Inhibition of MG-63 cell cycle progression by synthetic vitamin D₃ analogs mediated by p27, Cdk2, cyclin E, and the retinoblastoma protein

Abstract

Progression through eukaryotic cell division cycle is regulated by synergistic activities of both positive and negative regulatory factors. The active form of vitamin D₃ (1α,25(OH)₂D₃, 1,25D) and a number of its synthetic analogs have been shown to arrest cells in the G₁ phase of the cell cycle. In the present study, 1α,25(OH)₂D₃ and the analogs KH1060, EB1089, and CB1093 were used to study the mechanism of the cell cycle arrest and to compare the effectiveness of these compounds in human MG-63 osteosarcoma cells. The 20-epi analogs KH1060 and CB1093, as well as the 20-normal analog EB1089, were found to be more potent than 1α,25(OH)₂D₃ in inhibiting cell proliferation and arresting the MG-63 cells in the G₁ phase. These analogs were more active than 1α,25(OH)₂D₃ in increasing the cyclin dependent kinase inhibitor p27 protein levels (approximately 2.3–2.5-fold compared to 1α,25(OH)₂D₃) by both increasing its formation and decreasing its degradation rate. The increased p27 formation was accompanied by stabilization of binding of nuclear proteins to the Sp1+NF-Y responsive promoter region of the p27 gene. The increase in p27 protein levels and the simultaneous decrease in cyclin E protein levels was accompanied by decreased Cdk2 kinase activity, retinoblastoma (Rb) protein hypophosphorylation and, finally, cell cycle arrest in the G₁ phase. In summary, the analogs KH1060, EB1089, and CB1093 keep Rb protein in its growth-suppressing, hypophosphorylated form and prevent cell cycle progression through the restriction point. Therefore, these synthetic vitamin D₃ analogs may be potential candidates for treating diseases, where cell cycle regulation is needed.

Introduction

Most of the biological actions of 1α,25-dihydroxyvitamin D₃ (1α,25(OH)₂D₃, 1,25D) are mediated through the nuclear VDR, which is a member of structurally related steroid/thyroid hormone superfamily of ligand-dependent transcription factors [1]. VDR forms heterodimers with another member of this family, the retinoid X receptor, and regulates gene expression positively or negatively through binding to the vitamin D response elements (VDREs) in promoter regions of target genes. In addition to its role in mineral homeostasis, the active 1α,25(OH)₂D₃ affects growth and differentiation of different cell types, e.g., human breast, brain, colon, prostate, and skin cells [2], [3], [4], [5], [6], [7]. The potentially beneficial use of the hormone in treatment of hyperproliferative diseases is, however, compromised by hypercalcemia and hypercalciuria developing at therapeutic doses. This has evoked an increased interest in synthesis and evaluation of new 1α,25(OH)₂D₃ analogs that would retain the properties of the parent compound on cellular proliferation and differentiation while having reduced calcemic activity [2], [8], [9]. New analogs could be potentially useful in the treatment of, e.g., several types of cancer and skin disorders [2], [8]. One interesting group among the side chain analogs of 1α,25(OH)₂D₃ is the 20-epi analogs. The prolonged effects of these analogs quite likely result from their resistance against cell metabolism further leading to increased stabilization of the analog–receptor complex [10], [11].

Inhibition of cell proliferation is tightly linked to mechanisms that regulate cell cycle progression. Knowledge of mechanisms by which cell growth is arrested after exposure to 1α,25(OH)₂D₃ and its analogs is quite fragmentary and partly unclear. In most cases, the inhibition of human cell growth in response to 1α,25(OH)₂D₃ involves cell cycle arrest in the G₁ phase [12], [13], [14]. In dividing mammalian cells, cell cycle progression is regulated by sequential formation, activation, and subsequent inactivation of a series of cyclin dependent kinase (Cdk)–cyclin complexes. Cdk activities are regulated by phosphorylations and cyclin binding, as well as by the action of specific Cdk inhibitor (CKI) proteins ([15], [16], [17], [18] and references therein). In mammalian cells, complexes such as cyclin D/Cdk4,6 and cyclin E-Cdk2 are catalytically active during the G₁ phase and the main function of these complexes is phosphorylation of the retinoblastoma (Rb) protein. Indeed, many of these events have been identified during 1α,25(OH)₂D₃-induced inhibition of human cell proliferation. In a variety of human cells, the 1α,25(OH)₂D₃-induced cell cycle arrest is mediated by CKIs including p21 (Cip1/Waf1) and p27 (Kip1) [14], [19], [20], [21]. There are also studies showing that 1α,25(OH)₂D₃ signaling is mediated by the cell cycle regulatory proteins Rb and p57 (Kip2) and by the induction of apoptosis [22], [23], [24]. In addition to the inhibitory role of the Cip/Kip family proteins on cell cycle progression p21, and especially p27 can also be found in a stimulatory role in promoting activation of cyclin D₁-Cdk4,6 in proliferative cells [18], [25].

The 20-epi analogs, especially KH1060 and CB1093, as well as EB1089, an analog with a natural side chain orientation at carbon-20, have already shown their therapeutical potencies as antileukemic compounds and inhibitors of tumor cell growth as well as differentiation-inducing agents [23], [26], [27], [28], [29]. The present study was undertaken, first, to analyze several candidates of cell cycle controlling proteins possibly responsible for the G₁ arrest caused by 1α,25(OH)₂D₃ in human bone cells. Second, we were also interested in comparing these effects with those of selected 1α,25(OH)₂D₃ analogs. We used human MG-63 osteosarcoma cells which share many features of normal human osteoblasts and, therefore, should provide useful information relevant to osteoblastic proliferative events. Our results indicated that, in the osteoblastic cells, 1α,25(OH)₂D₃ signaling appears to be mediated through the Rb pathway, as cells treated with 1α,25(OH)₂D₃ showed decreased amounts of hyperphosphorylated Rb protein. Consistent with its effects on Rb protein, 1α,25(OH)₂D₃ reduced Cdk2 activity through increased p27, but not p21 protein levels. The selected analogs, KH1060, EB1089, and CB1093, were clearly more potent than the parent compound in their effects on these cell cycle control mechanisms.