pumping action was used to obtain milk, which was then stored at –20 C until assayed. The milk was diluted 1:100 with distilled water and analyzed for calcium content using a Model 2380 Atomic Absorption Spectrophotometer (PerkinElmer, Norwalk, CT). Because much of the calcium content of milk is bound to proteins (32), and to control for differing milk concentrations, the protein content of milk was measured using the BCA Protein Assay Kit (Pierce, Rockford, IL). Final results are expressed as micrograms calcium per microgram protein.
Duodenal calcium absorption
We adapted the method of Fleet and colleagues (33) to measure duodenal calcium absorption in vivo during mid-lactation and in nonpregnant mice. Food was removed from the cages 3 h before the experiment to empty the duodenum and jejunum. Under anesthesia, an abdominal incision was made and a closed duodenal loop was created by tying two ligatures, one at the pyloric juncture and the second one about 1.5 cm distal to the first one. Forty microliters of 45Ca buffer (150 nM NaCl, 2 mm CaCl2, 30 mM Tris-HCl, and 16 µCi 45Ca/ml at pH 7.4) were injected into the closed loop (45Ca was obtained from Promega/Fisher Scientific Ltd., Burlington, Ontario, Canada). After 10 min, the loop was inspected to ensure that no areas of ischemia were evident. The loop was excised, solubilized in Scintigest (Promega/Fisher), and then counted on a liquid scintillation counter. To measure the total 45Ca activity administered, and to control for the effect of biological tissue to quench the apparent level of radioactivity, additional scintillation vials were prepared in which the same amount of 45Ca and an identical length of nonradioactive duodenum were added. Efficiency of 45Ca absorption was expressed as (1 – 45Ca remaining in each loop/amount of 45Ca injected into the loop) x 100.
Biomechanical analysis
Left hind limbs were stored at –20 C after first removing the soft tissues and then thawed to room temperature in normal saline before analysis. Cortical bone strength was assayed using an Instron Series 3340 electromechanical test instrument (Instron, Norwood, MA). In brief, the femur was immobilized at each end in a fixture, and a load cell with maximum capacity of 10 N of force was positioned 2–3 cm above the mid-shaft of the tibia. The force required to break the femur (failure) was recorded, as was the load-displacement curve from which other biomechanical parameters could be derived.
Rescue experiments
Beginning on the day of delivery and for 21 d of lactation, selected WT and Ctcgrp null mice received once daily im injections of 0.9% saline (vehicle), salmon calcitonin (Rhone-Poulenc Rorer, Collegeville, PA), or rat CGRP (American Peptide Co., Sunnyvale, CA). Mice were briefly anesthetized with isoflurane to receive the injection. Bone densitometry was used every second day throughout the experiment to precisely determine the excursion of BMC.
Microcomputed tomography (µCT) analysis
Right hind limbs and lumbar spines were harvested, adjacent flesh was removed, and then the samples were fixed in paraformaldehyde for 18 h. The bones were then rinsed and stored in 10x PBS at room temperature until analysis. The quantitative and qualitative µCT analysis for each femur and for the first and second lumbar vertebrae was completed at the Centre for Bone and Periodontal Research at McGill University using the SkyScan 1072 (SkyScan, Aartselaar, Belgium), as previously described (34).
Mammary tissue collection and storage
Mammary glands were harvested after a rapid cervical dislocation during late pregnancy, mid-lactation, and late lactation. Lymph nodes were removed and mammary tissue was washed with water containing 0.1% diethylpyrocarbonate. Tissues were either snap-frozen in liquid nitrogen and then placed at –70 C or placed immediately in 10% formalin [37% formaldehyde, 10x PBS (pH 7.3), and distilled water].
RNA extraction and cDNA synthesis
RNA was extracted from whole mammary glands using an RNeasy Midi Kit (QIAGEN, Mississauga, Ontario, Canada). Approximately 100 g of mouse mammary gland yielded 0.6–1.3 µg RNA/µl solution. cDNA synthesis was carried out with the SuperScript III First-Strand Synthesis System (Invitrogen Canada Inc., Burlington, Ontario, Canada).
Real-time quantitative RT-PCR
cDNA synthesis was performed using the Superscript III First-Strand Synthesis System (Invitrogen) as previously described (35). Quantitative real-time RT-PCR data and analysis were obtained using the ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Foster City, CA), also as previously described (35). For the calcitonin receptor, the calcium receptor, and the vitamin D receptor, TaqMan Gene Expression Assays were used that included predesigned primers and probes for optimal amplification (Applied Biosystems) along with TaqMan Universal PCR Master Mix (Applied Biosystems). All gene expression assays have a FAM reporter dye at the 5' end of the TaqMan MGB probe and a nonfluorescent quencher at the 3' end of the probe. Other gene-specific primers used for this study were as follows: for PTHrP, forward 5'-TCCACACAGCCGAAATCAGAGCTA-3' and reverse 5'-TTCTCCTGTTCTCTGCGTTTCCCA-3'; for 1--hydroxylase, forward 5'-CCAGAGCGCTGTAGTTTCTCATCA-3' and reverse 5'ATGAAGGTTTCTGTGTCAGGAGGG-3'; for calcium transport protein subtype 1 (CaT1), forward 5'-ATCGATGGCCCTGCGAACT-3' and reverse 5'-CAGAGTAGAGGCCATCTTGTTGCTG-3' (36); and for epithelial calcium channel (ECaC or CaT2), forward 5'-ATTGACGGACCTGCCAATTACAGAG-3' and reverse 5'-GTGTTCAACCCGTAAGAACCAACGGTC-3' (36). These primers were used along with SYBR Green PCR Master Mix (Applied Biosystems). The real-time PCR thermal cycler protocol included one cycle at 50 C for 2 min, one cycle at 95 C for 10 min, and 40 cycles of 95 C for 15 sec and 1 min at 60.0 C. For primers using SYBR Green PCR Master Mix, the real-time PCR program was the same but without the first cycle at 50 C for 2 min. All samples were analyzed in triplicate. Relative expression ratios were representative of the threshold cycle (the PCR cycle at which an increase in reporter fluorescence is above a baseline signal) normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and compared with the WT.
Statistical analysis
Data were analyzed using SYSTAT 5.2.1 for Macintosh (SYSTAT Inc., Evanston, IL). ANOVA was used for the initial analysis; a post hoc test (Tukey’s) was used to determine which pairs of means differed significantly from each other. Real-time PCR results were analyzed by the 2–
CT method where the target and reference are amplified in separate wells (37). Two-tailed probabilities are reported, and all data are presented as mean ± SE.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The difference was present from early in lactation; WT mice lost 4.5 ± 3.6 and 15.7 ± 3.7% of baseline BMC by 7 and 14 d of lactation, respectively, whereas Ctcgrp null mice lost 10.1 ± 3.1 and 29.6 ± 3.1% of baseline BMC at the same time points. Both genotypes restored their BMC to their respective baseline values after weaning (Fig. 1A), although full recovery took several more days for Ctcgrp null mice (Fig. 1B). Closer examination of the BMC excursion indicated that the lactational losses were much more profound in the trabecular-rich vertebral bodies, such that Ctcgrp null mice lost more than half of their baseline BMC in the spine (Fig. 1C). At the hind limb, which contains relatively more cortical bone than the vertebral bodies, the losses were proportionately less, but Ctcgrp null mice still lost twice the WT value (Fig. 1D). The losses were not a result of any difference in litter sizes, which did not require any culling and were 8.9 ± 0.9 in WT and 7.6 ± 1.2 in Ctcgrp null (P = not significant).
The enhanced skeletal demineralization during lactation indicates increased bone resorption in the Ctcgrp null mice. Consistent with this, the urinary excretion of deoxypyridinoline/creatinine (a marker of bone resorption) was increased significantly in Ctcgrp null vs. WT mice during lactation (Fig. 2A). On the other hand, no difference was observed in the serum level of osteocalcin (a marker of bone formation) between Ctcgrp null and WT mice at any time point (Fig. 2B).
|
); however, Ctcgrp null mice were transiently hypercalcemic during the first 24–48 h postpartum at 1.40 ± 0.02 mmol/liter vs. 1.26 ± 0.02 mmol/liter in WT mice (P < 0.001). Urinary excretion of calcium normally decreases during lactation, likely because of the effect of PTHrP to enhance urinary calcium reabsorption. Ctcgrp null and WT mice both showed the same trend to decreased urinary calcium excretion during lactation but with no significant differences between the genotypes at any time point (Fig. 3B).
|
). Although PTH was also reduced during pregnancy in Ctcgrp null mice, PTH became significantly elevated in Ctcgrp null mice during lactation and may have contributed to the enhanced demineralization observed in these mice (Fig. 3C). The elevation in PTH occurred during the interval of normocalcemia and not during the initial 24–48 h postpartum in which the ionized calcium was transiently elevated. In a separate analysis of serum obtained at 24–48 h (to correspond with the transient hypercalcemia), PTH levels were suppressed to the limit of detection (1.6 pg/ml) in both WT and Ctcgrp null mice (P = not significant). Estradiol is normally suppressed during lactation as a consequence of hyperprolactinemia, and low estradiol contributes to the skeletal resorption during this time period. We found that estradiol levels at mid-lactation were suppressed to the level of sensitivity of the estradiol assay (8 pg/ml) in all samples from both WT and Ctcgrp null mice (P = not significant). Prolactin levels (Fig. 3D) were also not significantly different between Ctcgrp null and WT mice at any time point.
We assayed duodenal 45Ca absorption at mid-lactation in WT and Ctcgrp null mice to rule out impaired intestinal absorption as an indirect cause of the enhanced skeletal resorption in Ctcgrp null mice. Efficiency of absorption was greater than 80% at 10 min in both WT and Ctcgrp null mice and not significantly different from the nonpregnant WT value (Fig. 4A).
|
).
PTHrP in lactating mammary tissue and circulation
Given the established importance of mammary-derived PTHrP to stimulate skeletal demineralization during lactation, we measured both mRNA and protein levels of this hormone. Within mammary tissue, PTHrP mRNA was significantly elevated as assessed by real-time quantitative RT-PCR (Fig. 5A), and the protein level was visibly increased by immunohistochemistry (Fig. 5B). Within the circulation, the plasma PTHrP level showed no significant difference during lactation compared with WT, with most values clustered near the assay detection limit of 2 pmol/liter (Fig. 5C).
|
-hydroxylase, and the vitamin D receptor. In addition, mRNAs for the epithelial calcium channels (CaT1 and CaT2) were also examined. There was no difference in the expression of any of these mRNAs between WT and Ctcgrp null mice, whereas CaT1 and CaT2 were not detected (data not shown).
Skeletal strength and microarchitecture
We also examined the consequences of enhanced demineralization during lactation on skeletal microarchitecture of the femur and lumbar vertebrae and on the strength of the tibia. µCT measurements demonstrated that Ctcgrp null mice had greater trabecular thickness and more closely spaced trabeculae at baseline compared with WT in both femora and lumbar vertebrae (Table 1). Ctcgrp null mice experienced more marked trabecular thinning and reached an equivalent level to WT at the end of lactation (Table 1). Trabecular spacing increased significantly in Ctcgrp null femora and approached the level of the WT at baseline, whereas trabecular spacing did not change significantly within vertebrae (Table 1). The µCT results also indicate a greater trabecular bone volume at baseline in Ctcgrp null mice vs. WT. Whereas WT trabecular bone volume increased significantly during lactation (a normal finding), the changes were blunted in Ctcgrp null bones at both the femora and vertebrae (Table 1).
|
). Not included in the calculations were two Ctcgrp null femurs that were evidently fragile because they broke upon being placed into the device and thus could not be subjected to the three-point bend test. Additional intrinsic and extrinsic bone parameters are shown in Table 2 that suggest that cortical bone strength is not significantly different between WT and Ctcgrp null mice at any time point.
|
, or both. Preliminary dose-finding experiments indicated that a mean dose of 10 IU salmon calcitonin daily would prevent the excessive losses of lactation in Ctcgrp null mice; larger doses reduced the normal losses during lactation, whereas 1 IU daily had minimal effect (not shown). Starting at the day of delivery and continuing for 21 d, WT and Ctcgrp null mice were treated with daily injections of calcitonin or vehicle, and several Ctcgrp null mice were treated with an equimolar dose (255 µg) of CGRP. The mean losses of BMC observed in saline or CGRP-treated mice were similar, showing enhanced losses approximately double that of WT (Fig. 6). In contrast, treatment with calcitonin resulted in lactational losses that were no different from those observed in saline-treated WT mice (Fig. 6). Calcitonin treatment was without visible effect in WT mice (Fig. 6).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, confirming that the skeletal phenotype is specific to loss of calcitonin and not to loss of CGRP. Finally, Ctcgrp null mice fully regained to their baseline BMC after weaning, demonstrating the skeleton’s tremendous ability to repair itself from a greater than 50% deficit.
Physiological role for calcitonin during lactation
Our confirmation of a physiological role for calcitonin during lactation brings closure to a hypothesis that had been raised more than 30 yr ago but left unresolved (21, 22, 23, 24). The inconsistent results among these early investigations likely reflected the confounding effects of thyroid ablation and thyroid hormone replacement, but especially the previously unrecognized fact that extrathyroidal sources of calcitonin existed (25, 27, 29). Only through the use of a gene ablation model has it become possible to examine the effect of complete absence of calcitonin on skeletal metabolism during pregnancy and lactation.
Ctcgrp null mice have enhanced skeletal resorption as demonstrated by the large drop in BMC from either prepregnancy or peak of pregnancy to end of lactation, the trabecular thinning and increased trabecular separation on µCT, and the elevated excretion of urinary deoxypyridinoline/creatinine. The transient hypercalcemia in the first 24 h postpartum may be a consequence of enhanced bone resorption flooding the maternal circulation with calcium at a time when milk production is not fully established; the ionized calcium was normal throughout the rest of lactation.
Calcitonin evidently plays a central role in modulating skeletal responsiveness during lactation, a role that may not be limited to its direct actions on osteoclasts. The presence of calcitonin and its receptor in mammary tissue (25) and pituitary (29, 40) indicates that there are several possible routes through which absence of calcitonin may lead to enhanced bone resorption in Ctcgrp null mice. We will consider the osteoclast, pituitary, and mammary tissue in turn.
Calcitonin, osteoclasts, and increased trabecular bone resorption
During lactation, enhanced osteoclast-mediated bone resorption is driven by a hormonal milieu in which circulating PTH is low, PTHrP is elevated, estradiol is low, and calcitonin is elevated (1, 3, 4). In Ctcgrp null mice, circulating PTH was elevated, mammary gland expression of PTHrP was up-regulated above normal, estradiol was low, and calcitonin was absent. The circulating level of PTHrP was not significantly increased in Ctcgrp null mice, but the assay sensitivity (2 pmol/liter) was likely insufficient to detect a true difference given that most values were clustered near the detection limit. PTH and PTHrP are equivalent in their ability to activate the PTH/PTHrP receptor on osteoblasts and communicate through the RANK-RANKL-OPG pathway to stimulate osteoclast-mediated bone resorption (41, 42); in contrast, calcitonin acts directly on osteoclasts to inhibit bone resorption (18). Thus, the mechanism for enhanced bone resorption in the Ctcgrp null mice may be simply a result of exaggerated stimulation of PTH/PTHrP receptor-mediated bone resorption by both PTH and PTHrP, coupled with absence of the effect of calcitonin to directly inhibit the activity of osteoclasts. In turn, pituitary and mammary tissue may both contribute to the up-regulation in PTH and PTHrP as described below.
Calcitonin, pituitary, and prolactin
With respect to the pituitary, both pharmacological administration of calcitonin (43) and targeted overexpression of calcitonin within pituitary lactotrophs have been found to inhibit production of prolactin (44). Therefore, loss of calcitonin might be expected to lead to enhanced synthesis and release of prolactin, which in turn would cause more profound ovarian suppression, lower estradiol, and enhanced bone resorption. Low estradiol (45) and high prolactin (15, 46, 47, 48, 49) are factors that stimulate PTHrP production by mammary tissue in vivo. However, we found no significant difference in mean prolactin levels at any time point in Ctcgrp null vs. WT mice, and estradiol levels were undetectable in both genotypes during mid-lactation. Prolactin is released in pulsatile fashion during lactation (16), but it was not possible to collect multiple serum samples from each mouse to assess peak levels of this hormone. Our results indicate that the proposed role of calcitonin to regulate pituitary lactotrophs and systemic prolactin levels may not be a physiologically important pathway.
Calcitonin in mammary tissue
The presence of calcitonin and its receptor within mammary tissue (25) may indicate a local regulatory role for calcitonin in which the loss of it leads to enhanced PTHrP production. The calcium receptor has been shown to locally regulate the production of PTHrP within mammary tissue as well as the calcium and fluid content of milk (50). We did not observe a change in calcium receptor mRNA expression by real-time quantitative RT-PCR, and thus the pathway by which absence of calcitonin leads to stimulation of PTHrP may be independent of the calcium receptor.
The factors that control milk production are not well understood, although it is known that the calcium receptor, PTHrP, prolactin, and suckling all contribute to the regulation of the calcium and fluid content of milk (12, 50, 51). The numerous calcium regulatory genes that are expressed by lactating mammary tissue include the calcium receptor, calcitonin receptor, 1--hydroxylase, and vitamin D receptor (25, 50, 52). We did not observe any alteration in the expression of these genes within mammary tissue; the only consistent finding was an elevation in PTHrP in mammary tissue that lacks calcitonin. No previous publications have examined CaT1 or CaT2 mRNA expression in lactating mammary tissue, although CaT1 protein had previously been detected by immunohistochemistry in nonlactating human breast tissue (53). We did not detect CaT1 or CaT2 mRNA in lactating mammary tissue of WT or Ctcgrp null lactating mice, which suggests that the epithelial calcium channels are not involved in milk production.
Secondary hyperparathyroidism in Ctcgrp null mice
Normally PTH is suppressed during lactation in humans and rodents (1), but we observed a significant elevation in PTH in Ctcgrp null mice during mid-lactation and to a lesser degree during late lactation. The cause of the elevated PTH may be an indirect consequence of loss of calcitonin. It has previously been demonstrated that larger litter sizes or a restricted calcium intake will induce maternal hypocalcemia and secondary hyperparathyroidism during lactation (54, 55, 56). In such cases, loss of calcium into milk may overwhelm the mother’s ability to maintain her own blood calcium, resulting in hypocalcemia that in turn acts through the calcium receptor to induce PTH release. Ctcgrp null mice have low urinary calcium excretion, normal duodenal calcium absorption, and enhanced skeletal resorption; these factors should obviously lead to increased availability of calcium for milk production. The fact that PTH is elevated in Ctcgrp null mice may indicate that the loss of calcium into milk exceeds the requirement to maintain a normal blood calcium in the mother. Milk production by mammary tissue is controlled by both local and systemic factors (16, 57) and can induce hypocalcemia and tetany (milk fever) in lactating animals (58, 59); therefore, it is conceivable that loss of calcitonin within mammary tissue leads to an up-regulation in milk production that, in turn, challenges the ability of the mother to maintain her own serum calcium.
Confounding influences on bone strength
Ctcgrp null mice lost twice as much mineral content as WT siblings, and yet skeletal strength (as assessed by the three-point bend test) was not significantly different from WT at the end of lactation. This likely reflects several factors. First, the three-point bend test examines the strength of cortical bone, which undergoes very modest resorption during lactation (1, 60). Second, it is trabecular bone that is rapidly mobilized during lactation because of its large surface area and vascularity, factors that enable it to be resorbed by osteoclasts on multiple surfaces (60, 61). Third, absence of calcitonin from conception leads to increased trabecular bone mass, as indicated in the original report (30) and in our study in which the Ctcgrp null skeleton had thicker and more densely packed trabeculae compared with WT. Thus, even a 50% drop in skeletal mass was well tolerated by the Ctcgrp null mice because the large losses resulted in a trabecular thickness that was equivalent to the level of WT mice at the end of lactation. Finally, pregnancy and lactation are normally associated with other changes in skeletal architecture (increased bone volume and expansion of periosteal surfaces while endosteal bone is being lost) that will offset the weakening effect that skeletal demineralization during lactation would be expected to induce (62). Ctcgrp null mice have greater trabecular and total bone volumes than WT at baseline; although these volumes increased substantially in WT mice, the changes were comparably modest in Ctcgrp null mice. The larger bone volumes of the Ctcgrp null mice likely compensated for the loss of trabecular bone during lactation.
Relevance to lactation in humans
A few case series examined male and female thyroidectomized individuals who received thyroid hormone replacement and reported conflicting results as to whether such patients developed a long-term deficit in bone mass or an increase in fracture risk relative to matched individuals with intact thyroids (63, 64, 65, 66, 67). No study has examined the bone mass of thyroidectomized women after a cycle of pregnancy and lactation, and of course such women would retain the ability to make calcitonin in pituitary and mammary tissue and would not be fully calcitonin deficient (26). Vertebral compression fractures occasionally occur during lactation in otherwise healthy young women, a condition termed osteoporosis of lactation (68). In such cases, excess bone resorption during lactation is presumed to have led to increased fragility that resolves after weaning. Our results raise the possibility that at least some of these cases might represent the human equivalent of absence of calcitonin or its receptor.
Skeletal recovery after weaning
In contrast to the demineralization that occurs during lactation, the factors that are required to stimulate recovery of the skeleton after weaning remain unknown. Our results indicate that calcitonin is not required for skeletal recovery to be achieved in full, although its absence may contribute to a delay in recovery. Calcitonin is required in the short term during lactation to prevent excessive resorption of the skeleton and likely attendant fragility, but it is not required in the long term because the skeleton recovers fully despite its absence and despite the increased mineral deficit induced by lactation. Resolving the mechanism by which the skeleton recovers remains a clinically important and relevant task because the skeleton is able to recover so completely and rapidly from the large losses of skeletal mineral content that are induced by lactation. Ctcgrp null mice may be an ideal model to examine skeletal recovery because the factors stimulating recovery are likely to be up-regulated more in these mice compared with normal.
Breast, brain, and bone circuit
In summary, evidence from recent studies has pointed to the central role of mammary tissue to stimulate skeletal resorption during lactation by producing PTHrP under the control of the calcium receptor and also to the central role of the pituitary to secrete prolactin and, thereby, suppress estradiol and stimulate PTHrP (the breast, brain, and bone circuit, Fig. 7). The data in our manuscript indicate that calcitonin plays a key role by modulating the extent of skeletal resorption during lactation. Absence of calcitonin in the systemic circulation will lead to increased osteoclast activity through loss of its direct effects via the calcitonin receptor to inhibit osteoclast number and function. However, because the calcitonin receptor is also expressed in mammary tissue and pituitary (two tissues that also produce calcitonin), it is possible that calcitonin acts at all three sites to modulate lactation: directly to inhibit osteoclasts, within pituitary to inhibit prolactin release, and within mammary tissue to modulate PTHrP or milk production. Additional work is needed to determine the relative importance of these three physiological loops.
|
|
Footnotes |
|---|
Results from this work were presented in part at the 2003 American Society for Bone and Mineral Research annual conference for which J.P.W. received a Young Investigator Award and at the Canadian Society for Endocrinology and Metabolism conferences for which J.P.W. received the Antoni Nalecz Award, the Harold Copp Award for best abstract, and two travel awards.
The authors have no conflicts of interest to disclose.
First Published Online May 4, 2006
Abbreviations: BMC, Bone mineral content; BMD, bone mineral density; CaT1, calcium transport protein subtype 1; CGRP, calcitonin gene-related peptide-; µCT, microcomputed tomography; WT, wild type.
Received December 19, 2005.
Accepted for publication April 21, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT method. Methods 25:402–408[CrossRef][Medline]
