The Role of RASSF1A in Uveal Melanoma

Abstract

Purpose.: RASSF1A inactivation in uveal melanoma (UM) is common and methylation-induced. We investigated the effect of RASSF1A re-expression on the UM phenotype in vivo and in vitro.

Methods.: The phenotypic effect of methylation-induced inactivation of RASSF1A in UM was explored using a stable RASSF1A-expressing UM-15 clone. RASSF1A expression was assessed using QRT-PCR. Proliferation was evaluated in vitro using MTT assays. Additionally, athymic NOD/SCID mice were injected subcutaneously or intraocularly with RASSF1A-expressing and –non-expressing UM-15 clones, and euthanized when tumors reached a volume of 1500 mm³, or at 56 or 46 days, respectively. Tumor tissues, eyes, and livers were analyzed histologically.

Results.: In vitro analysis confirmed the lack of RASSF1A expression and full methylation of the RASSF1A promoter region in the UM-15 cell line, which was reversible following treatment with 5-Aza-2-deoxycytidine. Cells expressing exogenous RASSF1A showed slower proliferation than controls and regained sensitivity to cisplatin. Compared to mice injected with control cells, mice treated with UM-15 cells expressing exogenous RASSF1A did not acquire intraocular tumors, and their subcutaneous tumors were relatively delayed and small. Neither group had liver metastases.

Conclusions.: UM cells reduced tumorigenicity in the presence of activated RASSF1A. RASSF1A apparently has an important role in the development of UM, and its reactivation might be applied in the development of new treatments.

Introduction

Uveal melanoma (UM) is the most common form of primary eye cancer in adults, with an annual incidence of 6–7 cases per million per year. ¹ It accounts for 80% of all noncutaneous melanomas and 13% of all deaths caused by melanoma. The tumor carries up to 50% 5-year mortality from metastasis. ²  

Uveal and cutaneous melanomas originate from the same precursor cell, the melanocyte, which migrates from the neural crest to the respective sites during embryonic development. ^3,4 However, their biological and clinical behaviors differ. ² Additionally, although alterations of chromosomes 1 and 6 are common to both tumors, aberrations, such as monosomy of chromosome 3 and gain of 8q, in addition to other aberrations, typically are found only in UM. ^5–7 Following the identification of late genetic events in UM progression and metastasis, such as loss of chromosome 3, ⁸ researchers directed attention to the early initiating events leading to malignant transformation and development of a clinically detectable tumor. ⁹ Studies revealed that a mutation in the alpha subunit of the heterotrimeric G gene (GNAQ) was present in almost half of all UMs examined, ^9–15 and that UM metastatic spread was related to mutations in the BRCA associated protein 1 (BAP1) gene on chromosome 3. ¹⁶ However, more data were needed to clarify the still poorly characterized tumorigenesis of UM. ⁹  

Aberrant promoter hypermethylation of CpG islands is thought to have an important role in the inactivation of tumor suppressor genes (TSGs) in cancer. ^17,18 In a study of 86 metastatic specimens of cutaneous melanoma, Hoon et al. identified 4 TSGs that frequently were inactivated: retinoic acid receptor-beta2 (RAR-beta2, 70%), RAS association domain family protein 1A (RASSF1A, 57%), O6-methylguanine DNA methyltransferase (MGMT, 34%), and death-associated protein kinase (DAPK, 19%). ¹⁹ Hypermethylation of MGMT, RASSF1A, and DAPK was significantly lower in primary melanomas than in metastatic melanomas, whereas the rate of RAR-beta2 hypermethylation was 70% in both types. Prompted by these data, we sought, in an earlier study, to elucidate the role of epigenetic events in UM by investigating the methylation status of these genes and the newly described methylator phenotype TSG panel. ²⁰ The results showed that RAR-beta2, MGMT, and DAPK promoter hypermethylation was uncommon in UM, and all samples were negative for the CpG methylator phenotype. Similar findings also were reported by another group. ²¹ However, promoter hypermethylation for the RASSF1A gene was present in a significant proportion of the samples, in accordance with the study of Maat et al. ²²  

Besides cutaneous melanoma, the RASSF1A promoter gene is known to be extremely common in cancers of the breast, head and neck, and lung. ²³ In general, an epigenetic mechanism underlies its aberrant methylation, as opposed to somatic inactivating mutations, which occur rarely. ²³ Furthermore, RASSF1A lies on the 3p21.3 region of chromosome 3, which frequently is rearranged in UM, making it a candidate TSG in this tumor. ^24,25 The RASSF1A gene contains two CpG islands that are susceptible to inactivating methylation, spanning the promoter and the first exon gene regions. 

The aim of the our study was to examine the influence of exogenous expression of RASSF1A on the UM cell phenotype in vitro and in vivo. 

Methods

The study protocol was approved by the national and institutional review boards. The study was conducted in adherence with the Declaration of Helsinki. 

Cell Lines

The UM-15 cell line and all 20 paraffin-embedded UM samples were classified and provided by the Laboratory of Ophthalmic Pathology, Hadassah-Hebrew University Medical Center, Jerusalem. 

[Karyotype for UM-15 cell line: 53 ∼ 56 < 2n > ,XY,+X,+der(1;13)(p10;q10),+2,+der(3)t(3;5)(p13;q15),i(5)(p10)x2,+i(5)(q10),add(8)(p22)x2,der(11)t(1;11)(p15;q10),dup(14)(q11.2q22),+del(16)(q10),der(19)t(3;19)(p13;p13.3),der(20)t(17;20)(q21;q13.3),+22,+add(22)(q13)(cp16)]. 

Immunohistochemistry Staining for RASSF1A

UM-15 cells were seeded on 22 × 22 mm glass coverslips in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum and grown to 50−60% confluence. The cells then were transfected with 1 μg of each expression plasmid and incubated for 14−18 hours. Following incubation, the cells were fixed in 3% paraformaldehyde solution (PFA) for 20 minutes at room temperature and permeabilized using 0.5% Triton X-100 solution for 10 minutes. Coverslips were blocked with 3% horse serum for 30 minutes at room temperature. They then were incubated with primary mouse monoclonal [3F3] anti-RASSF1A antibody (ab23950; Abcam, Cambridge, UK) in a humidifying chamber for 1 hour at room temperature, and stained with Cy5- and fluorescein isothiocyanate (FITC)-conjugated secondary antibodies (1:100; Jackson Immunoresearch, West Grove, PA) for 1 hour at room temperature. Histological sections of kidney tissue were used as a positive control for RASSF1A expression. 

Fluorescence In Situ Hybridization (FISH) for Monosomy 3

Using standard techniques, FISH was performed with pericentromeric chromosome enumeration probes (CEP3) labeled with SpectrumOrange (Abbott Molecular, Abbott Park, IL). The UM-15 cell line was counterstained with 4′,6-diamidino-2-phenylindole Antifade. 

DNA Extraction

DNA was extracted from microdissected tumor material and the UM-15 cell line, as described previously. ²⁰ In brief, 10-mm section slides stained with hematoxylin and eosin were reviewed by a pathologist. Thereafter, areas containing only tumor were separated by microdissection from 5 consecutive 10-mm unstained paraffin sections of each block using a No. 11 surgical blade. Following deparaffinization, the microdissected tissues were incubated overnight in 1% sodium dodecyl sulfate (SDS) and Proteinase K 0.5 mg/ml. DNA was purified by phenol-chloroform extraction and ethanol precipitation, and dissolved in 50 mL of distilled water. 

Bisulfite Modification

Methylation detection using the bisulfite treatment method has been described previously. ²⁰ In brief, 1–2 μg of genomic DNA were denatured in sodium hydroxide (NaOH, 0.3 M) for 15 minutes at 37°C. Cytosines were sulfonated in sodium bisulfite 3.12 M (Sigma, St. Louis, MO), and hydroquinone 5 mM (Sigma) for 16 hours at 50°C. The DNA samples then were purified (Wizard DNA Clean-Up System; Promega, Madison, WI), desulfonated in NaOH (0.3 M), precipitated in ethanol, and suspended in water. 

Methylation-Specific PCR (MSP)

The primers and probes used in the MSP study are shown in Table 1. The 3′DNA was sequenced by mixing bisulfite-treated DNA (100 ng) with 50 pmoles of each primer in a reaction buffer (50 μL) containing dNTPs (200 μM each) and AmpliTaq Gold Taq polymerase (Applied Biosystems, Inc., Foster City, CA) at 95°C for 10 minutes followed by 94°C for 30 seconds, 55°C for 30 seconds, and 74°C for 1 minute, for 35 cycles. The PCR products were analyzed on 1.5% agarose gel. 

Table 1. 

 

List of All Primers and Probes Used in This Study

	Forward 5′-3′	Probe 5′-3′
MSP primers:
RASSF1A_UM  (unmethylated)	CCC ATA CTT CAA CTT TAA AC
RASSF1A_M  (methylated)	GCG TTG AAG TCG GGG TTC	6FAM-ACA AAC GCG AAC CGA ACG AAA CCA-TAMRA
β-Actin (MSP)	TGG TGA TGG AGG AGG TTT AGT AAG T	ACC ACC ACC CAA CAC ACA ATA ACA AAC ACA
Expression primers:
RASSF1A	GCA GTG CGC GCA TTG CAA GT
GAPDH	ACCACAGTCCATGCCATCAC

Table 1. 

 

Extended

Reverse 5′-3′
GGT GTT GAA GTT GGG GTT TG
CCC GTA CTT CGC TAA CTT TAA ACG
AAC CAA TAA AAC CTA CTC CTC CCT TAA
AGG CTC GTC CAC GTT CGT GT
TCC ACC ACC CTG TTG CTG TA

Real-Time Quantitative Methylation-Specific PCR (QMS-PCR)

The sodium-bisulfite-treated genomic DNA was analyzed with the ABI Prism 7900 Sequence Detection System (Applied Biosystems), as described previously. ²⁰ Amplifications were carried out in 96-well plates, each containing patient samples and water blanks, and positive and negative controls. For the positive controls, we used DNA extracted from leukocytes of healthy individuals that were methylated in vitro with SssI methyltransferase (New England Biolabs Inc., Beverly, MA); serial dilutions of this DNA were used to construct the standard curves on each plate. The relative degree of methylation of each DNA sample was defined as the ratio between the value of the gene of interest and the value of the internal reference gene (gene of interest:reference gene × 1000), as described previously. ²⁶ For the negative control, we used DNA extracted from a healthy individual. The primers used in the QMS-PCR study are shown in Table 1. The 3′DNA was sequenced by mixing bisulfite-treated DNA (100 ng) with 50 pmoles of each primer in a reaction buffer (50 μL) containing dNTPs (200 μM each) and AmpliTaq Gold Taq polymerase (Applied Biosystems) at 95°C for 10 minutes followed by 94°C for 30 seconds, 55°C for 30 seconds, and 74°C for 1 minutes, for 35 cycles. The PCR products were separated on 1.5% agarose gel. 

cDNA Preparation

Total RNA was isolated using TRIzol reagent (Invitrogen, Life Technologies, Carlsbad, CA), according to the manufacturer's protocol. Before cDNA synthesis, RNA was treated with DNase I. The cDNA synthesis was carried using a random hexamer (Amersham Biosciences, Buckinghamshire, UK) and Moloney murine leukemia virus (M-MLV)-reverse transcriptase (Promega). 

Analysis of mRNA Expression of RASSF1A by Quantitative Real-Time Polymerase Chain Reaction (QRT-PCR)

Two-stage QRT-PCR was used to evaluate RASSF1A expression. RASSF1A cDNA input levels were normalized against human glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and analyses were done with the Sequence Detection System (Prism 7000; Applied Biosystems). The primers for RASSF1A and GAPDH are shown in Table 1. 

Reactions were performed in a 20-μL volume containing 4 μL cDNA, 0.5 μM each of the forward and reverse primers, and buffer included in the Master Mix (SYBR^R Green I; Applied Biosystems). Cycling conditions consisted of an initial denaturation step of 95°C for 10 minutes followed by 40 cycles of 1-minute denaturation at 95°C and 1 minute of annealing and extension at 60°C. Duplicate QRT-PCR reactions were performed for each gene to minimize individual tube variability, and an average was taken for each time point. 

Re-Expression of RASSF1A by 5-Aza-2-Deoxycytidine Treatment

UM-15 cells were treated by different concentrations (1, 3, 5, 10 μM) of the demethylating agent 5-Aza-2′-deoxycytidine (5-Aza-CdR) for 4 days, 5 × 10⁶ cells for each concentration. The cells were fed with fresh medium supplemented daily with the drug. Control cells received mock treatment (no treatment). At the end of the treatment period, the medium was removed, and the expression pattern of RASSF1A was tested. The methylation status was analyzed by QRT-PCR, as described above. 

Construction of RASSF1A-Expressing UM-15 Clone

The construction of a stable RASSF1A clone was performed as described previously. ^27,28 In brief, a UM-15 cell line showing no RASSF1A expression was used for the transfection study. Human wild-type RASSF1A cDNA (kindly provided by Reinhard Dammann, Institute for Human Genetics, Martin Luther University, Halle-Wittenberg, Halle/Saale, Germany) was cloned into pcDNA3.1. The RASSF1A-containing plasmid and the empty vector were introduced into the UM-15 cells using Lipofectamine 2000 (Invitrogen). A stable clone of RASSF1A transfectants was obtained by sustained G418 selection (50 μg/mL). The expression of exogenous RASSF1A was confirmed by quantitative QRT-PCR. The same cell line transfected with empty pcDNA3.1 vector served as the control. 

Cell Proliferation Assays

The proliferation rate of melanoma cells treated with cisplatin was determined by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma). UM cell lines were cultured in 96-well plates (1X10⁶ cells/well) with and without cisplatin (0.25 or 0.5 μg/mL) for 24 hours (N = 3 for each cell line). The cells then were treated with a 20 μL aliquot of MTT (5 mg/mL) for 4 hours and solubilized in 200 μL of dimethyl sulfoxide (DMSO). Absorbance was measured spectrophotometrically at 570 nm. The inhibition rate (IR) of cell growth was calculated by the following formula: IR = 1 – (value in experimental groups/value in control group) × 100%. 

Mice

We used 59 athymic NOC/SCID mice 5–6 weeks old (44 female, 15 male) in the study. The mice were purchased from Jackson Laboratories (Bar Harbor, ME) and housed in a barrier facility (kindly provided by Nadir Askenasy, Frankel Laboratory, Center for Stem Cell Research, Schneider Children's Medical Center of Israel, Petach Tikva, Israel) under a 14-hour light/10-hour dark cycle with standard chow and water ad libitum. The mice were managed in accordance with the NIH Guidelines on Laboratory Animal Welfare and the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmology and Vision Research. All protocols were approved and monitored by the Animal Care Committee of Rabin Medical Center. 

Subcutaneous Cell Transplantation

Of the mice studied 18 were deeply anesthetized with ketamine 80 mg/kg and Xylazine 4 mg/kg (Kepro B.V., Barneveld, The Netherlands) and divided into 3 equal groups, each of which received 1 × 10⁶ cells, as follows: (1) RASSF1A-pcDNA3.1-transfected UM-15 cells (RASSF1A expression was validated before injection), (2) parental UM-15 cells transfected with empty pcDNA3.1 vector (control 1), and (3) non-transfected parental UM-15 cells (control 2). 

The resulting tumors were measured with calipers 3 times per week. On the basis of these values, tumor volume was calculated using the modified ellipsoidal formula: 1/2(length × width²). ²⁹ The mice were euthanized 56 days after cell transplantation, and the liver tissue, eyes, and subcutaneous tumors were analyzed histologically and molecularly for RASSF1A mRNA expression. This experiment was done in duplicate. 

Intraocular Cell Transplantation

As described previously, UM-15 cells with (group 1) or without (groups 2, 3) exogenous RASSF1A expression were transplanted intravitreally to the right eye of athymic NOD/SCID mice (n = 4 in each group). The left eye served as a control. Briefly, mice were placed under anesthesia, and the cells (5 × 10⁵ cells/3.0 μL) were injected with a syringe (Hamilton, Reno, NV) fitted with a 27-gauge needle. After 46 days, the mice were euthanized, and histologic analysis was performed on eye and liver (for metastasis) tissues under a light microscope. This experiment was done in duplicate. 

Statistical Analysis

Group differences in subcutaneous tumor size were analyzed by a square-root transformation and then a log transformation of the data to achieve a Gaussian distribution. A P value of less than 0.05 was considered statistically significant. 

Results

In Vitro Experiments

RASSF1A Methylation Status and Expression in UM-15 Cell Line (Figs. 1–3).

Promoter hypermethylation of the RASSF1A gene was detected in the UM-15 cell line (Fig. 1B). In addition, analysis of the promoter methylation status in 20 UM samples revealed that all were methylated. 

There was no mRNA transcription of the RASSF1A gene in the UM-15 cells. RASSF1A re-expressed after treatment with 5-Aza-CdR for 4 days (Figs. 1A, 1C). On analysis by QRT-PCR, the level of RASSF1A expression correlated with the increasing doses of 5-Aza-CdR demethylating agent up to 5 μM (Figs. 1C, 1D, 3A). Accordingly, there was no expression of the RASSF1A protein in the UM-15 cells, and it re-expressed after treatment with 5-Aza-CdR for 4 days (Fig. 2). 

Figure 1. 

 

RASSF1A expression. (A) RASSF1A expression is detected only in the positive control (PC, RNA extracted from peripheral blood of a healthy individual) and not in the UM-15 cell line (M-15). (B) RASSF1A methylation is detected in the UM-15 cell line and the positive control (PC, SSSI-treated DNA) but not in the negative control (NC, DNA from a healthy individual). (C) Following RASSF1A reactivation with increasing concentrations of 5-Aza-CdR, RASSF1A expression increases accordingly. 5-aza, 5-Aza-2′-deoxycytidine; 5-Aza-CdR, mock-no treatment, without the active agent 5-aza; GAPDH, reference gene, glyceraldehyde-3-phosphate dehydrogenase. (D) QRT-PCR results showing the relative expression levels of RASSF1A following increasing doses of 5-Aza-2′-deoxycytidine (5-Aza-CdR) treatment in the UM 15 cell line.

Figure 2. 

 

Immunohistochemistry staining for RASSF1A protein in the UM-15 cell line before and after treatment with 5-Aza-CdR. (A) Human kidney tissue was used as positive control staining for RASSF1A (red staining). (B) Untransfected UM-15 cell line showed negative staining for RASSF1A. (C) Following the addition of the demethylating agent 5-Aza-CdR, immunostaining of UM-15 cell line is positive (red staining) in all cells.

Figure 3. 

 

RASSF1A expression in cell lines and subcutaneous tumors. (A) RASSF1A is expressed (24 Ct cycles) in exogenous cell line transfected with RASSF1A containing plasmid; RASSF1A is not expressed (<34 Ct cycles) in the cell line transfected with empty vector. GAPDH served as the reference gene for both. (B) Same results for subcutaneous tumors originating from the RASSF1A-expressing cell line (24 Ct cycles); negative results for tumors originating from the cell line transfected with empty vector.

UM Cell-Line Phenotype.

The provided karyotype of the UM-15 cell line did not show monosomy of chromosome 3. We confirmed this observation using FISH analysis of chromosome 3 copy number. To explore the effect of RASSF1A re-expression on the UM-15 cell-line phenotype, cells were transfected with a plasmid containing the ORF sequence of human RASSF1A. The stable clone was maintained in complete medium supplemented with G418. Total RNA was isolated and subjected to QRT-PCR. The results confirmed that the clone was expressing the transfected RASSF1A gene (Fig. 3). The expression of RASSF1A in cell lines transfected with the RASSF1A-containing plasmid was greater by 21,076-fold than in the original UM-15 cell line and greater by 30,083-fold than in the empty plasmid transfected cell line. 

Cell Proliferation and Cisplatin Cytotoxicity.

Cell proliferation was estimated by comparing the initial and final amount of cells. The cells with exogenous expression of RASSF1A had a slower rate of proliferation than both the nontransfected cells and the cells transfected with empty vector plasmid. Given earlier findings of the relative resistance of UM cells (as well as skin melanoma) to cisplatin treatment in vivo, ^30,31 we investigated the cisplatin resistance of UM cells expressing exogenous RASSF1A. On MTT assay, the UM-15 nontransfected (original) cells and the cells containing empty vector were more resistant to cisplatin than the cells expressing RASSF1A, with twice the proliferation rate (85% and 84% vs. 69% for RASSF1A-expressing cells, Fig. 4). 

Figure 4. 

 

MTT assay results showing decreased proliferation rate in the presence of cisplatinum for the RASSF1A-expressing cell line compared to the non-RASSF1A-expressing (UM-15 and empty vector) cells.

In Vivo Studies

Subcutaneous UM-15 Cell Transplantation.

Tumor Size.

Subcutaneous tumors developed in all 3 groups (n = 17). Tumor growth was slower in mice injected with the RASSF1A-expressing cell line (n = 5, detected by day 39) than in the mice injected with parental UM (n = 6, detected by day 32) or empty vector (n = 6, detected by day 28, Fig. 5A). The tumors in the last group, which grew more slowly, also were smaller (NS). Macroscopically, all of the tumors in the parental UM and empty vector groups were elevated, smooth, elastic, and darkly pigmented. Two of the 5 tumors in the RASSF1A-expressing cells were amelanotic. 

Figure 5. 

 

Subcutaneous tumor size (mm) following injection of 3 cell lines: RASSF1A-expressing cells, cells transfected with empty vector, and original UM-15. (A) All mice injected subcutaneously had tumors (n = 17). Tumors were detected by day 28 in all mice injected with empty vector (n = 6), by day 32 in mice injected with parental UM (n = 6), and by day 39 in mice injected with RASSF1A-expressing cell line (n = 5). Tumor growth was slower in mice injected with RASSF1A-expressing cells than in the other groups (NS). (B) Histologic findings of melanoma epithelioid cells with a low mitotic index from a subcutaneous tumor originated from original (non RASSF1A-expressing) UM cells. (C) RASSF1A positive immunostaining of subcutaneous tumor originated from RASSF1A-expressing UM cells (n = 2, red, ×40).

Tumor Histology.

Histologic study revealed melanoma epithelioid cells with a low mitotic index (Fig. 5B). When exogenous expressing UM-15 cell line was injected, the tumor expressed RASSF1A (Fig. 5C, n = 2). No metastases were detected in the liver of any of the mice in the study. 

RASSF1A Expression Level in Subcutaneous Tumors.

QRT-PCR analysis showed no RASSF1A expression in tumor tissue from mice injected with UM-15 paternal cells (n = 6) or empty vector (n = 6). As expected, RASSF1A was expressed in all subcutaneous tumors that grew from RASSF1A-expressing UM-15 cells (n = 5, Figs. 3B, 5C). 

Intraocular UM Cell Transplantation.

Tumor Evaluation.

At 46 days after cell transplantation, mice injected with the stable RASSF1A-expressing clone (n = 4) showed no tumor on fundus examination or transscleral light illumination. Mice injected with cells containing empty vector (n = 4) or nontransfected parental cells (n = 4) acquired intraocular tumors. 

Ocular Histology.

Following enucleation, the presence of tumors was confirmed by transscleral light illumination, before the eyes were embedded in paraffin. Hematoxylin and eosin staining revealed epithelioid cells with low mitotic index in all retinal layers, with cells invading the sclera. In mice intravitreally injected with RASSF1A-expressing UM-15 cells, no tumor was detected funduscopically or histologically (Fig. 6A). However, histologic sections from eyes injected intravitreally with cells transfected with empty vector or parental UM cells revealed tumors invading all retinal layers and the sclera (Figs. 6B, 6C). 

Figure 6. 

 

(A) Normal retina. No tumors are detected following intraocular injection of UM-15 cell expressing exogenous RASSF1A (×5, n = 4). (B) Histological section of an eye following intravitreal injection of parental UM cells (×5, n = 4). Note the tumor invading all retinal layers. (C) Histological section of an eye injected with UM cells transfected with empty vector (n = 4). Note the large invasive melanoma.

Discussion

Our study shows that RASSF1A expression suppresses UM tumorigenesis and RASSF1A is silenced in the UM-15 cell line. Re-expression of RASSF1A in UM-15 cells reverses the tumoral behavior, as indicated by the slower proliferation rate and regain of sensitivity to cisplatin. 

RASSF1A is a tumor suppressor gene and is inactivated commonly through promoter hypermethylation in many human cancers, including UM. ^20,21 Our finding of RASSF1A methylation in UM-15 cells is in agreement with studies in other cell lines derived from primary tumors of breast and ovarian cancer. ^32,33 In addition, the methylation of p16INK4a was shown to be more common in UM cell lines than in other primary tumors. ^34–38 These observations suggest that tumors with RASSF1A methylation are more oncogenic and more likely to be established as cell lines. They also are consistent with the report of loss of heterogeneity of the primary tumor in cell-line cultures. ³⁹  

The RASSF1A transcript encodes a COOH-terminal RAS-association domain ⁴⁰ and serine residue 131, which serves as a putative phosphorylation target for the ataxia-telangiectasia mutation (ATM). Thus, RASSF1A could act as a link between the RAS and ATM-regulated pathways. ⁴¹ Shivakumar et al. reported that RASSF1A functions as a negative regulator of cell proliferation. ⁴² It apparently blocks cell-cycle progression from the G1 to the S phase by controlling the entry at the retinoblastoma restriction point and inhibiting cyclin D1 protein accumulation at the post-transcriptional level. ⁴² Therefore, hypermethylation-induced loss of RASSF1A expression could lead to a reduction in G1/S-phase cell-cycle control. In epithelial cells derived from lung and breast tumor, the reintroduction of RASSF1A expression resulted in growth arrest that was correlated with reduced cyclin D1 protein accumulation; by contrast, iRNA-mediated inhibition of RASSF1A expression resulted in abnormal accumulation of native cyclin D1. ⁴²  

Using an immortalized cell line, the same researchers reported that oncogenic RAS did not alter RASSF1A-induced growth-inhibitory effects. ⁴² However, in human mammary epithelial cells, the RASSF1A effects dominated the oncogenic RAS effects. Thus, loss of RASSF1A may be a determining step for oncogenic transformation in the absence of RAS-activating mutations. Moreover, overexpression of RASSF1A promotes the formation of stable microtubules and blocks RAS-activated genomic instability, suggesting a potential role of the RASSF1A protein in the maintenance of spindle function and genomic stability. ⁴³  

RASSF1A has been reported to inhibit tumor formation in nude mice. ⁴² Therefore, to analyze further the role of RASSF1A reactivation in vivo, we compared tumorigenesis between mice injected subcutaneously or intraocularly with UM-cells expressing/not expressing reactivated RASSF1A. Intraocular tumors developed only in the absence of RASSF1A expression. 

RASSF1A gene is located on chromosome 3p21.3, and its absence or inactivation has been suggested as a contributing factor for UM tumor formation and progression. ²² RASSF1A methylation, frequently observed in UM tumors, could be the second hit in the classical model for TSG inactivation when coupled with monosomy 3, which is the change observed most consistently in UM. The lack of intraocular tumor formation after intraocular injection of RASSF1A-expressing cell lines might point to an increased relevance of RASSF1A to the development of UM in the eye. 

The UM-15 cell line showed lack of monosomy of chromosome 3 and, although we confirmed the cell line karyotype by FISH, we could not rule out duplications of the remaining chromosome 3. Therefore, the study is limited in measurement of the effect of monosomy 3 and its relation to RASSF1A expression on intraocular tumor development. 

Epigenetic modification of gene expression is an important mechanism in tumor development, and may be reversed by specific treatment. The frequent methylation of the RASSF1A gene in UM cell lines (91%) suggests that RASSF1A also has a role in UM pathogenesis. ²² Loss of one copy of chromosome 3 (monosomy 3) has been reported in approximately 50% of all UMs and is associated with the metastatic behavior of the tumor. Given the location of RASSF1A on the p21.3 region of chromosome 3, it might serve as a TSG whose silencing by methylation represents the second hit after monosomy occurs. Although methylation of RASSF1A may not be held wholly responsible for UM development, it could be a contributing factor in tumor formation and progression. This assumption is supported by the high methylation in the primary tumor. One study of the correlation of UM methylation and survival yielded positive findings, though only at trend level. ²² The same study also suggested that the RASSF1A protein is a potential tumor marker in UM. 

In conclusion, RASSF1A has an important role in the biologic behavior of UM cells and the development of UM, specifically in the eye. Its activation might be applied to the development of future treatments. 

Acknowledgments

Reinhard Dammann of the Institute for Human Genetics, Martin Luther University, Halle-Wittenberg, Germany, kindly provided the plasmid encoding human RASSF1A. 

Footnotes

 Supported in part by the Zanvyl and Isabelle Krieger Fund, Baltimore, Maryland; Eldor-Metzner Clinician Scientist Award, Chief Scientist, Israel Ministry of Health (NG-C, Grant no. 3–3741), The Israel Science Foundation (ISF, NG-C, 1371/08), Legacy Heritage Clinical Research Initiative of the Israel Science Foundation (YC, Grant no. 1716/08), and the Cancer Biology Research Center, Sackler Faculty of Medicine and Faculty of Life Sciences (YC and NG-C), Tel Aviv University, Tel Aviv, Israel.

Footnotes

 Disclosure: O. Dratviman-Storobinsky, None; Y. Cohen, None; S. Frenkel, None; E. Merhavi-Shoham, None; S. Dadon-Bar El, None; N. Binkovsky, None; J. Pe'er, None; N. Goldenberg-Cohen, None