Fax: (011) 91 04428254180
The role of nitric oxide synthases and nitrotyrosine in retinoblastoma
Article first published online: 7 MAR 2005
Copyright © 2005 American Cancer Society
Volume 103, Issue 8, pages 1701–1711, 15 April 2005
How to Cite
Adithi, M., Nalini, V. and Krishnakumar, S. (2005), The role of nitric oxide synthases and nitrotyrosine in retinoblastoma. Cancer, 103: 1701–1711. doi: 10.1002/cncr.20961
- Issue published online: 4 APR 2005
- Article first published online: 7 MAR 2005
- Manuscript Accepted: 16 DEC 2004
- Manuscript Revised: 27 NOV 2004
- Manuscript Received: 2 JUN 2004
- Indian Council for Medical Research. Grant Numbers: reference 5/4/6/6/2003-2004-NCD-II, Iris code 2003-000810
- endothelial nitric oxide synthase;
- inducible nitric oxide synthase;
To investigate the potential involvement of the nitric oxide (NO) pathway in retinoblastoma, the authors correlated immunoreactivity for endothelial nitric oxide synthase (eNOS), inducible nitric oxide synthase (iNOS), and nitrotyrosine (NT) with the degree of tumor invasiveness in retinoblastoma.
eNOS, iNOS, and NT reactivity was evaluated by immunohistochemistry in 34 archival retinoblastoma specimens and in a human Y79 retinoblastoma cell line. The tumors were divided into 2 groups: Group A tumors (n = 17 tumors) with no invasion and Group B tumors (n = 17 tumors) with invasion of the choroid, optic nerve, and/or orbit. The expression levels of eNOS, iNOS, and NT were correlated with invasiveness of the tumors.
In Group A tumors (n = 17 tumors) without invasion, eNOS was positive in 17 of 17 tumors (100%), iNOS was positive in 14 of 17 tumors (82%), and NT was positive in 17 of 17 tumors (100%). In Group B tumors (n = 17 tumors) with invasion, eNOS was positive in 17 of 17 tumors (100%), iNOS was positive in 16 of 17 tumors (94%), and NT was positive in 17 of 17 tumors (100%). The invasive cohort showed significantly higher expression of iNOS (P < 0.0001) and NT (P < 0.020) compared with the noninvasive cohort. Y79 cells also expressed eNOS, iNOS, and NT; and nonneoplastic retina was positive for eNOS, iNOS, and NT.
Taken together, the results suggested that retinoblastomas can produce NO. The roles of NO in the biology of retinoblastoma and in the prognosis for patients with retinoblastoma remain to be established. Cancer 2005. © 2005 American Cancer Society.
Retinoblastoma is the most common intraocular malignancy in children and almost uniformly is fatal when left untreated. Traditionally, patients with retinoblastoma have undergone enucleation.1 Because the tumor occurs in children, enucleation is particularly devastating and disfiguring. Recently, a number of therapeutic alternatives have been adopted. These include radiation, cryotherapy, chemotherapy, or various combinations of these three therapies.2–7 Such treatments are more effective with smaller tumors and often are associated with unwanted side effects, such as a higher incidence of secondary tumors. Thus, alternative therapeutic approaches are needed. A number of newer strategies have been tested recently, including the use of vitamin D analogs,8–11 viral therapy,12, 13 suicide gene therapy,14 and targeted therapy.15 In addition, a recent study showed that retinoblastoma cells respond to cytokine exposure and that interferon γ and/or tumor necrosis factor α induce apoptosis, which leads to cell death, raising the possibility that cytokine therapy may be an effective treatment for retinoblastoma.16 However, current research is directed toward novel treatment modalities to improve the effectiveness of a conservative approach.
There has been much recent interest in the role that nitric oxide (NO) plays in both tumor growth and dissemination.17 NO is a small messenger molecule that was discovered first as a potent vasodilator, known as the endothelium-derived relaxing factor, which is produced and released by vascular endothelial cells.18, 19 Now, it is well established that NO has several diverse biologic functions and is produced by many cell types other than endothelium.20, 21 The NO radical is generated by the action of the enzyme NO synthase (NOS). There are three distinct isoforms of this enzyme that are encoded by three different genes. Two of the NOS isoforms are constitutive and calcium/calmodulin dependent–the endothelial and neuronal types (eNOS and nNOS, respectively); the third isoform is inducible NOS (iNOS) and is not dependent on calcium/calmodulin for its enzymatic action. The produced NO can react with free radicals, such as superoxide anions, to form peroxynitrite, which is a potent nitrating agent. Peroxynitrite can cause oxidation of DNA and membrane phospholipids in addition to nitration of free or protein-associated tyrosines, producing nitrotyrosine (NT). Thus, the occurrence of NT in tissues has been measured as a marker of peroxynitrite formation.22, 23
Recently, investigators have studied the expression levels and the activity of iNOS in human malignancies. An increased level of iNOS expression and/or activity has been found in tumor cells from gynecologic malignancies,24 in the stroma of breast carcinoma,25 and in tumor cells from head and neck carcinoma.26, 27 Although iNOS expression and activity have also been demonstrated in colorectal carcinoma, the results remain controversial.28 There is no information available about the expression of NOS in retinoblastoma. Therefore, we studied the expression of eNOS, iNOS, and NT in a small series of retinoblastomas and also in a retinoblastoma cell line and correlated the results with differentiation and invasion of the tumors.
MATERIALS AND METHODS
This study was reviewed and approved by the local ethics committee at our institute, and the committee deemed that it conformed to the generally accepted principles of research, in accordance to the Helsinki Declaration.
Inclusion and Exclusion Criteria
The inclusion criteria required that all patients had undergone enucleation and had a minimum follow-up of 24 months. The criteria excluded patients who had received preoperative chemotherapy, because this would influence the interpretation of immunohistochemistry results.
There were 34 formalin fixed, paraffin embedded tumors from 34 patients, comprising 14 females and 20 males. The median patient age was 3 years at the time of enucleation.
Tumor samples from the 34 patients were available between 1997 and 2002. The tumors were divided into two groups. Group A included tumors without invasion of the choroid, optic nerve, or orbit and/or without metastasis. Group B included tumors with invasion of the choroid, optic nerve, or orbit and/or with metastasis. The tumors were analyzed immunohistochemically for the expression of eNOS, iNOS, and NT.
Retinoblastoma sections were graded microscopically and were divided into three groups according to the predominant pattern of differentiation.29 All tumor slides were reviewed, and the level of choroidal invasion was classified as either focal or diffuse. For the optic nerve, prelaminar invasion, postlaminar invasion, and invasion of the surgical end of the optic nerve were analyzed.29 The invasion of tumor cells into the orbit was also investigated in the sections.
Invasion and differentiation
Among the 17 tumors with invasion of the choroid, optic nerve, and/or orbit, there were 11 tumors with choroidal invasion, including 8 tumors with diffuse choroidal invasion and 3 tumors with focal choroidal invasion. Among the eight tumors with diffuse choroidal invasion, one tumor had prelaminar invasion, and two tumors had postlaminar invasion of the optic nerve. Among the three tumors with focal choroidal invasion, one tumor had prelaminar invasion. There were six tumors with invasion into the optic nerve, including two tumors that invaded the prelaminar portion of the optic nerve, two tumors that invaded the postlaminar portion of the optic nerve, and two tumors that invaded the surgical end of the optic nerve. Among the 34 retinoblastomas, there were 19 poorly differentiated tumors, 6 moderately differentiated tumors, and 9 well differentiated tumors.
The human Y79 retinoblastoma cell line (National Center for Cell Sciences, Pune, India) was cultured in RPMI 1640 medium and supplemented with 20% fetal calf serum, 0.1% ciprofloxacin, 2 mM L-glutamine, 1 mM sodium pyruvate, and 4.5% dextrose and was grown in suspension. Passages 22 and 23 were used for making a cell block for immunohistochemistry. All chemicals were obtained from HiMedia Laboratories, Pvt. Limited (Mumbai, India).
Immunohistochemical detection of NOS isoforms and NT on paraffin wax embedded tumor sections and on Y79 cell line sections was performed using a rabbit polyclonal antibody raised against human eNOS (1:50 dilution; Santa Cruz Biotechnology Inc., Santa Cruz, CA), a rabbit polyclonal antibody developed against human iNOS (1:50 dilution; Santa Cruz Biotechnology Inc.), and a mouse monoclonal antibody developed against human NT (1:25 dilution; Santa Cruz Biotechnology Inc.). Staining for NOS and NT was visualized with the streptavidin-biotin immunoperoxidase technique using the Dako LSAB+ system with horseradish peroxidase (DakoCytomation, Glostrup, Denmark) according to the manufacturer's instructions. For control sections, the same procedure was performed in the absence of the primary antibody. At least one control was undertaken in each experiment.
Microscopic evaluation and statistical analysis
The slides were examined by two observers (M.A. and S.K.) using conventional light microscopy. The levels of immunoreactivity and the cell types that exhibited eNOS, iNOS, and NT immunoreactivity were recorded for each slide. The amount of immunostaining was evaluated as + for immunostaining in > 5% but less than one-third of tumor cells, ++ for immunostaining in more than one-third but less than two-thirds of tumor cells, and +++ for immunostaining in more than two-thirds of tumor cells.30
The data were analyzed using nonparametric methods. The associations of eNOS, iNOS, and NT with tumor differentiation were determined with the Jonckheere–Terpstra test and the Mann–Whitney U test was used to determine associations with tumor invasion. Correlations between the expression of eNOS and iNOS and between the expression of iNOS and NT were determined using the Kendall τ-b test. For statistical purposes, moderately differentiated and well differentiated tumors were grouped together and were compared with poorly differentiated tumors. The correlation coefficients reported are Spearman ρ values. The analyses were performed using the SPSS software package for Windows (version 9.0; SPSS Inc., Chicago, IL).
|Patient no.||Age||Gender||Histopathologic diagnosis||Immunoreactivitya|
|1||1 yr, 6 mos||Male||OD: WD||++||++||++|
|2||4 yrs||Male||OD: PD||+++||++||++|
|3||5 yrs||Male||OS: PD||+++||+++||++|
|4||4 mos||Male||OS: WD||+++||++||+|
|5||2 yrs, 2 mos||Male||OD: WD||+++||+||+|
|6||3 yrs||Female||OS: WD||+||+||++|
|7||3 yrs, 6 mos||Female||OD: PD||+||+||+|
|8||8 yrs||Male||OS: PD||+||Neg||+|
|9||1 yr||Female||OD: WD||++||+||++|
|10||3 yrs||Female||OD: PD||+++||+||++|
|11||1 yr||Female||OD: PD||+++||+||+|
|12||4 yrs||Female||OS: PD||+++||Neg||+|
|13||1 yrs, 6 mos||Male||OS: WD||+||Neg||++|
|14||1 yr, 4 mos||Male||OD: WD||+++||+||++|
|15||1 yr, 6 mos||Male||OD: MD||+||+||+|
|16||7 mos||Male||OD: WD||+||+||+++|
|17||4 yrs, 6 mos||Male||OS: MD||+++||++||+++|
|Patient no.||Age||Gender||Histopathologic diagnosis||Immunoreactivitya|
|1||3 yrs||Male||OS, PD: Diffuse Ch invasion, post lam ON invasion||+++||+++||+++|
|2||4 yrs||Female||OS, PD: Post lam ON invasion||+++||++||+++|
|3||2 yrs, 10 mos||Male||OS, PD: Diffuse Ch invasion, pre lam ON invasion||+++||+++||++|
|4||21 yrs||Female||OD, PD: Pre and post lam ON invasion, surgical end involved||++||+++||++|
|5||1 yr||Female||OD, PD: Diffuse Ch invasion||+++||+++||+++|
|6||4 yrs||Female||OD, WD: Pre lam ON invasion||++||++||++|
|7||7 yrs||Male||OD, PD: Diffuse Ch invasion||++||++||+|
|8||4 yrs||Female||OD, MD: Focal Ch invasion||+||Neg||++|
|9||4 yrs||Female||OD, MD: Focal Ch and pre lam ON invasion||++||+++||+++|
|10||1 yrs, 6 mos||Male||OD, PD: Pre lam ON invasion||++||+||++|
|11||5 mos||Male||OS, MD: Focal Ch invasion||+++||+++||+++|
|12||3 yrs||Male||OS, PD: Diffuse Ch and post lam ON invasion||+++||+++||++|
|13||2 yrs||Female||OS, PD: Surgical end involved||++||+++||+++|
|14||3 yrs, 6 mos||Male||OS, PD: Diffuse Ch invasion||+++||+++||+++|
|15||4 yrs||Male||OS, PD: Diffuse Ch invasion||+||++||++|
|16||7 yrs||Male||OS, MD: Post lam ON invasion||++||++||+|
|17||2 yrs, 6 mos||Female||OS, PD: Diffuse Ch invasion||++||++||+++|
NOS and NT Immunoreactivity in Ocular Tissues and in Nonneoplastic Retina
eNOS, iNOS, and NT were expressed in the nonneoplastic retina. eNOS was expressed in normal corneal epithelium and endothelium. iNOS and NT were negative in normal corneas. ENOS, iNOS, and NT were expressed in the ganglion cell layer, in the inner and outer nuclear layers of nonneoplastic retinal tissue, and in the retinal pigment epithelial tissue; however, all three were negative in the optic nerve (Figs. 1A, 2A, and 3A [inset]).
eNOS Immunoreactivity in Retinoblastoma
Group A tumors
Among 17 noninvasive retinoblastomas, there were 6 tumors with + staining, 2 tumors with ++ staining, and 9 tumors with +++ staining.
Group B tumors
Among 17 invasive retinoblastomas, there were 2 tumors with + staining, 8 tumors with ++ staining, and 7 tumors with +++ staining. There was no significant difference in the expression of eNOS between the two groups (Mann–Whitney U test; P = 0.865).
iNOS Immunoreactivity in Retinoblastoma
Group A tumors
Among 17 noninvasive retinoblastomas, there were 9 tumors with + staining, 4 tumors with ++ staining, and 1 tumor with +++ staining. Three tumors were negative for iNOS.
Group B tumors
Among 17 invasive retinoblastomas, there was 1 tumor with + staining, 6 tumors with ++ staining, and 9 tumors with +++ staining. One tumor was negative for iNOS. The invasive cohort showed significantly higher expression of iNOS (Mann–Whitney U test; P < 0.0001).
NT Immunoreactivity in Retinoblastoma
Figure 3A–D shows NT immunoreactivity in invasive and noninvasive tumors and in normal retinal tissues.
Group A tumors
Among 17 noninvasive retinoblastomas, there were 7 tumors with + staining, 8 tumors with ++ staining, and 2 tumors with +++ staining.
Group B tumors
Among 17 invasive retinoblastomas, there were 2 tumors with + staining, 7 tumors with ++ staining, and 8 tumors with +++ staining. The invasive cohort showed significantly higher expression of NT (Mann–Whitney U test; P = 0.020).
eNOS, iNOS, and NT Immunoreactivity with Differentiation of the Tumors
iNOS immunoreactivity was significantly higher in poorly-differentiated tumors compared with moderately differentiated tumors and well differentiated tumors (JT test; P = 0.048), whereas there was no significant difference in immunoreactivity for eNOS or NT among the Group A and Group B tumors (P = 0.151 and P = 0.695, respectively).
Correlation between expression levels of eNOS, iNOS, and NT
The correlation between the expression of eNOS and iNOS and between the expression of iNOS and NT did not reach statistical significance (Kendall τ B test; correlation coefficient = 0.389 and 0.456, respectively).
Expression levels of eNOS, iNOS, and NT in Y79 retinoblastoma cells
The Y79 cells were arranged in typical grape-like clusters. The expression of eNOS was +, the expression of iNOS was ++, and the expression of NT was +++ in the Y79 cells (Fig. 4–4C). The expression levels of eNOS, iNOS, and NT in Group A tumors (Fig. 5A) and in Group B tumors (Fig. 5B) are represented as scatterplots in Figure 5.
In the current study, there were two important observations. First, we observed the expression of iNOS, eNOS, and NT immunoreactivity in retinal tissues. The expression of iNOS and eNOS suggests a biologic function of NOS in the normal retina. It has been found that human retinal tissues express RNA for both eNOS and iNOS. The discovery of NOS activity in the developing chick tectum in proximity to the retinal axons suggests that NOS also may play a role in retinal development.31 In the retina, NOS is found in the normal retinal neurons, pigment epithelium, amacrine and ganglion cells, nerve fibers in the outer and inner plexiform layers, and photoreceptor ellipsoids. The normal optic nerve is devoid of NOS. It is likely that NO is a neurotransmitter within the retina, although its significance in visual transduction has not been elucidated fully.31
Second, we observed that eNOS, iNOS, and NT were expressed in retinoblastoma. When their expression levels were correlated with invasiveness, eNOS was expressed in both Group A tumors (without invasion) and Group B tumors (with invasion); however, the expression levels of iNOS and NT were significantly higher in tumors with invasion. The human Y79 retinoblastoma cell line also exhibited positivity for eNOS, iNOS, and NT, with iNOS and NT expressed more strongly than eNOS. This report documents for the first time the in situ expression of NOS isoforms in human retinoblastomas. Thus, retinoblastoma joins the list of tumors that express NO.24–28
A number of activities may contribute to the tumor-enhancing effects of NO. NO is converted to peroxynitrite in the presence of the superoxide anion. This highly potent, toxic molecule causes DNA damage,32 increased angiogenesis and blood flow,33 prevention of apoptotic cell death,34 and suppression of the immune system.35 Conversely, numerous reports also indicate that NO can inhibit neoplasia. NO is cytotoxic to tumor cells36 and can decrease tumor growth and metastasis in vivo.37–41
The inhibitory effects of NO on tumorigenesis have been associated with antioxidant actions,42 with the inhibition of angiogenesis43 and platelet aggregation,44 and with the enhancement of vasodilatation,45 differentiation,46 and apoptosis.34, 47 This apparent paradox, in which NO both enhances and inhibits tumorigenesis, has been attributed to several factors, including local NO concentrations, cell type, cellular genetics, and redox status.48–52 Low levels of NO appear to increase the tumor-promoting effects of NO, whereas high levels are cytostatic/cytotoxic. Thus, the question becomes not whether iNOS promotes or prevents malignancy—it appears to have the capability to do both—but, rather, what role iNOS plays during the natural course of carcinogenesis in vivo. Most available data suggest that iNOS is present at levels that promote tumor development.
Neoplastic transformation involves an accumulation of genetic mutations that enables the tumor to usurp control of the cell cycle, overcome growth-inhibitory signals, circumvent apoptosis, avoid senescence, and recruit a blood supply. Several factors contribute to the aggressiveness of retinoblastoma, such as early genetic events, including increased copy numbers of chromosome 6p and 1q, and late events, such as high levels of telomerase activity, loss of chromosome 1p, and p53 inactivation. In addition, increased angiogenesis, deregulation of cell-to-cell adhesion molecules, and changes in integrins contribute to the aggressiveness of the tumor.29 In our previous work53 on 232 eyes that were enucleated for retinoblastoma at our institute, we observed a greater incidence of choroidal invasion and optic nerve involvement in Asian-Indian children than was reported among children from the west. We hypothesized that this may have been due to delays in diagnosis, different biologic tumor behavior,54 or both.
Additional studies at our institute have shown that there are multiple mechanisms in retinoblastoma that contribute to their invasiveness: Invasive retinoblastoma had 1) decreased metastasis-suppressor protein nm23-H1 and nm-H255; 2) decreased tetraspanin protein KAI1/CD82,56 which contributes to increased motility of tumor cells and to invasion; 3) decreased/absent Fas expression,57 which allows tumor cells to escape from Fas ligand (FasL)-mediated tumor cell death; 4) elevated FasL expression in invading tumors that may enable them to lyse Fas-expressing tumor-infiltrating lymphocytes58; and 5) decreased human leukocyte antigens Class I and II and the antigen-processing molecules of the Class I pathway,59 which allows the escape of tumor cells from cytotoxic T lymphocyte-mediated cell death while maintaining resistance to natural killer cell lysis. Hence, these deregulations allow tumors to gain time for the accumulation of critical mutations and/or derangements in the expression of malignancy causing genes, which will provide these tumors with uncontrollable invasive potential.
Retinoblastoma tumor cells also express the multiple drug resistance proteins, P-glycoprotein and lung resistance-related protein (LRP), which helps in the intracellular transport of substrates, a critical step in tumor cell growth and survival.60 Thus, the progression of the malignant cell from initial neoplastic transformation to distant metastasis is a result of successive molecular events that provide the tumor cell with a growth and survival advantage with respect to adjacent cells. These events are just beginning to be understood; hence, much about the factors that contribute to tumor survival and aggressiveness in retinoblastoma remains to be clarified.
There are three ways in which expression of NOS may contribute to the survival of tumor cells and aggression in retinoblastoma. First, because the tumors cannot exceed 1–2 mm3 in volume without developing new blood vessels within the eye, where space and vascularity are limited, tumors must produce angiogenic factors at an early point in their development.61 Studies have shown that vascular endothelial growth factor (VEGF) mRNA and protein are expressed in retinoblastoma. VEGF is hypoxia-inducible in retinoblastoma cells. This suggests that focal hypoxia could act as a stimulus for VEGF production thus contributing to tumor growth by stimulating the formation of a vascular stroma.62, 63 Thus, iNOS would enhance the nutrition of tumor cells, whereas increased VEGF would make more blood vessels available for tumor cells to invade. Second, NO has been implicated in the expression of matrix metalloproteases (MMPs), which are highly homologous proteolytic enzymes involved in the degradation of basement membrane and other extracellular components. It is believed that MMPs down-regulate the synthesis of their natural inhibitors, the tissue inhibitors of matrix metalloproteases (TIMPs).64 Previously, Surti et al.65 observed that MMP-2 and MMP-9 were positive in invasive retinoblastomas compared with noninvasive tumors. Thus, it is likely that the raised iNOS concentration facilitates a more aggressive tumor phenotype by its additional actions on MMPs. Third, it has been shown that endogenously produced NO can phosphorylate and inactivate the tumor suppressor pRb to cause a significant increase in cell proliferation. This has been observed in the T47D breast carcinoma cell line. This effect was reversed when cells were treated with L-NG-monomethyl-arginine (L-NMMA), an NO inhibitor,66 providing direct evidence for its role in manipulating the cell cycle.
To our knowledge, the results of the current study demonstrated for the first time that eNOS, iNOS, and NT are expressed in retinoblastoma. Currently, the role of NO in tumor biology remains poorly understood. In the past few years, data regarding the promoting effects of iNOS on tumor development in vivo have been mounting. A consistent association between up-regulation of iNOS in carcinomas of the bladder, prostate, oral cavity, and esophagus has been observed. Moreover, deregulation appears to occur during early neoplastic progression in these organs, suggesting that intervention with iNOS inhibitors may be a viable chemopreventive strategy. The chemopreventive effects of iNOS inhibitors have been demonstrated in preclinical colon carcinoma models.67 However, additional studies clearly are needed to determine the role of the NO/iNOS pathway in tumorigenesis per se and to establish the utility of iNOS inhibitors as chemoprevention agents. The complex biologic actions of this ubiquitous signaling molecule will necessitate careful experimentation to assess risks and benefits adequately and to identify the most appropriate cohorts for preventive intervention.
- 65Matrix metalloproteinases in retinoblastoma: correlation with metastatic behavior [abstract 1421]. Invest Ophthalmol Vis Sci. 2003; 44: E-1421., , , .