Cortical and trabecular bone are equally affected in rats with renal failure and secondary hyperparathyroidism

The incidence and prevalence of chronic kidney disease (CKD) continues to rise, with rates doubling in people aged 65 years or older [1]. The pathological manifestations of the disease include renal hyperparathyroidism, cardiovascular disease, vascular calcification, and renal osteodystrophy, all of which can alter bone mineral metabolism resulting in increased fracture risk and other skeletal abnormalities [2‐5]. Patients with moderate CKD can have up to a 2.4-fold increase in fracture occurrence compared to mild CKD or healthy patients [6].

CKD causes a dysregulation of mineral homeostasis that negatively alters bone mineralization parameters and leads to skeletal abnormalities [7, 8]. Several rodent models have been developed for CKD to evaluate the consequence and underlying mechanisms for bone metabolism changes in response to moderate CKD. In a 5/6-nephrectomy mouse model of CKD, the material properties of cortical bone within the tibial diaphysis were affected [9]. Similarly, a 5/6-nephrectomy rat model of CKD altered bone mineral and material properties and led to hyperparathyroidism [10]. In another model of CKD where rats underwent thyroparathyroidectomy and partial nephrectomy, cortical bone mineral properties in the femur were altered resulting in reduced mechanical strength [11]. In patients with CKD, common complications include hyperphosphatemia and secondary renal hyperparathyroidism (rHPT), in which disturbances in the homeostasis of calcium, phosphate, and vitamin D lead to elevated parathyroid hormone (PTH) levels and bone loss. Elevated PTH levels lead to reduced cortical [3] and trabecular bone mineral densities [12] as well as pronounced changes in trabecular microstructure [12‐16].

In the absence of disease, calcium and phosphorus levels are maintained through a complex interaction between the kidneys, intestine, bones, and parathyroid glands. Calcium-sensing receptors in the parathyroid gland detect changes in serum calcium and regulate the synthesis and secretion of PTH [17]. Increases in PTH levels, as in CKD, occur as a response to increased renal phosphorus excretion and the downregulation of calcium sensing receptors that lead to sustained activation of PTH, thus causing detrimental effects on bone [18]. The disruption of phosphate homeostasis in CKD induces secondary hyperparathyroidism. Furthermore, higher levels of fibroblast growth factor 23 (FGF23), a bone-derived hormone that regulates phosphorous and 1,25[OH]₂D3, may in fact precede changes in PTH levels [19]. Calcium, 1,25[OH]₂D3, and PTH stimulate FGF23 secretion via negative endocrine feedback loops [20, 21]. Thus, PTH and FGF23 have been implicated as key players in mediating the observed changes in skeletal metabolism during kidney failure.

The effects of chronic renal failure on bone metabolism have been studied with nephrectomy animal models, and the 5/6 nephrectomy model is commonly used to induce severe secondary hyperparathyroidism. The aim of this study was to evaluate the effect of high phosphorus diet on the trabecular and cortical bone in rats with kidney failure and advanced secondary hyperparathyroidism using DXA scanning and micro-CT. We also aimed to determine the relationship between skeletal changes and serum levels of PTH, FGF23, and sclerostin.

All rats survived to the completion of the study. There was no significant difference in body weight (Fig. 1b) between groups, but gain in total body length was reduced by 7.6% (p < 0.05) in the Nx-Phos vs. control and by 6.5% (p < 0.05) in the Nx-Phos vs. Nx-Control (Fig. 1c). See Table 1 for anthropometric measurements.

As expected, both femur and lumbar bone mineral density (BMD) increased across time in all 3 groups of rats. However, the lumbar BMD was significantly lower in the Nx-Phos compared to Nx-Control and control groups at 2 and 4 weeks post-surgery (p < 0.01; Fig. 2a). After 2 weeks, femur BMD was significantly reduced in the Nx-Phos group compared to the other groups (p < 0.05; Fig. 2b) but not at 4 weeks post-surgery

Micro-CT analysis showed that nephrectomy triggered significant cortical bone loss. The volumetric bone mineral density (vBMD) in the fifth lumbar vertebrae was reduced by 4.5% in Nx-Control (p < 0.01) and 4.2% in Nx-Phos (p < 0.01) vs. Control (Fig. 3a). Femur cortical vBMD was also reduced by 2.6% in Nx-Control (p < 0.01) and 3% in Nx-Phos (p < .01) vs. Control group (Fig. 3b). Micro-CT analyses of trabecular bone of L5 vertebrae revealed that trabecular bone volume adjusted for tissue volume (BV/TV) was significantly reduced in Nx-Phos compared to Nx-Control and Control animals (p < 0.01; Fig. 4a–c). The reduced trabecular BV was caused by significant reductions in trabecular number (Fig. 4d) and increases in trabecular separation (Fig. 4e). In contrast, trabecular thickness was not different among the three groups of rats (Fig. 4f). The loss of trabecular elements reduced the number of interconnections between trabeculae as reflected by a > 25% reduction (p < 0.01) in connectivity density in the Nx-Phos compared to Nx-Control and Control groups (Fig. 4g). Accordingly, structure model index (SMI), which indicates relative prevalence of rod-like to plate-like structures, was increased in Nx-Phos animals (Fig. 4h). Similar to lumbar vertebrae, femurs of Nx-Phos animals exhibited significantly reduced trabecular BV that was caused by changes in trabecular number and thickness (Fig. 5a–g). Consistent with the changes in trabecular architecture, connectivity density was decreased while SMI was increased in the femurs of Nx-Phos animals (Fig. 5h–i). In contrast to bone volume (BV), measured tissue volume (TV) was not different between the three groups at the lumbar vertebrae and femurs (Fig. 4b and Fig. 5c).

Histomorphometric measurements at the secondary spongiosa of the tibia revealed increased bone resorption (TRAP-positive surface) caused by nephrectomy (Table 2). TRAP-positive osteoclast surface adjusted for bone surface (Oc.S/BS) was increased by 177% (p < .01) in Nx-C rats compared to Control. The Oc.S/BS was also increased by 72% (p < .05) in Nx-Control compared to Nx-Phos. Similarly, the osteoclast perimeter (Oc.Pm) was also increased in Nx-Control compared to Control rats, but did not reach statistical significance (p = .08).

Serum levels of calcium (Fig. 6a) and sclerostin (Fig. 6b) were unaffected in all groups at days 19 and 26 after nephrectomy. Serum phosphorus (p < 0.01; Fig. 6c), creatinine (p = .05; Fig. 6d), and FGF23 (p < 0.01; Fig. 6e) levels were significantly elevated in Nx-Phos animals at day 26 post-surgery. In addition, nephrectomy induced a significant increase in serum creatinine and iPTH at day 19 post-surgery. The increases in serum PTH levels at day 26 were significant for the Nx-Phos but not for Nx-Control animals (p < 0.01; Fig. 6f). Serum levels of iPTH and FGF23 were negatively associated with trabecular vBMD (r = − 0.78, r ²  = 0.61, p < 0.05; r = − 0.78, r ²  = 0.61, p < .05; Fig. 7a & b, respectively) at day 26 after nephrectomy. Serum levels of phosphorus trended towards a positive association with iPTH (r = 0.54, r ²  = 0.29, p = 0.10; Fig. 7c) at day 26 after nephrectomy. A positive relationship between iPTH and FGF23 (r = 0.75, r ²  = 0.56, p < 0.05; Fig. 7d) was also observed at day 26 after nephrectomy. Interestingly, serum levels of sclerostin did not show significant correlation with bone parameters (trabecular or cortical bone), PTH, or FGF23 (data not shown).

Animal models of CKD use nephrectomy (partial or complete) as a method to induce early renal failure to accelerate endocrine dysfunction and bone turnover. Previous studies have shown the cortical region of bone to be the most susceptible to resorption in CKD [29, 30]. In this study, we investigated the effects of a high phosphorus diet on bone in a rat model of CKD and found that high phosphorus diet with advanced secondary hyperparathyroidism enhanced trabecular bone loss in the lumbar vertebrae and femur.

High levels of phosphorus have been found to increase vascular calcification, reduce cortical bone density and thickness, and cause severe trabecular bone alterations [2]. In a rat model of 5/6 nephrectomy, high phosphate reduced trabecular bone volume in the femur [31]. Similarly, we found that high phosphorus diet led to a significant change in bone microarchitecture as indicated by reduced trabecular bone volume, reduced trabecular number, and increased trabecular separation at day 35. These changes were also present in the lumbar vertebrae where the reductions in trabecular bone volume were caused primarily by increased trabecular separation. The changes in trabecular bone parameters cannot be explained based on systemic changes in somatic growth as neither body weight nor femur length were significantly different among the three groups.

To further explore the mechanisms behind trabecular bone loss, we examined the relationship between serum levels of PTH and skeletal alterations induced by high phosphorus diet. We found that high phosphorus diet increased PTH levels at days 19 and 26 and that increased PTH could have contributed to the increased bone resorption and reduced trabecular bone mass. PTH is known to enhance both bone formation and resorption [32]. In addition to exerting anabolic effects on osteoblasts, PTH treatment increases expression of receptor activator of nuclear factor kappa-B ligand, a key regulator of osteoclastic differentiation and activation [32]. PTH also promotes the conversion of 25-hydroxy vitamin D3, an inactive form of vitamin D3, into active 1,25-dihydroxy vitamin D3 (1,25[OH]₂D3) by the activation of 1-α-hydroxylase in kidneys [33] leading to absorption of calcium and phosphorous in the gut. Several studies have investigated how PTH and 1,25[OH]₂D affect serum phosphate via calcium regulation [34, 35]. Our data shows that nephrectomy alone increased serum PTH levels. However, this increase in PTH was further augmented by high phosphorus diet and confirmed by a positive correlative relationship. We also do report that serum levels of iPTH and FGF23 were negatively associated with trabecular bone volume; however, the issue of whether elevated PTH contributes to both cortical bone loss caused by nephrectomy and trabecular bone loss caused by high phosphorus diet in the nephrectomy rats remains to be evaluated. There are also data to suggest that there may be alternate mechanisms by which high phosphate affects trabecular bone, regardless of PTH levels [31].

To explore alternate mechanisms, we also examined the relationship between serum levels of FGF23 and skeletal parameters in the combined data from the 3 groups of animals. FGF23 has been reported to play a role in balancing mineral ion homeostasis and bone mineralization through the regulation of phosphorus [36, 37]. FGF23 is known to be secreted by osteocytes and regulate phosphorus homeostasis by reducing the reactivity of sodium-phosphate co-transporters in the proximal renal tubules and affecting 1,25[OH]₂D metabolism and absorption by decreasing 1-α-hydroxyase activity [37‐40], thus excreting excess phosphate through the kidney. In CKD and renal failure, FGF23 concentrations rise to excrete excess phosphate as a result of diminished renal function, leading to high bone turnover. We found that high phosphorus diet increased serum FGF23 levels at day 26. Additionally, high levels of serum FGF23 were strongly associated with trabecular bone loss. In our study, elevations in FGF23 were associated with increased phosphate and PTH levels, and all parameters were negatively associated with bone density, as other studies have reported [41, 42]. Sclerostin, like FGF23, is mainly secreted by osteocytes [43, 44]. Sclerostin decreases bone formation by inhibiting terminal differentiation of osteoblasts and promoting their apoptosis [45] in part via blockade of Wnt signaling pathways by binding to LRP5/6 receptors [45, 46]. In our studies, we did not find significant changes in serum levels of sclerostin in nephrectomy rats fed with normal or high phosphorus diet compared to control animals. However, we cannot exclude the possibility that local expression of sclerostin in osteocytes is altered by nephrectomy and/or high phosphorus diet. A recent study has also identified that FGF23/Klotho complex induces Dkk1 expression and thereby inactivates Wnt/β-catenin signaling in osteoblasts. Thus, impairment of Wnt/β-catenin signaling by high FGF23 levels may provide a mechanism for inhibition of bone formation in CKD [47]. In addition, the creatinine levels observed in our study are consistent with what others have reported in similar rat models of CKD [48‐50].

Interestingly, nephrectomy induced significant cortical bone loss and higher osteoclast activity in the secondary spongiosa of the tibia, independent of phosphorus intake. Impaired kidney function may have altered the mineral or collagen composition of the cortical bone matrix [11, 51], which, in combination with PTH levels, may increase porosity [7]. In fact, higher PTH levels may predict cortical deterioration [3]. In addition, serum calcium levels were not reduced; this lack of an effect can be attributed to secondary hyperparathyroidism in nephrectomy rats.

Our study investigated the negative effects of renal failure and its relationship with phosphorus, FGF23, PTH, sclerostin, and bone. We found that high phosphorus diet adversely affects trabecular bone, whereas nephrectomy induced cortical bone loss. Future studies are needed to investigate the molecular mechanisms by which high phosphorus diet affects trabecular bone mass in models of CKD.