- Top of page
- Materials and Methods
ABSTRACT: The aim of this study was to evaluate the presence and distribution of retinoid × receptors (RXRs) α, β, and γ in normal, hyperplastic (nodular, basal cell, and atrophic hyperplasia), and carcinomatous human prostates in order to elucidate the relationship among these receptors and the onset and development of prostatic adenocarcinoma. RXRα and RXRγ were immunodetected in all samples of normal, nodular, and basal cell hyperplasia, as well as carcinomatous prostates. In atrophic glands, the expression of both receptors was found in 22.5% of samples. Positive immunostaining for RXRβ was observed in 53.3% of normal prostates, 100% of samples showed basal cell hyperplasia, and were negative in nodular and atrophic hyperplasia. In prostatic adenocarcinoma, only 3 of 25 samples (the 3 diagnosed as well-differentiated) were positive for RXRβ. Results suggest that diminished RXRβ expression might be related to prostate cancer progression and because the responsiveness to retinoic acid treatments depends on the expression of different receptors, it is important to study their expression before therapy.
Retinoids have recently attracted the interest of many researchers in various fields, particularly in developmental biology, dermatology, and oncology. Among retinoids, it is well known that retinoic acid plays a critical role in inhibiting proliferation and in stimulating the differentiation of a wide spectrum of cell types (Strickland and Mahdavi, 1978; Lotan, 1980). Thus, Chopra and Wilkoff (1979) showed that retinoic acid induces in vitro terminal differentiation of mouse prostate cancer cells. In this sense, Dahiya et al (1994) reported that retinoids either reverse or prevent chemically induced premalignant changes in rodent prostate by antiangiogenesis, antiproliferative action, or both. Furthermore, it has been reported that retinoids could inhibit tumor formation and metastasis (Edwards, 1992). Considering that malignant transformation of normal epithelium results from the loss or disruption of normal differentiation mechanisms, retinoids have been used for prevention and treatment of various epithelial cancers (Lotan and Nicolson, 1977; Lippman et al, 1987; Lotan et al, 1990). Retinoids have also shown antiproliferative and differentiation-inducing activity in cultured cell lines (Blutt et al, 1997).
Studies on the molecular mechanisms of retinoid action have revealed that retinoic acid binds to two receptor types: retinoic acid receptors (RARs) and retinoid × receptors (RXRs; Petrovick et al, 1987; Mangelsdorf et al, 1992). Both receptor types belong to the steroid/thyroid hormone nuclear receptor superfamily, and are characterized by their ligand- and DNA-binding abilities, and also by their possible dimerization partners (Kastner et al, 1994; Mangelsdorf and Evans, 1995). Each class of receptor is composed of three gene products (RAR α, β, and γ; and RXR α, β, and γ), the transcription of which results in several isoforms as a result of the action of different promoters and messenger RNA (mRNA) splicing (Kastner et al, 1994; Brocard et al, 1996). Two forms of retinoic acid, named all-trans-retinoic acid (ATRA) and 9-cis-retinoic acid (9-cis-RA), can bind to RARs, but only 9-cis-RA is able to bind to RXRs (Mangelsdorf et al, 1992).
In addition to the occurrence of different ligands and receptors, the complexity of retinoid signaling is increased by the possible formation of different homodimer and heterodimer receptors. These dimers can bind to different hormone response elements (HREs) in the promoters of certain genes, and act as transcription factors (Kastner et al, 1994; Mangelsdorf and Evans, 1995). In general, when an HRE is bound to a nuclear hormone receptor, this may either activate or repress the transcription, depending on the presence of ligand, cell type, promoter, response element, or other signals (Vos et al, 1997).
Few studies have examined the expression of RXRs in the human prostate. Using immunohistochemistry, Kikugawa et al (2000) detected much more expression of RXRα and RXRγ than of RXRβ in human prostatic adenocarcinoma cells. In addition, Lotan et al (2000) found RXRα and RXRγ mRNA expression in normal and carcinomatous human prostates. However, these authors observed that the intensity of the in situ hybridization signal was much weaker for RXRβ than for the other receptor probes.
Further studies are needed to investigate the possible role of these receptors in the physiological behavior of human prostatic cells. Thus, the aim of this study was to evaluate the presence and distribution of RXR α, β, and γ in normal, hyperplastic and carcinomatous human prostates, using immunohistochemistry and Western blot analysis, in order to elucidate the relationship among these receptors and the onset and development of prostatic adenocarcinoma.
- Top of page
- Materials and Methods
Retinoic acid effects are mediated by specific RAR and RXR subtypes and modulated by binding proteins at the tissue level (Gudas et al, 1994; Guiguere, 1994; Hong and Itri, 1994). These receptors may play a role in controlling cell proliferation and apoptosis, and the aberrant expression of one or more RARs and RXRs might result in abrogated retinoid signaling and greater cell transformation in several cancer types, including prostate, breast, and lung (Liu et al, 1996; Xu et al, 1997; Campbell et al, 1998). The identification of a molecular alteration in the RARα gene in human acute promyelocytic leukemia and the report of a lower expression of RARβ in lung cancer (de The et al, 1990; Gebert et al, 1991) agree with this hypothesis.
Gyftopoulos et al (2000) studied the distribution of RARα in neoplastic and nonneoplastic human prostate. They observed a low expression of this receptor in hyperplastic prostate tissue, in which 3 of 24 cases were completely negative. In contrast, RARα-positive cells were present in all cases of prostatic carcinoma.
The expression of retinoid receptors may depend on the level of retinoids in prostate tissue. Pasquali et al (1996) reported that prostate cancer tissues have five to eight times less retinoic acid than normal prostate or those exhibiting BPH. In primary cultures of prostate cells, ATRA is implicated in the control of growth and induction of apoptosis, and these effects are mediated by specific RAR subtypes, RARα and RARβ. In these cells, the expression of messenger RARβ was increased, whereas bcl-2 protein levels were decreased (Pasquali et al, 1999). However, in a clinical trial, Trump et al (1997) concluded that ATRA was not active in patients with hormone refractory prostate cancer. These authors proposed that the failure of this agent in hormone refractory prostate cancer might be related to a failure of drug delivery and associated with enhanced mechanisms of ATRA clearance, which occur within a few days of beginning ATRA treatment.
The synthetic retinoid N-(4-hydroxyphenyl)retinamide (4-HPR) has been shown to induce apoptosis in various malignant cells including human prostate carcinoma cell lines; this induction is mediated by nuclear RARs; by increasing the reactive oxygen species activity; expression of p53, p21, and c-jun genes; and decreasing the expression of the c-myc gene (Sun et al, 1999). However, in patients treated with 4-HPR for 28 days before radical prostatectomy, this synthetic retinoid was ineffective because retinoic acid concentrations in serum and in prostates were not significantly altered (Thaller et al, 2000).
To bind DNA, RARs require heterodimerization with RXRs; the latter can act as homodimer or heterodimer partners of a number of nuclear receptors (Mangelsdorf and Evans, 1995). Because of the role of RXRs as pivotal mediators in several signaling pathways, the aim of this work was to study by immunohistochemistry the presence and distribution of RXRs in normal, hyperplastic, and carcinomatous prostates.
In this study we found both nuclear and cytoplasmic locations for the three types of receptors in some cases. It is known that retinoid receptors belong to the class of receptors (ie, thyroid hormone receptors, retinoid receptors, PPAR, etc) that are constitutively found in the nucleus, regardless of whether the ligand is bound or not bound to the receptor (Reichrath et al, 1997). However, some studies suggest that the intracellular location of retinoic acid nuclear receptors may be regulated by retinoic acid and protein kinase C (Tahayato et al, 1993; Weis et al, 1994; Akmal et al, 1998). In this sense, Akmal et al (1998) reported that depletion of vitamin A leads to a change in the location of RARα from the nucleus to the cytoplasm in rat germ cells. Moreover, down-regulation of protein kinase C, a molecule that is not a ligand for these receptors, is able to increase the cytoplasmic location of RARα in COS-7 cells (Tahayato et al, 1993). Also, Liu et al (2000) encountered RXRα in the cytoplasm and nucleus of LAPC-4 cells and PC3 cells and, after the cells were treated with the RXR ligand LG1069, cytoplasmic RXRα translocated to the nucleus. Moreover, it has been shown that some nuclear receptors (steroids receptors) are found as an inactive cytoplasmic form in a complex with heat shock proteins (Pratt and Toft, 1997). Thus, it is possible that either inactive retinoic acid nuclear receptors were forming a similar complex together with heat shock proteins or they are located in the cytoplasm in ligand absence.
We observed that expression of RXRα and RXRγ decreased in atrophic hyperplasia in comparison to normal prostatic tissue. However, in nodular and basal cell hyperplasia, the expression was maintained. In prostatic adenocarcinoma, RXRα and RXRγ immunoexpression did not suffer variation compared with normal tissue. The absence of changes in the expression of these receptors could suggest that they do not play a significant role in prostate carcinogenesis. This hypothesis is in agreement with the weak inhibition of prostate cancer cell proliferation induced by RXRα selective synthetic ligands (Vos et al, 1997).
In situ hybridization studies by Lotan et al (2000) reported a selective and significant reduction of RXRβ mRNA in prostate cancer and in normal prostate tissue adjacent to carcinoma. Also, Kikuwaga et al (2000) detected a lower expression of RXRβ protein in prostatic cancer tissue. We have also observed a reduction of RXRβ (it was expressed only in three cases classified as well-differentiated carcinomas of 25 prostatic carcinoma samples); however, Kikuwaga et al (2000) observed RXRβ expression in moderate and poorly differentiated carcinomas.
This reduced expression of RXRβ could be involved in the onset of prostate carcinogenesis and has also been related (in advanced disease) to the ineffectiveness of retinoic acid treatment in some patients with androgen-in-dependent or -dependent prostate cancer (Trump et al, 1997; Kelly et al, 2000), but this hypothesis remains to be investigated.
We have also observed that RXRβ was expressed in 8 of 15 (53.3%) normal prostates and in 32% of the hyperplastic prostates studied, all of them diagnosed as basal cell hyperplasia. These data, together with those observed for the other RXRs, lead us to suggest that patients with basal cell hyperplasia are potential targets for receiving treatment with retinoic acid due to the presence of the three types of receptors. However, in those patients who suffer atrophic and nodular hyperplasia, it is probable that this treatment will be unsuccessful because of the low amount of the three types of RXRs (and a complete absence of RXRβ).
At present, many of the markers in use are not useful for a precise and early diagnosis of prostatic adenocarcinoma. Qiu et al (1999) have proposed that the loss of RARβ expression is an early event associated with esophageal carcinogenesis and the status of squamous differentiation. They have also suggested that loss of this receptor is a common event across cancers of different sites and etiologies.
We propose that the study of RXRβ expression, together with RARs in prostatic tissue, could be considered in the near future as key factors in the responsiveness to retinoic acid-based therapies. However, further studies are needed to elucidate the mechanisms of these receptors in order to improve their usefulness in prostate cancer.