- Email: [email protected]
Hypercalcemia produced by parathyroid hormone suppresses experimental autoimmune encephalomyelitis in female but not male mice
Archives of Biochemistry and Biophysics 442 (2005) 214–221 www.elsevier.com/locate/yabbi
Hypercalcemia produced by parathyroid hormone suppresses experimental autoimmune encephalomyelitis in female but not male mice Terrence F. Meehan, Janeen Vanhooke, Jean Prahl, Hector F. DeLuca ¤ Department of Biochemistry, University of Wisconsin-Madison, 433 Babcock Drive, Madison, WI 53706-1544, USA Received 20 July 2005 Available online 2 September 2005
Abstract Besides its role in regulating serum levels of calcium and phosphorus, 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3) has potent eVects on the immune system and suppresses disease in several animal models of autoimmune disorders including experimental autoimmune encephalomyelitis (EAE), a mouse model of multiple sclerosis. While the amount of 1,25-(OH)2D3 needed to prevent EAE is dependent on the gender of the mouse and amount of calcium available in the diet, the minimum levels of 1,25-(OH)2D3 suYcient to prevent disease cause hypercalcemia. To test if hypercalcemia independent of high levels of 1,25-(OH)2D3 can suppress EAE, we used a 25-hydroxyvitamin D3-1-hydroxylase (1-hydroxylase) knockout mouse strain. Because these 1-hydroxylase knockout mice lack the parathyroid hormone (PTH)-regulated enzyme that synthesizes 1,25-(OH)2D3, hypercalcemia from increased bone turnover was created by continuous administration of PTH without changing the circulating levels of 1,25-(OH)2D3. This PTH-mediated hypercalcemia generated after EAE induction prevented disease in female mice but not male mice. When hypercalcemia was prevented by diet manipulation, PTH administration no longer prevented EAE. We conclude that hypercalcemia is able to prevent EAE after disease induction in female mice. 2005 Elsevier Inc. All rights reserved. Keywords: 1; 25-Dihydroxyvitamin D3; Calcium; Experimental autoimmune encephalomyelitis; Parathyroid hormone; Multiple sclerosis; 25-Hydroxyvitamin D3-1a-hydroxylase; CYP27B1; Phosphorus; Autoimmune; Sex eVect
The active metabolite of vitamin D, 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3),1 acts as a hormone to regulate calcium and phosphorous levels in blood by increasing calcium and phosphorus absorption from the duodenum of the small intestine, and by working in concert with parathyroid hormone (PTH) to release calcium and phosphorus from bone, and to increase the reabsorption of calcium in the kidney [1]. In addition to its role in mineral homeostasis, 1,25-(OH)2D3 has wide ranging impacts on the immune *
Corresponding author. Fax: +1 608 262 7122. E-mail address: [email protected] (H.F. DeLuca). 1 Abbreviations used: 1,25-(OH)2D3, 1,25-dihydroxyvitamin D3; 1-hydroxylase, 25-hydroxy-vitamin D3-1-hydroxylase; EAE, experimental autoimmune encephalomyelitis; mg/dl, milligrams per deciliter; MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; PTH, parathyroid hormone. 0003-9861/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2005.08.011
system including inhibitory eVects on maturation of dendritic cells, the skewing of CD4+ T-cells upon activation to a less inXammatory Th2 cytokine proWle, and promoting hematopoietic cells to diVerentiate into a monocyte/macrophage pathway [2–5]. These eVects are reXected in the immunosuppressive ability of 1,25-(OH)2D3 in a number of animal models for autoimmune disorders including experimental autoimmune encephalomyelitis (EAE), non-obese diabetes, inXammatory bowel disease, arthritis, and lupus [6–11]. However, it is not clear if this suppression of disease is entirely dependent on the vitamin D hormone, or if the concurrent rise in serum calcium upon administration of 1,25-(OH)2D3 has any impact. The possibility of hypercalcemia being immunosuppressive is most clear in the EAE mouse model of multiple sclerosis (MS). Multiple sclerosis is CD4+ T-cell mediated autoimmune disease that results in demyelination of axonal tracts in the
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
central nervous system causing a wide range of neurological symptoms [12]. Epidemiology studies have revealed genetic and environmental factors that modify disease incidence and progression including a gender bias where women are more likely to develop disease than men, remission of MS during pregnancy, and a geographic distribution that suggests a correlation between MS prevalence and vitamin D deWciency [13–15]. Recently, a signiWcant inverse correlation was found between vitamin D supplementation and the risk of developing MS in a study involving a large group of women [16]. While the event in MS that leads to activation and proliferation of these disease eVector T-cells is unknown, a similar disease can be induced in rodents by immunization of antigens present in the central nervous system. Besides diVerent triggering mechanisms, these EAE animal models share many characteristics of MS. This includes an “activation phase” where antigen presenting cells process the immunized antigen, migrate to the lymph nodes and present immunodominant peptides to naive T-cells; and an “eVector phase” where CD4+ T-cells that recognize antigen proliferate and cross the blood–brain barrier to lead an inXammatory attack that results in demyelinated lesions. 1,25-(OH)2D3 has been shown to prevent the occurrence and progression of this disease in diVerent EAE models in both rats and mice [6,7,17]. In mice, it was established that the dose of 1,25-(OH)2D3 needed to suppress disease was dependent on the sex of the animal and the amount of calcium present in the diet. Females required smaller doses of 1,25-(OH)2D3 to suppress disease than males and that less hormone was needed when the amount of calcium in the diet was increased [18]. In either gender, complete suppression of the disease was not observed until severe hypercalcemia (serum calcium levels >12 mg/dl) was established. This led us to develop a hypothesis that hypercalcemia, dependent or independent of 1,25-(OH)2D3, might account for the action of 1,25(OH)2D3 in preventing disease onset. To separate the eVects of hypercalcemia and 1,25(OH)2D3 in vivo is diYcult due to the tight regulation of the synthesis of 1,25-(OH)2D3 by serum calcium levels and PTH. PTH, an 84 amino acid peptide, is secreted by the parathyroid gland within seconds of detection of a lowering of serum calcium by the calcium sensing receptor, a G-protein coupled receptor [19–21]. The active N-terminal fragment of this hormone binds to its receptor in the kidney and increases the activity of the renal cytochrome P450 enzyme CYP27B1, the 25-hydroxy-vitamin D3-1-hydroxylase (1hydroxylase) [22]. This enzyme converts the biologically inactive circulating form of vitamin D3, 25-hydroxyvitamin D3, to 1,25-(OH)2D3. This in vivo regulation is mimicked by subcutaneous injections of PTH in mammals where increased levels of circulating 1,25-(OH)2D3 and increased bone mass are observed without hypercalcemia [23]. Hypercalcemia can be generated by chronically administering PTH by osmotic pumps; however, there are inconsistent results whether this treatment leads to increased or decreased levels of 1,25-(OH)2D3 [23–27].
215
Our research group recently developed a 1-hydroxylase knockout mouse strain similar to the two strains previously reported by other groups [28,29, J.L. Vanhooke, J.M. Prahl, C. Kimmel-Jehan, M. Mendelsohn, E.W. Danielson, K.D. Healy, and H.F. DeLuca, unpublished results]. All three strains of mice are unable to synthesize the active metabolite of vitamin D and display a phenotype similar to the human disorder vitamin D-dependency rickets type I (VDDR-1) including hypocalcemia, hyperparathyoidism, and rickets. The phenotype in these mice can be rescued by daily administration of 1,25-(OH)2D3 or by a high calcium diet containing lactose [30,31]. The development of these 1-hydroxylase knockout strains allows for administration of PTH without altering levels of 1,25-(OH)2D3. When administered continuously, PTH has catabolic actions on bone that leads to increases in serum calcium. These catabolic action appears to be dependent on the presence of 1,25-(OH)2D3 since this eVect is abolished in vitamin D-deWcient animals [32] and in 1-hydroxylase knockout mice (unpublished results). In this paper, we report the development of mouse model of hypercalcemia that is independent of increases of 1,25-(OH)2D3 and the impact this metabolic state has on EAE induced by immunization with an encephalogenic peptide. The hypercalcemia was achieved by giving 1-hydroxylase knockout mice, a small amount of 1,25(OH)2D3 that is insuYcient to prevent EAE, but could rescue the VDDR-1 phenotype and aid in PTH-mediated bone catabolism, and then dosing with pharmacological amounts of the bioactive fragment of human PTH. This PTH-mediated hypercalcemia led to protection against EAE after disease induction in female mice but not male mice. The implications of these Wndings and future uses of our vitamin D independent hypercalcemia model are discussed. Materials and methods Animals and diet 25-Hydroxyvitamin D3-1-hydroxylase knockout mice were recently generated in our laboratory by targeted deletion of exons 1–8. 1-Hydroxylase knockout mice were backcrossed seven to nine times onto a C57Bl/6 strain and genotyped by PCR analysis with primers speciWc for exons 3 (forward) and exons 4 and 5 (reverse). All mice were fed puriWed diets based on those described by Cantorna et al. [7]. Homozygous 1-hydroxylase knockout mice were weaned onto a rescue diet containing 0.47% calcium and 5 ng of 1,25-(OH)2D3 per 4 g of diet. Starting at 5–6 weeks of age, 1-hydroxylase knockout mice received 4 g per day of an experimental diet containing 0.87% calcium and diVering amounts of 1,25-(OH)2D3. These diets were made and replaced three times a week for the duration of the studies. Mice in the PTH or vehicle groups received 1 ng of 1,25-(OH)2D3 per day except in one study where mice in both groups received a 20% lactose diet containing 2%
216
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
calcium and 1.25 % phosphorus with no 1,25-(OH)2D3 supplementation. A second group was switched 10 days after EAE induction from the 1 ng of 1,25-(OH)2D3 per day diet to a diet containing 50 ng (female mice) or 200 ng (male mice) of 1,25-(OH)2D3 per day. A third group of 1hydroxylase knockout mice received 37.5 ng of 1,25(OH)2D3 per day from 5 weeks of age until the end of the study. Experimental autoimmune encephalomyelitis induction EAE was induced in the 1-hydroxylase knockout mice backcrossed on the C57Bl/6 strain by subcutaneous immunization in the rear Xanks with 200 g of the immunodominant peptide to myelin oligodendrocyte glycoprotein (MOG) (MEVGWYRSPFSRVVHLYRNGK) [33]. The peptide was synthesized at the University of Wisconsin Biotechnology Center using standard 9-Xuorenyl-methoxycarbonyl chemistry and was judged to be 92% pure by reverse-phase HPLC. The peptide was dissolved in Freund’s complete adjuvant containing 4 mg/ml of heat-killed mycobacterium tuberculosis H37a (Difco Laboratories, Detroit, MI). On the day of immunization and 48 h later, mice were injected intraperitoneally with 200 ng of Bordetella pertussis toxin dissolved in PBS (List Biological Laboratories, Campbell, CA). Mice were examined daily for clinical signs of EAE and scored as follows: 0, no paralysis; 1, loss of tail tone; 2, hindlimb weakness; 3, hindlimb paralysis; 4, forelimb paralysis; 5, moribund or dead.
Perkin–Elmer atomic absorption spectrometer. Serum was diluted in 0.1% LaCl3 for this determination [34]. Serum phosphorus levels were determined as previously described [35]. BrieXy, 25 l of serum was added to 4 ml of a hydrochloric acid solution containing ammonium molybdate and the resulting product of unreduced phosphomolybdate was determined by spectrophotometric measurement (Unicam UV1, Thermo-Spectronic, Rochester, NY). All values were compared to those of reference values measured on the same day. Histopathology Spinal cords from 1-hydroxylase immunized with MOG peptide were removed and Wxed in 10% phosphatebuVered formalin. ParaYn-embedded tissue sections were prepared and stained in luxol fast blue and periodic acid schiVs reagent by the University of Wisconsin Veterinary Science Department. Statistical analysis Statistical analysis was performed using a statistical program for the Macintosh computer (Graphpad Prism). The two-tailed Fisher exact probability test was performed on incidence rates and the unpaired two group Student’s t test was used on other measurements. Values of P < 0.05 were considered statistically signiWcant. Results
Administration of PTH The bioactive, human fragment of PTH (peptides 1–34) was purchased from Bachem (Bachem Californa, Torrence, CA). The peptide was dissolved to a concentration of 1 mg/ ml in a vehicle containing 150 mM NaCl, 1 mM HCl, and 2% heat inactivated sera, as described previously [23]. The sera used were from 1-hydroxylase knockout mice fed a rescue diet and was matched by sex to the group of mice receiving treatment. PTH was administered chronically using Alzet osmotic minipumps model 1002 (Durect Corporation, Cupertino, CA). These pumps deliver 0.25 l/h of solution for a 15 day period. Nine days after EAE induction, mice were weighed and the reservoirs of the pumps were Wlled with vehicle or PTH diluted to deliver 110 g per kilogram bodyweight per day. The pumps were placed in microcentrifuge tubes containing saline and were primed at 37 °C overnight. The next morning, the pumps were placed subcutaneously in the upper back of mice anesthetized with isoXurane. At the end of the study, successful delivery of the solutions was assessed by measuring the remaining volume of liquids in the reservoirs. Serum mineral analysis Blood samples were centrifuged at 3500 rpm (1096g) for 15 min. Serum calcium values were determined using a
Oral feeding of 1,25-(OH)2D3 prevents EAE in 1-hydroxylase knockout mice In the Wrst study, we investigated the eVects of oral administration of 1,25-(OH)2D3 on EAE in mice that lack the 1-hydroxylase gene. Because these mice are unable to produce 1,25-(OH)2D3 [29], a small amount of the hormone must be given to maintain normal serum mineral homeostasis. Pilot studies indicated that 1 ng per day of 1,25-(OH)2D3 given to mice fed a 0.87% calcium diet was suYcient to rescue the mice from abnormal mineral homeostasis while being below the minimal dose needed to prevent EAE (data not shown). To test the ability of 1,25(OH)2D3 to prevent EAE, a 2nd group received 37.5 ng per day of 1,25-(OH)2D3 starting 2 weeks prior to immunization. To demonstrate the ability of 1,25-(OH)2D3 to prevent disease after disease induction, female mice received 50 ng per day of 1,25-(OH)2D3 and male mice received 200 ng per day of 1,25-(OH)2D3 starting at 10 days after immunization, which is before the Wrst clinical signs of EAE are observed. Male mice required more 1,25-(OH)2D3 to protect against EAE, which is consistent with a previous report [18]. Immunization with 200 g of a peptide corresponding to the immunodominant epitope of MOG induced EAE in the 1-hydroxylase knockout mice fed the rescue diet containing 1 ng of 1,25-(OH)2D3 per day (Fig. 1, Table 1). The
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
217
Fig. 1. EAE in 1-hydroxylase knockout mice fed diVering amounts of 1,25-(OH)2D3. At 6 weeks of age, 1-hydroxylase knockout mice were switched to a 0.87% calcium diet containing either 1 ng or 37.5 ng of 1,25-(OH)2D3 per day. Two weeks later, mice were immunized with 200 g of MOG35–55 peptide along with an intraperitoneal injection of pertussis toxin at the time of immunization and 48 h later. Ten days after immunization, one group of mice fed 1 ng of 1,25-(OH)2D3 was switched to a diet containing 50 ng of 1,25-(OH)2D3. Mice were scored for EAE once a day for 24 days. The higher amounts of 1,25-(OH)2D3 protected against EAE while the 1 ng per day amount did not. A statistical analysis appears in Table 1.
Table 1 1,25-(OH)2D3 protects against MOG-induced EAE Incidence
Mean severity
Mean day of onset
Serum calcium (mg%)
Both sexes 1 ng 1,25-(OH)2D3 per day 37.5 ng 1,25-(OH)2D3 per day 50 or 200 ng 1,25-(OH)2D3 starting day 10
91% n D 10/11 0%a n D 0/6 15%a n D 2/13
2.0 § 0.7 — 1.5 § 0.7
16 § 3 — 15 § 2
9.9 § 0.6 13.0 § 0.7a 11.9 § 0.7a
Female 1 ng 1,25-(OH)2D3 per day 37.5 ng 1,25-(OH)2D3 per day 50 ng 1,25-(OH)2D3 starting day 10
100% n D 6/6 0%a n D 0/3 10%a n D 1/10
1.8 § 0.8 — 1b
16 § 3 — 16b
9.7 § 0.5 12.1 § 0.4a 12.0 § 0.6a
Male 1 ng 1,25-(OH)2D3 day 37.5 ng 1,25-(OH)2D3 per day 200 ng 1,25-(OH)2D3 starting day 10
86% n D 6/7 0%a n D 0/3 17%a n D 1/6
2.5 § 0.4 — 2b
15 § 3 — 13b
10.2 § 0.6 13.9 § 0.7a 13.0 § 0.9a
C57Bl/6 mice were switched to a 0.87% calcium diet containing the indicated amounts of 1,25-(OH)2D3 at 6 weeks of age or 10 days post-immunization. Mice were injected subcutaneously with 400 g of MOG35–55 peptide along with an intraperitoneal injection of 200 ng of pertussis toxin at the time of immunization and 48 h later. Mice were scored for EAE once a day for 24 days. At the end of the study, mice were sacriWced and serum calcium concentrations determined by atomic absorption spectrometry. a P < 0.05 compared to mice fed with 1 ng of 1,25-(OH)2D3. b Only one animal sick.
group receiving the 37.5 ng of 1,25-(OH)2D3 prior to immunization was completely protected from developing EAE while mice receiving 1,25-(OH)2D3 10 days after immunization had signiWcantly lower incidence of EAE. Both groups of mice that were protected against EAE also had elevated levels of serum calcium. Continuous administration of PTH increases serum calcium levels and prevents EAE in female mice but not male mice To determine the role of hypercalcemia in the prevention of EAE by pharmacological doses of 1,25-(OH)2D3, we studied the eVect of raising serum calcium levels by continuous administration of PTH. PTH raises serum calcium levels by increasing bone turnover and by increasing
the activity of the 1-hydroxylase enzyme which leads to increased circulating levels of 1,25-(OH)2D3 [23]. In mice that lack the 1-hydroxylase gene, the later eVect of PTH is negated while the former can still occur if small amounts of 1,25-(OH)2D3 are supplied to the mice exogenously. Dosing studies indicated that administration of 110 g of the human active fragment of PTH per kilogram body weight led to consistent hypercalcemia (data not shown). 1-Hydroxylase knockout mice receiving the rescue diet of 1 ng of 1,25-(OH)2D3 per day were immunized with 200 g of MOG peptide. Ten days after immunization, osmotic pumps containing PTH or vehicle were delivered subcutaneously to the upper back. Mice were bled periodically and calcium concentrations determined. Mice receiving PTH had serum calcium concentrations
218
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
Fig. 2. EVects of PTH-treatment on EAE in 1-hydroxylase knockout mice. At 6 weeks of age, 1-hydroxylase knockout mice were switched to a diet containing 0.87% calcium and 1 ng of 1,25-(OH)2D3 per day. Two weeks later, mice were immunized with 200 g of MOG35–55 peptide along with an intraperitoneal injection of pertussis toxin at the time of immunization and 48 h later. 10 days after immunization, osmotic pumps containing 110 g/kg/day of PTH or vehicle were placed subcutaneously in the upper back. Blood was drawn periodically and mice were scored daily for signs of EAE. (A) Levels of serum calcium were determined as described and the averages § the standard deviation were plotted with a statistical analysis appearing in Table 2. (B) Mean EAE scores of 1-hydroxylase knockout mice treated with vehicle or PTH were plotted. PTH-treatment protects against EAE in female but not male mice.
Table 2 PTH-mediated hypercalcemia protects against EAE in female but not male mice Po4 (mg%)
Incidence
Mean severity
Mean day of onset
Serum calcium (mg%)
Female Vehicle + 1 ng 1,25 D3 PTH + 1 ng 1,25 D3 Vehicle + 20% lactose PTH + 20% lactose
75% n D 9/12 0%a n D 0/13 56% n D 9/16 46% n D 7/15
2.7 § 1.1 — 1.4 § 0.5a 1.5 § 0.7a
17 § 4 — 18 § 4 18 § 4
10.2 § 0.5 13.4 § 1.2a 9.3 § 0.5 9.9 § 1.0
6.7 § 1.3 5.8 § 1.5 6.3 + 1.0 6.0 + 1.5
Male Vehicle + 1 ng 1,25 D3 PTH + 1 ng 1,25 D3
73% n D 11/15 72% n D 13/18
2.3 § 1.1 2.5 § 0.9
14 § 3 15 § 2
10.4 § 0.3 13.2 § 1.6a
7.4 § 1.0 5.1 § 1.3a
At 6 weeks of age, 1-hydroxylase knockout mice were switched to a diet containing either 0.87% calcium and 1 ng 1,25-(OH)2D3 or 2% calcium and 20% lactose. At 8 weeks of age, mice were injected subcutaneously with 400 g of MOG35–55 peptide along with an intraperitoneal injection of 200 ng of pertussis toxin at the time of immunization and 48 h later. Ten days after immunization mice received osmotic pumps that released 110 g/kg/day of PTH or vehicle. Mice were scored for EAE once a day for 24 days. At the end of the study, mice were sacriWced and serum calcium and phosphorus concentrations determined as described in the materials and methods section. a P < 0.05 compared to vehicle-treated + 1 ng 1,25-(OH)2D3 mice of the same sex.
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
219
averaging over 13 mg/dl for the 2-week lifespan of the pumps while mice receiving vehicle had normal serum calcium levels (Fig. 2A). There was no statistical diVerence in serum calcium values between male and female mice within a treatment group at any of the time points measured. Surprisingly, hypercalcemia induced by PTH completely prevented EAE in female mice while having no eVect on EAE incidence or progression in male mice (Fig. 2B). Male mice had an incidence of 73% in the vehicle treated group versus 72% in the PTH-treated group while vehicle treated female mice had a 75% incidence rate as compared to no incidence in PTH-treated females (Table 2). In addition to high serum calcium values, male mice receiving PTH also had lowered serum concentrations of inorganic phosphorus. This is an indication of the well-characterized eVect of PTH on renal handling of phosphorus [1]. A similar trend was observed in female mice treated with PTH and 1,25-(OH)2D3 though statistical signiWcance was not reached. PTH treatment in absence of hypercalcemia does not prevent EAE in female mice In the previous experiments, 1 ng of 1,25-(OH)2D3 per day was included in the diet to normalize serum calcium and to insure PTH had catabolic actions on bone. When no 1,25-(OH)2D3 is present, PTH is unable to mediate bonebreakdown [32]. To test if PTH itself independent of hypercalcemia was able to prevent EAE, we implanted pumps containing PTH or vehicle in female 1-hydroxylase knockout mice fed a diet containing high amounts of calcium and lactose but no 1,25-(OH)2D3. Previous work has demonstrated that this diet is suYcient to maintain mineral homeostasis in the absence of vitamin D and in the 1-hydroxylase knockout mouse [30]. Our Wndings with vehicle-treated mice in Fig. 3A support this. The mean serum calcium levels for the vehicle-treated mice fall within the normal range of 9–10 mg/dl. The PTH-treated mice have a similar range of serum calcium levels, which is further evidence that 1,25-(OH)2D3 is needed in PTH-mediated bone catabolism. EAE was induced in these mice 10 days before treatment was started and their disease progression is shown in Fig. 3B with a statistical analysis in Table 2. PTH-treatment itself in the absence of hypercalcemia is unable to prevent EAE (56%-vehicle-treated vs. 46% PTH-treated). Incidence and severity of disease in both groups of mice fed a lactose diet was less compared to mice fed the non-lactose diet is a protective eVect we have noted previously with the lactose diet [36]. Histopathology Histopathologic analysis revealed signiWcant reduction in inXammation and demyelinating lesions in the spinal cord of mice treated with PTH (Fig. 4). Demyelination was more severe in vehicle-treated animals where much less staining of myelin is evident.
Fig. 3. EVect of PTH-treatment on EAE in 1-hydroxylase knockout mice without 1,25-(OH)2D3 supplementation. At 6 weeks of age, female 1hydroxylase knockout mice were switched to a diet containing 2.0% calcium and 20% lactose. Two weeks later, mice were immunized with 200 g of MOG35–55 peptide along with an intraperitoneal injection of pertussis toxin at the time of immunization and 48 h later. Ten days after immunization, osmotic pumps containing 110 g/kg/day of PTH or vehicle were placed subcutaneously in the upper back. Blood was drawn periodically and mice were scored daily for signs of EAE. (A) Levels of serum calcium were determined as described and the averages § the standard deviation were plotted. (B) Mean EAE scores of 1-hydroxylase knockout mice treated with vehicle or PTH were plotted. PTH-treatment does not protects against EAE without 1,25-(OH)2D3. A statistical analysis appears in Table 2.
Discussion The immunosuppressive ability of the active metabolite of vitamin D has been clearly established in a number of animal models for autoimmune disorders. However, the amount of 1,25-(OH)2D3 needed to prevent disease also causes hypercalcemia, a dangerous condition that results in calciWcation of soft tissues. In the EAE model of MS, the amount of 1,25-(OH)2D3 needed to prevent disease is dependent on the amount of calcium in the diet and gender
220
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
Fig. 4. EVect of PTH-treatment on demyelination of spinal cord. Spinal cords were collected from representative mice of each group at the end of the study and perfused with 10% phosphate buVered formalin before being embedded in paraYn. Sections were stained in luxol fast blue and periodic acid schiV’s reagent. Vehicle-treated mice (A) display perivascular inXammation and demyelination, while PTH-treated mice (B) demonstrate no signs of disease.
of the mouse. In either gender, complete suppression of the disease by the natural hormone is not observed until hypercalcemia is established [18]. This led us to develop a hypothesis that hypercalcemia, dependent or independent of vitamin D, is immunosuppressive. To test this, we developed a model using constant administration of the active fragment of human PTH that generates hypercalcemia from bone turnover while keeping the in vivo levels of 1,25(OH)2D3 at low and constant amounts. This PTH-mediated hypercalcemia led to protection against EAE in female mice but not male mice. When hypercalcemia was prevented by not administering 1,25-(OH)2D3 to the 1hydroxylase knockout animals, the same dose of PTH had no eVect on EAE. Several research groups including our own are trying to develop analogs of 1,25-(OH)2D3 that are immunosuppressive without causing hypercalcemia. The present work
though, suggests that hypercalcemia itself is a potent suppressor of EAE in females and may explain the observation that 25 times less 1,25-(OH)2D3 is needed to prevent EAE in female mice than male mice on a high calcium diet [18]. Hypercalcemia in female mice is enough to suppress EAE after disease induction regardless of which hormone is used to generate the hypercalcemia. However, both the hypercalcemia and 1,25-(OH)2D3 are needed to prevent disease in males. This suggests that there is more than one mechanism of immune suppression by 1,25-(OH)2D3-treatment and consequently complicates the search for an eVective analog that can be used clinically. Recently, Zhu et al. [37] reported that dietary calcium and 1,25-(OH)2D3 had independent immunosuppressive eVects upon the TNF- signaling pathway in a mouse model of inXammatory bowel disease. The authors found an inverse correlation between disease severity and serum calcium levels, which gives further evidence that non-calcemic analogs of 1,25-(OH)2D3 may not be eVective in suppressing autoimmune diseases. There are important diVerences in protection of EAE between 1,25-(OH)2D3-mediated hypercalcemia and PTHmediated hypercalcemia. Dietary calcium must be present for 1,25-(OH)2D3 to exert an immunosuppressive eVect [18] while the calcium source in PTH-mediated hypercalcemia in the 1-hydroxylase knockout mouse is bone. Another possibility we considered is that PTH, separate from the eVects of bone turnover and hypercalcemia, may itself be immunosuppressive. However, our results argue against PTH directly suppressing EAE. PTH-treatment was unable to prevent EAE in male mice, and when hypercalcemia was avoided by not giving 1,25-(OH)2D3, PTH was no longer able to prevent EAE in female mice. One possibility is that PTH is enhancing the immunosuppressive eVects of 1,25(OH)2D3 in this EAE model. However, numerous published reports by our group and others indicate that 1,25-(OH)2D3 only prevents EAE when causing high serum calcium levels [18,38–41]; hypercalcemia is the common denominator. By demonstrating that hypercalcemia protects against EAE in females regardless of whether PTH or 1,25-(OH)2 D3 is mediating the eVect, the third calcium regulating hormone must be considered, namely calcitonin. Calcitonin is a 32-amino acid peptide secreted by the C cells of the thyroid in response to high serum calcium concentrations. Fentomolar concentrations of this hormone inhibit bone resorption by osteoclasts, and the more potent salmon calcitonin is used clinically to treat bone resorption disorders including Paget’s disease and osteoporosis [42–44]. Basal levels of calcitonin in women are lower than men, and women exhibit 4–5 times lower calcitonin secretory capacity then men upon calcium infusion [45]. Interestingly, serum calcitonin levels rise during pregnancy and remain high through lactation, which is the same time many MS patients go into remission [46,47]. Lymphocytes are known to have highaYnity receptors for calcitonin [48]. Because calcitonin increases the synthesis of the 1-hydroxylase enzyme [49], we hypothesized in the past that the hormone’s anti-inXammatory ability maybe secondary to increased synthesis of
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
the active metabolite of vitamin D. With our current Wndings in the 1-hydroxylase knockout mice, the role of calcitonin in blocking inXammatory processes should be reconsidered. By developing a model of hypercalcemia that was independent of increases in 1,25-(OH)2D3, we hoped to Wnd a simple answer implicating hypercalcemia or 1,25-(OH)2D3 as the causative agent in suppression of EAE. Instead our results reXect the complex etiology of this disease where several factors impinge on the immune system to give diVerent outcomes. This work suggests development of nonhypercalcemic analogs of 1,25-(OH)2D3 that completely suppress EAE will be diYcult not just because hypercalcemia appears to have immunosuppressive eVects in females, but also because of the varied impact such an analog would have on PTH and calcitonin secretion. A better understanding is needed about the relationship between the immune system and calcium homeostasis regulation to aid in design of 1,25-(OH)2D3 analogs with clinical potential. Thus, this work serves as a Wrst step in dissecting apart the immunosuppressive eVects of altered mineral homeostasis and the hormones involved. Acknowledgment This work has been funded in part by the Wisconsin Alumni Research Foundation. References [1] M.R. Haussler, G.K. WhitWeld, C.A. Haussler, J.C. Hsieh, P.D. Thompson, S.H. Selznick, C.E. Dominguez, P.W. Jurutka, J. Bone Miner. Res. 13 (1998) 325–349. [2] G. Penna, L. Adorini, J. Immunol. 164 (2000) 2405–2411. [3] M.D. GriYn, W. Lutz, V.A. Phan, L.A. Bachman, D.J. McKean, R. Kumar, Proc. Natl. Acad. Sci. USA 98 (2001) 6800–6805. [4] A. Boonstra, F.J. Barrat, C. Crain, V.L. Heath, V.H.F. Savelkoul, A. O’Garra, J. Immunol. 167 (2001) 4974–4980. [5] H. Tanaka, E. Abe, C. Miyaura, Y. Shiina, T. Suda, Biochem. Biophys. Res. Commun. 117 (1983) 86–92. [6] J.M. Lemire, D.C. Archer, J. Clin. Invest. 87 (1991) 1103–1107. [7] M.T. Cantorna, C.E. Hayes, H.F. DeLuca, Proc. Natl. Acad. Sci. USA 93 (1996) 7861–7864. [8] C. Mathieu, J. Laureys, H. Sobis, M. Vandeputte, M. Waer, R. Bouillon, Diabetes 41 (1992) 1491–1495. [9] M.T. Cantorna, C. Munsick, C. Bemiss, B.D. Mahon, J. Nutr. 130 (2000) 2648–2652. [10] M.T. Cantorna, C.E. Hayes, H.F. DeLuca, J. Nutr. 128 (1998) 68–72. [11] J.M. Lemire, A. Ince, M. Takashima, Autoimmunity 12 (1992) 143– 148. [12] L. Metz, Semin. Neurol. 18 (1998) 389–395. [13] P. Duquette, J. Pleines, M. Girard, L. Charest, M. Senecal-Quevillon, C. Masse, Can. J. Neurol. Sci. 19 (1992) 466–471. [14] K. Birk, S.C. Smeltzer, R. Rudick, Semin. Neurol. 8 (1988) 205–213. [15] C.E. Hayes, Proc. Nutr. Soc. 59 (2000) 531–535. [16] K.L. Munger, S.M. Zhang, E. O’Reilly, M.A. Hernan, M.J. Olek, W.C. Willett, A. Ascherio, Neurology 62 (2004) 60–65.
221
[17] E. Garcion, L. Sindji, S. Nataf, P. Brachet, F. Darcy, C.N. MonteroMenei, Acta Neuropathol. (Berl.) 105 (2003) 438–448. [18] M.T. Cantorna, J. Humpal-Winter, H.F. DeLuca, J. Nutr. 129 (1999) 1966–1971. [19] E.M. Brown, R. Leombruno, J. Thatcher, M. Burrowes, Endocrinology 116 (1985) 1123–1132. [20] A. Mithal, O. Kifor, I. Kifor, P. Vassilev, R. Butters, K. Krapcho, R. Simin, F. Fuller, S.C. Hebert, E.M. Brown, Endocrinology 136 (1995) 3087–3092. [21] J. Fox, H. Heath III, Am. J. Physiol. 240 (1981) E649–E655. [22] H.L. Brenza, C. Kimmel-Jehan, F. Jehan, T. Shinki, S. Wakino, H. Anazawa, T. Suda, H.F. DeLuca, Proc. Natl. Acad. Sci. USA 95 (1998) 1387–1391. [23] H. Dobnig, R.T. Turner, Endocrinology 138 (1997) 4607–4612. [24] S.C. Kukreja, P.A. York, C. Nalbantian-Brandt, D.H. Shevrin, M.J. Favus, Am. J. Physiol. 256 (1989) E309–E314. [25] H. Tsuji, C. Cawthorn, B. Ecarot, J. Bone Miner. Res. 11 (1996) 1234– 1240. [26] P. Jaeger, W. Jones, M. Kashgarian, R. Baron, T.L. Clemens, G.V. Segre, J.P. Hayslett, Am. J. Physiol. 252 (1987) E790–E798. [27] K. Endo, K. Katsumata, M. Hirata, T. Masaki, N. Kubodera, T. Nakamura, K. Ikeda, E. Ogata, J. Bone Miner. Res. 15 (2000) 175–181. [28] D.K. Panda, D. Miao, M.L. Tremblay, J. Sirois, R. Farookhi, G.N. Hendy, D. Goltzman, Proc. Natl. Acad. Sci. USA 98 (2001) 7498–7503. [29] O. Dardenne, J. Prud’homme, A. Arabian, F.H. Glorieux, R. St-Arnaud, Endocrinology 142 (2001) 3135–3141. [30] O. Dardenne, J. Prud’homme, S.A. Hacking, F.H. Glorieux, R. Snst-Arnaud, Bone 32 (2003) 332–340. [31] O. Dardenne, J. Prudhomme, S.A. Hacking, F.H. Glorieux, R. St Arnaud, J. Bone Miner. Res. 18 (2003) 637–643. [32] H. Rasmussen, H. Deluca, C. Arnaud, C. Hawker, M. Vonstedingk, J. Clin. Invest. 42 (1963) 1940–1946. [33] C.C. Bernard, T.G. Johns, A. Slavin, M. Ichikawa, C. Ewing, J. Liu, J. Bettadapura, J. Mol. Med. 75 (1997) 77–88. [34] W. Slavin, Methods Enzymol. 158 (1988) 117–145. [35] K. Itaya, M. Ui, Clin. Chim. Acta 14 (1966) 361–366. [36] T.F. Meehan, H.F. DeLuca, Arch. Biochem. Biophys. 408 (2002) 200–204. [37] Y. Zhu, B.D. Mahon, M. Froicu, M.T. Cantorna, Eur. J. Immunol. 35 (2005) 217–224. [38] D.D. Branisteanu, M. Waer, H. Sobis, S. Marcelis, M. Vandeputte, R. Bouillon, J. Neuroimmunol. 61 (1995) 151–160. [39] F. Mattner, S. Smiroldo, F. Galbiati, M. Muller, P. Di Lucia, P.L. Poliani, G. Martino, P. Panina-Bordignon, L. Adorini, Eur. J. Immunol. 30 (2000) 498–508. [40] E. van Etten, D.D. Branisteanu, A. Verstuyf, M. Waer, R. Bouillon, C. Mathieu, Transplantation 69 (2000) 1932–1942. [41] E. Van Etten, D.D. Branisteanu, L. Overbergh, R. Bouillon, A. Verstuyf, C. Mathieu, Bone 32 (2003) 397–404. [42] M. Zaidi, H.K. Datta, B.S. Moonga, I. MacIntyre, J. Endocrinol. 126 (1990) 473–481. [43] F.R. Singer, Calcif. Tissue Int. 49 (Suppl. 2) (1991) S7–S8. [44] M. Azria, D.H. Copp, J.M. Zanelli, Calcif. Tissue Int. 57 (1995) 405–408. [45] R.D. Tiegs, J.J. Body, J.M. Barta, H. Heath III, J. Bone Miner. Res. 1 (1986) 339–349. [46] M. Whitehead, G. Lane, O. Young, S. Campbell, G. Abeyasekera, C.J. Hillyard, I. MacIntyre, K.G. Phang, J.C. Stevenson, Br. Med. J. (Clin. Res. Ed.) 283 (1981) 10–12. [47] S.U. Toverud, C.W. Cooper, P.L. Munson, Endocrinology 103 (1978) 472–479. [48] J.J. Body, F. Glibert, S. Nejai, G. Fernandez, A. Van Langendonck, A. Borkowski, J. Clin. Endocrinol. Metab. 71 (1990) 675–681. [49] A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, Y. Kawaguchi, T. Hosoya, S. Kato, Endocrinology 140 (1999) 2224–2231.
Hypercalcemia produced by parathyroid hormone suppresses experimental autoimmune encephalomyelitis in female but not male mice Terrence F. Meehan, Janeen Vanhooke, Jean Prahl, Hector F. DeLuca ¤ Department of Biochemistry, University of Wisconsin-Madison, 433 Babcock Drive, Madison, WI 53706-1544, USA Received 20 July 2005 Available online 2 September 2005
Abstract Besides its role in regulating serum levels of calcium and phosphorus, 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3) has potent eVects on the immune system and suppresses disease in several animal models of autoimmune disorders including experimental autoimmune encephalomyelitis (EAE), a mouse model of multiple sclerosis. While the amount of 1,25-(OH)2D3 needed to prevent EAE is dependent on the gender of the mouse and amount of calcium available in the diet, the minimum levels of 1,25-(OH)2D3 suYcient to prevent disease cause hypercalcemia. To test if hypercalcemia independent of high levels of 1,25-(OH)2D3 can suppress EAE, we used a 25-hydroxyvitamin D3-1-hydroxylase (1-hydroxylase) knockout mouse strain. Because these 1-hydroxylase knockout mice lack the parathyroid hormone (PTH)-regulated enzyme that synthesizes 1,25-(OH)2D3, hypercalcemia from increased bone turnover was created by continuous administration of PTH without changing the circulating levels of 1,25-(OH)2D3. This PTH-mediated hypercalcemia generated after EAE induction prevented disease in female mice but not male mice. When hypercalcemia was prevented by diet manipulation, PTH administration no longer prevented EAE. We conclude that hypercalcemia is able to prevent EAE after disease induction in female mice. 2005 Elsevier Inc. All rights reserved. Keywords: 1; 25-Dihydroxyvitamin D3; Calcium; Experimental autoimmune encephalomyelitis; Parathyroid hormone; Multiple sclerosis; 25-Hydroxyvitamin D3-1a-hydroxylase; CYP27B1; Phosphorus; Autoimmune; Sex eVect
The active metabolite of vitamin D, 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3),1 acts as a hormone to regulate calcium and phosphorous levels in blood by increasing calcium and phosphorus absorption from the duodenum of the small intestine, and by working in concert with parathyroid hormone (PTH) to release calcium and phosphorus from bone, and to increase the reabsorption of calcium in the kidney [1]. In addition to its role in mineral homeostasis, 1,25-(OH)2D3 has wide ranging impacts on the immune *
Corresponding author. Fax: +1 608 262 7122. E-mail address: [email protected] (H.F. DeLuca). 1 Abbreviations used: 1,25-(OH)2D3, 1,25-dihydroxyvitamin D3; 1-hydroxylase, 25-hydroxy-vitamin D3-1-hydroxylase; EAE, experimental autoimmune encephalomyelitis; mg/dl, milligrams per deciliter; MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; PTH, parathyroid hormone. 0003-9861/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2005.08.011
system including inhibitory eVects on maturation of dendritic cells, the skewing of CD4+ T-cells upon activation to a less inXammatory Th2 cytokine proWle, and promoting hematopoietic cells to diVerentiate into a monocyte/macrophage pathway [2–5]. These eVects are reXected in the immunosuppressive ability of 1,25-(OH)2D3 in a number of animal models for autoimmune disorders including experimental autoimmune encephalomyelitis (EAE), non-obese diabetes, inXammatory bowel disease, arthritis, and lupus [6–11]. However, it is not clear if this suppression of disease is entirely dependent on the vitamin D hormone, or if the concurrent rise in serum calcium upon administration of 1,25-(OH)2D3 has any impact. The possibility of hypercalcemia being immunosuppressive is most clear in the EAE mouse model of multiple sclerosis (MS). Multiple sclerosis is CD4+ T-cell mediated autoimmune disease that results in demyelination of axonal tracts in the
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
central nervous system causing a wide range of neurological symptoms [12]. Epidemiology studies have revealed genetic and environmental factors that modify disease incidence and progression including a gender bias where women are more likely to develop disease than men, remission of MS during pregnancy, and a geographic distribution that suggests a correlation between MS prevalence and vitamin D deWciency [13–15]. Recently, a signiWcant inverse correlation was found between vitamin D supplementation and the risk of developing MS in a study involving a large group of women [16]. While the event in MS that leads to activation and proliferation of these disease eVector T-cells is unknown, a similar disease can be induced in rodents by immunization of antigens present in the central nervous system. Besides diVerent triggering mechanisms, these EAE animal models share many characteristics of MS. This includes an “activation phase” where antigen presenting cells process the immunized antigen, migrate to the lymph nodes and present immunodominant peptides to naive T-cells; and an “eVector phase” where CD4+ T-cells that recognize antigen proliferate and cross the blood–brain barrier to lead an inXammatory attack that results in demyelinated lesions. 1,25-(OH)2D3 has been shown to prevent the occurrence and progression of this disease in diVerent EAE models in both rats and mice [6,7,17]. In mice, it was established that the dose of 1,25-(OH)2D3 needed to suppress disease was dependent on the sex of the animal and the amount of calcium present in the diet. Females required smaller doses of 1,25-(OH)2D3 to suppress disease than males and that less hormone was needed when the amount of calcium in the diet was increased [18]. In either gender, complete suppression of the disease was not observed until severe hypercalcemia (serum calcium levels >12 mg/dl) was established. This led us to develop a hypothesis that hypercalcemia, dependent or independent of 1,25-(OH)2D3, might account for the action of 1,25(OH)2D3 in preventing disease onset. To separate the eVects of hypercalcemia and 1,25(OH)2D3 in vivo is diYcult due to the tight regulation of the synthesis of 1,25-(OH)2D3 by serum calcium levels and PTH. PTH, an 84 amino acid peptide, is secreted by the parathyroid gland within seconds of detection of a lowering of serum calcium by the calcium sensing receptor, a G-protein coupled receptor [19–21]. The active N-terminal fragment of this hormone binds to its receptor in the kidney and increases the activity of the renal cytochrome P450 enzyme CYP27B1, the 25-hydroxy-vitamin D3-1-hydroxylase (1hydroxylase) [22]. This enzyme converts the biologically inactive circulating form of vitamin D3, 25-hydroxyvitamin D3, to 1,25-(OH)2D3. This in vivo regulation is mimicked by subcutaneous injections of PTH in mammals where increased levels of circulating 1,25-(OH)2D3 and increased bone mass are observed without hypercalcemia [23]. Hypercalcemia can be generated by chronically administering PTH by osmotic pumps; however, there are inconsistent results whether this treatment leads to increased or decreased levels of 1,25-(OH)2D3 [23–27].
215
Our research group recently developed a 1-hydroxylase knockout mouse strain similar to the two strains previously reported by other groups [28,29, J.L. Vanhooke, J.M. Prahl, C. Kimmel-Jehan, M. Mendelsohn, E.W. Danielson, K.D. Healy, and H.F. DeLuca, unpublished results]. All three strains of mice are unable to synthesize the active metabolite of vitamin D and display a phenotype similar to the human disorder vitamin D-dependency rickets type I (VDDR-1) including hypocalcemia, hyperparathyoidism, and rickets. The phenotype in these mice can be rescued by daily administration of 1,25-(OH)2D3 or by a high calcium diet containing lactose [30,31]. The development of these 1-hydroxylase knockout strains allows for administration of PTH without altering levels of 1,25-(OH)2D3. When administered continuously, PTH has catabolic actions on bone that leads to increases in serum calcium. These catabolic action appears to be dependent on the presence of 1,25-(OH)2D3 since this eVect is abolished in vitamin D-deWcient animals [32] and in 1-hydroxylase knockout mice (unpublished results). In this paper, we report the development of mouse model of hypercalcemia that is independent of increases of 1,25-(OH)2D3 and the impact this metabolic state has on EAE induced by immunization with an encephalogenic peptide. The hypercalcemia was achieved by giving 1-hydroxylase knockout mice, a small amount of 1,25(OH)2D3 that is insuYcient to prevent EAE, but could rescue the VDDR-1 phenotype and aid in PTH-mediated bone catabolism, and then dosing with pharmacological amounts of the bioactive fragment of human PTH. This PTH-mediated hypercalcemia led to protection against EAE after disease induction in female mice but not male mice. The implications of these Wndings and future uses of our vitamin D independent hypercalcemia model are discussed. Materials and methods Animals and diet 25-Hydroxyvitamin D3-1-hydroxylase knockout mice were recently generated in our laboratory by targeted deletion of exons 1–8. 1-Hydroxylase knockout mice were backcrossed seven to nine times onto a C57Bl/6 strain and genotyped by PCR analysis with primers speciWc for exons 3 (forward) and exons 4 and 5 (reverse). All mice were fed puriWed diets based on those described by Cantorna et al. [7]. Homozygous 1-hydroxylase knockout mice were weaned onto a rescue diet containing 0.47% calcium and 5 ng of 1,25-(OH)2D3 per 4 g of diet. Starting at 5–6 weeks of age, 1-hydroxylase knockout mice received 4 g per day of an experimental diet containing 0.87% calcium and diVering amounts of 1,25-(OH)2D3. These diets were made and replaced three times a week for the duration of the studies. Mice in the PTH or vehicle groups received 1 ng of 1,25-(OH)2D3 per day except in one study where mice in both groups received a 20% lactose diet containing 2%
216
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
calcium and 1.25 % phosphorus with no 1,25-(OH)2D3 supplementation. A second group was switched 10 days after EAE induction from the 1 ng of 1,25-(OH)2D3 per day diet to a diet containing 50 ng (female mice) or 200 ng (male mice) of 1,25-(OH)2D3 per day. A third group of 1hydroxylase knockout mice received 37.5 ng of 1,25(OH)2D3 per day from 5 weeks of age until the end of the study. Experimental autoimmune encephalomyelitis induction EAE was induced in the 1-hydroxylase knockout mice backcrossed on the C57Bl/6 strain by subcutaneous immunization in the rear Xanks with 200 g of the immunodominant peptide to myelin oligodendrocyte glycoprotein (MOG) (MEVGWYRSPFSRVVHLYRNGK) [33]. The peptide was synthesized at the University of Wisconsin Biotechnology Center using standard 9-Xuorenyl-methoxycarbonyl chemistry and was judged to be 92% pure by reverse-phase HPLC. The peptide was dissolved in Freund’s complete adjuvant containing 4 mg/ml of heat-killed mycobacterium tuberculosis H37a (Difco Laboratories, Detroit, MI). On the day of immunization and 48 h later, mice were injected intraperitoneally with 200 ng of Bordetella pertussis toxin dissolved in PBS (List Biological Laboratories, Campbell, CA). Mice were examined daily for clinical signs of EAE and scored as follows: 0, no paralysis; 1, loss of tail tone; 2, hindlimb weakness; 3, hindlimb paralysis; 4, forelimb paralysis; 5, moribund or dead.
Perkin–Elmer atomic absorption spectrometer. Serum was diluted in 0.1% LaCl3 for this determination [34]. Serum phosphorus levels were determined as previously described [35]. BrieXy, 25 l of serum was added to 4 ml of a hydrochloric acid solution containing ammonium molybdate and the resulting product of unreduced phosphomolybdate was determined by spectrophotometric measurement (Unicam UV1, Thermo-Spectronic, Rochester, NY). All values were compared to those of reference values measured on the same day. Histopathology Spinal cords from 1-hydroxylase immunized with MOG peptide were removed and Wxed in 10% phosphatebuVered formalin. ParaYn-embedded tissue sections were prepared and stained in luxol fast blue and periodic acid schiVs reagent by the University of Wisconsin Veterinary Science Department. Statistical analysis Statistical analysis was performed using a statistical program for the Macintosh computer (Graphpad Prism). The two-tailed Fisher exact probability test was performed on incidence rates and the unpaired two group Student’s t test was used on other measurements. Values of P < 0.05 were considered statistically signiWcant. Results
Administration of PTH The bioactive, human fragment of PTH (peptides 1–34) was purchased from Bachem (Bachem Californa, Torrence, CA). The peptide was dissolved to a concentration of 1 mg/ ml in a vehicle containing 150 mM NaCl, 1 mM HCl, and 2% heat inactivated sera, as described previously [23]. The sera used were from 1-hydroxylase knockout mice fed a rescue diet and was matched by sex to the group of mice receiving treatment. PTH was administered chronically using Alzet osmotic minipumps model 1002 (Durect Corporation, Cupertino, CA). These pumps deliver 0.25 l/h of solution for a 15 day period. Nine days after EAE induction, mice were weighed and the reservoirs of the pumps were Wlled with vehicle or PTH diluted to deliver 110 g per kilogram bodyweight per day. The pumps were placed in microcentrifuge tubes containing saline and were primed at 37 °C overnight. The next morning, the pumps were placed subcutaneously in the upper back of mice anesthetized with isoXurane. At the end of the study, successful delivery of the solutions was assessed by measuring the remaining volume of liquids in the reservoirs. Serum mineral analysis Blood samples were centrifuged at 3500 rpm (1096g) for 15 min. Serum calcium values were determined using a
Oral feeding of 1,25-(OH)2D3 prevents EAE in 1-hydroxylase knockout mice In the Wrst study, we investigated the eVects of oral administration of 1,25-(OH)2D3 on EAE in mice that lack the 1-hydroxylase gene. Because these mice are unable to produce 1,25-(OH)2D3 [29], a small amount of the hormone must be given to maintain normal serum mineral homeostasis. Pilot studies indicated that 1 ng per day of 1,25-(OH)2D3 given to mice fed a 0.87% calcium diet was suYcient to rescue the mice from abnormal mineral homeostasis while being below the minimal dose needed to prevent EAE (data not shown). To test the ability of 1,25(OH)2D3 to prevent EAE, a 2nd group received 37.5 ng per day of 1,25-(OH)2D3 starting 2 weeks prior to immunization. To demonstrate the ability of 1,25-(OH)2D3 to prevent disease after disease induction, female mice received 50 ng per day of 1,25-(OH)2D3 and male mice received 200 ng per day of 1,25-(OH)2D3 starting at 10 days after immunization, which is before the Wrst clinical signs of EAE are observed. Male mice required more 1,25-(OH)2D3 to protect against EAE, which is consistent with a previous report [18]. Immunization with 200 g of a peptide corresponding to the immunodominant epitope of MOG induced EAE in the 1-hydroxylase knockout mice fed the rescue diet containing 1 ng of 1,25-(OH)2D3 per day (Fig. 1, Table 1). The
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
217
Fig. 1. EAE in 1-hydroxylase knockout mice fed diVering amounts of 1,25-(OH)2D3. At 6 weeks of age, 1-hydroxylase knockout mice were switched to a 0.87% calcium diet containing either 1 ng or 37.5 ng of 1,25-(OH)2D3 per day. Two weeks later, mice were immunized with 200 g of MOG35–55 peptide along with an intraperitoneal injection of pertussis toxin at the time of immunization and 48 h later. Ten days after immunization, one group of mice fed 1 ng of 1,25-(OH)2D3 was switched to a diet containing 50 ng of 1,25-(OH)2D3. Mice were scored for EAE once a day for 24 days. The higher amounts of 1,25-(OH)2D3 protected against EAE while the 1 ng per day amount did not. A statistical analysis appears in Table 1.
Table 1 1,25-(OH)2D3 protects against MOG-induced EAE Incidence
Mean severity
Mean day of onset
Serum calcium (mg%)
Both sexes 1 ng 1,25-(OH)2D3 per day 37.5 ng 1,25-(OH)2D3 per day 50 or 200 ng 1,25-(OH)2D3 starting day 10
91% n D 10/11 0%a n D 0/6 15%a n D 2/13
2.0 § 0.7 — 1.5 § 0.7
16 § 3 — 15 § 2
9.9 § 0.6 13.0 § 0.7a 11.9 § 0.7a
Female 1 ng 1,25-(OH)2D3 per day 37.5 ng 1,25-(OH)2D3 per day 50 ng 1,25-(OH)2D3 starting day 10
100% n D 6/6 0%a n D 0/3 10%a n D 1/10
1.8 § 0.8 — 1b
16 § 3 — 16b
9.7 § 0.5 12.1 § 0.4a 12.0 § 0.6a
Male 1 ng 1,25-(OH)2D3 day 37.5 ng 1,25-(OH)2D3 per day 200 ng 1,25-(OH)2D3 starting day 10
86% n D 6/7 0%a n D 0/3 17%a n D 1/6
2.5 § 0.4 — 2b
15 § 3 — 13b
10.2 § 0.6 13.9 § 0.7a 13.0 § 0.9a
C57Bl/6 mice were switched to a 0.87% calcium diet containing the indicated amounts of 1,25-(OH)2D3 at 6 weeks of age or 10 days post-immunization. Mice were injected subcutaneously with 400 g of MOG35–55 peptide along with an intraperitoneal injection of 200 ng of pertussis toxin at the time of immunization and 48 h later. Mice were scored for EAE once a day for 24 days. At the end of the study, mice were sacriWced and serum calcium concentrations determined by atomic absorption spectrometry. a P < 0.05 compared to mice fed with 1 ng of 1,25-(OH)2D3. b Only one animal sick.
group receiving the 37.5 ng of 1,25-(OH)2D3 prior to immunization was completely protected from developing EAE while mice receiving 1,25-(OH)2D3 10 days after immunization had signiWcantly lower incidence of EAE. Both groups of mice that were protected against EAE also had elevated levels of serum calcium. Continuous administration of PTH increases serum calcium levels and prevents EAE in female mice but not male mice To determine the role of hypercalcemia in the prevention of EAE by pharmacological doses of 1,25-(OH)2D3, we studied the eVect of raising serum calcium levels by continuous administration of PTH. PTH raises serum calcium levels by increasing bone turnover and by increasing
the activity of the 1-hydroxylase enzyme which leads to increased circulating levels of 1,25-(OH)2D3 [23]. In mice that lack the 1-hydroxylase gene, the later eVect of PTH is negated while the former can still occur if small amounts of 1,25-(OH)2D3 are supplied to the mice exogenously. Dosing studies indicated that administration of 110 g of the human active fragment of PTH per kilogram body weight led to consistent hypercalcemia (data not shown). 1-Hydroxylase knockout mice receiving the rescue diet of 1 ng of 1,25-(OH)2D3 per day were immunized with 200 g of MOG peptide. Ten days after immunization, osmotic pumps containing PTH or vehicle were delivered subcutaneously to the upper back. Mice were bled periodically and calcium concentrations determined. Mice receiving PTH had serum calcium concentrations
218
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
Fig. 2. EVects of PTH-treatment on EAE in 1-hydroxylase knockout mice. At 6 weeks of age, 1-hydroxylase knockout mice were switched to a diet containing 0.87% calcium and 1 ng of 1,25-(OH)2D3 per day. Two weeks later, mice were immunized with 200 g of MOG35–55 peptide along with an intraperitoneal injection of pertussis toxin at the time of immunization and 48 h later. 10 days after immunization, osmotic pumps containing 110 g/kg/day of PTH or vehicle were placed subcutaneously in the upper back. Blood was drawn periodically and mice were scored daily for signs of EAE. (A) Levels of serum calcium were determined as described and the averages § the standard deviation were plotted with a statistical analysis appearing in Table 2. (B) Mean EAE scores of 1-hydroxylase knockout mice treated with vehicle or PTH were plotted. PTH-treatment protects against EAE in female but not male mice.
Table 2 PTH-mediated hypercalcemia protects against EAE in female but not male mice Po4 (mg%)
Incidence
Mean severity
Mean day of onset
Serum calcium (mg%)
Female Vehicle + 1 ng 1,25 D3 PTH + 1 ng 1,25 D3 Vehicle + 20% lactose PTH + 20% lactose
75% n D 9/12 0%a n D 0/13 56% n D 9/16 46% n D 7/15
2.7 § 1.1 — 1.4 § 0.5a 1.5 § 0.7a
17 § 4 — 18 § 4 18 § 4
10.2 § 0.5 13.4 § 1.2a 9.3 § 0.5 9.9 § 1.0
6.7 § 1.3 5.8 § 1.5 6.3 + 1.0 6.0 + 1.5
Male Vehicle + 1 ng 1,25 D3 PTH + 1 ng 1,25 D3
73% n D 11/15 72% n D 13/18
2.3 § 1.1 2.5 § 0.9
14 § 3 15 § 2
10.4 § 0.3 13.2 § 1.6a
7.4 § 1.0 5.1 § 1.3a
At 6 weeks of age, 1-hydroxylase knockout mice were switched to a diet containing either 0.87% calcium and 1 ng 1,25-(OH)2D3 or 2% calcium and 20% lactose. At 8 weeks of age, mice were injected subcutaneously with 400 g of MOG35–55 peptide along with an intraperitoneal injection of 200 ng of pertussis toxin at the time of immunization and 48 h later. Ten days after immunization mice received osmotic pumps that released 110 g/kg/day of PTH or vehicle. Mice were scored for EAE once a day for 24 days. At the end of the study, mice were sacriWced and serum calcium and phosphorus concentrations determined as described in the materials and methods section. a P < 0.05 compared to vehicle-treated + 1 ng 1,25-(OH)2D3 mice of the same sex.
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
219
averaging over 13 mg/dl for the 2-week lifespan of the pumps while mice receiving vehicle had normal serum calcium levels (Fig. 2A). There was no statistical diVerence in serum calcium values between male and female mice within a treatment group at any of the time points measured. Surprisingly, hypercalcemia induced by PTH completely prevented EAE in female mice while having no eVect on EAE incidence or progression in male mice (Fig. 2B). Male mice had an incidence of 73% in the vehicle treated group versus 72% in the PTH-treated group while vehicle treated female mice had a 75% incidence rate as compared to no incidence in PTH-treated females (Table 2). In addition to high serum calcium values, male mice receiving PTH also had lowered serum concentrations of inorganic phosphorus. This is an indication of the well-characterized eVect of PTH on renal handling of phosphorus [1]. A similar trend was observed in female mice treated with PTH and 1,25-(OH)2D3 though statistical signiWcance was not reached. PTH treatment in absence of hypercalcemia does not prevent EAE in female mice In the previous experiments, 1 ng of 1,25-(OH)2D3 per day was included in the diet to normalize serum calcium and to insure PTH had catabolic actions on bone. When no 1,25-(OH)2D3 is present, PTH is unable to mediate bonebreakdown [32]. To test if PTH itself independent of hypercalcemia was able to prevent EAE, we implanted pumps containing PTH or vehicle in female 1-hydroxylase knockout mice fed a diet containing high amounts of calcium and lactose but no 1,25-(OH)2D3. Previous work has demonstrated that this diet is suYcient to maintain mineral homeostasis in the absence of vitamin D and in the 1-hydroxylase knockout mouse [30]. Our Wndings with vehicle-treated mice in Fig. 3A support this. The mean serum calcium levels for the vehicle-treated mice fall within the normal range of 9–10 mg/dl. The PTH-treated mice have a similar range of serum calcium levels, which is further evidence that 1,25-(OH)2D3 is needed in PTH-mediated bone catabolism. EAE was induced in these mice 10 days before treatment was started and their disease progression is shown in Fig. 3B with a statistical analysis in Table 2. PTH-treatment itself in the absence of hypercalcemia is unable to prevent EAE (56%-vehicle-treated vs. 46% PTH-treated). Incidence and severity of disease in both groups of mice fed a lactose diet was less compared to mice fed the non-lactose diet is a protective eVect we have noted previously with the lactose diet [36]. Histopathology Histopathologic analysis revealed signiWcant reduction in inXammation and demyelinating lesions in the spinal cord of mice treated with PTH (Fig. 4). Demyelination was more severe in vehicle-treated animals where much less staining of myelin is evident.
Fig. 3. EVect of PTH-treatment on EAE in 1-hydroxylase knockout mice without 1,25-(OH)2D3 supplementation. At 6 weeks of age, female 1hydroxylase knockout mice were switched to a diet containing 2.0% calcium and 20% lactose. Two weeks later, mice were immunized with 200 g of MOG35–55 peptide along with an intraperitoneal injection of pertussis toxin at the time of immunization and 48 h later. Ten days after immunization, osmotic pumps containing 110 g/kg/day of PTH or vehicle were placed subcutaneously in the upper back. Blood was drawn periodically and mice were scored daily for signs of EAE. (A) Levels of serum calcium were determined as described and the averages § the standard deviation were plotted. (B) Mean EAE scores of 1-hydroxylase knockout mice treated with vehicle or PTH were plotted. PTH-treatment does not protects against EAE without 1,25-(OH)2D3. A statistical analysis appears in Table 2.
Discussion The immunosuppressive ability of the active metabolite of vitamin D has been clearly established in a number of animal models for autoimmune disorders. However, the amount of 1,25-(OH)2D3 needed to prevent disease also causes hypercalcemia, a dangerous condition that results in calciWcation of soft tissues. In the EAE model of MS, the amount of 1,25-(OH)2D3 needed to prevent disease is dependent on the amount of calcium in the diet and gender
220
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
Fig. 4. EVect of PTH-treatment on demyelination of spinal cord. Spinal cords were collected from representative mice of each group at the end of the study and perfused with 10% phosphate buVered formalin before being embedded in paraYn. Sections were stained in luxol fast blue and periodic acid schiV’s reagent. Vehicle-treated mice (A) display perivascular inXammation and demyelination, while PTH-treated mice (B) demonstrate no signs of disease.
of the mouse. In either gender, complete suppression of the disease by the natural hormone is not observed until hypercalcemia is established [18]. This led us to develop a hypothesis that hypercalcemia, dependent or independent of vitamin D, is immunosuppressive. To test this, we developed a model using constant administration of the active fragment of human PTH that generates hypercalcemia from bone turnover while keeping the in vivo levels of 1,25(OH)2D3 at low and constant amounts. This PTH-mediated hypercalcemia led to protection against EAE in female mice but not male mice. When hypercalcemia was prevented by not administering 1,25-(OH)2D3 to the 1hydroxylase knockout animals, the same dose of PTH had no eVect on EAE. Several research groups including our own are trying to develop analogs of 1,25-(OH)2D3 that are immunosuppressive without causing hypercalcemia. The present work
though, suggests that hypercalcemia itself is a potent suppressor of EAE in females and may explain the observation that 25 times less 1,25-(OH)2D3 is needed to prevent EAE in female mice than male mice on a high calcium diet [18]. Hypercalcemia in female mice is enough to suppress EAE after disease induction regardless of which hormone is used to generate the hypercalcemia. However, both the hypercalcemia and 1,25-(OH)2D3 are needed to prevent disease in males. This suggests that there is more than one mechanism of immune suppression by 1,25-(OH)2D3-treatment and consequently complicates the search for an eVective analog that can be used clinically. Recently, Zhu et al. [37] reported that dietary calcium and 1,25-(OH)2D3 had independent immunosuppressive eVects upon the TNF- signaling pathway in a mouse model of inXammatory bowel disease. The authors found an inverse correlation between disease severity and serum calcium levels, which gives further evidence that non-calcemic analogs of 1,25-(OH)2D3 may not be eVective in suppressing autoimmune diseases. There are important diVerences in protection of EAE between 1,25-(OH)2D3-mediated hypercalcemia and PTHmediated hypercalcemia. Dietary calcium must be present for 1,25-(OH)2D3 to exert an immunosuppressive eVect [18] while the calcium source in PTH-mediated hypercalcemia in the 1-hydroxylase knockout mouse is bone. Another possibility we considered is that PTH, separate from the eVects of bone turnover and hypercalcemia, may itself be immunosuppressive. However, our results argue against PTH directly suppressing EAE. PTH-treatment was unable to prevent EAE in male mice, and when hypercalcemia was avoided by not giving 1,25-(OH)2D3, PTH was no longer able to prevent EAE in female mice. One possibility is that PTH is enhancing the immunosuppressive eVects of 1,25(OH)2D3 in this EAE model. However, numerous published reports by our group and others indicate that 1,25-(OH)2D3 only prevents EAE when causing high serum calcium levels [18,38–41]; hypercalcemia is the common denominator. By demonstrating that hypercalcemia protects against EAE in females regardless of whether PTH or 1,25-(OH)2 D3 is mediating the eVect, the third calcium regulating hormone must be considered, namely calcitonin. Calcitonin is a 32-amino acid peptide secreted by the C cells of the thyroid in response to high serum calcium concentrations. Fentomolar concentrations of this hormone inhibit bone resorption by osteoclasts, and the more potent salmon calcitonin is used clinically to treat bone resorption disorders including Paget’s disease and osteoporosis [42–44]. Basal levels of calcitonin in women are lower than men, and women exhibit 4–5 times lower calcitonin secretory capacity then men upon calcium infusion [45]. Interestingly, serum calcitonin levels rise during pregnancy and remain high through lactation, which is the same time many MS patients go into remission [46,47]. Lymphocytes are known to have highaYnity receptors for calcitonin [48]. Because calcitonin increases the synthesis of the 1-hydroxylase enzyme [49], we hypothesized in the past that the hormone’s anti-inXammatory ability maybe secondary to increased synthesis of
T.F. Meehan et al. / Archives of Biochemistry and Biophysics 442 (2005) 214–221
the active metabolite of vitamin D. With our current Wndings in the 1-hydroxylase knockout mice, the role of calcitonin in blocking inXammatory processes should be reconsidered. By developing a model of hypercalcemia that was independent of increases in 1,25-(OH)2D3, we hoped to Wnd a simple answer implicating hypercalcemia or 1,25-(OH)2D3 as the causative agent in suppression of EAE. Instead our results reXect the complex etiology of this disease where several factors impinge on the immune system to give diVerent outcomes. This work suggests development of nonhypercalcemic analogs of 1,25-(OH)2D3 that completely suppress EAE will be diYcult not just because hypercalcemia appears to have immunosuppressive eVects in females, but also because of the varied impact such an analog would have on PTH and calcitonin secretion. A better understanding is needed about the relationship between the immune system and calcium homeostasis regulation to aid in design of 1,25-(OH)2D3 analogs with clinical potential. Thus, this work serves as a Wrst step in dissecting apart the immunosuppressive eVects of altered mineral homeostasis and the hormones involved. Acknowledgment This work has been funded in part by the Wisconsin Alumni Research Foundation. References [1] M.R. Haussler, G.K. WhitWeld, C.A. Haussler, J.C. Hsieh, P.D. Thompson, S.H. Selznick, C.E. Dominguez, P.W. Jurutka, J. Bone Miner. Res. 13 (1998) 325–349. [2] G. Penna, L. Adorini, J. Immunol. 164 (2000) 2405–2411. [3] M.D. GriYn, W. Lutz, V.A. Phan, L.A. Bachman, D.J. McKean, R. Kumar, Proc. Natl. Acad. Sci. USA 98 (2001) 6800–6805. [4] A. Boonstra, F.J. Barrat, C. Crain, V.L. Heath, V.H.F. Savelkoul, A. O’Garra, J. Immunol. 167 (2001) 4974–4980. [5] H. Tanaka, E. Abe, C. Miyaura, Y. Shiina, T. Suda, Biochem. Biophys. Res. Commun. 117 (1983) 86–92. [6] J.M. Lemire, D.C. Archer, J. Clin. Invest. 87 (1991) 1103–1107. [7] M.T. Cantorna, C.E. Hayes, H.F. DeLuca, Proc. Natl. Acad. Sci. USA 93 (1996) 7861–7864. [8] C. Mathieu, J. Laureys, H. Sobis, M. Vandeputte, M. Waer, R. Bouillon, Diabetes 41 (1992) 1491–1495. [9] M.T. Cantorna, C. Munsick, C. Bemiss, B.D. Mahon, J. Nutr. 130 (2000) 2648–2652. [10] M.T. Cantorna, C.E. Hayes, H.F. DeLuca, J. Nutr. 128 (1998) 68–72. [11] J.M. Lemire, A. Ince, M. Takashima, Autoimmunity 12 (1992) 143– 148. [12] L. Metz, Semin. Neurol. 18 (1998) 389–395. [13] P. Duquette, J. Pleines, M. Girard, L. Charest, M. Senecal-Quevillon, C. Masse, Can. J. Neurol. Sci. 19 (1992) 466–471. [14] K. Birk, S.C. Smeltzer, R. Rudick, Semin. Neurol. 8 (1988) 205–213. [15] C.E. Hayes, Proc. Nutr. Soc. 59 (2000) 531–535. [16] K.L. Munger, S.M. Zhang, E. O’Reilly, M.A. Hernan, M.J. Olek, W.C. Willett, A. Ascherio, Neurology 62 (2004) 60–65.
221
[17] E. Garcion, L. Sindji, S. Nataf, P. Brachet, F. Darcy, C.N. MonteroMenei, Acta Neuropathol. (Berl.) 105 (2003) 438–448. [18] M.T. Cantorna, J. Humpal-Winter, H.F. DeLuca, J. Nutr. 129 (1999) 1966–1971. [19] E.M. Brown, R. Leombruno, J. Thatcher, M. Burrowes, Endocrinology 116 (1985) 1123–1132. [20] A. Mithal, O. Kifor, I. Kifor, P. Vassilev, R. Butters, K. Krapcho, R. Simin, F. Fuller, S.C. Hebert, E.M. Brown, Endocrinology 136 (1995) 3087–3092. [21] J. Fox, H. Heath III, Am. J. Physiol. 240 (1981) E649–E655. [22] H.L. Brenza, C. Kimmel-Jehan, F. Jehan, T. Shinki, S. Wakino, H. Anazawa, T. Suda, H.F. DeLuca, Proc. Natl. Acad. Sci. USA 95 (1998) 1387–1391. [23] H. Dobnig, R.T. Turner, Endocrinology 138 (1997) 4607–4612. [24] S.C. Kukreja, P.A. York, C. Nalbantian-Brandt, D.H. Shevrin, M.J. Favus, Am. J. Physiol. 256 (1989) E309–E314. [25] H. Tsuji, C. Cawthorn, B. Ecarot, J. Bone Miner. Res. 11 (1996) 1234– 1240. [26] P. Jaeger, W. Jones, M. Kashgarian, R. Baron, T.L. Clemens, G.V. Segre, J.P. Hayslett, Am. J. Physiol. 252 (1987) E790–E798. [27] K. Endo, K. Katsumata, M. Hirata, T. Masaki, N. Kubodera, T. Nakamura, K. Ikeda, E. Ogata, J. Bone Miner. Res. 15 (2000) 175–181. [28] D.K. Panda, D. Miao, M.L. Tremblay, J. Sirois, R. Farookhi, G.N. Hendy, D. Goltzman, Proc. Natl. Acad. Sci. USA 98 (2001) 7498–7503. [29] O. Dardenne, J. Prud’homme, A. Arabian, F.H. Glorieux, R. St-Arnaud, Endocrinology 142 (2001) 3135–3141. [30] O. Dardenne, J. Prud’homme, S.A. Hacking, F.H. Glorieux, R. Snst-Arnaud, Bone 32 (2003) 332–340. [31] O. Dardenne, J. Prudhomme, S.A. Hacking, F.H. Glorieux, R. St Arnaud, J. Bone Miner. Res. 18 (2003) 637–643. [32] H. Rasmussen, H. Deluca, C. Arnaud, C. Hawker, M. Vonstedingk, J. Clin. Invest. 42 (1963) 1940–1946. [33] C.C. Bernard, T.G. Johns, A. Slavin, M. Ichikawa, C. Ewing, J. Liu, J. Bettadapura, J. Mol. Med. 75 (1997) 77–88. [34] W. Slavin, Methods Enzymol. 158 (1988) 117–145. [35] K. Itaya, M. Ui, Clin. Chim. Acta 14 (1966) 361–366. [36] T.F. Meehan, H.F. DeLuca, Arch. Biochem. Biophys. 408 (2002) 200–204. [37] Y. Zhu, B.D. Mahon, M. Froicu, M.T. Cantorna, Eur. J. Immunol. 35 (2005) 217–224. [38] D.D. Branisteanu, M. Waer, H. Sobis, S. Marcelis, M. Vandeputte, R. Bouillon, J. Neuroimmunol. 61 (1995) 151–160. [39] F. Mattner, S. Smiroldo, F. Galbiati, M. Muller, P. Di Lucia, P.L. Poliani, G. Martino, P. Panina-Bordignon, L. Adorini, Eur. J. Immunol. 30 (2000) 498–508. [40] E. van Etten, D.D. Branisteanu, A. Verstuyf, M. Waer, R. Bouillon, C. Mathieu, Transplantation 69 (2000) 1932–1942. [41] E. Van Etten, D.D. Branisteanu, L. Overbergh, R. Bouillon, A. Verstuyf, C. Mathieu, Bone 32 (2003) 397–404. [42] M. Zaidi, H.K. Datta, B.S. Moonga, I. MacIntyre, J. Endocrinol. 126 (1990) 473–481. [43] F.R. Singer, Calcif. Tissue Int. 49 (Suppl. 2) (1991) S7–S8. [44] M. Azria, D.H. Copp, J.M. Zanelli, Calcif. Tissue Int. 57 (1995) 405–408. [45] R.D. Tiegs, J.J. Body, J.M. Barta, H. Heath III, J. Bone Miner. Res. 1 (1986) 339–349. [46] M. Whitehead, G. Lane, O. Young, S. Campbell, G. Abeyasekera, C.J. Hillyard, I. MacIntyre, K.G. Phang, J.C. Stevenson, Br. Med. J. (Clin. Res. Ed.) 283 (1981) 10–12. [47] S.U. Toverud, C.W. Cooper, P.L. Munson, Endocrinology 103 (1978) 472–479. [48] J.J. Body, F. Glibert, S. Nejai, G. Fernandez, A. Van Langendonck, A. Borkowski, J. Clin. Endocrinol. Metab. 71 (1990) 675–681. [49] A. Murayama, K. Takeyama, S. Kitanaka, Y. Kodera, Y. Kawaguchi, T. Hosoya, S. Kato, Endocrinology 140 (1999) 2224–2231.