Secondary hyperparathyroidism is a frequent complication of chronic renal failure characterized by parathyroid hyperplasia and enhanced synthesis and secretion of parathyroid hormone (PTH) [1-3]. High circulating PTH levels causes osteitis fibrosa and bone loss, typical features of renal osteodystrophy, as well as cardiovascular complications, which increase mortality in renal failure patients [4-6].

The three main direct causes of hyperparathyroidism are hypocalcemia, hyperphosphatemia and 1,25-dihydroxyvitamin D [1,25(OH)₂D₃] deficiency. Hyperphosphatemia and 1,25(OH)₂D₃ deficiency also enhance parathyroid function indirectly by lowering serum calcium (Ca) [1-3,7]. As renal disease disease progresses, a reduction in parathyroid content of the Ca sensing receptor (CaSR) and the vitamin D receptor (VDR) renders the parathyroid glands more resistant to suppression of cell growth and PTH synthesis in response to Ca and 1,25(OH)₂D₃[8-10]. Lack of an appropriate parathyroid cell line and the rapid de-differentiation of primary cultures of parathyroid cells have impeded direct characterization of the pathogenic mechanisms underlying uremia induced parathyroid hyperplasia. To address these issues, we used the only experimental approach available at present, the 5/6 nephrectomized rat model. The results of these studies have implicated increases in parathyroid co-expression of the growth promoter TGFa and its receptor, the EGFR, in uremia induced- enhancement of proliferative activity and gland size[11,12] . In addition, they have revealed novel insights into the antiproliferative actions of vitamin D therapy in arresting growth driven by overexpressed EGFR. This review present the most recent experimental evidence on the relevance of 1,25(OH)₂D₃ downregulation of the autocrine TGF/EGFR growth loop in the control of parathyroid hyperplasia in renal failure.

Enhancement of TFG/EGFR co-expression is an early event in parathyroid hyperplasia of renal failure. Similar to hyperparathyroidism in humans [13], TGF expression is higher in uremic rats compared to normal controls [11]. More importantly, when uremia-induced parathyroid hyperplasia is worsened by a high P [11] or low Ca intake [14], there is a temporal relationship between increases in parathyroid TGFa, the enlargement of the parathyroid glands and proliferating activity. Furthermore, suppression of uremia-induced parathyroid hyperplasia by high dietary Ca (2% Ca diet) or P restriction prevented the increases in parathryorid TGF induced by the onset of renal failure by day 7 after 5/6 nephrectomy [11,12]. These finding suggest a role for TGF as a mitogenic stimulus in the parathyroid glands, triggered by uremia and further enhanced by high dietary P [11] or low Ca intake [14].

Since enhanced co-expression of TGF and EGFR associates with more aggressive growth in normal and transformed tissues [15], we examined whether uremia and dietary Ca and P manipulations also modulate parathyroid EGFR expression. Quantification of TGF and EGFR immunostaining, as previously described [11], demonstrated an average increase of 2.3-fold above normal by high dietary P or low Ca intake by 7 days after the onset of uremia. Conversely, similar to their efficacy in preventing increases in parathyroid TGF content, high dietary Ca and P restriction prevented uremia-induced increases in EGFR content. The strong association between enhanced TGF/EGFR co-expression and high proliferative activity does not constitute substantial evidence of its contribution to uremia induced parathyroid hyperplasia. To directly address the pathophysiological relevance of enchanced TGF/EGFR co-expression on parathyroid-cell growth, we utilized the EGFR-tyrosine kinase inhibitor AG1478, effective in inhibiting the autocrine TGF/EGFR-growth loop in vitro [16] and in vivo [17,18]. .

Enhanced co-expression of TGF and EGFR is a major pathogenic mechanism for parathyroid hyperplasia. Similar to most EGFR-tyrosine kinase inhibitors (TKIs), AG1478 is a small molecule, highly selective for EGFR-tyrosine kinase [19], that competes with ATP binding and reversibly inhibits tyrosine trans-phosphorylation, thereby blocking downstream signaling.

5/6 nephrectomized rats (180-200 g body weight), fed a high P diet (1.2 % P), received intraperitoneal injections of either vehicle (1ml of DMSO: PBS; 1:1) or AG1478 (25 mg/kg body weight, every other day for one week). This dose is half of that effective to arrest the growth of aggressively growing tumors in mice (given daily) and was further adjusted considering a metabolic mouse: rat ratio of 2. This dose had no adverse effect as judged by no differences in body weight, serum creatinine, BUN and pH. Ionized Ca, total Ca and P levels were similar in the uremic control group and the AG1478–treated animals. AG1478 reduced the enlargement of the parathyroids glands and the proliferative activity by 60% compared to uremic, high P-controls. These findings demonstrate that enhanced TGF/EGFR co-expression is a main contributor to uremia-induced parathyroid hyperplasia.

The antiproliferative properties of prophylactic vitamin D therapy in the parathyroid glands involve downregulation of TGF/EGFR co-expression. The efficacy of prophyactic administration of either 1,25-dihydroxyvitamin D or the less calcemic vitamin D analog 19-Nor-1,25(OH)₂D₂ (at doses of 4 ng or 30 ng, respectively) to arrest parathyroid cell growth and consequently, the enlargement of the parathyroid glands induced by uremia and high dietary P, associates with prevention of the increases in TGF [11,12] and EGFR expression (Figure 1). Several reports demonstrate a direct and cell specific regulation of EGFR mRNA by 1,25-dihydroxyvitamin D [21,22]. Since TGF activation of EGFR induces both TGFa- and EGFR-gene expression [23,24], it is also possible that 1,25(OH)₂D₃-inhibition of EGFR activation [25] mediates the suppressive effects of the sterol on TGF and EGFR expression

1,25(OH)₂D₃ antiproliferative properties in EGFR overexpressing cells involve downregulation of both classical EGFR-growth signals from the cell membrane and EGFR-transactivaction of the cyclin D1 gene. The efficacy of vitamin D therapy to arrest TGF-driven hyperplastic growth in established secondary hyperparathyroidism and psoriasis [26,27] suggests that 1,25(OH)₂D₃ could downregulate TGF/EGFR-growth signals.

Recent studies from our laboratory [25] in the human epidermoid carcinoma cell line A431, overexpressing TGF and EGFR [28], and in NR6 cells, normal human embryonic cells overexpressing exclusively EGFR, demonstrate that 1,25(OH)₂D₃ treatment reduced growth by decreasing EGFR activation. In A431 cells, the main mechanism involved in 1,25D-inhibition of EGFR activation is induction of EGFR accumulation in early endosomes by prevention of its recycling to the plasma membrane [25]. As a consequence, there is reduced membrane bound EGFR and a marked reduction in EGFR activation by tyrosine phosphorylation upon ligand binding. The ability of 1,25(OH)₂D₃ to inhibit autocrine EGFR activation in EGFR-overexpressing cells results in reduced phosphorylation of ERK1/2, the main downstream-growth signal. In aggressively growing tumors driven by enhanced TGF/EGFR co-expression, prevention of nuclear P-ERK1/2 translocation is used as an index of the efficacy of anti-EGFR therapy [17]. The potency of 1,25(OH)₂D₃ to downregulate TGF/EGFR autocrine signals in A431 cells is similar to that of EGFR-tyrosine kinase inhibitors as shown by the lack in nuclear translocation of phosphorylated ERK1/2 in A431 cells treated with 1,25(OH)₂D₃[25].

In A431 cells, 1,25(OH)₂D₃ inhibits classical EGFR-growth signals from the plasma membrane, common to most tyrosine-kinase receptors, and the novel EGFR nuclear actions as a transactivator (co-activator) of the cyclin D1 gene, an important contributor to parathyroid hyperplasia in humans. The mechanism appears to involve impaired nuclear translocation of the EGFR, possibly as a consequence of the stasis of the receptor in the endocytic compartment, since total EGFR levels are not affected by 1,25(OH)₂D₃ treatment[25].

1,25(OH)₂D₃ potentiates the growth arrest induced by TKI in EGFR-overexpressing cells. Carcinogenic cells overexpressing EGFR subvert the G1 to S transition by decreasing the levels of cyclin dependent kinase inhibitor p27 and inducing cyclin D1, the two cell cycle regulators more frequently associated with high proliferating activity in the parathyroid glands of renal failure patients. Reports on the mechanisms mediating the inhibition of EGFR-driven growth by AG1478 demonstrate that up-regulation of p27 protein levels is mandatory for the efficacy of TKI therapy [16,29]. Since 1,25(OH)₂D₃ induces p21 and p27 expression in several tissues including the parathyroid glands [11,30] and reduces cyclin D1 protein and mRNA levels in colon carcinoma we examined potential synergistic effects of the sterol on TKI therapy.

Dose response studies on the efficacy of EGFR-TKI to suppress A431 growth demonstrated that maximal inhibition was achieved after a 20h exposure to 1 to 3 uM AG1478. Simultaneous treatment with submaximal (0.1 mM) and maximal inhibitory (1 mM) doses of AG1478 and 100 nM 1,25(OH)₂D₃ resulted in an additional suppression of the growth induced by AG1478 alone, as shown by the reduction in thymidine incorporation to DNA in Table 1.

These results demonstrate that mechanisms unrelated to inhibition of EGFR-growth signals also contribute to 1,25(OH)₂D₃ suppression of growth in EGFR-overexpressing cells. They also suggest the potential benefits of combined therapy with TKIs and 1,25(OH)₂D₃ vitamin D in the control of hyperplastic growth driven by EGFR overexpression.

The 1,25(OH)₂D₃ synthesized in the kidney by mitochondrial 1-hydroxylase is transported in the blood by carrier proteins. Although vitamin D binding protein (DBP) is the main carrier, 1,25(OH)₂D₃ also binds albumin and lipoproteins [31]. Recently, the reports showing that the free form of 1,25(OH)₂D₃ triggers biological responses after entering target cells by simple diffusion have been challenged by the demonstration that, in renal proximal tubular cells, 25-hydroxyvitamin D uptake occurs through receptor (megalin)-mediated endocytosis of the 25(OH)D bound to plasma DBP DBP [32]. A similar endocytosis could mediate the cellular uptake of 1,25(OH)₂D₃-bound to DBP or lipoproteins. Uremia-induced reduction in megalin expression may constitute a mechanism for vitamin D resistance independent of abnormalities in VDR content or function. Once inside the cell, 1,25(OH)₂D₃ can be inactivated by mitochondrial 24-hydroxylase or it can bind the VDR. Ligand binding activates the VDR to translocate from the cytosol to the nucleus where it heterodimerizes with its partner the retinoid X receptor, RXR. The VDR/RXR complex binds specific sequences in the promoter region of target genes, called vitamin D response elements (VDRE), and recruits basal transcription factors and co-regulator molecules to either increase or suppress the rate of gene transcription by RNA-polymerase II. Numerous genes, transcriptionally induced or suppressed by the 1,25D/VDR complex, are relevant for the efficacy of 1,25(OH)₂D₃ therapy in renal failure. The biological actions resulting from vitamin D-regulation of their expression include: (a) the classical vitamin D-control of calcium homeostasis in bone, intestine, and the kidney; (b) regulation of the rates of 1,25(OH)₂D₃ synthesis and catabolism; (c) suppression of PTH synthesis; (d) modulation of immune responses and (e) suppression of cell proliferation [31]. Several mechanisms have been identified as responsible for the reduced efficacy of vitamin D to control the expression of these genes in renal failure.

It can be easily inferred from the diagram depicting 1,25(OH)₂D₃/VDR-control of gene transcription (Fig. 1) that, in the parathyroid glands, the magnitude of 1,25(OH)₂D₃/VDR inhibition of PTH synthesis and/or parathyroid growth arrest (induction of p21) is determined mainly by the intracellular levels of both 1,25(OH)₂D₃ and the VDR. It is well known that in renal failure, serum 1,25(OH)₂D₃ decreases with the progressive reduction in glomerular filtration rates as renal function deteriorates. In addition to low serum 1,25(OH)₂D₃, the immunohistochemical study by Fukuda and collaborators [9] of nodular and diffuse hyperplastic human parathyroid glands demonstrate a marked reduction in VDR content, particularly in areas of more aggressive nodular growth. One reason for reduced VDR levels is the low serum 1,25(OH)₂D₃ since the sterol increases VDR mRNA levels and protein stability, the latter by preventing VDR degradation by the proteasome complex, thus prolonging the half life of 1,25(OH)₂D₃-bound VDR over that of the apo-VDR (unbound receptor) [33]. The actual contribution of 1,25(OH)₂D₃ levels to VDR content in vivo was demonstrated in uremic rats by a strong correlation between serum 1,25(OH)₂D₃ and parathyroid VDR content [34]. This suggests a potential for 1,25(OH)₂D₃ therapy to correct the reduced parathyroid VDR content in renal failure patients. In fact, the reduced VDR content in the parathyroid glands of uremic rats could be increased to the levels in normal animals by administration of either 1,25(OH)₂D₃ or its analog 22-oxa-calcitriol[34] .

In addition to impaired formation of the 1,25(OH)₂D₃/VDR complex, resulting from the combination of decreases in 1,25(OH)₂D₃ synthesis and parathyroid VDR content, abnormalities in steps downstream from ligand binding to the VDR were demonstrated by the studies shown in figure 2, comparing 1,25(OH)₂D₃ action in peripheral blood monocytes from normal individuals and hemodialysis patients. In the presence of a similar VDR content, the binding of endogenous VDR/RXR complex to DNA is markedy impaired in uremia, leading to an 80% inhibition of the ability of exogenous 1,25(OH)₂D₃ to induce 24-hydroxylase gene transcription. Other factors contribute to impairing 1,25(OH)₂D₃/VDR-control of gene transcription. (a) Reduced RXR. Studies in unilaterally nephrectomized rats demonstrated a reduction in the content of a 50 kDa RXR isoform in cell extracts from the remnant kidney. This decrease in RXR results in a reduction of the binding of the endogenous VDR/RXR heterodimer to the VDRE of the mouse osteopontin promoter. A similar reduction of RXR content in the parathyroid glands in these rats could explain their enhanced serum PTH levels in the absence of hypocalcemia or hypophosphatemia[35]; (b) Accumulation of uremic toxins. Ultrafiltrate from uremic plasma causes a dose dependent inhibition of VDR/RXR binding to VDRE and 1,25(OH)₂D₃/VDR-transactivating function [36]; (c) Increases in parathyroid calreticulin. Calreticulin is a cytosolic protein that binds integrins in the plasma membrane and the DNA-binding domain of nuclear receptors, including the VDR, thus interfering with receptor mediated transactivation. Hypocalcemia, commonly present in renal failure, and caused by both low 1,25(OH)₂D₃ or hyperphosphatemia, enhances nuclear levels of parathyroid calreticulin. In vitro studies demonstrate that increases in calreticulin inhibit VDR/RXR binding to VDRE in a dose dependent manner and totally abolish 1,25(OH)₂D₃ suppression of PTH gene transcription [37]; (d) activation of VDR-unrelated pathways interfering with 1,25(OH)₂D₃ signaling. In human monocyte and macrophages, cytokine activation markedly inhibits 1,25(OH)₂D₃-VDR gene transcription[38]. Activation by the cytokine gamma interferon of its signaling molecule, Stat1, induces physical interactions between Stat1and the DNA binding domain of the VDR, thus impairing VDR/RXR binding to VDRE and gene transcription . The higher levels of inflammatory cytokines after hemodialysis could contribute to vitamin D resistance.

Little is known at present on how renal failure affects the last and most critical step in 1,25(OH)₂D₃/VDR-mediated transactivation or transrepression. Binding of the VDR/RXR heterodimer to the VDRE of genes induced by vitamin D (Fig. 3, top panel) begins the recruitment of co-activator molecules that act synergistically with the VDR to markedly amplify 1,25(OH)₂D₃-transactivation[39-41]. These co-activators possess histone acetyl transferase activity (SRC-1 and CBP/p300) that unfold and expose the DNA. The recruitment of a second complement of transcriptional coactivators, including DRIP205 and TRIP, favors the assembly of the pre-initiation complex and potentiates 1,25D/VDR–induction of transcription. In transcriptional repression, the VDR/RXR complex bound to negative VDRE, that is, of genes that are suppressed by vitamin D, recruits co-repressors of the family of HDAC2 thus preventing DNA exposure and binding of TATA binding protein to initiate transcription (Fig. 3, bottom panel). The complex interactions of the VDR/RXR with VDRE and co-regulators in VDR- transactivation or transrepression of vitamin D-responsive genes suggests that, in uremia, vitamin D resistance may also result from a decreased expression of essential coactivator- or co-repressors molecules or from defective recruitment of these molecules by the VDR. Uremia-induced activation of VDR-unrelated signaling pathways could also interfere the recruitment of co-regulator molecules to the VDR-transcriptional pre-initiation complex.

In view of the numerous inhibitory interactions triggered by uremia, what are the mechanisms responsible for the efficacy of therapy with vitamin D or its less calcemic analogs?. Studies by Takeyama and collaborators [42] demonstrate that the ligand bound to the VDR dictates which co-activator is recruited by the VDR. This selective recruitment of co-activators has double implications for therapy. If a target cell expresses the coactivators required by either 1,25(OH)₂D₃ or its analog, both vitamin D metabolites will elicit similar efficacy. However, if the coactivator required by the analog is limiting or absent in a target tissue, the analog will elicit weaker potency than the parent hormone. This mechanism could certainly contribute to analog selectivity. Furthermore, analog recruitment to the VDR-transcription initiation complex of a co-activator more potent than that recruited by 1,25(OH)₂D₃ could result in higher potency of the analog in eliciting a biological response. In fact, preliminary studies by Takeyama and collaborators (J. Bone Min Res 16:S433, 2001) laboratory demonstrated that analog specific recruitment of co-repressor molecules mediate their differential transrepression potency of the human PTH gene. Extensive research is mandatory before extrapolating these findings in vitro to VDR-transcriptional potency in the parathyroid glands in vivo.

In summary, early therapeutic interventions with 1,25(OH)₂D₃ or its analogs in renal failure results not only in prevention of 1,25(OH)₂D₃ deficiency but also of the reduction in cellular VDR content thus improving formation of the 1,25D-VDR complex. Adequate prevention of hypocalcemia, hyperphosphatemia and the accumulation of uremic toxins will improve VDR/RXR binding to the VDRE of target genes. A better understanding of the role of nuclear co-activator/repressors in VDR-mediated transactivation should help design better VDR-ligands in recruiting the most effective co-regulator molecules thus maximizing vitamin D efficacy in controlling parathyroid hyperfunction.