Increased PHOSPHO1 and alkaline phosphatase expression during the anabolic bone response to intermittent parathyroid hormone delivery

Abstract The administration of intermittent parathyroid hormone (iPTH) is anabolic to the skeleton. Recent studies with cultured osteoblasts have revealed that the expression of PHOSPHO1, a bone‐specific phosphatase essential for the initiation of mineralisation, is regulated by PTH. Therefore, this study sought to determine whether the bone anabolic response to iPTH involves modulation of expression of Phospho1 and of other enzymes critical for bone matrix mineralisation. To mimic iPTH treatment, primary murine osteoblasts were challenged with 50 nM PTH for 6 h in every 48 h period for 8 days (4 cycles), 14 days (7 cycles) and 20 days (10 cycles) in total. The expression of both Phospho1 and Smpd3 was almost completely inhibited after 4 cycles, whereas 10 cycles were required to stimulate a similar response in Alpl expression. To explore the in vivo role of PHOSPHO1 in PTH‐mediated osteogenesis, the effects of 14‐ and 28‐day iPTH (80 µg/kg/day) administration was assessed in male wild‐type (WT) and Phospho1−/− mice. The expression of Phospho1, Alpl, Smpd3, Enpp1, Runx2 and Trps1 expression was enhanced in the femora of WT mice following iPTH administration but remained unchanged in the femora of Phospho1−/− mice. After 28 days of iPTH administration, the anabolic response in the femora of WT was greater than that noted in Phospho1−/− mice. Specifically, cortical and trabecular bone volume/total volume, as well as cortical thickness, were increased in femora of iPTH‐treated WT but not in iPTH‐treated Phospho1−/− mice. Trabecular bone osteoblast number was also increased in iPTH‐treated WT mice but not in iPTH‐treated Phospho1−/− mice. The increased levels of Phospho1, Alpl, Enpp1 and Smpd3 in WT mice in response to iPTH administration is consistent with their contribution to the potent anabolic properties of iPTH in bone. Furthermore, as the anabolic response to iPTH was attenuated in mice deficient in PHOSPHO1, this suggests that the osteoanabolic effects of iPTH are at least partly mediated via bone mineralisation processes.


| INTRODUCTION
Parathyroid hormone (PTH) is synthesised and secreted by the chief cells of the parathyroid glands to maintain serum calcium (Ca2+) concentrations within a very narrow range. This function is essential to prevent disturbance to a range of cellular functions and is mediated by interactions with the PTH1 receptor in key target tissues such as bone, kidney and intestine. 1,2 PTH also maintains phosphate homoeostasis by several mechanisms including the promotion of osteocyte synthesis of fibroblast growth factor 23 (FGF23) which, in conjunction with cofactor klotho and via the renal FGF receptor, promotes phosphaturia by downregulating sodium-phosphate cotransporter (Npt2a and Npt2c) expression. 3 In addition to increasing osteoclast bone resorption to mobilise Ca2+, PTH also has profound effects on the cells of the osteoblast lineage and is now regarded as an important hormone regulating bone remodelling. 4 The pleiotropic effects of PTH on bone are dictated by the frequency of the exposure, and the mechanisms involved have been progressively uncovered, aiding our understanding of the osteo-anabolic effects of intermittent administration of PTH (iPTH). The pioneering studies by Reeve et al. 5 revealed increased bone formation in osteoporotic women in response to daily exposure to low-dose PTH 1-34 (100 μg/day human PTH). Extensive clinical trials have also confirmed the efficacy of daily PTH  injection in reducing vertebral and nonvertebral fractures and increasing bone mineral density (BMD) in postmenopausal osteoporotic women. [6][7][8] In contrast, continuous delivery results in increased bone loss and increased porosity principally in the cortical compartment. [9][10][11] This bone loss is associated with increased bone remodelling where the enhanced expression of receptor activator of nuclear factor-ƙB ligand (RANKL) expression and the decreased expression of the RANKL decoy receptor, osteoprotegerin, ensure that any increases in bone formation are dwarfed by a prevailing bone resorption response. 12,13 The capacity of continuous or iPTH delivery to finely tune osteoblast and osteocyte gene expression is likely to contribute to the pleiotropic effects of PTH on the skeleton. 10,14 The osteoblast and its progenitors have reproducibly been shown to be the primary in vivo target for PTH. iPTH administration promotes osteoblastogenesis and the treatment of cultured bone marrow cells and calvarial osteoblasts by iPTH results in increased expression of Runt-related transcription factor 2 (Runx2), type I collagen procollagen (Col1a1), osteocalcin (Bglap) and tissue-nonspecific alkaline phosphatase (Alpl). 10,15,16 iPTH administration also reduces osteoblast apoptosis, reactivates quiescent bone-lining cells and downregulates sclerostin expression. 17,18 Whilst the effects of PTH on the expression of osteoblast transcription and differentiation factors have been studied widely, less is known about the impact of PTH on the expression of genes that encode mineralisation-regulating enzymes. Such knowledge would help clarify the function of PTH on the mineralisation process per se rather than osteoblast differentiation and matrix production during bone formation. Although hyperparathyroidism is associated with increased bone remodelling and decreased BMD, the bone formation response that occurs is predominated by osteoid production rather than the formation of a true mineralised bone matrix. 19 Furthermore, rats receiving an anabolic PTH regimen for 3 weeks have increased osteoid surface, thickness and volume which is consistent with an increased rate of bone formation. 20 The initiation and propagation of the mineral phase within newly formed osteoid is dependent upon the interplay between PHOS-PHO1 and TNAP which are required for the generation of Pi within matrix vesicles (MVs) and the hydrolysis of the mineralisation inhibitor, PPi, respectively. 21-23 nSMase2 catalyses the hydrolysis of sphingomyelin within MV membranes to produce phosphocholine; a recognised substrate for PHOSPHO1 24,25 whereas ENPP1 generates PP i from nucleoside triphosphates. 26  is recognised to increase osteoblast number and promote bone formation. A critical component of this anabolic response is the mineralisation of the osteoid matrix but the effects iPTH has on the expression of mineralisation regulating enzymes is unclear. This study's principal finding was that iPTH administration increased the expression of Phospho1, Alpl, Enpp1 and Smpd3 in vivo which was consistent with a bone anabolic PTH regimen. The data also disclosed that in the absence of PHOSPHO1, the iPTH anabolic response was dampened, suggesting that amplified Phospho1 expression is a prerequisite for a full iPTHmediated bone anabolic response.
in MC3T3-C14 osteoblast-like cells after 24 h of continuous PTH exposure whereas Enpp1 expression by differentiated IDG-SW3 osteocyte-like cells was decreased after 3-6 h of PTH treatment. 27,28 These data confirm previous studies reporting the decreased expression of Phospho1, Smpd3 and other mineralisation-associated genes by PTH (up to 24 h exposure) in both osteocyte and bone marrow stromal cell lines. 29,30 Whilst the downregulation of Phospho1 and Smpd3 in osteoblasts in response to continuous PTH may provide mechanistic insight for the observed lower BMD in hyperparathyroidism, these studies are limited by the use of cell lines and do not provide insight into the anabolic effects of iPTH on the expression of mineralisation regulating enzymes. This study, therefore, sought to determine whether iPTH alters the expression of Phospho1, Smpd3 and Alpl in primary osteoblasts and murine bones. We also completed studies in     which the greyscale range was set on were 0-2.34 g/cm 3 .
Metaphyseal trabecular bone of the proximal tibia was assessed in a 1000 μm section, 5% of the total bone length below the first appearance of a trabecular 'bridge' connecting the two primary spongiosa bone islands. 34 Cortical bone was assessed in a 500 μm section at 37% of the total bone length from the reference starting slice (first appearance of the medial tibial condyles). To assess BMD, BMD phantoms were used to calibrate the CTAn software. BMD phantoms of known calcium hydroxyapatite mineral densities of 0.25 and 0.75 g/cm 3 were scanned and reconstructed using the same parameters as used for bone samples.

| Calvarial osteoblast isolation and cell culture
Osteoblasts were isolated from calvaria of 3-5 days old WT mice following established procedures. 35,36 In brief, excised calvaria were digested with 1 mg/ml collagenase type II for 10 min followed

| RNA extraction and RT-qPCR analysis
Total RNA was extracted from cells or whole bones using the RNeasy mini kit (Qiagen) according to the manufacturer's The effects of iPTH on the expression of mineralisation regulating enzymes during osteoblast matrix mineralisation. PTH (50 nM) was administered to primary osteoblasts in cycles (6 h PTH treatment in every 48 h period for up to 20 days). For cycle 4:  This dosing regimen resulted in a rapid increase in Pthr1 (p < .001; Figure 3A) and Phospho1 (p < .01; Figure 3B) expression compared with vehicle-treated animals. The expression of Alpl (p < .05; Figure 3C) and Smpd3 (p < .001; Figure 3D) were likewise significantly increased in response to PTH treatment. Sost expression displayed a trend towards inhibition by PTH exposure, but this did not reach statistical significance ( Figure 3E). The increased expression of the osteoblast transcription factors, Runx2 (p < .05; Figure 3F) and Trps1 (p < .05; Figure 3G) in response to a single PTH dose is consistent with its bone anabolic actions. Osteoblast number on the trabecular bone surface was similar between the control and PTH-administered mice ( Figure 4A).

| Fourteen-day iPTH administration increases Phospho1, Alpl and Smpd3 expression in WT mice only
Having shown that a single delivery of PTH rapidly increases Phospho1, Alpl and Smpd3 expression within femora, we next determined if the anabolic bone response typical of an iPTH regimen was also associated with changes to the expression of these key regulators of bone mineralisation. Also, as PHOSPHO1 is recognised to be essential for the initiation of the mineralisation process, we also aimed to establish if amplified Phospho1 expression in response to iPTH was a prerequisite for a bone anabolic response by examining whether the genetic ablation of PHOSPHO1 attenuated the iPTH anabolic response.
In the diaphysis of WT animals, iPTH administration for 14 days induced a 3.1-fold increase in Phospho1 expression compared to vehicle-treated control animals (p < .05; Figure 5A). Similarly, Pthr1 (2.0-fold; p < .001; Figure 5B), Alpl (3.4-fold; p < .001; Figure 5C), Smpd3 (2.9-fold; p < .001; Figure 5D) and Enpp1 deficiency ( Figure 5B-E). Runx2 (2.6-fold; p < .01; Figure 5F) and Trps1 (2.4-fold; p < .01; Figure 5G) expression was increased in response to iPTH administration in WT but not Phospho1 KO mice resulting in a significant interaction between PTH treatment and genotype ( Figure 5F,G). As expected, WT mice receiving iPTH expressed less Sost compared to vehicle-treated WT mice (p < .05; Figure 5H). Furthermore, the basal expression of Sost mRNA in Phospho1 KO mice was significantly reduced compared to WT mice but there was no further decrease in Sost expression in response to iPTH administration (p < .05; Figure 5H). When the data are compared to the results of the single 6 h PTH in vivo study (Figure 3), the effects of iPTH on gene expression in WT mice were all accentuated with 14 days iPTH treatment ( Figure 5).
To determine if the changes in gene expression were reflected in a bone anabolic response, the effects of 14-day iPTH exposure on bone microarchitecture, geometry and BMD within both WT and Phospho1 KO tibiae was determined ( Table 2) were also limited although there was a clear genotype effect on many of the measured parameters (Table 2).  Therefore, we next administered iPTH to WT and Phospho1 KO mice for 28 days to determine if there was a modified response to iPTH administration in the absence of PHOSPHO1. The percentage trabecular bone volume (BV/TV) (p < .001; Figure 6B) was increased in iPTH-treated WT mice but not in iPTH-treated Phospho1 KO mice but there was no significant overall interaction between PTH and genotype. Trabecular number, thickness and BMD were increased whereas trabecular pattern factor was decreased by iPTH in WT mice and these responses were similarly observed in Phospho1 KO mice ( Figure 6C,D,F,G). Trabecular separation in WT and Phospho1 KO mice was not affected by iPTH ( Figure 6D). After 28 daily injections of PTH, the anabolic response in cortical parameters was now evident in WT mice. Cortical BV/TV (p < .05; Figure 6I) and cortical thickness (p < .05; Figure 6J) were increased in iPTH-treated WT but not in iPTH-treated Phospho1 KO mice and for cortical thickness there was a significant interaction between PTH treatment and genotype. Total cortical tissue area was increased only in iPTHtreated Phospho1 KO mice ( Figure 6L; p < .05). Cortical area, medullary area and BMD in both WT and Phospho1 KO mice were not affected by iPTH but there was a significant genotype effect on all three parameters ( Figure 6K,M,N). Osteoblast number on the trabecular bone surface was increased (p < .05) by iPTH in WT but not Phospho1 KO mice but there was no significant overall interaction between PTH and genotype ( Figure 4B). These findings show that the anabolic bone response to iPTH is linked to changes in the expression of these key regulators of mineralisation and are attenuated by the genetic ablation of PHOSPHO1.

| DISCUSSION
Cells of the osteoblast lineage are responsible for the osteo-anabolic effects of iPTH. iPTH reduces osteoblast apoptosis and promotes the commitment and differentiation of mesenchymal cells and osteoblast precursors to progress through the osteoblast lineage. 38 Lineage tracing experiments confirmed that iPTH exposure induces the reactivation of bone-lining cells to functional osteoblasts and delays the transition of mature osteoblasts to bone-lining cells. 18 Osteocytes are also targets of PTH where it inhibits sclerostin expression to promote bone formation through the WNT/β-catenin signalling pathway. 39,40 In addition to these established pathways by which iPTH promotes BMD it is conceivable that iPTH may directly alter the expression levels of genes that impact on the mineralisation process. [27][28][29][30] The deletion of Smpd3, 24,41 Phospho1 32,42,43 or Alpl 44,45 results in a hypomineralised skeleton, whereas in mice with an ablation of both Phospho1 and Alpl there is a complete absence of a mineralised skeleton. 31 The aims of this study were, therefore, to fully assess the effects of iPTH delivered in vitro and in vivo on the expression of Phospho1, Alpl and Smpd3 to better understand the response of osteoblasts to an anabolic PTH regimen. The expression T A B L E 2 Micro-CT analysis from bones of wild-type and Phospho1 −/− mice treated with vehicle and iPTH for 14 days  Despite the lack of change in osteoblast numbers after 6 h of PTH treatment, the expression of both Runx2 and Trps1 was upregulated, with both known to regulate Smpd3 and Phospho1 expression in osteoblasts. 48,49 Trps1, a GATA transcription factor which primarily acts to repress genes known to be associated with the bone mineralisation process. 48,50 Mutations in the Trps1 gene lead to the human condition tricho-rhino-phalangeal syndrome which is characterised by craniofacial and skeletal dysplasias and siRNA mediated knockdown of Trps1 in a pre-odontoblastic cell line leads to the suppression of Phospho1, Alpl and Smpd3 expression. 48 Also, Runx2, an osteoblast transcription factor upstream of osterix expression, is essential for osteoblastogenesis and its overexpression in mouse limb bud cultures leads to a >3-fold increase in Phospho1 and Smpd3 expression. 49  KO mice. Alternatively, as PHOSPHO1 has a critical role in the initiation of the mineralisation process, its presence may be a prerequisite for the bone to elicit an anabolic response to iPTH. On this basis, it is tempting to speculate that the actions of iPTH on bone are at least partly dependent on the targeting of the mineralisation phase of bone formation. 51 The presence of a poorly mineralised bone may be compounded by the inability of iPTH to elevate Alpl and Smpd3 expression in the absence of Phospho1 as both are essential for bone mineralisation. 24,44 How the absence of PHOSPHO1 influences osteoblast expression of Smpd3 and Alpl in response to iPTH is unclear. However, as BMP2 stimulates both Smpd3 and Alpl expression, an action blunted by PTH, there may be, in a Phospho1 deficient state, complex cross-talk between PTH and BMP2 signalling. 27,52 is also of note that cortical porosity appeared to be exacerbated in Phospho1 KO mice treated by iPTH-treatment ( Figure 6H). Enhanced cortical porosity is normally associated with In summary, the response of Phospho1, Alpl and Smpd3 expression by cultured osteoblasts to iPTH was at odds with that noted in vivo. The increased Phospho1, Alpl, Enpp1 and Smpd3 expression in bone from mice subjected to long-term iPTH treatment was consistent with an anabolic PTH regimen. In the absence of PHOSPHO1, the iPTH anabolic response was attenuated, suggesting that amplified Phospho1 expression in response to iPTH is a prerequisite for an anabolic response.