Do Cyclooxygenase-2 Knockout Mice Have Primary Hyperparathyroidism?
http://www.100md.com
内分泌学杂志 2005年第4期
Departments of Medicine (M.X., S.C., L.R., C.P.) and Orthopedics (F.L., D.A., G.G., B.K.-J.), University of Connecticut Health Center, Farmington, Connecticut 06030; and Department of Medicine, McGill University Health Center and McGill University (D.G.), Montréal, Canada 3A 1A1
Address all correspondence and requests for reprints to: Dr. Carol Pilbeam, Room AM047, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, Connecticut 06030. E-mail: pilbeam@nso.uchc.edu.
Abstract
The absence of cyclooxygenase-2 (COX-2) activity in vitro reduces differentiation of both bone-forming and bone-resorbing cells. To examine the balance of COX-2 effects on bone in vivo, we studied COX-2 knockout (KO) and wild-type (WT) mice. After weaning, KO mice died 4 times faster than WT mice, consistent with reports of progressive renal failure in KO mice. Among KO mice killed at 4 months of age, some had renal failure with marked secondary hyperparathyroidism, but others appeared healthy. On the assumption that renal failure was not inevitable in COX-2 KO mice and that phenotypic differences might increase with age, we studied KO mice surviving to 10 months of age with serum creatinine levels similar to those of WT mice. In 10-month-old male KO mice, serum calcium and PTH, but not phosphorus, levels were increased compared with those in WT mice. 1,25-Dihydroxyvitamin D3 levels were markedly elevated in KO mice. Skeletal analysis showed small nonsignificant decreases in cortical bone density by BMD and either an increase (distal femur, by microcomputed tomography) or no difference (distal femur, by static histomorphometry) in trabecular bone density in KO mice. There was a trend toward increased percent osteoblastic and osteoclastic surfaces, and on dynamic histomorphometry, the rates of trabecular bone formation and mineral apposition were increased in KO mice relative to WT mice. Similar trends were observed for most parameters in 10-month-old female COX-2 KO mice. However, rates of trabecular bone formation and mineral apposition were increased in 10-month-old WT females compared with males and did not increase further in female KO mice. These data suggest that COX-2 KO mice with intact renal function have primary hyperparathyroidism, and that effects of increased PTH and 1,25-dihydroxyvitamin D3 to increase bone turnover may compensate for the absence of COX-2.
Introduction
THERE ARE TWO enzymes for cyclooxygenase (COX) encoded by separate genes, COX-1 and COX-2 (1, 2, 3). In most tissues, COX-1 is constitutively expressed, whereas COX-2 is transiently induced. COX-2 is highly expressed in bone and is regulated by multiple local and systemic factors important in bone metabolism (4). In the absence of COX-2, little prostaglandin E2 (PGE2) is produced by cultured osteoblasts or cultured bone marrow cells (5, 6, 7). Agonists shown to induce COX-2 expression and PGE2 production in osteoblastic cells include cytokines [IL-1 (8, 9, 10), TNF- (11), and IL-6 (12)], growth factors [TGF- (13, 14), TGF-? (14), bone morphogenetic protein-2 (6), and basic fibroblast growth factor (15)], fetal bovine serum (16), systemic hormones [PTH (8, 17) and 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) (5)], extracellular calcium (7), and fluid shear stress or mechanical loading (18, 19, 20). PGs themselves induce COX-2 expression and therefore can amplify PG responses to other agonists (21).
PGs produced by COX-2 can be potent stimulators of both bone resorption and bone formation (4). In addition, the many factors that induce COX-2 expression generally have their own independent effects on bone turnover. It is therefore difficult to predict the effects of COX-2 expression on bone mass in vivo. Several studies using mice with disruption of the COX-2 gene, COX-2 knockout (KO) mice, have suggested that bone mass or density is decreased in COX-2 KO relative to WT mice (22, 23). However, it is unclear whether COX-2 KO mice can be used for studying the direct effects of COX-2 absence on bone turnover. The developers of COX-2 KO mice reported that KO mice have nephropathy severe enough to decrease the average life span to approximately 3–4 months (24, 25). A recent study reported that COX-2 KO mice have progressive renal deterioration, with 20% of mice dying between 7–23 wk of age (26). Loss of functioning renal mass in mice, as in humans, is expected to decrease 1,25(OH)2D3 production, thereby decreasing serum calcium levels, and decreased serum calcium levels are expected to increase PTH secretion. This secondary hyperparathyroidism could increase bone turnover, making it difficult to separate out the direct effects of the absence of COX-2 on bone from those secondary to nephropathy.
In the present study we studied differences in skeletal phenotype due to COX-2 absence, apart from renal dysfunction if possible. Although COX-2 KO mice did die faster than WT mice, renal failure did not appear to be universal. Initially, we killed mice at 4 months of age and found several with renal failure associated with marked secondary hyperparathyroidism, whereas others appeared relatively healthy. On the assumption that most COX-2 KO mice with progressive renal dysfunction would die within the first 6–8 months of life and that phenotypic differences between COX-2 WT and KO mice might be most evident in older mice, we then studied 10-month-old mice. Our results suggest that COX-2 KO mice without functional renal deterioration have primary hyperparathyroidism with increased levels of 1,25(OH)2D3. The skeletal differences between COX-2 WT and KO mice are small, and we speculate that the effects of hyperparathyroidism on bone turnover compensate for the absence of COX-2 in bone.
Materials and Methods
Animals
The COX-2 KO mice used in this study were developed at University of North Carolina on a C57BL/6, 129SV background (25). Because COX-2 KO females are infertile (27) and because we wanted to obtain both WT and KO in the same litter, we bred COX-2 WT and KO mice for experiments by mating mice heterozygous for the disrupted COX-2 allele (COX-2 HET mice). To establish founders for the breeding colonies, we backcrossed the original C57BL/6,129SV HET mice onto a pure C57BL/6 background (backcrossed more than nine generations). Colonies were maintained by mating the C57BL/6 HET founders with C57BL/6 WT mice from The Jackson Laboratory (Bar Harbor, ME), which were refreshed every 6 months to avoid genetic drift. We tried to generate COX-2 KO mice by crossing COX-2 HET mice on a C57BL/6 background, but found few or no KO mice in the litters. A similar observation was made by Dr. Scott Morham (personal communication), who developed the COX-2 KO mice. Hence, the C57BL/6 COX-2 HET mice were outbred to 129P3/J WT mice to maintain the production of KO mice. The final breeding protocol for COX-2 KO and WT mice used in this study is shown in Fig. 1. Mice were genotyped as described previously (5). All animal protocols were approved by the animal care and use committee of University of Connecticut Health Center (Farmington, CT).
FIG. 1. Breeding scheme for generating COX-2 WT and KO mice from COX-2 HET mice.
Analytical procedures on serum
Mice were euthanized at 4 months (±2 wk) or 10 months (±3 wk) of age. Mice were killed in groups containing eight or more mice of both genotypes upon reaching the appropriate age. Blood was obtained by heart puncture at the time of death after anesthesia with gaseous carbon dioxide. Mice were not fasted. Blood was allowed to clot at room temperature, and the serum was removed after centrifugation of samples at 3000 rpm for 20 min. Serum from each animal was divided into aliquots and frozen at –20 C. There was no repeated freezing and thawing of specimens before measurements.
Serum creatinine, phosphorus, and calcium measurements were measured using kits (procedures 555, 360-UV, and 587, respectively) from Sigma-Aldrich Corp. (St. Louis, MO) following the manufacturer’s directions. Five microliters of serum were needed to assay calcium or phosphorus, and 30 μl serum were needed to assay creatinine. PTH was assayed on 25 μl serum using a Mouse Intact PTH ELISA Kit from Immutopics (San Clemente, CA). This is a two-site assay, recognizing epitopes within the midregion/C-terminal portion and epitopes within the N-terminal portion of the PTH peptide. The manufacturer reports an intraassay coefficient of variation (CV) of 2.5–3.9% and an interassay CV of 7.8–8.9%. 1,25(OH)2D3 was assayed on 100 μl serum by immunoextraction followed by 125I RIA using a kit from Immunodiagnostics Systems (IDS Ltd., Tyne and Wear, UK). The manufacturer reports an intraassay CV of 5–8% and an interassay CV of 9–10%.
Skeletal imaging
Femurs and lumbar vertebrae were dissected free of connective tissue at the time of death and stored in 70% ethanol. Areal bone mineral density (BMD) and bone mineral content (BMC) of the femur and of the first through fifth lumbar vertebrae (L1–L5) were measured using a PIXImus2 densitometer (LUNAR, GE Medical Systems, Madison, WI). Trabecular morphometry within the metaphyseal region of the distal femur and the centrum of the third lumbar vertebrae (L3) was quantified using x-ray microcomputed tomography (μCT20, Scanco Medical AG, Bassersdorf, Switzerland). Volumetric analysis regions for measuring trabecular bone volume fraction (BV/TV) were selected within the endosteal borders to include the central 80% of vertebral height and the secondary spongiosa of femoral metaphyses located 760 μm (5% of length) distal to the physis.
Bone histomorphometry
For static histomorphometry, femurs were fixed in 4% paraformaldehyde at 4 C, decalcified in 15% EDTA, dehydrated in progressive concentrations of ethanol, cleared in xylene, and embedded in paraffin. Five-micrometer sections were cut, stained for tartrate-resistant acid phosphatase to visualize osteoclasts, and counterstained with hematoxylin. Measurements included cortical width, percent BV/TV, percent osteoblast surface, percent osteoclast surface, and trabecular spacing and number.
For dynamic histomorphometry, mice received calcein (Sigma-Aldrich Corp.; 10 mg/kg body weight) 7 d before death and xylenol orange (Sigma-Aldrich Corp.; 90 mg/kg body weight) 2 d before death, administered ip. Femurs were dissected free of tissue and fixed in 70% ethanol at the time of death, then dehydrated in increasing concentrations of ethanol, cleared in xylene, and embedded undecalcified in methyl methacrylate. Five-micrometer-thick longitudinal serial sections were cut on a Polycut S microtome (Reichert-Jung, Nussloch, Germany) with a D profile knife (Delaware Diamond Knives Corp., Wilmington, DE).
Both static and dynamic measurements were made in a blinded, nonbiased manner using the OsteoMeasure computerized image analysis system (OsteoMetrics, Inc., Atlanta, GA) interfaced with an Optiphot Nikon microscope (Nikon, Inc., Melville, NY). The terminology and units used are those recommended by the histomorphometry nomenclature committee of the American Society for Bone and Mineral Research (28). All measurements were confined to the growth plate-metaphyseal junction. Osteoblasts were identified as cuboidal cells lining the trabecular perimeter. Osteoclasts were identified as multinucleated cells on the trabecular bone surface. In static measurements, cortical width was measured in the diaphysis 4000 μm from the growth plate. Dynamic measurements included single and double-labeled surfaces, mineral apposition rate (MAR), and bone formation rate (BFR/BS). In addition, a mineral apposition rate of 0.3 μm/d was assigned to samples that had single labels, but no double labels (29).
Renal histology
Mouse kidneys were halved; fixed in 10% neutral-buffered formalin; processed through 70%, 95%, and absolute ethanol; embedded in a paraffin block; and sectioned at 5-μm intervals. The paraffin sections were deparaffinized in toluene and rehydrated through graded alcohols to distilled water, then stained with hematoxylin and eosin.
Statistics
Analysis of data was performed using SigmaStat for Windows, version 2.03 (San Rafael, CA). When the data were distributed normally, means of groups were compared by ANOVA, with post hoc testing of pairwise multiple comparisons by Bonferroni’s method. When the data failed to meet the normality test, Kruskal-Wallis ANOVA on ranks was performed, followed by Dunn’s test for all pairwise multiple comparisons.
Results
Mortality of COX-2 KO mice
Previous studies had suggested that renal abnormalities in COX-2 KO mice were present shortly after birth and resulted in death by 3–4 months of age (24, 25, 26). We wanted to study the COX-2 KO phenotype in mice without compromised renal function so that the effects of COX-2 disruption on skeletal phenotype could be studied without being complicated by the secondary hyperparathyroidism associated with renal failure. We hypothesized that COX-2 KO mice surviving to 4 months of age might be free of renal abnormalities. When we found some 4-month-old KO mice with renal failure, we expanded the study to include mice killed at 9–10 months of age. Cohorts of COX-2 WT and KO mice were identified at 2 months of age and set aside to be killed at 4 or 10 months of age.
The estimated mortality in these cohorts from 2 months of age is shown in Table 1. Deaths were generally reported by the animal care facility, and animals were removed without additional investigation. The total numbers of mice do not reflect the distribution of COX-2 genotypes born because some mice died or were removed before 2 months of age. From these data, it is estimated that 0.5–1.5% of WT or HET mice died per month and 4–6% of KO mice died per month from 2–10 months of age. Hence, the death rate among KO mice was about 4 times greater than the rate among WT or HET mice. Contrary to our initial hypothesis, there was not a decrease in the death rate among KO mice after 4 months of age.
TABLE 1. Deaths in mice with both COX-2 alleles intact (WT), one COX-2 allele disrupted (HET), or both COX-2 alleles disrupted (KO)
Serum creatinine, calcium, phosphorus, and PTH levels in COX-2 WT and KO mice
Body weights and serum data for the 4-month-old mice are shown in Table 2. Two of the KO mice killed at 4 months of age had serum creatinine levels greater than 1.5 mg/dl. Data for these mice were excluded from the summary in Table 2 and are shown in Table 3. These mice looked sick at time of death. They had low body weight, low serum calcium, high serum phosphorus, and extremely high PTH levels. Thus, these mice appeared to have renal failure with secondary hyperparathyroidism.
TABLE 2. Comparison of body weights and serum parameters in COX-2 WT and KO mice killed at 4 months of age
TABLE 3. COX-2 KO male mice, killed at 4 months of age, with apparent renal failure
After exclusion of these two mice, the mean serum creatinine level was still 30% higher in COX-2 KO males compared with WT males (Table 2). Fewer females were examined, and although not statistically significant, a similar trend toward higher serum creatinine was seen in female KO mice. PTH levels in KO mice also tended to be higher than those in WT mice, and this difference was statistically significant in female mice. It was difficult to explain the trend toward higher PTH levels in KO male mice as being due to renal failure in view of the low phosphorus level in KO males, but the role of renal dysfunction in the elevated PTH in KO females was unclear.
We examined gross renal histology from some healthy 4-month-old WT and KO mice. Although some KO kidneys had lymphocytic infiltrates not seen in WT kidneys, we found no differences in glomerular or tubular structures between WT and KO kidneys (data not shown). It seemed, therefore, that renal disease affecting COX-2 KO mice was not universal. Hence, we decided to examine older mice, with the hope that most or all KO mice surviving to 10 months of age might have intact renal function, and other metabolic abnormalities might become more evident.
Body weights and serum parameters in 10-month-old COX-2 WT and KO mice are shown in Table 4. Weights tended to be lower in KO mice compared with WT mice, but this difference was statistically significant only for females (8%). There was a nonstatistically significant trend for creatinine levels to be higher (10%) in KO mice compared with WT mice. Weights and creatinine levels were normally distributed around the medians, as indicated by the box plots shown in Fig. 2. Box plots for serum calcium, phosphorus, and PTH data are shown in Fig. 3. There was a small, but significant, increase (8%) in mean serum calcium in COX-2 KO male mice compared with WT male mice, but variability was also greater in KO mice (Fig. 3A). WT females had higher calcium and phosphorus levels than WT males, but there was no difference between WT and KO females. The mean PTH level was 50% higher in 10-month-old COX-2 KO males compared with WT males (Table 4). There was also a trend for female KO mice to have higher PTH levels than WT mice. Unlike the other serum parameters, PTH values for males, females, or both combined, as shown in Fig. 3C, were not normally distributed.
TABLE 4. Comparison of body weights and serum parameters measured in COX-2 WT and KO mice killed at 10 months of age
FIG. 2. Body weights and serum creatinine levels measured in 10 COX-2 WT and KO mice killed at 10 months of age. Boxes are bounded by the 25th and 75th percentiles, bars indicate the 10th and 90th percentiles, and the population median is indicated by the line within the box.
FIG. 3. Serum parameters measured in COX-2 WT and KO mice killed at 10 months of age. Boxes are bounded by the 25th and 75th percentiles, bars indicate the 10th and 90th percentiles, and the population median is indicated by the line within the box.
Only one mouse was excluded from the summary of mice killed at 10 months of age. This was a female mouse with a serum PTH level of 624 ng/ml, which was an outlier (>2 SD from the mean). Although this mouse had low body weight (19.6 g), the highest serum creatinine in the female KO group (0.81 mg/dl), and high phosphorus (12.4 mg/dl), consistent with renal dysfunction, its serum calcium (12.0 mg/dl) was high. Inclusion of this mouse in the summary data would have made the PTH difference between WT and KO females statistically significant. Regression of serum calcium (dependent variable) for 10-month-old male and female KO mice against PTH (independent variable) showed a small, significant, positive correlation (r = 0.35; P < 0.05; n = 43; data not shown). No other statistical significant correlations between pairs of parameters or between weights and parameters were found.
Serum 1,25(OH)2D3 levels in COX-2 WT and KO mice
Overall, the serum data from the 10-month-old mice suggested primary, rather than secondary, hyperparathyroidism in COX-2 KO mice. In an effort to clarify this, we tested whether the KO mice had impaired 1,25(OH)2D3 synthesis, which is expected in chronic renal dysfunction, or increased 1,25(OH)2D3 levels, which is expected in primary hyperparathyroidism. We measured 1,25(OH)2D3 levels for all mice for which adequate serum remained after the above assays were completed (Table 5). In 4-month-old female mice, there was no difference between WT and KO mice, but there were few KO data, and variability was high. In 4-month-old males and 10-month-old males and females, KO mice had 1,25(OH)2D3 levels 1.8- to 2.3-fold greater than the WT mice.
TABLE 5. Comparison of serum 1,25(OH)2D3 (picograms per deciliter) values measured in COX-2 WT and KO mice
Because we had no serum for measurement of 1,25(OH)2D3 levels from the few mice identified with definite renal dysfunction, we followed a small group of female mice as they continued to age, and when a KO mouse appeared to become sick at 12 months of age, we killed it. At the same time, we killed seven female WT mice ranging in age from 13–16 months. Serum from these mice was processed simultaneously, and results are shown in Table 6. The sick KO mouse had a serum creatinine level of 1.04, with lower body weight, lower serum calcium, and higher serum phosphorus than the WT mice. Although the sick KO mouse had an elevated PTH level, its 1,25(OH)2D3 level was very low, as expected for renal failure.
TABLE 6. Comparison of a group (n = 7) of COX-2 WT female mice, killed at 13–16 months of age, with a sick COX-2 KO female mouse, killed at 12 months of age
The kidneys from this sick KO mouse were smaller than those from WT mice (Fig. 4). As we had noticed frequently in kidneys from 4- and 10-month-old KO mice that had no structural abnormalities (data not shown), the KO mouse had lymphocytic infiltrates especially evident around blood vessels (Fig. 4). There was also evidence for enlarged renal tubules and dilated calices in the sick KO mouse.
FIG. 4. Sections of kidneys, stained with hematoxylin and eosin, from a 13-month-old female COX-2 WT mouse (A) and a 12-month-old female COX-2 KO mouse (B) with serum parameters consistent with renal failure and secondary hyperparathyroidism. Higher magnification shows enlarged tubules and lymphocytic infiltration in the KO kidney (D) compared with the WT kidney (C).
Skeletal analysis of COX-2 WT and KO mice: imaging
BMD was measured on dissected left femurs and lumbar vertebrae (L1–L5) from 4- and 10-month-old mice (Table 7). BMD tended to be lower in COX-2 KO mice compared with WT mice in all groups. The difference was statistically significant for 4-month-old female lumbar vertebral BMD (12%) and 10-month-old female femur BMD (9%). In some groups, vertebral (4-month-old female KO mice) and femoral (10-month-old WT females) BMDs were higher (6%) in females than in males. Lumbar BMD, but not femur BMD, decreased about 9% with age in WT mice, suggesting that trabecular, but not cortical, bone mass decreased with age.
TABLE 7. Comparison of BMD (milligrams per square centimeter), measured by Dexa, in COX-2 WT and KO mice
The 10-month-old female COX-2 KO mouse, excluded from the summary analysis because of possible renal dysfunction, as discussed above, had low femur BMD (58.4 mg/cm2) compared with the remaining mice in this group (65.3 mg/cm2). Femur BMD was also measured in one of the 4-month-old KO mice excluded for renal failure, and this value (59.9 mg/cm2) was low compared with that in the remaining group (67.6 mg/cm2). These observations suggest that renal abnormalities associated with effects of COX-2 absence can decrease BMD.
μCT analyses of the BV/TV in the distal femur and in the L3 vertebral body are shown in Table 8. Unlike BMD, which measures largely cortical bone, BV/TV did not tend to decrease in KO mice compared with WT mice. In fact, there was a significant increase in BV/TV in the distal femur (26%) in 10-month-old male KO mice compared with WT male mice. Also, in contrast to the BMD comparisons, 10-month-old females had 40–50% lower BV/TV in the distal femur than males. There was also a significant decrease in femur BV/TV at 10 months of age compared with 4 months of age for both males (56–68%) and females (44–48%).
TABLE 8. Comparison of BV/TV (percentage), measured by μCT, in COX-2 WT and KO mice
Skeletal analysis of COX-2 WT and KO mice: histomorphometry
Static histomorphometric analyses on the distal femur showed decreased cortical width in 4-month-old male KO mice compared with WT (9%), but not in 10-month-old males (Table 9). There was also a significant decrease in cortical width in 10-month-old KO females compared with WT females (9%). No effect of genotype on percent BV/TV was seen. Female mice tended to have lower BV/TV than males in all groups, with a significant difference (35%) between WT females and males at 10 months of age. There was a trend toward higher osteoblast and osteoclast surfaces in KO mice compared with WT mice. The difference (40–60%) was significant for percent osteoclast surface in females.
TABLE 9. Comparison of static histomorphometric parameters measured in the distal femur in COX-2 WT and KO mice
Dynamic histomorphometry of the distal femur is shown in Table 10. There was a significant increase of 55% in MAR and 70% in BFR/BS in 10-month-old KO male mice compared with WT mice, but not in female mice. Interestingly, both MAR and BFR/BS were 1.5- to 2.6-fold higher in females than in males at 10 months of age. This difference appeared to be largely due to the lack of age-related decreases in MAR and BFR/BS in females. There were no differences between genotypes in the other parameters measured (data not shown).
TABLE 10. Comparison of dynamic histomorphometric parameters measured in the distal femur in COX-2 WT and KO mice
Discussion
We found no increase in mortality among COX-2 HET mice, but there was high mortality among COX-2 KO mice, as reported in other studies (24, 25, 26, 30). One study reported that 35% of COX-2 KO mice died with a patent ductus arteriosus within 48 h of birth (30). Other studies have not found a high incidence of neonatal death among COX-2 KO mice, but have found shortened life span secondary to renal dysplasia (25, 26). Renal development in COX-2 KO mice appeared normal until postnatal d 10, after which there was progressive renal architectural disruption and functional deterioration, and about 20% of KO mice died between 7 and 23 wk of age from renal failure (26). We found that approximately 40% of COX-2 KO mice died between 2 and 10 months of age. All of these studies used mice in a mixed C57BL/6 and 129 genetic background, and it is unclear what impact various degrees of inbreeding may have had on the results. For example, C57BL/6 mice lack type IIA secreted phospholipase A2; as a result, PG production associated with induction of COX-2 in these mice depends largely on cytosolic phospholipase A2 (31). In COX-2 KO mice backcrossed into an outbred CD-1 background, our preliminary observations suggest that there is no increased mortality (data not shown). Hence, the severity of problems associated with the absence of COX-2 could be increased in a pure C57BL/6 background, and this might account for our observation that backcrossing cox-2–/– onto a C57BL/6 genetic background resulted in decreased numbers of COX-2 KO mice in litters after weaning.
As expected, some COX-2 KO mice had evidence of renal tubular impairment at the time of death. These mice had decreased serum calcium, increased serum phosphorus, and markedly elevated PTH, similar to the changes seen in secondary hyperparathyroidism in humans (32, 33, 34, 35), dogs (36), rats (37, 38, 39, 40, 41), and mice (42). The mechanisms proposed to account for secondary hyperparathyroidism in renal failure include 1) phosphorus retention, which inhibits the enzyme 25-hydroxyvitamin D-1-hydroxylase (1-hydoxylase) that converts 25(OH)2D3 to 1,25(OH)2D3, 2) reduction of 1,25(OH)2D3 as a result of phosphorus retention and decreased renal mass, 3) increased parathyroid hyperplasia due to phosphorus retention and reduced 1,25(OH)2D3, 4) resistance to the actions of 1,25(OH)2D3 and PTH as a result of reduced numbers of vitamin D receptors and calcium sensors, and 5) subsequent decrease in serum calcium resulting from all of the above (32, 33, 34, 35). Thus, decreased 1,25(OH)2D3 should be an important factor in the development of secondary hyperparathyroidism. Although we could find no studies of 1,25(OH)2D3 in renal failure in mice, there are several studies in rats showing the same relationships as those in humans (37, 38, 40, 41). We were able to measure serum 1,25(OH)2D3 in one of the COX-2 KO mice with evident renal failure and found it to be very low compared with WT mice and KO mice without renal failure.
In both male and female COX-2 KO mice with relatively intact renal function, there was still a trend for mean PTH levels to be higher than levels in WT mice at all ages. This difference was statistically significant for 4-month-old KO female mice and 10-month-old male KO mice. In addition, in no group was the serum calcium level in KO mice lower than that in WT mice. Hence, the higher PTH levels in KO mice were not driven by decreased serum calcium, suggesting that the KO mice had primary hyperparathyroidism. Although serum calcium was only significantly elevated in 10-month-old male KO mice compared with WT mice, serum calcium could be said to be inappropriately high, perhaps corresponding to normocalcemic primary hyperparathyroidism in humans (43).
As another indication of whether the increased PTH levels were due to renal insufficiency or to primary hyperparathyroidism, we measured 1,25(OH)2D3 levels. PTH stimulates the conversion of 25(OH)D3 to 1,25(OH)2D3 by 1-hydroxylase, and in primary hyperparathyroidism, elevated PTH levels are associated with increased serum calcium, decreased or normal serum phosphorus, and increased 1,25(OH)2D3 levels (44, 45). 1,25(OH)2D3 levels were increased 1.8-fold in male KO mice at 4 months of age and in both male and female KO mice at 10 months of age. We had too few measurements from 4-month-old females for an adequate assessment of 1,25(OH)2D3. Decreased serum phosphorus can also stimulate conversion of 25(OH)D3 to 1,25(OH)2D3 by 1-hydroxylase. Although serum phosphorus was low in 4-month-old male COX-2 KO mice compared with WT mice, it was not low in 10-month-old KO male or female mice and thus cannot account for the elevated 1,25(OH)2D3 level in these mice. The absence of COX-2 could have a direct effect on 1-hydroxylase activity. One study has shown a biphasic effect of PGE2 on 1,25(OH)2D3 production (46), but several others have shown PGE2 to increase 1-hydroxylase activity (47, 48). PGE2 has also been shown to increase both 1-hydroxylase activity and vitamin D receptors in a human myeloid leukemia cell line (49). If the absence of PGs produced by COX-2 decreased vitamin D receptors, there might be a compensatory increase in 1,25(OH)2D3 production, but this would not explain the trend toward higher PTH levels in KO mice.
Many studies in transgenic mice have now shown that the PTH-vitamin D-calcium regulation is similar in mice and humans (42, 50, 51, 52, 53, 54). The simplest explanation consistent with the data is that COX-2 KO mice have primary hyperparathyroidism. Early renal insufficiency in the KO mice might have increased variability in the measurements. A study in humans of the influence of renal function on the features of primary hyperparathyroidism found that PTH and serum calcium were higher and 1,25(OH)2D3 levels were lower in patients with renal insufficiency and primary hyperparathyroidism compared with those in patients with primary hyperparathyroidism alone (55).
PGs have been shown to affect parathyroid function. Early studies of cells and fragments from bovine PTG showed that PTG cells secreted PGs (6-keto-PGF1>PGF2>PGE2) and that PGF2 could inhibit, and PGE2 could stimulate, PTH secretion (56, 57). This modulation of PTH secretion by endogenous PGs might depend on COX-2 expression. Extracellular Ca2+ is a potent inducer of COX-2 in kidney and bone cells (7, 58). In the kidney, PGE2 and extracellular Ca2+ can inhibit the same pool of adenylyl cyclase via different mechanisms, leading to the suggestion that PGE2 increases the sensitivity of the cells to Ca2+ so that effects can be seen at physiological concentrations of extracellular Ca2+ (59). Hence, absence of COX-2-produced PGE2 might decrease the Ca2+ sensitivity of PTG cells, leading to increased PTH secretion.
The absence of COX-2 expression might be expected to affect both bone resorption and formation. PGE2 is abundantly produced by osteoblasts in response to multiple agonists, largely as a result of the induction of COX-2. PGE2 is a potent stimulator of bone resorption and can mediate in part the resorptive activities of other agonists (4). When the specific role of COX-2 was examined in marrow cultures from COX-2 KO mice, the ability of 1,25(OH)2D3 or PTH to stimulate osteoclast formation was reduced 60–70% in KO cultures compared with WT cultures (5). PTH injected above the calvariae caused hypercalcemia in COX-2 WT mice, but not in COX-2 KO mice (5). In contrast, PGE2 can also stimulate bone formation. In vivo studies in rats have demonstrated a potent anabolic effect of exogenous PGE2 (60, 61). Exogenous PGE2 stimulates osteoblastic differentiation in marrow stromal and primary calvarial cell cultures (62), and PGE2 injected in vivo enhanced osteoblastic differentiation in explanted marrow stromal cells (63). Fresh serum also induces COX-2 and endogenous PG production in cultured osteoblastic cells (16), and osteoblastic differentiation in cultured marrow stromal cells and primary calvarial cells from COX-2 KO mice was delayed compared with that in WT cultures (64, 65).
Few studies have examined the balance of effects of endogenous PGs in vivo on bone metabolism or bone mass in the absence of inflammatory conditions. One study of 3-month-old male COX-2 KO mice in a mixed C57BL/6, DBA background reported that COX-2 KO cortical bones were thinner and more porous than WT cortical bones with decreased elastic modulus compared with WT mice (23). Another recent study used COX-2 KO mice reported to have increased mortality as well as significantly decreased body weights, BMC (normalized to weights), and volumetric BMD of long bones compared with WT mice (22). This study concluded that the anabolic response of bone to mechanical loading was not affected by the absence of COX-2. Our data along with those from previous studies (24, 25, 26) suggest that COX-2 KO mice with low body weight and high mortality are likely to have impaired renal function and elevated PTH levels. These high PTH values could have marked effects on bone turnover. A study examining BMD of the whole body in humans taking nonsteroidal antiinflammatory drugs (NSAIDs) and aspirin concluded that the combination of relative COX-2-selective NSAIDs and aspirin, which inhibits COX-1 activity, was associated with higher BMD (66). However, the subjects in this study were taking NSAIDs for a variety of reasons, such as rheumatoid arthritis, and may have had increased presence of inflammatory mediators, which could stimulate both resorption and COX-2 production of PGs.
In the current study, BMD analysis, which measures predominantly cortical bone, suggested that COX-2 KO mice had small decreases (<10%) in cortical bone density compared with WT mice. Cortical bone width measured by histomorphometry also tended to be decreased in COX-2 KO mice. Trabecular bone volume in KO mice was either no different from that in WT mice or increased relative to that in WT mice. Measurements of osteoblastic and osteoclastic surfaces suggested increased turnover in KO mice. Both trabecular bone formation rate and mineral apposition rate were significantly elevated in 10-month-old male KO mice compared with WT mice, but not in female mice. However, formation and mineral apposition rates did not decrease with age in WT females as they did in males and were higher in females than in males at 10 months of age. Perhaps the effects of mild hyperparathyroidism are more apparent in 10-month-old male KO mice because rates of formation had decreased. Both males and females lost trabecular bone volume with age, although females had lower trabecular bone volume than males at both 4 and 10 months of age. Higher bone turnover in WT females than males at 10 months of age might explain the observation that serum calcium and phosphorus decreased in WT males from 4–10 months of age, but not in females. In general, the skeletal analyses, particularly in COX-2 KO males, are consistent with increased bone turnover associated with hyperparathyroidism.
We had the statistical power to detect differences of 8–10% or greater in many of the measurements and cannot rule out smaller differences between WT and KO mice. Variability in measurements could come from several sources. Because we needed to outbreed to obtain KO mice, our experimental mice came from C57BL/6,129P3/JxC57BL/6,129P3/J crosses. The inherent variability in the skeletal phenotype with the variable genetic background of the litters of such crosses may limit the ability to define small differences. It is important to note, however, that published studies on COX-2 KO mice have used mice bred in a similar mixed background (22, 23, 24, 25, 26, 27, 65). We have generated C57BL/6 COX-2 HET mice and 129P3/J COX-2 HET mice by backcrossing into each background, and the offspring of these mice, which will have a more defined genetic background than the mice in our current study, will be studied in the future. Preliminary observations of the COX-2 KO offspring of these mice show the same increased mortality as that seen in this study; thus, these mice probably have similar renal abnormalities (data not shown). Because the severity of the effects of absent COX-2 expression may depend on the relative amount of a particular genetic background, we have also bred the COX-2 HET genotype into the outbred CD-1 background for future studies. Preliminary observations suggest that COX-2 KO mice on this background do not have increased mortality after weaning (data not shown). Another source of variability might be varying degrees of early renal insufficiency in COX-2 KO mice.
In view of the dramatic effects of COX-2 absence on osteoblastic and osteoclastic differentiation in vitro, why are there no marked effects on bone mass in vivo? It could be argued that the effects seen in vitro are the result of factors that induce COX-2 expression transiently, such as addition of fresh serum to cultures or addition of PTH, and that under unstressed conditions in vivo or under conditions of continuous agonist secretion, a transiently inducible gene such as COX-2 would not be expected to be highly expressed. However, mechanical loading of bone occurs regularly, and mechanical loading can induce COX-2 and COX-2-derived PGs, which are thought to be important mediators of the anabolic response to mechanical loading (67). In vitro studies suggest that the absence of COX-2 should reduce both resorption and formation. If the net balance of resorption and formation remained unchanged, there might be only a delay in returning to steady state conditions after perturbation or challenge. Such an effect would fit with the delayed healing seen after fracture in COX-2 KO mice or in mice in which COX-2 activity has been inhibited by NSAIDs (68, 69). However, bone turnover is not decreased in the COX-2 KO mice. Hence, another explanation for the difference between in vitro and in vivo results is that the effects of absent COX-2 on bone turnover are compensated by increased PTH and/or 1,25(OH)2D3, both of which can stimulate bone resorption.
It is also possible that COX-1 plays a compensatory role in the absence of COX-2 (70). Although COX-1 is constitutively expressed, we have seen little PGE2 production in cells cultured from COX-2 KO mice (5, 6, 14), and we have not seen differential mRNA expression of COX-1 in tissues from COX-2 KO mice (data not shown). However, we cannot rule out some compensation in vivo. Because double COX-2 KO, COX-1 KO mice are not expected to survive, secondary to failure of the patent ductus arteriosus to close, we are generating COX-2 KO mice that are heterozygous for COX-1 for future study.
In conclusion, our study suggests that although COX-2 KO mice can have severe renal dysfunction, associated with marked secondary hyperparathyroidism, about 60% of COX-2 KO mice that reach 2 months of age survive to 10 months of age with relatively normal renal function. These surviving mice appear to have primary hyperparathyroidism, with elevated PTH and 1,25(OH)2D3 levels and increased rates of trabecular bone formation and mineral apposition, which may compensate for the absence of COX-2 in preserving bone mass.
Acknowledgments
We thank Dr. Kevin Claffey and Nancy Ryan for performing the renal histology, and Miren Gratton for performing the 1,25(OH)2D3 assays.
References
Herschman HR, Reddy ST, Xie W 1997 Function and regulation of prostaglandin synthase-2. Adv Exp Med Biol 407:61–66
Smith WL, Langenbach R 2001 Why there are two cyclooxygenase isozymes. J Clin Invest 107:1491–1495
Smith WL, DeWitt DL, Garavito RM 2000 Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem 69:145–182
Pilbeam CC, Harrison JR, and Raisz LG 2002 Prostaglandins and bone metabolism. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. New York: Academic Press; 979–994
Okada Y, Lorenzo JA, Freeman AM, Tomita M, Morham SG, Raisz LG, and Pilbeam CC 2000 Prostaglandin G/H synthase-2 is required for maximal formation of osteoclast-like cells in culture. J Clin Invest 105:823–832
Chikazu D, Li X, Kawaguchi H, Sakuma Y, Voznesensky OS, Adams DJ, Xu M, Hoshio K, Katavic V, Herschman HR, Raisz LG, and Pilbeam CC 2002 Bone morphogenetic protein 2 induces cyclo-oxygenase 2 in osteoblasts via a Cbfal binding site: role in effects of bone morphogenetic protein 2 in vitro and in vivo. J Bone Miner Res 17:1430–1440
Choudhary S, Wadhwa S, Raisz LG, Alander C, Pilbeam CC 2003 Extracellular calcium is a potent inducer of cyclo-oxygenase-2 in murine osteoblasts through an ERK signaling pathway. J Bone Miner Res 18:1813–1824
Kawaguchi H, Raisz LG, Voznesensky OS, Alander CB, Hakeda Y, Pilbeam CC 1994 Regulation of the two prostaglandin G/H synthases by parathyroid hormone, interleukin-1, cortisol and prostaglandin E2 in cultured neonatal mouse calvariae. Endocrinology 135:1157–1164
Min YK, Rao Y, Okada Y, Raisz LG, Pilbeam CC 1998 Regulation of prostaglandin G/H synthase-2 expression by interleukin-1 in human osteoblast-like cells. J Bone Miner Res 13:1066–1075
Harrison JR, Kelly PL, Pilbeam CC 2000 Involvement of C/EBP transcription factors in prostaglandin G/H synthase-2 gene induction by interleukin-1 in osteoblastic MC3T3–E1 cells. J Bone Miner Res 15:1138–1146
Kawaguchi H, Nemoto K, Raisz LG, Harrison JR, Voznesensky OS, Alander CB, Pilbeam CC 1996 Interleukin-4 inhibits prostaglandin G/H synthase-2 and cytosolic phospholipase A2 induction in neonatal mouse parietal bone cultures. J Bone Miner Res 11:358–366
Tai H, Miyaura C, Pilbeam CC, Tamura T, Ohsugi Y, Koishihara Y, Kubodera N, Kawaguchi H, Raisz LG, Suda T 1997 Transcriptional induction of cyclooxygenase-2 in osteoblasts is involved in interleukin-6-induced osteoclast formation. Endocrinology 138:2372–2379
Harrison JR, Lorenzo JA, Kawaguchi H, Raisz LG, Pilbeam CC 1994 Stimulation of prostaglandin E2 production by interleukin-1 and transforming growth factor- in osteoblastic MC3T3–E1 cells. J Bone Miner Res 9:817–823
Pilbeam CC, Rao Y, Voznesensky O, Kawaguchi H, Alander C, Raisz LG, Herschman H 1997 Transforming growth factor-?1 regulation of prostaglandin G/H syntase-2 expression in osteoblastic MC3T3–E1 cells. Endocrinology 138:4672–4682
Kawaguchi H, Pilbeam CC, Gronowicz G, Abreu C, Fletcher BS, Herschman HR, Raisz LG, Hurley MM 1995 Regulation of inducible prostaglandin G/H synthase mRNA levels and promoter activity by basic fibroblast growth factor in osteoblastic cells. J Clin Invest 96:923–930
Pilbeam CC, Kawaguchi H, Hakeda Y, Voznesensky O, Alander CB, Raisz LG 1993 Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic MC3T3–E1 cells. J Biol Chem 268:25643–25649
Tetradis S, Pilbeam CC, Liu Y, Herschman HR, Kream BE 1997 Parathyroid hormone increases prostaglandin G/H synthase-2 transcription by a cyclic adenosine 3',5'-monophosphate-mediated pathway in murine osteoblastic MC3T3–E1 cells. Endocrinology 138:3594–3600
Klein-Nulend J, Burger EH, Semeins CM, Raisz LG, Pilbeam CC 1997 Pulsating fluid flow stimulates prostaglandin release and inducible prostaglandin G/H synthase mRNA expression in primary mouse bone cells. J Bone Miner Res 12:45–51
Pavalko FM, Chen NX, Turner CH, Burr DB, Atkinson S, Hsieh YF, Qiu J, Duncan RL 1998 Fluid shear-induced mechanical signaling in MC3T3–E1 osteoblasts requires cytoskeleton-integrin interactions. Am J Physiol 275:C1591–C1601
Wadhwa S, Godwin SL, Peterson DR, Epstein MA, Raisz LG, Pilbeam CC 2002 Fluid flow induction of cyclo-oxygenase 2 gene expression in osteoblasts is dependent on an extracellular signal-regulated kinase signaling pathway. J Bone Miner Res 17:266–274
Pilbeam CC, Raisz LG, Voznesensky O, Alander CB, Delman BN, Kawaguchi K 1994 Autoregulation of inducible prostaglandin G/H synthase in osteoblastic cells by prostaglandins. J Bone Miner Res 10:406–414
Alam I, Warden SJ, Robling AG, Turner CH 2003 Knockout of cyclooxygenase-2 (COX-2) in mice results in a skeletal phenotype but does not influence mechanical responsiveness. J Bone Miner Res 18:S45
Chen Q, Rho JY, Fan Z, Laulederkind SJ, Raghow R 2003 Congenital lack of COX-2 affects mechanical and geometric properties of bone in mice. Calcif Tissue Int 73:387–392
Dinchuk JE, Car BD, Focht RJ, Johnston JJ, Jaffee BD, Covington MB, Contel NR, Eng VM, Collins RJ, Czerniak PM 1995 Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature 378:406–409
Morham SG, Langenbach R, Loftin CD, Tiano HF, Vouloumanos N, Jennette JC, Mahler JF, Kluckman KD, Ledford A, Lee CA 1995 Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell 83:473–482
Norwood VF, Morham SG, Smithies O 2000 Postnatal development and progression of renal dysplasia in cyclooxygenase-2 null mice. Kidney Int 58:2291–2300
Lim H, Paria BC, Das SK, Dinchuk JE, Langenbach R, Trzaskos JM, Dey SK 1997 Multiple female reproductive failures in cyclooxygenase 2-deficient mice. Cell 91:197–208
Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR 1987 Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 2:595–610
Folders J, Shih MS, Parfitt AM 1990 Frequency distributions of tetracycline-based measurements: Implications for the interpretation of bone formation indices in the absence of double-labeled surfaces. J Bone Miner Res 5:1063–1067
Loftin CD, Trivedi DB, Tiano HF, Clark JA, Lee CA, Epstein JA, Morham SG, Breyer MD, Nguyen M, Hawkins BM, Goulet JL, Smithies O, Koller BH, Langenbach R 2001 Failure of ductus arteriosus closure and remodeling in neonatal mice deficient in cyclooxygenase-1 and cyclooxygenase-2. Proc Natl Acad Sci USA 98:1059–1064
Murakami M, Kuwata H, Amakasu Y, Shimbara S, Nakatani Y, Atsumi G, Kudo I 1997 Prostaglandin E2 amplifies cytosolic phospholipase A2- and cyclooxygenase-2-dependent delayed prostaglandin E2 generation in mouse osteoblastic cells. Enhancement by secretory phospholipase A2. J Biol Chem 272:19891–19897
Silver J, Kilav R, Sela-Brown A, Naveh-Many T 2000 Molecular mechanisms of secondary hyperparathyroidism. Pediatr Nephrol 14:626–628
Slatopolsky E, Brown A, Dusso A 2001 Role of phosphorus in the pathogenesis of secondary hyperparathyroidism. Am J Kidney Dis 37:S54–S57
Kifor O, Kifor I, Brown EM 2002 Signal transduction in the parathyroid. Curr Opin Nephrol Hypertens 11:397–402
Rodriguez M, Canalejo A, Garfia B, Aguilera E, Almaden Y 2002 Pathogenesis of refractory secondary hyperparathyroidism. Kidney Int 80(Suppl):155–160
Gerber B, Hassig M, Reusch CE 2003 Serum concentrations of 1,25-dihydroxycholecalciferol and 25-hydroxycholecalciferol in clinically normal dogs and dogs with acute and chronic renal failure. Am J Vet Res 64:1161–1166
Ishii H, Wada M, Furuya Y, Nagano N, Nemeth EF, Fox J 2000 Daily intermittent decreases in serum levels of parathyroid hormone have an anabolic-like action on the bones of uremic rats with low-turnover bone and osteomalacia. Bone 26:175–182
Chin J, Miller SC, Wada M, Nagano N, Nemeth EF, Fox J 2000 Activation of the calcium receptor by a calcimimetic compound halts the progression of secondary hyperparathyroidism in uremic rats. J Am Soc Nephrol 11:903–911
Denda M, Finch J, Brown AJ, Nishii Y, Kubodera N, Slatopolsky E 1996 1,25-Dihydroxyvitamin D3 and 22-oxacalcitriol prevent the decrease in vitamin D receptor content in the parathyroid glands of uremic rats. Kidney Int: 50:34–39
Slatopolsky E, Cozzolino M, Lu Y, Finch J, Dusso A, Staniforth M, Wein Y, Webster J 2003 Efficacy of 19-Nor-1,25-(OH)2D2 in the prevention and treatment of hyperparathyroid bone disease in experimental uremia. Kidney Int 63:2020–2027
Murayama A, Takeyama K-I, Kitanaka S, Kodera Y, Kawaguchi Y, Hosoya T, Kato S 1999 Positive and negative regulations of the renal 25-hydroxyvitamin D3 1-hydroxylase gene by parathyroid hormone, calcitonin, and 1,25(OH)2D3 in intact animals. Endocrinology 140:2224–2231
Moser M, Dahmen S, Kluge R, Grone H, Dahmen J, Kunz D, Schorle H, Buettner R 2003 Terminal renal failure in mice lacking transcription factor AP-2?. Lab Invest 83:571–578
Maruani G, Hertig A, Paillard M, Houillier P 2003 Normocalcemic primary hyperparathyroidism: evidence for a generalized target-tissue resistance to parathyroid hormone. J Clin Endocrinol Metab 88:4641–4648
Brown EM 2002 The pathophysiology of primary hyperparathyroidism. J Bone Miner Res 17(Suppl 2):N24–N29
Christakos S 2002 Vitamin D gene regulation. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. New York: Academic Press; 573–586
Wark JD, Taft JL, Michelangeli VP, Veroni MC, Larkins RG 1984 Biphasic action of prostaglandin E2 on conversion of 25-hydroxyvitamin D3 to 1,25-dihydroxyvitamin D3 in chick renal tubules. Prostaglandins 27:453–463
Yamada M, Matsumoto T, Takahashi N, Suda T, Ogata E 1983 Stimulatory effect of prostaglandin E2 on 1,25-dihydroxyvitamin D3 synthesis in rats. Biochem J 216:237–240
Hasanoglu A, Ercan ZS, Buyan N, Memioglu N, Hasanoglu E 1997 Parathormone, 1,25 dihydroxyvitamin D and prostaglandin E2 correlation in children with idiopathic hypercalciuria. Prostaglandins Leukotrienes Essent Fatty Acids 56:235–237
Smith SJ, Green LM, Hayes ME, Mawer EB 1999 Prostaglandin E2 regulates vitamin D receptor expression, vitamin D-24-hydroxylase activity and cell proliferation in an adherent human myeloid leukemia cell line (Ad-HL60). Prostaglandins Other Lipid Mediat 57:73–85
Imanishi Y, Hosokawa Y, Yoshimoto K, Schipani E, Mallya S, Papanikolaou A, Kifor O, Tokura T, Sablosky M, Ledgard F, Gronowicz G, Wang TC, Schmidt EV, Hall C, Brown EM, Bronson R, Arnold A 2001 Primary hyperparathyroidism caused by parathyroid-targeted overexpression of cyclin D1 in transgenic mice. J Clin Invest 107:1093–1102
Gonzalez EA, Lund RJ, Martin KJ, McCartney JE, Tondravi MM, Sampath TK, Hruska KA 2002 Treatment of a murine model of high-turnover renal osteodystrophy by exogenous BMP-7. Kidney Int 61:1322–1331
Dardenne O, Prud’homme J, Hacking SA, Glorieux FH, St Arnaud R 2003 Correction of the abnormal mineral ion homeostasis with a high-calcium, high-phosphorus, high-lactose diet rescues the PDDR phenotype of mice deficient for the 25-hydroxyvitamin D-1-hydroxylase (CYP27B1). Bone 32:332–340
Healy KD, Zella JB, Prahl JM, DeLuca HF 2003 Regulation of the murine renal vitamin D receptor by 1,25-dihydroxyvitamin D3 and calcium. Proc Natl Acad Sci USA 100:9733–9737
Miao D, He B, Lanske B, Bai XY, Tong XK, Hendy GN, Goltzman D, Karaplis AC 2004 Skeletal abnormalities in Pth-null mice are influenced by dietary calcium. Endocrinology 145:2046–2053
Yamashita H, Noguchi S, Uchino S, Watanabe S, Murakami T, Ogawa T, Masatsugu T, Takamatsu Y, Miyatake E, Yamashita H 2003 Influence of renal function on clinico-pathological features of primary hyperparathyroidism. Eur J Endocrinol 148:597–602
Brown EM, Swartz SL 1985 Production of prostaglandins by dispersed cells and fragments from bovine parathyroid glands. Prostaglandins 29:35–46
Gardner DG, Brown EM, Attie MF, Aurbach GD 1981 Effects of prostaglandins on adenosine 3',5'-monophosphate content and adenylate cyclase activity in dispersed bovine parathyroid cells. Endocrinology 109:1545–1551
Mangat H, Peterson LN, Burns KD 1997 Hypercalcemia stimulates expression of intrarenal phospholipase A2 and prostaglandin H synthase-2 in rats. Role of angiotensin II AT1 receptors. J Clin Invest 100:1941–1950
Jesus Ferreira MC, Helies-Toussaint C, Imbert-Teboul M, Bailly C, Verbavatz JM, Bellanger AC, Chabardes D 1998 Co-expression of a Ca2+-inhibitable adenylyl cyclase and of a Ca2+-sensing receptor in the cortical thick ascending limb cell of the rat kidney. Inhibition of hormone-dependent cAMP accumulation by extracellular Ca2+. J Biol Chem 273:15192–15202
Jee WS, Ma YF 1997 The in vivo anabolic actions of prostaglandins in bone. Bone 21:297–304
Suponitzky I, Weinreb M 1998 Differential effects of systemic prostaglandin E2 on bone mass in rat long bones and calvariae. J Endocrinol 156:51–57
Flanagan AM, Chambers TJ 1992 Stimulation of bone nodule formation in vitro by prostaglandins E1 and E2. Endocrinology 130:443–448
Weinreb M, Suponitzky I, Keila S 1997 Systemic administration of an anabolic dose of PGE2 in young rats increases the osteogenic capacity of bone marrow. Bone 20:521–526
Okada Y, Tomita M, Gronowicz G, Kawaguchi H, Sohn J, Tanaka Y, Morimoto I, Nakamura T, Raisz LG, Pilbeam CC 2000 Effects of cyclooxygenase-2 gene disruption on osteoblastic function. J Bone Miner Res 15:S217
Zhang X, Schwarz EM, Young DA, Puzas JE, Rosier RN, O’Keefe RJ 2002 Cyclooxygenase-2 regulates mesenchymal cell differentiation into the osteoblast lineage and is critically involved in bone repair. J Clin Invest 109:1405–1415
Carbone LD, Tylavsky FA, Cauley JA, Harris TB, Lang TF, Bauer DC, Barrow KD, Kritchevsky SB 2003 Association between bone mineral density and the use of nonsteroidal anti-inflammatory drugs and aspirin: impact of cyclooxygenase selectivity. J Bone Miner Res 18:1795–1802
Lanyon LE 1992 Control of bone architecture by functional load bearing. J Bone Miner Res 7:S369–S375
Einhorn TA 2002 Do inhibitors of cyclooxygenase-2 impair bone healing? J Bone Miner Res 17:977–978
Simon AM, Manigrasso MB, O’Connor JP 2002 Cyclo-oxygenase 2 function is essential for bone fracture healing. J Bone Miner Res 17:963–976
Kirtikara K, Morham SG, Raghow R, Laulederkind SJ, Kanekura T, Goorha S, Ballou LR 1998 Compensatory prostaglandin E2 biosynthesis in cyclooxygenase 1 or 2 null cells. J Exp Med 187:517–523(Manshan Xu, Shilpa Choudh)
Address all correspondence and requests for reprints to: Dr. Carol Pilbeam, Room AM047, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, Connecticut 06030. E-mail: pilbeam@nso.uchc.edu.
Abstract
The absence of cyclooxygenase-2 (COX-2) activity in vitro reduces differentiation of both bone-forming and bone-resorbing cells. To examine the balance of COX-2 effects on bone in vivo, we studied COX-2 knockout (KO) and wild-type (WT) mice. After weaning, KO mice died 4 times faster than WT mice, consistent with reports of progressive renal failure in KO mice. Among KO mice killed at 4 months of age, some had renal failure with marked secondary hyperparathyroidism, but others appeared healthy. On the assumption that renal failure was not inevitable in COX-2 KO mice and that phenotypic differences might increase with age, we studied KO mice surviving to 10 months of age with serum creatinine levels similar to those of WT mice. In 10-month-old male KO mice, serum calcium and PTH, but not phosphorus, levels were increased compared with those in WT mice. 1,25-Dihydroxyvitamin D3 levels were markedly elevated in KO mice. Skeletal analysis showed small nonsignificant decreases in cortical bone density by BMD and either an increase (distal femur, by microcomputed tomography) or no difference (distal femur, by static histomorphometry) in trabecular bone density in KO mice. There was a trend toward increased percent osteoblastic and osteoclastic surfaces, and on dynamic histomorphometry, the rates of trabecular bone formation and mineral apposition were increased in KO mice relative to WT mice. Similar trends were observed for most parameters in 10-month-old female COX-2 KO mice. However, rates of trabecular bone formation and mineral apposition were increased in 10-month-old WT females compared with males and did not increase further in female KO mice. These data suggest that COX-2 KO mice with intact renal function have primary hyperparathyroidism, and that effects of increased PTH and 1,25-dihydroxyvitamin D3 to increase bone turnover may compensate for the absence of COX-2.
Introduction
THERE ARE TWO enzymes for cyclooxygenase (COX) encoded by separate genes, COX-1 and COX-2 (1, 2, 3). In most tissues, COX-1 is constitutively expressed, whereas COX-2 is transiently induced. COX-2 is highly expressed in bone and is regulated by multiple local and systemic factors important in bone metabolism (4). In the absence of COX-2, little prostaglandin E2 (PGE2) is produced by cultured osteoblasts or cultured bone marrow cells (5, 6, 7). Agonists shown to induce COX-2 expression and PGE2 production in osteoblastic cells include cytokines [IL-1 (8, 9, 10), TNF- (11), and IL-6 (12)], growth factors [TGF- (13, 14), TGF-? (14), bone morphogenetic protein-2 (6), and basic fibroblast growth factor (15)], fetal bovine serum (16), systemic hormones [PTH (8, 17) and 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) (5)], extracellular calcium (7), and fluid shear stress or mechanical loading (18, 19, 20). PGs themselves induce COX-2 expression and therefore can amplify PG responses to other agonists (21).
PGs produced by COX-2 can be potent stimulators of both bone resorption and bone formation (4). In addition, the many factors that induce COX-2 expression generally have their own independent effects on bone turnover. It is therefore difficult to predict the effects of COX-2 expression on bone mass in vivo. Several studies using mice with disruption of the COX-2 gene, COX-2 knockout (KO) mice, have suggested that bone mass or density is decreased in COX-2 KO relative to WT mice (22, 23). However, it is unclear whether COX-2 KO mice can be used for studying the direct effects of COX-2 absence on bone turnover. The developers of COX-2 KO mice reported that KO mice have nephropathy severe enough to decrease the average life span to approximately 3–4 months (24, 25). A recent study reported that COX-2 KO mice have progressive renal deterioration, with 20% of mice dying between 7–23 wk of age (26). Loss of functioning renal mass in mice, as in humans, is expected to decrease 1,25(OH)2D3 production, thereby decreasing serum calcium levels, and decreased serum calcium levels are expected to increase PTH secretion. This secondary hyperparathyroidism could increase bone turnover, making it difficult to separate out the direct effects of the absence of COX-2 on bone from those secondary to nephropathy.
In the present study we studied differences in skeletal phenotype due to COX-2 absence, apart from renal dysfunction if possible. Although COX-2 KO mice did die faster than WT mice, renal failure did not appear to be universal. Initially, we killed mice at 4 months of age and found several with renal failure associated with marked secondary hyperparathyroidism, whereas others appeared relatively healthy. On the assumption that most COX-2 KO mice with progressive renal dysfunction would die within the first 6–8 months of life and that phenotypic differences between COX-2 WT and KO mice might be most evident in older mice, we then studied 10-month-old mice. Our results suggest that COX-2 KO mice without functional renal deterioration have primary hyperparathyroidism with increased levels of 1,25(OH)2D3. The skeletal differences between COX-2 WT and KO mice are small, and we speculate that the effects of hyperparathyroidism on bone turnover compensate for the absence of COX-2 in bone.
Materials and Methods
Animals
The COX-2 KO mice used in this study were developed at University of North Carolina on a C57BL/6, 129SV background (25). Because COX-2 KO females are infertile (27) and because we wanted to obtain both WT and KO in the same litter, we bred COX-2 WT and KO mice for experiments by mating mice heterozygous for the disrupted COX-2 allele (COX-2 HET mice). To establish founders for the breeding colonies, we backcrossed the original C57BL/6,129SV HET mice onto a pure C57BL/6 background (backcrossed more than nine generations). Colonies were maintained by mating the C57BL/6 HET founders with C57BL/6 WT mice from The Jackson Laboratory (Bar Harbor, ME), which were refreshed every 6 months to avoid genetic drift. We tried to generate COX-2 KO mice by crossing COX-2 HET mice on a C57BL/6 background, but found few or no KO mice in the litters. A similar observation was made by Dr. Scott Morham (personal communication), who developed the COX-2 KO mice. Hence, the C57BL/6 COX-2 HET mice were outbred to 129P3/J WT mice to maintain the production of KO mice. The final breeding protocol for COX-2 KO and WT mice used in this study is shown in Fig. 1. Mice were genotyped as described previously (5). All animal protocols were approved by the animal care and use committee of University of Connecticut Health Center (Farmington, CT).
FIG. 1. Breeding scheme for generating COX-2 WT and KO mice from COX-2 HET mice.
Analytical procedures on serum
Mice were euthanized at 4 months (±2 wk) or 10 months (±3 wk) of age. Mice were killed in groups containing eight or more mice of both genotypes upon reaching the appropriate age. Blood was obtained by heart puncture at the time of death after anesthesia with gaseous carbon dioxide. Mice were not fasted. Blood was allowed to clot at room temperature, and the serum was removed after centrifugation of samples at 3000 rpm for 20 min. Serum from each animal was divided into aliquots and frozen at –20 C. There was no repeated freezing and thawing of specimens before measurements.
Serum creatinine, phosphorus, and calcium measurements were measured using kits (procedures 555, 360-UV, and 587, respectively) from Sigma-Aldrich Corp. (St. Louis, MO) following the manufacturer’s directions. Five microliters of serum were needed to assay calcium or phosphorus, and 30 μl serum were needed to assay creatinine. PTH was assayed on 25 μl serum using a Mouse Intact PTH ELISA Kit from Immutopics (San Clemente, CA). This is a two-site assay, recognizing epitopes within the midregion/C-terminal portion and epitopes within the N-terminal portion of the PTH peptide. The manufacturer reports an intraassay coefficient of variation (CV) of 2.5–3.9% and an interassay CV of 7.8–8.9%. 1,25(OH)2D3 was assayed on 100 μl serum by immunoextraction followed by 125I RIA using a kit from Immunodiagnostics Systems (IDS Ltd., Tyne and Wear, UK). The manufacturer reports an intraassay CV of 5–8% and an interassay CV of 9–10%.
Skeletal imaging
Femurs and lumbar vertebrae were dissected free of connective tissue at the time of death and stored in 70% ethanol. Areal bone mineral density (BMD) and bone mineral content (BMC) of the femur and of the first through fifth lumbar vertebrae (L1–L5) were measured using a PIXImus2 densitometer (LUNAR, GE Medical Systems, Madison, WI). Trabecular morphometry within the metaphyseal region of the distal femur and the centrum of the third lumbar vertebrae (L3) was quantified using x-ray microcomputed tomography (μCT20, Scanco Medical AG, Bassersdorf, Switzerland). Volumetric analysis regions for measuring trabecular bone volume fraction (BV/TV) were selected within the endosteal borders to include the central 80% of vertebral height and the secondary spongiosa of femoral metaphyses located 760 μm (5% of length) distal to the physis.
Bone histomorphometry
For static histomorphometry, femurs were fixed in 4% paraformaldehyde at 4 C, decalcified in 15% EDTA, dehydrated in progressive concentrations of ethanol, cleared in xylene, and embedded in paraffin. Five-micrometer sections were cut, stained for tartrate-resistant acid phosphatase to visualize osteoclasts, and counterstained with hematoxylin. Measurements included cortical width, percent BV/TV, percent osteoblast surface, percent osteoclast surface, and trabecular spacing and number.
For dynamic histomorphometry, mice received calcein (Sigma-Aldrich Corp.; 10 mg/kg body weight) 7 d before death and xylenol orange (Sigma-Aldrich Corp.; 90 mg/kg body weight) 2 d before death, administered ip. Femurs were dissected free of tissue and fixed in 70% ethanol at the time of death, then dehydrated in increasing concentrations of ethanol, cleared in xylene, and embedded undecalcified in methyl methacrylate. Five-micrometer-thick longitudinal serial sections were cut on a Polycut S microtome (Reichert-Jung, Nussloch, Germany) with a D profile knife (Delaware Diamond Knives Corp., Wilmington, DE).
Both static and dynamic measurements were made in a blinded, nonbiased manner using the OsteoMeasure computerized image analysis system (OsteoMetrics, Inc., Atlanta, GA) interfaced with an Optiphot Nikon microscope (Nikon, Inc., Melville, NY). The terminology and units used are those recommended by the histomorphometry nomenclature committee of the American Society for Bone and Mineral Research (28). All measurements were confined to the growth plate-metaphyseal junction. Osteoblasts were identified as cuboidal cells lining the trabecular perimeter. Osteoclasts were identified as multinucleated cells on the trabecular bone surface. In static measurements, cortical width was measured in the diaphysis 4000 μm from the growth plate. Dynamic measurements included single and double-labeled surfaces, mineral apposition rate (MAR), and bone formation rate (BFR/BS). In addition, a mineral apposition rate of 0.3 μm/d was assigned to samples that had single labels, but no double labels (29).
Renal histology
Mouse kidneys were halved; fixed in 10% neutral-buffered formalin; processed through 70%, 95%, and absolute ethanol; embedded in a paraffin block; and sectioned at 5-μm intervals. The paraffin sections were deparaffinized in toluene and rehydrated through graded alcohols to distilled water, then stained with hematoxylin and eosin.
Statistics
Analysis of data was performed using SigmaStat for Windows, version 2.03 (San Rafael, CA). When the data were distributed normally, means of groups were compared by ANOVA, with post hoc testing of pairwise multiple comparisons by Bonferroni’s method. When the data failed to meet the normality test, Kruskal-Wallis ANOVA on ranks was performed, followed by Dunn’s test for all pairwise multiple comparisons.
Results
Mortality of COX-2 KO mice
Previous studies had suggested that renal abnormalities in COX-2 KO mice were present shortly after birth and resulted in death by 3–4 months of age (24, 25, 26). We wanted to study the COX-2 KO phenotype in mice without compromised renal function so that the effects of COX-2 disruption on skeletal phenotype could be studied without being complicated by the secondary hyperparathyroidism associated with renal failure. We hypothesized that COX-2 KO mice surviving to 4 months of age might be free of renal abnormalities. When we found some 4-month-old KO mice with renal failure, we expanded the study to include mice killed at 9–10 months of age. Cohorts of COX-2 WT and KO mice were identified at 2 months of age and set aside to be killed at 4 or 10 months of age.
The estimated mortality in these cohorts from 2 months of age is shown in Table 1. Deaths were generally reported by the animal care facility, and animals were removed without additional investigation. The total numbers of mice do not reflect the distribution of COX-2 genotypes born because some mice died or were removed before 2 months of age. From these data, it is estimated that 0.5–1.5% of WT or HET mice died per month and 4–6% of KO mice died per month from 2–10 months of age. Hence, the death rate among KO mice was about 4 times greater than the rate among WT or HET mice. Contrary to our initial hypothesis, there was not a decrease in the death rate among KO mice after 4 months of age.
TABLE 1. Deaths in mice with both COX-2 alleles intact (WT), one COX-2 allele disrupted (HET), or both COX-2 alleles disrupted (KO)
Serum creatinine, calcium, phosphorus, and PTH levels in COX-2 WT and KO mice
Body weights and serum data for the 4-month-old mice are shown in Table 2. Two of the KO mice killed at 4 months of age had serum creatinine levels greater than 1.5 mg/dl. Data for these mice were excluded from the summary in Table 2 and are shown in Table 3. These mice looked sick at time of death. They had low body weight, low serum calcium, high serum phosphorus, and extremely high PTH levels. Thus, these mice appeared to have renal failure with secondary hyperparathyroidism.
TABLE 2. Comparison of body weights and serum parameters in COX-2 WT and KO mice killed at 4 months of age
TABLE 3. COX-2 KO male mice, killed at 4 months of age, with apparent renal failure
After exclusion of these two mice, the mean serum creatinine level was still 30% higher in COX-2 KO males compared with WT males (Table 2). Fewer females were examined, and although not statistically significant, a similar trend toward higher serum creatinine was seen in female KO mice. PTH levels in KO mice also tended to be higher than those in WT mice, and this difference was statistically significant in female mice. It was difficult to explain the trend toward higher PTH levels in KO male mice as being due to renal failure in view of the low phosphorus level in KO males, but the role of renal dysfunction in the elevated PTH in KO females was unclear.
We examined gross renal histology from some healthy 4-month-old WT and KO mice. Although some KO kidneys had lymphocytic infiltrates not seen in WT kidneys, we found no differences in glomerular or tubular structures between WT and KO kidneys (data not shown). It seemed, therefore, that renal disease affecting COX-2 KO mice was not universal. Hence, we decided to examine older mice, with the hope that most or all KO mice surviving to 10 months of age might have intact renal function, and other metabolic abnormalities might become more evident.
Body weights and serum parameters in 10-month-old COX-2 WT and KO mice are shown in Table 4. Weights tended to be lower in KO mice compared with WT mice, but this difference was statistically significant only for females (8%). There was a nonstatistically significant trend for creatinine levels to be higher (10%) in KO mice compared with WT mice. Weights and creatinine levels were normally distributed around the medians, as indicated by the box plots shown in Fig. 2. Box plots for serum calcium, phosphorus, and PTH data are shown in Fig. 3. There was a small, but significant, increase (8%) in mean serum calcium in COX-2 KO male mice compared with WT male mice, but variability was also greater in KO mice (Fig. 3A). WT females had higher calcium and phosphorus levels than WT males, but there was no difference between WT and KO females. The mean PTH level was 50% higher in 10-month-old COX-2 KO males compared with WT males (Table 4). There was also a trend for female KO mice to have higher PTH levels than WT mice. Unlike the other serum parameters, PTH values for males, females, or both combined, as shown in Fig. 3C, were not normally distributed.
TABLE 4. Comparison of body weights and serum parameters measured in COX-2 WT and KO mice killed at 10 months of age
FIG. 2. Body weights and serum creatinine levels measured in 10 COX-2 WT and KO mice killed at 10 months of age. Boxes are bounded by the 25th and 75th percentiles, bars indicate the 10th and 90th percentiles, and the population median is indicated by the line within the box.
FIG. 3. Serum parameters measured in COX-2 WT and KO mice killed at 10 months of age. Boxes are bounded by the 25th and 75th percentiles, bars indicate the 10th and 90th percentiles, and the population median is indicated by the line within the box.
Only one mouse was excluded from the summary of mice killed at 10 months of age. This was a female mouse with a serum PTH level of 624 ng/ml, which was an outlier (>2 SD from the mean). Although this mouse had low body weight (19.6 g), the highest serum creatinine in the female KO group (0.81 mg/dl), and high phosphorus (12.4 mg/dl), consistent with renal dysfunction, its serum calcium (12.0 mg/dl) was high. Inclusion of this mouse in the summary data would have made the PTH difference between WT and KO females statistically significant. Regression of serum calcium (dependent variable) for 10-month-old male and female KO mice against PTH (independent variable) showed a small, significant, positive correlation (r = 0.35; P < 0.05; n = 43; data not shown). No other statistical significant correlations between pairs of parameters or between weights and parameters were found.
Serum 1,25(OH)2D3 levels in COX-2 WT and KO mice
Overall, the serum data from the 10-month-old mice suggested primary, rather than secondary, hyperparathyroidism in COX-2 KO mice. In an effort to clarify this, we tested whether the KO mice had impaired 1,25(OH)2D3 synthesis, which is expected in chronic renal dysfunction, or increased 1,25(OH)2D3 levels, which is expected in primary hyperparathyroidism. We measured 1,25(OH)2D3 levels for all mice for which adequate serum remained after the above assays were completed (Table 5). In 4-month-old female mice, there was no difference between WT and KO mice, but there were few KO data, and variability was high. In 4-month-old males and 10-month-old males and females, KO mice had 1,25(OH)2D3 levels 1.8- to 2.3-fold greater than the WT mice.
TABLE 5. Comparison of serum 1,25(OH)2D3 (picograms per deciliter) values measured in COX-2 WT and KO mice
Because we had no serum for measurement of 1,25(OH)2D3 levels from the few mice identified with definite renal dysfunction, we followed a small group of female mice as they continued to age, and when a KO mouse appeared to become sick at 12 months of age, we killed it. At the same time, we killed seven female WT mice ranging in age from 13–16 months. Serum from these mice was processed simultaneously, and results are shown in Table 6. The sick KO mouse had a serum creatinine level of 1.04, with lower body weight, lower serum calcium, and higher serum phosphorus than the WT mice. Although the sick KO mouse had an elevated PTH level, its 1,25(OH)2D3 level was very low, as expected for renal failure.
TABLE 6. Comparison of a group (n = 7) of COX-2 WT female mice, killed at 13–16 months of age, with a sick COX-2 KO female mouse, killed at 12 months of age
The kidneys from this sick KO mouse were smaller than those from WT mice (Fig. 4). As we had noticed frequently in kidneys from 4- and 10-month-old KO mice that had no structural abnormalities (data not shown), the KO mouse had lymphocytic infiltrates especially evident around blood vessels (Fig. 4). There was also evidence for enlarged renal tubules and dilated calices in the sick KO mouse.
FIG. 4. Sections of kidneys, stained with hematoxylin and eosin, from a 13-month-old female COX-2 WT mouse (A) and a 12-month-old female COX-2 KO mouse (B) with serum parameters consistent with renal failure and secondary hyperparathyroidism. Higher magnification shows enlarged tubules and lymphocytic infiltration in the KO kidney (D) compared with the WT kidney (C).
Skeletal analysis of COX-2 WT and KO mice: imaging
BMD was measured on dissected left femurs and lumbar vertebrae (L1–L5) from 4- and 10-month-old mice (Table 7). BMD tended to be lower in COX-2 KO mice compared with WT mice in all groups. The difference was statistically significant for 4-month-old female lumbar vertebral BMD (12%) and 10-month-old female femur BMD (9%). In some groups, vertebral (4-month-old female KO mice) and femoral (10-month-old WT females) BMDs were higher (6%) in females than in males. Lumbar BMD, but not femur BMD, decreased about 9% with age in WT mice, suggesting that trabecular, but not cortical, bone mass decreased with age.
TABLE 7. Comparison of BMD (milligrams per square centimeter), measured by Dexa, in COX-2 WT and KO mice
The 10-month-old female COX-2 KO mouse, excluded from the summary analysis because of possible renal dysfunction, as discussed above, had low femur BMD (58.4 mg/cm2) compared with the remaining mice in this group (65.3 mg/cm2). Femur BMD was also measured in one of the 4-month-old KO mice excluded for renal failure, and this value (59.9 mg/cm2) was low compared with that in the remaining group (67.6 mg/cm2). These observations suggest that renal abnormalities associated with effects of COX-2 absence can decrease BMD.
μCT analyses of the BV/TV in the distal femur and in the L3 vertebral body are shown in Table 8. Unlike BMD, which measures largely cortical bone, BV/TV did not tend to decrease in KO mice compared with WT mice. In fact, there was a significant increase in BV/TV in the distal femur (26%) in 10-month-old male KO mice compared with WT male mice. Also, in contrast to the BMD comparisons, 10-month-old females had 40–50% lower BV/TV in the distal femur than males. There was also a significant decrease in femur BV/TV at 10 months of age compared with 4 months of age for both males (56–68%) and females (44–48%).
TABLE 8. Comparison of BV/TV (percentage), measured by μCT, in COX-2 WT and KO mice
Skeletal analysis of COX-2 WT and KO mice: histomorphometry
Static histomorphometric analyses on the distal femur showed decreased cortical width in 4-month-old male KO mice compared with WT (9%), but not in 10-month-old males (Table 9). There was also a significant decrease in cortical width in 10-month-old KO females compared with WT females (9%). No effect of genotype on percent BV/TV was seen. Female mice tended to have lower BV/TV than males in all groups, with a significant difference (35%) between WT females and males at 10 months of age. There was a trend toward higher osteoblast and osteoclast surfaces in KO mice compared with WT mice. The difference (40–60%) was significant for percent osteoclast surface in females.
TABLE 9. Comparison of static histomorphometric parameters measured in the distal femur in COX-2 WT and KO mice
Dynamic histomorphometry of the distal femur is shown in Table 10. There was a significant increase of 55% in MAR and 70% in BFR/BS in 10-month-old KO male mice compared with WT mice, but not in female mice. Interestingly, both MAR and BFR/BS were 1.5- to 2.6-fold higher in females than in males at 10 months of age. This difference appeared to be largely due to the lack of age-related decreases in MAR and BFR/BS in females. There were no differences between genotypes in the other parameters measured (data not shown).
TABLE 10. Comparison of dynamic histomorphometric parameters measured in the distal femur in COX-2 WT and KO mice
Discussion
We found no increase in mortality among COX-2 HET mice, but there was high mortality among COX-2 KO mice, as reported in other studies (24, 25, 26, 30). One study reported that 35% of COX-2 KO mice died with a patent ductus arteriosus within 48 h of birth (30). Other studies have not found a high incidence of neonatal death among COX-2 KO mice, but have found shortened life span secondary to renal dysplasia (25, 26). Renal development in COX-2 KO mice appeared normal until postnatal d 10, after which there was progressive renal architectural disruption and functional deterioration, and about 20% of KO mice died between 7 and 23 wk of age from renal failure (26). We found that approximately 40% of COX-2 KO mice died between 2 and 10 months of age. All of these studies used mice in a mixed C57BL/6 and 129 genetic background, and it is unclear what impact various degrees of inbreeding may have had on the results. For example, C57BL/6 mice lack type IIA secreted phospholipase A2; as a result, PG production associated with induction of COX-2 in these mice depends largely on cytosolic phospholipase A2 (31). In COX-2 KO mice backcrossed into an outbred CD-1 background, our preliminary observations suggest that there is no increased mortality (data not shown). Hence, the severity of problems associated with the absence of COX-2 could be increased in a pure C57BL/6 background, and this might account for our observation that backcrossing cox-2–/– onto a C57BL/6 genetic background resulted in decreased numbers of COX-2 KO mice in litters after weaning.
As expected, some COX-2 KO mice had evidence of renal tubular impairment at the time of death. These mice had decreased serum calcium, increased serum phosphorus, and markedly elevated PTH, similar to the changes seen in secondary hyperparathyroidism in humans (32, 33, 34, 35), dogs (36), rats (37, 38, 39, 40, 41), and mice (42). The mechanisms proposed to account for secondary hyperparathyroidism in renal failure include 1) phosphorus retention, which inhibits the enzyme 25-hydroxyvitamin D-1-hydroxylase (1-hydoxylase) that converts 25(OH)2D3 to 1,25(OH)2D3, 2) reduction of 1,25(OH)2D3 as a result of phosphorus retention and decreased renal mass, 3) increased parathyroid hyperplasia due to phosphorus retention and reduced 1,25(OH)2D3, 4) resistance to the actions of 1,25(OH)2D3 and PTH as a result of reduced numbers of vitamin D receptors and calcium sensors, and 5) subsequent decrease in serum calcium resulting from all of the above (32, 33, 34, 35). Thus, decreased 1,25(OH)2D3 should be an important factor in the development of secondary hyperparathyroidism. Although we could find no studies of 1,25(OH)2D3 in renal failure in mice, there are several studies in rats showing the same relationships as those in humans (37, 38, 40, 41). We were able to measure serum 1,25(OH)2D3 in one of the COX-2 KO mice with evident renal failure and found it to be very low compared with WT mice and KO mice without renal failure.
In both male and female COX-2 KO mice with relatively intact renal function, there was still a trend for mean PTH levels to be higher than levels in WT mice at all ages. This difference was statistically significant for 4-month-old KO female mice and 10-month-old male KO mice. In addition, in no group was the serum calcium level in KO mice lower than that in WT mice. Hence, the higher PTH levels in KO mice were not driven by decreased serum calcium, suggesting that the KO mice had primary hyperparathyroidism. Although serum calcium was only significantly elevated in 10-month-old male KO mice compared with WT mice, serum calcium could be said to be inappropriately high, perhaps corresponding to normocalcemic primary hyperparathyroidism in humans (43).
As another indication of whether the increased PTH levels were due to renal insufficiency or to primary hyperparathyroidism, we measured 1,25(OH)2D3 levels. PTH stimulates the conversion of 25(OH)D3 to 1,25(OH)2D3 by 1-hydroxylase, and in primary hyperparathyroidism, elevated PTH levels are associated with increased serum calcium, decreased or normal serum phosphorus, and increased 1,25(OH)2D3 levels (44, 45). 1,25(OH)2D3 levels were increased 1.8-fold in male KO mice at 4 months of age and in both male and female KO mice at 10 months of age. We had too few measurements from 4-month-old females for an adequate assessment of 1,25(OH)2D3. Decreased serum phosphorus can also stimulate conversion of 25(OH)D3 to 1,25(OH)2D3 by 1-hydroxylase. Although serum phosphorus was low in 4-month-old male COX-2 KO mice compared with WT mice, it was not low in 10-month-old KO male or female mice and thus cannot account for the elevated 1,25(OH)2D3 level in these mice. The absence of COX-2 could have a direct effect on 1-hydroxylase activity. One study has shown a biphasic effect of PGE2 on 1,25(OH)2D3 production (46), but several others have shown PGE2 to increase 1-hydroxylase activity (47, 48). PGE2 has also been shown to increase both 1-hydroxylase activity and vitamin D receptors in a human myeloid leukemia cell line (49). If the absence of PGs produced by COX-2 decreased vitamin D receptors, there might be a compensatory increase in 1,25(OH)2D3 production, but this would not explain the trend toward higher PTH levels in KO mice.
Many studies in transgenic mice have now shown that the PTH-vitamin D-calcium regulation is similar in mice and humans (42, 50, 51, 52, 53, 54). The simplest explanation consistent with the data is that COX-2 KO mice have primary hyperparathyroidism. Early renal insufficiency in the KO mice might have increased variability in the measurements. A study in humans of the influence of renal function on the features of primary hyperparathyroidism found that PTH and serum calcium were higher and 1,25(OH)2D3 levels were lower in patients with renal insufficiency and primary hyperparathyroidism compared with those in patients with primary hyperparathyroidism alone (55).
PGs have been shown to affect parathyroid function. Early studies of cells and fragments from bovine PTG showed that PTG cells secreted PGs (6-keto-PGF1>PGF2>PGE2) and that PGF2 could inhibit, and PGE2 could stimulate, PTH secretion (56, 57). This modulation of PTH secretion by endogenous PGs might depend on COX-2 expression. Extracellular Ca2+ is a potent inducer of COX-2 in kidney and bone cells (7, 58). In the kidney, PGE2 and extracellular Ca2+ can inhibit the same pool of adenylyl cyclase via different mechanisms, leading to the suggestion that PGE2 increases the sensitivity of the cells to Ca2+ so that effects can be seen at physiological concentrations of extracellular Ca2+ (59). Hence, absence of COX-2-produced PGE2 might decrease the Ca2+ sensitivity of PTG cells, leading to increased PTH secretion.
The absence of COX-2 expression might be expected to affect both bone resorption and formation. PGE2 is abundantly produced by osteoblasts in response to multiple agonists, largely as a result of the induction of COX-2. PGE2 is a potent stimulator of bone resorption and can mediate in part the resorptive activities of other agonists (4). When the specific role of COX-2 was examined in marrow cultures from COX-2 KO mice, the ability of 1,25(OH)2D3 or PTH to stimulate osteoclast formation was reduced 60–70% in KO cultures compared with WT cultures (5). PTH injected above the calvariae caused hypercalcemia in COX-2 WT mice, but not in COX-2 KO mice (5). In contrast, PGE2 can also stimulate bone formation. In vivo studies in rats have demonstrated a potent anabolic effect of exogenous PGE2 (60, 61). Exogenous PGE2 stimulates osteoblastic differentiation in marrow stromal and primary calvarial cell cultures (62), and PGE2 injected in vivo enhanced osteoblastic differentiation in explanted marrow stromal cells (63). Fresh serum also induces COX-2 and endogenous PG production in cultured osteoblastic cells (16), and osteoblastic differentiation in cultured marrow stromal cells and primary calvarial cells from COX-2 KO mice was delayed compared with that in WT cultures (64, 65).
Few studies have examined the balance of effects of endogenous PGs in vivo on bone metabolism or bone mass in the absence of inflammatory conditions. One study of 3-month-old male COX-2 KO mice in a mixed C57BL/6, DBA background reported that COX-2 KO cortical bones were thinner and more porous than WT cortical bones with decreased elastic modulus compared with WT mice (23). Another recent study used COX-2 KO mice reported to have increased mortality as well as significantly decreased body weights, BMC (normalized to weights), and volumetric BMD of long bones compared with WT mice (22). This study concluded that the anabolic response of bone to mechanical loading was not affected by the absence of COX-2. Our data along with those from previous studies (24, 25, 26) suggest that COX-2 KO mice with low body weight and high mortality are likely to have impaired renal function and elevated PTH levels. These high PTH values could have marked effects on bone turnover. A study examining BMD of the whole body in humans taking nonsteroidal antiinflammatory drugs (NSAIDs) and aspirin concluded that the combination of relative COX-2-selective NSAIDs and aspirin, which inhibits COX-1 activity, was associated with higher BMD (66). However, the subjects in this study were taking NSAIDs for a variety of reasons, such as rheumatoid arthritis, and may have had increased presence of inflammatory mediators, which could stimulate both resorption and COX-2 production of PGs.
In the current study, BMD analysis, which measures predominantly cortical bone, suggested that COX-2 KO mice had small decreases (<10%) in cortical bone density compared with WT mice. Cortical bone width measured by histomorphometry also tended to be decreased in COX-2 KO mice. Trabecular bone volume in KO mice was either no different from that in WT mice or increased relative to that in WT mice. Measurements of osteoblastic and osteoclastic surfaces suggested increased turnover in KO mice. Both trabecular bone formation rate and mineral apposition rate were significantly elevated in 10-month-old male KO mice compared with WT mice, but not in female mice. However, formation and mineral apposition rates did not decrease with age in WT females as they did in males and were higher in females than in males at 10 months of age. Perhaps the effects of mild hyperparathyroidism are more apparent in 10-month-old male KO mice because rates of formation had decreased. Both males and females lost trabecular bone volume with age, although females had lower trabecular bone volume than males at both 4 and 10 months of age. Higher bone turnover in WT females than males at 10 months of age might explain the observation that serum calcium and phosphorus decreased in WT males from 4–10 months of age, but not in females. In general, the skeletal analyses, particularly in COX-2 KO males, are consistent with increased bone turnover associated with hyperparathyroidism.
We had the statistical power to detect differences of 8–10% or greater in many of the measurements and cannot rule out smaller differences between WT and KO mice. Variability in measurements could come from several sources. Because we needed to outbreed to obtain KO mice, our experimental mice came from C57BL/6,129P3/JxC57BL/6,129P3/J crosses. The inherent variability in the skeletal phenotype with the variable genetic background of the litters of such crosses may limit the ability to define small differences. It is important to note, however, that published studies on COX-2 KO mice have used mice bred in a similar mixed background (22, 23, 24, 25, 26, 27, 65). We have generated C57BL/6 COX-2 HET mice and 129P3/J COX-2 HET mice by backcrossing into each background, and the offspring of these mice, which will have a more defined genetic background than the mice in our current study, will be studied in the future. Preliminary observations of the COX-2 KO offspring of these mice show the same increased mortality as that seen in this study; thus, these mice probably have similar renal abnormalities (data not shown). Because the severity of the effects of absent COX-2 expression may depend on the relative amount of a particular genetic background, we have also bred the COX-2 HET genotype into the outbred CD-1 background for future studies. Preliminary observations suggest that COX-2 KO mice on this background do not have increased mortality after weaning (data not shown). Another source of variability might be varying degrees of early renal insufficiency in COX-2 KO mice.
In view of the dramatic effects of COX-2 absence on osteoblastic and osteoclastic differentiation in vitro, why are there no marked effects on bone mass in vivo? It could be argued that the effects seen in vitro are the result of factors that induce COX-2 expression transiently, such as addition of fresh serum to cultures or addition of PTH, and that under unstressed conditions in vivo or under conditions of continuous agonist secretion, a transiently inducible gene such as COX-2 would not be expected to be highly expressed. However, mechanical loading of bone occurs regularly, and mechanical loading can induce COX-2 and COX-2-derived PGs, which are thought to be important mediators of the anabolic response to mechanical loading (67). In vitro studies suggest that the absence of COX-2 should reduce both resorption and formation. If the net balance of resorption and formation remained unchanged, there might be only a delay in returning to steady state conditions after perturbation or challenge. Such an effect would fit with the delayed healing seen after fracture in COX-2 KO mice or in mice in which COX-2 activity has been inhibited by NSAIDs (68, 69). However, bone turnover is not decreased in the COX-2 KO mice. Hence, another explanation for the difference between in vitro and in vivo results is that the effects of absent COX-2 on bone turnover are compensated by increased PTH and/or 1,25(OH)2D3, both of which can stimulate bone resorption.
It is also possible that COX-1 plays a compensatory role in the absence of COX-2 (70). Although COX-1 is constitutively expressed, we have seen little PGE2 production in cells cultured from COX-2 KO mice (5, 6, 14), and we have not seen differential mRNA expression of COX-1 in tissues from COX-2 KO mice (data not shown). However, we cannot rule out some compensation in vivo. Because double COX-2 KO, COX-1 KO mice are not expected to survive, secondary to failure of the patent ductus arteriosus to close, we are generating COX-2 KO mice that are heterozygous for COX-1 for future study.
In conclusion, our study suggests that although COX-2 KO mice can have severe renal dysfunction, associated with marked secondary hyperparathyroidism, about 60% of COX-2 KO mice that reach 2 months of age survive to 10 months of age with relatively normal renal function. These surviving mice appear to have primary hyperparathyroidism, with elevated PTH and 1,25(OH)2D3 levels and increased rates of trabecular bone formation and mineral apposition, which may compensate for the absence of COX-2 in preserving bone mass.
Acknowledgments
We thank Dr. Kevin Claffey and Nancy Ryan for performing the renal histology, and Miren Gratton for performing the 1,25(OH)2D3 assays.
References
Herschman HR, Reddy ST, Xie W 1997 Function and regulation of prostaglandin synthase-2. Adv Exp Med Biol 407:61–66
Smith WL, Langenbach R 2001 Why there are two cyclooxygenase isozymes. J Clin Invest 107:1491–1495
Smith WL, DeWitt DL, Garavito RM 2000 Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem 69:145–182
Pilbeam CC, Harrison JR, and Raisz LG 2002 Prostaglandins and bone metabolism. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. New York: Academic Press; 979–994
Okada Y, Lorenzo JA, Freeman AM, Tomita M, Morham SG, Raisz LG, and Pilbeam CC 2000 Prostaglandin G/H synthase-2 is required for maximal formation of osteoclast-like cells in culture. J Clin Invest 105:823–832
Chikazu D, Li X, Kawaguchi H, Sakuma Y, Voznesensky OS, Adams DJ, Xu M, Hoshio K, Katavic V, Herschman HR, Raisz LG, and Pilbeam CC 2002 Bone morphogenetic protein 2 induces cyclo-oxygenase 2 in osteoblasts via a Cbfal binding site: role in effects of bone morphogenetic protein 2 in vitro and in vivo. J Bone Miner Res 17:1430–1440
Choudhary S, Wadhwa S, Raisz LG, Alander C, Pilbeam CC 2003 Extracellular calcium is a potent inducer of cyclo-oxygenase-2 in murine osteoblasts through an ERK signaling pathway. J Bone Miner Res 18:1813–1824
Kawaguchi H, Raisz LG, Voznesensky OS, Alander CB, Hakeda Y, Pilbeam CC 1994 Regulation of the two prostaglandin G/H synthases by parathyroid hormone, interleukin-1, cortisol and prostaglandin E2 in cultured neonatal mouse calvariae. Endocrinology 135:1157–1164
Min YK, Rao Y, Okada Y, Raisz LG, Pilbeam CC 1998 Regulation of prostaglandin G/H synthase-2 expression by interleukin-1 in human osteoblast-like cells. J Bone Miner Res 13:1066–1075
Harrison JR, Kelly PL, Pilbeam CC 2000 Involvement of C/EBP transcription factors in prostaglandin G/H synthase-2 gene induction by interleukin-1 in osteoblastic MC3T3–E1 cells. J Bone Miner Res 15:1138–1146
Kawaguchi H, Nemoto K, Raisz LG, Harrison JR, Voznesensky OS, Alander CB, Pilbeam CC 1996 Interleukin-4 inhibits prostaglandin G/H synthase-2 and cytosolic phospholipase A2 induction in neonatal mouse parietal bone cultures. J Bone Miner Res 11:358–366
Tai H, Miyaura C, Pilbeam CC, Tamura T, Ohsugi Y, Koishihara Y, Kubodera N, Kawaguchi H, Raisz LG, Suda T 1997 Transcriptional induction of cyclooxygenase-2 in osteoblasts is involved in interleukin-6-induced osteoclast formation. Endocrinology 138:2372–2379
Harrison JR, Lorenzo JA, Kawaguchi H, Raisz LG, Pilbeam CC 1994 Stimulation of prostaglandin E2 production by interleukin-1 and transforming growth factor- in osteoblastic MC3T3–E1 cells. J Bone Miner Res 9:817–823
Pilbeam CC, Rao Y, Voznesensky O, Kawaguchi H, Alander C, Raisz LG, Herschman H 1997 Transforming growth factor-?1 regulation of prostaglandin G/H syntase-2 expression in osteoblastic MC3T3–E1 cells. Endocrinology 138:4672–4682
Kawaguchi H, Pilbeam CC, Gronowicz G, Abreu C, Fletcher BS, Herschman HR, Raisz LG, Hurley MM 1995 Regulation of inducible prostaglandin G/H synthase mRNA levels and promoter activity by basic fibroblast growth factor in osteoblastic cells. J Clin Invest 96:923–930
Pilbeam CC, Kawaguchi H, Hakeda Y, Voznesensky O, Alander CB, Raisz LG 1993 Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic MC3T3–E1 cells. J Biol Chem 268:25643–25649
Tetradis S, Pilbeam CC, Liu Y, Herschman HR, Kream BE 1997 Parathyroid hormone increases prostaglandin G/H synthase-2 transcription by a cyclic adenosine 3',5'-monophosphate-mediated pathway in murine osteoblastic MC3T3–E1 cells. Endocrinology 138:3594–3600
Klein-Nulend J, Burger EH, Semeins CM, Raisz LG, Pilbeam CC 1997 Pulsating fluid flow stimulates prostaglandin release and inducible prostaglandin G/H synthase mRNA expression in primary mouse bone cells. J Bone Miner Res 12:45–51
Pavalko FM, Chen NX, Turner CH, Burr DB, Atkinson S, Hsieh YF, Qiu J, Duncan RL 1998 Fluid shear-induced mechanical signaling in MC3T3–E1 osteoblasts requires cytoskeleton-integrin interactions. Am J Physiol 275:C1591–C1601
Wadhwa S, Godwin SL, Peterson DR, Epstein MA, Raisz LG, Pilbeam CC 2002 Fluid flow induction of cyclo-oxygenase 2 gene expression in osteoblasts is dependent on an extracellular signal-regulated kinase signaling pathway. J Bone Miner Res 17:266–274
Pilbeam CC, Raisz LG, Voznesensky O, Alander CB, Delman BN, Kawaguchi K 1994 Autoregulation of inducible prostaglandin G/H synthase in osteoblastic cells by prostaglandins. J Bone Miner Res 10:406–414
Alam I, Warden SJ, Robling AG, Turner CH 2003 Knockout of cyclooxygenase-2 (COX-2) in mice results in a skeletal phenotype but does not influence mechanical responsiveness. J Bone Miner Res 18:S45
Chen Q, Rho JY, Fan Z, Laulederkind SJ, Raghow R 2003 Congenital lack of COX-2 affects mechanical and geometric properties of bone in mice. Calcif Tissue Int 73:387–392
Dinchuk JE, Car BD, Focht RJ, Johnston JJ, Jaffee BD, Covington MB, Contel NR, Eng VM, Collins RJ, Czerniak PM 1995 Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature 378:406–409
Morham SG, Langenbach R, Loftin CD, Tiano HF, Vouloumanos N, Jennette JC, Mahler JF, Kluckman KD, Ledford A, Lee CA 1995 Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell 83:473–482
Norwood VF, Morham SG, Smithies O 2000 Postnatal development and progression of renal dysplasia in cyclooxygenase-2 null mice. Kidney Int 58:2291–2300
Lim H, Paria BC, Das SK, Dinchuk JE, Langenbach R, Trzaskos JM, Dey SK 1997 Multiple female reproductive failures in cyclooxygenase 2-deficient mice. Cell 91:197–208
Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR 1987 Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 2:595–610
Folders J, Shih MS, Parfitt AM 1990 Frequency distributions of tetracycline-based measurements: Implications for the interpretation of bone formation indices in the absence of double-labeled surfaces. J Bone Miner Res 5:1063–1067
Loftin CD, Trivedi DB, Tiano HF, Clark JA, Lee CA, Epstein JA, Morham SG, Breyer MD, Nguyen M, Hawkins BM, Goulet JL, Smithies O, Koller BH, Langenbach R 2001 Failure of ductus arteriosus closure and remodeling in neonatal mice deficient in cyclooxygenase-1 and cyclooxygenase-2. Proc Natl Acad Sci USA 98:1059–1064
Murakami M, Kuwata H, Amakasu Y, Shimbara S, Nakatani Y, Atsumi G, Kudo I 1997 Prostaglandin E2 amplifies cytosolic phospholipase A2- and cyclooxygenase-2-dependent delayed prostaglandin E2 generation in mouse osteoblastic cells. Enhancement by secretory phospholipase A2. J Biol Chem 272:19891–19897
Silver J, Kilav R, Sela-Brown A, Naveh-Many T 2000 Molecular mechanisms of secondary hyperparathyroidism. Pediatr Nephrol 14:626–628
Slatopolsky E, Brown A, Dusso A 2001 Role of phosphorus in the pathogenesis of secondary hyperparathyroidism. Am J Kidney Dis 37:S54–S57
Kifor O, Kifor I, Brown EM 2002 Signal transduction in the parathyroid. Curr Opin Nephrol Hypertens 11:397–402
Rodriguez M, Canalejo A, Garfia B, Aguilera E, Almaden Y 2002 Pathogenesis of refractory secondary hyperparathyroidism. Kidney Int 80(Suppl):155–160
Gerber B, Hassig M, Reusch CE 2003 Serum concentrations of 1,25-dihydroxycholecalciferol and 25-hydroxycholecalciferol in clinically normal dogs and dogs with acute and chronic renal failure. Am J Vet Res 64:1161–1166
Ishii H, Wada M, Furuya Y, Nagano N, Nemeth EF, Fox J 2000 Daily intermittent decreases in serum levels of parathyroid hormone have an anabolic-like action on the bones of uremic rats with low-turnover bone and osteomalacia. Bone 26:175–182
Chin J, Miller SC, Wada M, Nagano N, Nemeth EF, Fox J 2000 Activation of the calcium receptor by a calcimimetic compound halts the progression of secondary hyperparathyroidism in uremic rats. J Am Soc Nephrol 11:903–911
Denda M, Finch J, Brown AJ, Nishii Y, Kubodera N, Slatopolsky E 1996 1,25-Dihydroxyvitamin D3 and 22-oxacalcitriol prevent the decrease in vitamin D receptor content in the parathyroid glands of uremic rats. Kidney Int: 50:34–39
Slatopolsky E, Cozzolino M, Lu Y, Finch J, Dusso A, Staniforth M, Wein Y, Webster J 2003 Efficacy of 19-Nor-1,25-(OH)2D2 in the prevention and treatment of hyperparathyroid bone disease in experimental uremia. Kidney Int 63:2020–2027
Murayama A, Takeyama K-I, Kitanaka S, Kodera Y, Kawaguchi Y, Hosoya T, Kato S 1999 Positive and negative regulations of the renal 25-hydroxyvitamin D3 1-hydroxylase gene by parathyroid hormone, calcitonin, and 1,25(OH)2D3 in intact animals. Endocrinology 140:2224–2231
Moser M, Dahmen S, Kluge R, Grone H, Dahmen J, Kunz D, Schorle H, Buettner R 2003 Terminal renal failure in mice lacking transcription factor AP-2?. Lab Invest 83:571–578
Maruani G, Hertig A, Paillard M, Houillier P 2003 Normocalcemic primary hyperparathyroidism: evidence for a generalized target-tissue resistance to parathyroid hormone. J Clin Endocrinol Metab 88:4641–4648
Brown EM 2002 The pathophysiology of primary hyperparathyroidism. J Bone Miner Res 17(Suppl 2):N24–N29
Christakos S 2002 Vitamin D gene regulation. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. New York: Academic Press; 573–586
Wark JD, Taft JL, Michelangeli VP, Veroni MC, Larkins RG 1984 Biphasic action of prostaglandin E2 on conversion of 25-hydroxyvitamin D3 to 1,25-dihydroxyvitamin D3 in chick renal tubules. Prostaglandins 27:453–463
Yamada M, Matsumoto T, Takahashi N, Suda T, Ogata E 1983 Stimulatory effect of prostaglandin E2 on 1,25-dihydroxyvitamin D3 synthesis in rats. Biochem J 216:237–240
Hasanoglu A, Ercan ZS, Buyan N, Memioglu N, Hasanoglu E 1997 Parathormone, 1,25 dihydroxyvitamin D and prostaglandin E2 correlation in children with idiopathic hypercalciuria. Prostaglandins Leukotrienes Essent Fatty Acids 56:235–237
Smith SJ, Green LM, Hayes ME, Mawer EB 1999 Prostaglandin E2 regulates vitamin D receptor expression, vitamin D-24-hydroxylase activity and cell proliferation in an adherent human myeloid leukemia cell line (Ad-HL60). Prostaglandins Other Lipid Mediat 57:73–85
Imanishi Y, Hosokawa Y, Yoshimoto K, Schipani E, Mallya S, Papanikolaou A, Kifor O, Tokura T, Sablosky M, Ledgard F, Gronowicz G, Wang TC, Schmidt EV, Hall C, Brown EM, Bronson R, Arnold A 2001 Primary hyperparathyroidism caused by parathyroid-targeted overexpression of cyclin D1 in transgenic mice. J Clin Invest 107:1093–1102
Gonzalez EA, Lund RJ, Martin KJ, McCartney JE, Tondravi MM, Sampath TK, Hruska KA 2002 Treatment of a murine model of high-turnover renal osteodystrophy by exogenous BMP-7. Kidney Int 61:1322–1331
Dardenne O, Prud’homme J, Hacking SA, Glorieux FH, St Arnaud R 2003 Correction of the abnormal mineral ion homeostasis with a high-calcium, high-phosphorus, high-lactose diet rescues the PDDR phenotype of mice deficient for the 25-hydroxyvitamin D-1-hydroxylase (CYP27B1). Bone 32:332–340
Healy KD, Zella JB, Prahl JM, DeLuca HF 2003 Regulation of the murine renal vitamin D receptor by 1,25-dihydroxyvitamin D3 and calcium. Proc Natl Acad Sci USA 100:9733–9737
Miao D, He B, Lanske B, Bai XY, Tong XK, Hendy GN, Goltzman D, Karaplis AC 2004 Skeletal abnormalities in Pth-null mice are influenced by dietary calcium. Endocrinology 145:2046–2053
Yamashita H, Noguchi S, Uchino S, Watanabe S, Murakami T, Ogawa T, Masatsugu T, Takamatsu Y, Miyatake E, Yamashita H 2003 Influence of renal function on clinico-pathological features of primary hyperparathyroidism. Eur J Endocrinol 148:597–602
Brown EM, Swartz SL 1985 Production of prostaglandins by dispersed cells and fragments from bovine parathyroid glands. Prostaglandins 29:35–46
Gardner DG, Brown EM, Attie MF, Aurbach GD 1981 Effects of prostaglandins on adenosine 3',5'-monophosphate content and adenylate cyclase activity in dispersed bovine parathyroid cells. Endocrinology 109:1545–1551
Mangat H, Peterson LN, Burns KD 1997 Hypercalcemia stimulates expression of intrarenal phospholipase A2 and prostaglandin H synthase-2 in rats. Role of angiotensin II AT1 receptors. J Clin Invest 100:1941–1950
Jesus Ferreira MC, Helies-Toussaint C, Imbert-Teboul M, Bailly C, Verbavatz JM, Bellanger AC, Chabardes D 1998 Co-expression of a Ca2+-inhibitable adenylyl cyclase and of a Ca2+-sensing receptor in the cortical thick ascending limb cell of the rat kidney. Inhibition of hormone-dependent cAMP accumulation by extracellular Ca2+. J Biol Chem 273:15192–15202
Jee WS, Ma YF 1997 The in vivo anabolic actions of prostaglandins in bone. Bone 21:297–304
Suponitzky I, Weinreb M 1998 Differential effects of systemic prostaglandin E2 on bone mass in rat long bones and calvariae. J Endocrinol 156:51–57
Flanagan AM, Chambers TJ 1992 Stimulation of bone nodule formation in vitro by prostaglandins E1 and E2. Endocrinology 130:443–448
Weinreb M, Suponitzky I, Keila S 1997 Systemic administration of an anabolic dose of PGE2 in young rats increases the osteogenic capacity of bone marrow. Bone 20:521–526
Okada Y, Tomita M, Gronowicz G, Kawaguchi H, Sohn J, Tanaka Y, Morimoto I, Nakamura T, Raisz LG, Pilbeam CC 2000 Effects of cyclooxygenase-2 gene disruption on osteoblastic function. J Bone Miner Res 15:S217
Zhang X, Schwarz EM, Young DA, Puzas JE, Rosier RN, O’Keefe RJ 2002 Cyclooxygenase-2 regulates mesenchymal cell differentiation into the osteoblast lineage and is critically involved in bone repair. J Clin Invest 109:1405–1415
Carbone LD, Tylavsky FA, Cauley JA, Harris TB, Lang TF, Bauer DC, Barrow KD, Kritchevsky SB 2003 Association between bone mineral density and the use of nonsteroidal anti-inflammatory drugs and aspirin: impact of cyclooxygenase selectivity. J Bone Miner Res 18:1795–1802
Lanyon LE 1992 Control of bone architecture by functional load bearing. J Bone Miner Res 7:S369–S375
Einhorn TA 2002 Do inhibitors of cyclooxygenase-2 impair bone healing? J Bone Miner Res 17:977–978
Simon AM, Manigrasso MB, O’Connor JP 2002 Cyclo-oxygenase 2 function is essential for bone fracture healing. J Bone Miner Res 17:963–976
Kirtikara K, Morham SG, Raghow R, Laulederkind SJ, Kanekura T, Goorha S, Ballou LR 1998 Compensatory prostaglandin E2 biosynthesis in cyclooxygenase 1 or 2 null cells. J Exp Med 187:517–523(Manshan Xu, Shilpa Choudh)