Antioxidant Pattern in Uveal Melanocytes and Melanoma Cell Cultures

Abstract

purpose. To investigate the antioxidant status of cultured uveal melanocytes from patients with uveal melanoma and uveal melanoma cells to characterize some of the biochemical properties of these cells in respect to the normal cutaneous melanocytes.

methods. The fatty acid pattern of membrane phospholipids, intracellular vitamin E level, and superoxide dismutase (SOD) and catalase activities were studied in uveal melanocytes (n = 10) and uveal melanoma cell (n = 10) cultures, by gas chromatography mass spectrometry or by spectrophotometer.

results. Among the uveal melanocyte cultures, two groups were differentiated, according to catalase activity: group A with catalase values comparable to those of cutaneous ones and higher SOD activity and group B with catalase values 2 SD lower (P < 0.001) and lower SOD activity. Vitamin E concentration was not significantly different between melanoma cells and melanocytes, whereas a significantly higher percentage of polyunsaturated fatty acids was found in melanoma cells and the B group of melanocytes (P = 0.022). In uveal melanoma cells SOD activity was significantly lower than that detected in uveal melanocytes (P < 0.005).

conclusions. These results show a different pattern of antioxidants in uveal melanocytes with respect to cutaneous ones, possibly related to the anatomic distribution. However, as in cutaneous melanocytes, two subgroups were identified on the basis of the antioxidant pattern that could be the expression of a constitutional increased susceptibility to oxidative stress in some subjects. Moreover, an imbalance of the antioxidants was observed in melanoma cells, possibly related to the disease status and progression.

Despite diagnostic accuracy and the introduction of new conservative treatment modalities over the past decades, the survival rate in patients affected by uveal melanoma has not been substantially modified. ¹ Furthermore, the identification of risk factors for the development of this tumor has proved to be unsuccessful so far. Although clinical, epidemiologic, and biological observations support the hypothesis that UV light is an etiologic agent for cutaneous melanoma in white populations, ² no irrefutable evidence has appeared in favor of a role for sunlight in uveal melanoma. However, in both cutaneous and uveal melanomas, triggering factors related to the generation of free radical species, and subsequent oxidative stress, can be considered. Oxidative stress results from an imbalance between prooxidant agents and antioxidant systems. Wavelengths in the UVA range (320–400 nm; i.e., not directly absorbed by DNA but capable of generating reactive oxygen species; ROS) are considered to be effective in inducing cutaneous melanoma, at least in animal models. ² On the other hand, uveal vascular changes may significantly increase the amount of ROS generated. Therefore, the evaluation of the cellular antioxidant system could be crucial in understanding the pathway of tumorigenesis. 

Recently, we have shown the presence of an antioxidant system imbalance in cultured cutaneous melanoma cells compared with normal cutaneous melanocytes, and we suggested that this alteration could be related to the disease status and progression. ^{3 4} A similar alteration was detected in melanocyte cultures from the apparently normal skin of patients with melanoma. In the present study we investigated the antioxidant status of normal uveal melanocytes from patients with uveal melanoma and cultured uveal melanoma cells, to contribute to the characterization of the biochemical properties of these cells. Therefore, we studied the activity of the intracellular scavenger enzymes, superoxide dismutase (SOD) and catalase, the intracellular levels of vitamin E, and the pattern of fatty acids of cell membranes, and we have compared the results with those previously obtained in normal cutaneous melanocytes. ^{3 4}  

Methods

Xanthine, xanthine oxidase, and nitro blue of tetrazolium (NBT) were from Sigma (St. Louis, MO). Ham’s F-10 medium, fetal calf serum (FCS), and antibiotics were provided by GIBCO (Paisley, Scotland, UK). Butylated hydroxytoluene, tricosanoic acid (C23:0), N,O-bis(trimethylsilyl)trifluoro-acetamide and trimethyl chorsilane, sodium methoxide, and other reagents and solvents were from Merck AG (Darmastadt, Germany) and were of the highest purity grade. 

Cell Cultures

Normal uveal melanocytes from uveal melanoma patients (UM, n = 10) were collected from eyes, enucleated for large choroidal and ciliary body melanomas, at the site diametrically opposed to the tumor, and uveal melanoma cells (UMC, n = 10) were taken from the same eyes. Uveal melanocytes and uveal melanoma cells were isolated after mechanical and enzymatic dissection and cultured as previously described for cutaneous melanocytes. ^{3 4} Cells were cultured in Ham’s F-10 medium, with 5% FCS and 5000 UI/ml penicillin and 5000 ng/ml streptomycin, and in melanocyte cultures, bovine pituitary gland extract (50 μg/ml) was added. Subconfluent cultures at the fourth or fifth passage were studied. The batch of FCS was the same for all the experimental periods and was analyzed for fatty acid pattern and vitamin E level. 

Antioxidant Enzyme Assays

Cells (4 × 10⁶) were collected and sonicated in phosphate-buffered saline (pH 7.4; GIBCO; 1 ml) and centrifuged at 10,000g for 10 minutes at 4°C. Enzymatic activities were evaluated by a spectrophotometer (Beckman DU 70) on cell supernatants. Catalase activity was determined by the disappearance of hydrogen peroxide, ^{3 4} and SOD activity was determined by evaluating the inhibition of reduction of NBT by superoxide produced by xanthine–xanthine oxidase system. ^{3 4} One unit of catalase was defined as the amount that degrades 1 μmol of H₂O₂, and 1 unit of SOD was defined as the amount of enzyme that produces 50% inhibition of NBT reduction. The activities of cell supernatants were compared with the known purified enzymes and then calculated per 10⁶ cells. At least two determinations were performed on each supernatant, and experiments were repeated twice. The results are reported as the mean of different determinations and expressed as unit/10⁶ cells. For each group of cultures data are reported as mean ± SD. 

Vitamin E Analysis

Cells (4 × 10⁶) were extracted three times in hexane–ethanol (3:1) with 1% sodium dodecyl sulfate in the presence of 25 ng of γ and δ tocopherols as internal standards. Tocopherols were derivatized with N,O-bis-(trimethylsilyl)-trifluoroacetamide with 1% trimethyl chlorsilane as catalyst and were analyzed by gas chromatography mass spectrometry on SPB1 column (30 m × 0.20 μm ID, 0.25 mm, Supelchem) by a selected ion(s) monitoring technique. ^{3 4} Analyses were repeated twice in each extract, and a SD less than 1% was found. Results are expressed as nanograms per 10⁶ cells. 

Polyunsaturated Fatty Acid Analysis

Cell pellets were extracted twice in chloroform:methanol (1:1) in the presence of butylated hydroxy toluene (50 μg) as antioxidant and 25 μg tricosanoic acid ethyl ester as internal standard. The fatty acids of phospholipid fraction were trans-methylated with sodium methoxide in methanol and analyzed by gas chromatography mass spectrometry on capillary column (FFA-P, 60 m × 0.32 μm × 0.25 mm; Hewlett–Packard). The results were obtained after time integration of the chromatogram and final processing of the peak areas. The identity of each fatty acid was determined by comparing the mass spectrum of the peaks with those obtained using reference standards. ^{3 4}  

Statistical Analysis

Student’s t-test was used to determine the statistical significance. Statistical significance was accepted as P < 0.05. The correlation existing was studied by linear regression analysis and r values were calculated by the Pearson test. 

Morphologic Analysis

Cell cultures of normal uveal melanocytes and uveal melanoma cells were seeded into 2% gelatin–coated tissue Transwell andcultured in complete melanocyte medium. Seventy-two hours afterward cells were washed three times in PBS and incubated for 60 minutes at 4°C in fixative buffer (5% milk, 0.01% Tween-20, 0.3 M NaCl, and 2% glutaraldehyde). After fixative treatment, samples were postfixed in 1% osmium tetroxide in Veronal acetate buffer (pH 7.4) for 2 hours at 4°C, stained with uranyl acetate (5 mg/ml), dehydrated in acetone, and embedded in Epon 812. Thin sections were examined both unstained or poststained with uranyl acetate and lead hydroxide. 

Results

The clinical and histologic features of the lesions of the studied cases are reported in Table 1 . The cultures obtained were studied at subconfluence, at the fourth or fifth passage grown in medium containing the same batch of fetal calf serum, so that the amounts of vitamin E, external fatty acids, and essential elements were the same in each culture. The levels of antioxidants were expressed as units per 10⁶ cells or nanograms per 10⁶ cells, to minimize differences among results due to cell size and shape and to compare the values with those previously obtained in normal cutaneous melanocytes. 

Electron microscopic examinations revealed pure cultures of melanocytes and melanoma cells defined on the basis of the presence of characteristic markers such as melanosomes and melanin granules, respectively. No significant differences were observed among the cultures of melanocytes, whereas spindle and epithelioid cells were observed in melanoma cultures. 

Antioxidant Pattern

In uveal melanocytes mean catalase activity was 1.06 ± 0.81 U/10⁶ cells, a value not significantly different from that reported in cutaneous melanocytes (CM) (Table 2) , ^{3 4} but the elevated SD suggested a wide range of variability in this population. A similar situation was reported in a previous work of ours, in cultures of cutaneous melanocytes from the apparently normal skin of subjects with melanoma. ^{3 4}  

Two populations with different morphologic and biological proprieties were differentiated according to catalase activity. ^{3 4} Therefore, using the same scheme, two groups were identified among those with uveal melanocytes (UM): In the first one (UM-A, 5 of 10), catalase activity was 1.79 ± 0.42 U/10⁶ cells, a value similar to that observed in normal cutaneous melanocytes. On the contrary, in the second one (UM-B, 5 of 10), catalase activity was 2 SD lower (0.33 ± 0.065 U/10⁶ cells, P < 0.001; Table 2 ). Mean SOD activity in those with UM was 1.12 ± 0.38 U/10⁶ cells, significantly higher than that previously observed in normal cutaneous melanocytes; however, the mean in the UM-A group was 1.42 ± 0.22 U/10⁶ cells, whereas in UM-B it was significantly lower at 0.81 ± 0.21 U/10⁶ cells (P = 0.01, Table 2 ). 

Levels of vitamin E were 1.15 ± 0.15 ng/10⁶ cells in UM-A and 1.04 ± 0.73 ng/10⁶ cells in UM-B. Both values were significantly lower than those previously observed in normal cutaneous melanocytes (Table 2) . 

In UMC, SOD activity was 0.34 ± 0.11 U/10⁶ cells, significantly lower than that observed in the UM-A and UM-B groups, respectively (P < 0.005). In the same cells, catalase activity was 0.42 ± 0.18 U/10⁶ cells, a value significantly lower than that observed in UM-A but not dissimilar from that found in UM-B (Table 2) . 

In UMC mean vitamin E concentration was 3.09 ± 2.68 ng/10⁶ cells, higher but not statistically different from that observed in both UM-A and UM-B even if with a wider range of variability (Table 2) . A relationship was found between lipophilic and enzymatic antioxidants, with a significant direct correlation between SOD/catalase ratio and vitamin E level (R = 0.73, P = 0.015; Fig. 1 ). 

Fatty Acids of Cell Membranes

To study the peroxidizable components of cell membranes, the fatty acid pattern of membrane phospholipids was analyzed. In all the cultures, among saturated fatty acids, palmitic (C16:0) and stearic acids (C18:0) were the most representative; oleic acid (C18:1 n9) was the main monounsaturated and linolenic acid (C18:2 n6), C20:3 n6, arachidonic acid (C20:4 n6) and C22:6 n3 the polyunsaturated fatty acids (PUFAs). Arachidonic acid was the most representative long-chain PUFA present in phospholipid fraction. In the UM-A group, PUFA percentage, compared with total fatty acids analyzed, was 7.9% ± 4.37%, lower than those of cutaneous melanocytes, whereas in the UM-B group the PUFA percentage was 21.49% ± 7.02% (P < 0.001). A significant direct correlation was found between the PUFA percentage and the SOD/catalase ratio (r = 0.69), indicating that the increased concentration of peroxidizable compounds was associated with an imbalance in the enzymatic antioxidant activities (Fig. 2) . In UMC the PUFA percentage was 18.16% ± 5.02% (Table 2) , a value not statistically different from that observed in the UM-B group. 

Discussion

Our study demonstrates that cultured uveal melanocytes from patients with melanoma, compared with cultured normal cutaneous melanocytes, ^{3 4} have a higher level of enzymatic antioxidant activities. In particular, SOD activity appears to be significantly higher, possibly related to the high O₂ tension found in the choroidal district. SOD, in fact, is important in clinical situations such as in reperfusion after ischemia, where huge amounts of superoxide anion (O₂ ^−·) are generated and need to be dismutated by this enzymatic activity. 

More interestingly, as previously described in cutaneous melanocytes from melanoma patients, ^{3 4} two subgroups of uveal melanocytes were identified on the basis of catalase activity: one with values comparable to those observed in cutaneous melanocytes (UM-A) and the other with significantly lower values (UM-B). Consequently, in the B group, the SOD/catalase ratio, which is correlated with the susceptibility of the cells to a peroxidative stress, ⁵ was significantly increased compared with that of group A. The SOD/catalase ratio imbalance represents an alteration of the scavenger system. Antioxidants, in fact, interact in a complex fashion, so that changes in the concentration or activity in one component can affect the whole system. SOD dismutates superoxide anion radicals, generating hydrogen peroxide and oxygen. Catalase and glutathione peroxidase (GSH–Px) are the main enzymes involved in removing H₂O₂. ⁶ If the production of H₂O₂ overwhelms the activities of these latter enzymes, in the presence of transitional metals (Fe²⁺, Cu⁺), H₂O₂ becomes a substrate of the Fenton reaction, giving rise to extremely toxic and mutagenic hydroxyl radicals (HO^·). ⁶ Moreover, the imbalance of the SOD/catalase ratio was associated with an increased PUFA percentage in the cell membranes of UM-B, suggesting that these cells are more susceptible to the deleterious effects of prooxidants. In vitro, cutaneous melanocytes with alteration of the SOD/catalase ratio, in fact, undergo a significant proliferation after treatment with a low concentration of cumene hydroperoxide. ⁴ Moreover, cell cultures deficient in catalase activity showed an increased DNA alteration after exposure to peroxidizing agents. ⁷ Therefore, the imbalance of the intracellular antioxidants has been considered as a possible additional risk factor for the development of melanoma. ^{3 4}  

In uveal melanoma cells, instead, a significant decrease of SOD activity, compared with that of uveal melanocytes, was detected, whereas catalase activity, even if lower, was not significantly modified. An increased level of the polyunsaturated component of cell membranes and vitamin E concentration was observed. The higher vitamin E level is likely to be a compensatory mechanism adopted by the cells, at least in vitro, to reduce the intracellular oxidative events. In fact, in UMC a significant direct correlation was observed between the SOD/catalase ratio and the vitamin E level. These results are in agreement with previous data that demonstrated the correlation among the differentiation status, antioxidant systems, and percentage of PUFAs in cultured cells: the higher the proliferation rate, the less differentiated the cells; the lower the total antioxidant protection system, the higher the PUFA percentage. ⁸ However, no correlation was found between alteration of the antioxidant pattern and the melanoma cell type. 

The source of ROS for the development of cutaneous melanoma could be UV exposure, ² but uveal melanocytes, especially those embedded in the ciliary body, are reached by low amounts of UV radiation; therefore, different free radical sources should be considered. Uveal melanocytes are tightly connected with the vascular bed and the oxidative insults might be related to hemodynamic changes. Alterations in blood flow can induce free radical release, and free radical–mediated injury has been involved in the pathophysiological alterations observed during ischemia and reperfusion. ⁹  

Episodes of choroidal ischemia–reperfusion may take place many times during a lifetime without causing clinical manifestations, because of the reservoir in choroidal blood flow. ¹⁰  

In conclusion these data demonstrate that an alteration of the antioxidant pattern can be detected in uveal melanoma cells, as well as in cutaneous ones, possibly related to the disease status and progression. Moreover, data obtained from apparently normal uveal melanocytes suggest that in some subjects a constitutional imbalance of the antioxidant system, detectable by a decrease of catalase activity, can exist. This could be the basis for increased susceptibility to free radical–mediated damage and possibly to the development of uveal melanoma. As for skin melanoma, we can suggest that an individual predisposition together with environmental and general risk factors could play an important role in tumor onset and that the occurrence of the uveal melanoma might be the expression of acute repeated damages. ²  

 Submitted for publication November 13, 1998; revised June 10, 1999; accepted July 7, 1999.

 Commercial relationships policy: N.

 Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May, 1998.

 

Table 1.

 

Clinical and Histological Data

Patient	Age	Site	Tumor Size*	Cell Type
FM	58	Choroid	Large	Epithelioid
CC	69	Choroid	Large	Spindle
LF	25	Choroid	Large	Spindle
GD	36	Choroid	Large	Epithelioid
AA	73	Ciliary body	Large	Epithelioid
PC	71	Choroid	Large	Epithelioid
FDC	56	Choroid	Large	Mixed
AS	50	Choroid	Large	Spindle
LM	41	Choroid–ciliary body	Large	Epithelioid
GF	65	Choroid	Large	Epithelioid

*  According to COMS criteria.

Table 2.

 

Antioxidants and PUFA Percentage in Cutaneous Melanocytes, Uveal Melanocytes, and Uveal Melanoma Cells

	Vitamin E, ng/106 cells	SOD, U/106 cells	Cat, U/106 cells	SOD/Cat Ratio	PUFA, % Tot
CM	1.88 ± 0.39	0.45 ± 0.11	0.99 ± 0.34	0.51 ± 0.24	15.9 ± 4.2
UM-A
AS	1.19	1.55	2.3	0.67	5.88
FM	0.99	1.06	1.45	0.73	14.85
LF	1.29	1.66	1.6	1.03	3.19
GD	1	1.42	1.4	1.01	7
AA	1.3	1.45	2.2	0.65	9
Mean± SD	1.15 ± 0.15	1.42 ± 0.22	1.79 ± 0.42	0.81 ± 0.18	7.9 ± 4.37
UM-B
CC	0.36	0.57	0.29	1.96	23.27
PC	1.84	0.79	0.3	2.63	18.33
FDC	0.2	1.09	0.4	2.72	32.88
LM	1.2	0.66	0.41	1.6	15
GF	1.6	0.98	0.27	3.62	18
Mean± SD	1.04 ± 0.73	0.81 ± 0.21*	0.33 ± 0.065, †	2.5 ± 0.77, †	21.49 ± 7.02, †
UMC
FM	9.23	0.41	0.25	1.64	13.11
CC	3.33	0.48	0.64	0.75	20.98
LF	0.32	0.29	0.79	0.36	14.25
GD	1.1	0.49	0.6	0.81	14.89
AA	1.6	0.51	0.45	1.13	20.06
PC	5.43	0.19	0.28	0.67	29.62
FDC	2	0.25	0.29	0.86	19.93
AS	2.41	0.3	0.35	0.85	14.47
LM	1	0.24	0.35	0.68	14.5
GF	4.5	0.31	0.28	1.1	19.8
Mean± SD	3.09 ± 2.68	0.34 ± 0.11, †	0.42 ± 0.18, †	0.88 ± 0.34	18.16 ± 5.02, †

 Each result represents the mean of 2 experiments in duplicate.

*   P < 0.05 versus UM-A.

†   P < 0.001.

Figure 1.

 

Correlation between enzymatic and lipophilic antioxidants in UMC. A significant direct correlation was observed between the SOD/catalase (SOD/Cat) ratio and vitamin E (Vit.E) levels (P = 0.015, r = 0.73). Each point represents the mean of two or three replicates.

Figure 2.

 

Correlation between SOD/catalase (SOD/Cat) ratio and PUFA percentage in UM (closed symbols, UM-A group; open symbols, UM-B group). A significant direct correlation was observed between SOD/Cat ratio and PUFA percentage (P = 0.015, r = 0.69). Each point represents the mean of two or three replicates.