- Email: [email protected]
Effects of medium conditioned by retinal pigmented epithelial cells on neurotransmitter phenotype in retinoblastoma cells
207
Cancer Letters, 68 (1993) 207 - 213 Elsevier Scientific Publishers Ireland Ltd.
Effects of medium conditioned by retinal pigmented epithelial cells on neurotransmitter phenotype in retinoblastoma cells Lori K. Klaidmanb, Joyce Tombran-Tink”, Lincoln V. Johnsona ‘Department Toxicology,
of Anatomy
and Cell Biology,
School of Pharmacy,
(Received
18 November
1992)
(Accepted
16 December
1992)
School of Medicine,
Uniuersity of Southern
Summary We previously reported that medium conditioned by retinal pigmented epithelial cells can induce cellular differentiation in human retinoblastoma cells. Extensive neurite outgrowth, of neuronal marker increased expression molecules and decreased expression of glial marker molecules are characteristic of the dijjerentiated phenotype. In the studies described here, we examine whether modulations in the expression of potential neurotransmitter molecules, catecholamines and indolealkyl amines, might be associated with the differentiation of retinoblastoma cells. Concentrations of seroacid, S-methtonin, Shydroxyindoleacetic homovanillic acid, and 3oxytyrosine, methoxy-4-hydroxyphenylacetic acid in extracts of differentiated and undifferentiated retinoblastoma cells were assessed by HPLC. The results show that serotonin and its metabolite, Shydroxyindoleacetic acid, are characteristically present in undifferentiated cells. Dopa metabolites, 3-methoxytyrosine, homovanillic acid and 3-methoxy-4-hydroxyphenylacetic acid, are uniquely present in dijCorrespondence to: James D. Adams, Jr., Dept. Molecular Pharmacology and Toxicology, School of Pharmacy, University of Southern California, 1985 Zonal Ave PSC 510, Los Angeles, CA 90033, U.S.A.
0304-3835/93/$06.00 Printed and Published
James
0 1993 Elsevier Scientific Publishers in Ireland
D. Adams,
and bDepartment
California,
Los Angeles,
Jr.b and
of Molecular
Pharmacology
and
CA (U.S.A.)
jerentiated cells. It appears that differentiation of retinoblastoma cells induced by factors secreted by retinal pigmented epithelial cells involves a switch from a serotonergic phenotype to one dominated by metabolites of dopa. These findings may provide clues about the factors that control retinoblastoma cells and metastasis.
pigmented Keywords: retinoblastoma; epithelium; serotonin; dopa; differentiation
Introduction Retinoblastoma is a tumor of the eye that is seen most commonly in children under the age of 5 years old. Metastasis of the tumor is uncommon in infants [l] . Malignant spread of the tumor is more common in children of ages between 1 and 5 years old and has a very poor prognosis. Therefore, it is important for the neoplasm to be diagnosed early and treated without delay. Retinoblastoma is probably derived from a mixture of primitive, multipotential retinoblastic cells [6]. Cultured human retinoblastoma cells (Y79) can be induced to differentiate into cells with neuronal, glial or photoreceptor cell characteristics [4,5]. Undifferentiated Y79 cells Ireland Ltd.
208
have many biochemical characteristics typical of neuronal cells. They express neuronal marker molecules [5] and have uptake systems for dopamine and serotonin [ 111. In the retina, dopamine and serotonin have been implicated as neurotransmitters in a subclass of amacrine cells [2]. In a recent report, we showed that medium conditioned by cultured human retinal pigmented epithelial cells (RPE-CM) induces neuronal differentiation in Y79 cells [8]. A critical period of stimulation with RPE-CM prior to substratum attachment promotes a high frequency of neuronal differentiation in Y79 cells. Extensive neurite outgrowth, increased expression of neuronal marker molecules and decreased expression of glial marker molecules are associated with phenotypic changes and differentiation of Y79 cells [8]. The differentiation inducing activity present in RPE-CM is associated with a 50 -55 kDa pigmented epithelial differentiating factor (PEDF) that can be isolated and purified from RPE-CM 19,lo]. A childhood tumor related to retinoblastoma is neuroblastoma, which can be monitored in its metastatic form due to the elaboration of norepinephrine by the cancer cells. When neuroblastoma becomes apparent in children under the age of 1, regression is common, which involves either spontaneous cytolysis of the tumor cells or differentiation into ganglioneuroma or ganglioneuroblastoma cells. Retinoblastoma behaves similarly, being much more metastatic when present in children older than the age of 1, as mentioned above. The retina is not completely formed at birth, but requires a few months of development postnatalIy. It is possible that a differentiation factor that becomes less abundant in more differentiated retina leads to the maturation of aggressive retinoblastoma tumors. A differentiation factor, such as PEDF, that is present in immature retina may protect the retina from retinoblastoma by inducing cellular differentiation of retinoblastoma cells. We therefore became interested in the ability of PEDF to alter the neurotransmitter phenotype in retinoblastoma
cells. Possible alterations in neurotransmitter
phenotype may provide a valuable clinical marker of the differentiation of the retinoblastoma, which could correlate with the potential of the tumor to metastasize. Materials and Methods Y79 retinoblastoma cells were obtained from the American Type Culture Collection in Rockville, MD, and maintained in culture as described previously [8]. Human RPE-CM and control non-conditioned media were kindly provided by Dr. Dean Bok of the Jules Stein Eye Institute at UCLA. All other reagents were obtained from Sigma Chemical Co. Cell culture
Human Y79 retinoblastoma cells were induced to differentiate into neuron-like cells by ,human RPE-CM as described previously [8]. The cells were stimulated by growth in 50% RPE-CM for 2 weeks, then seeded onto glass coverslips coated with 250 pg/ml of poly-Dlysine. Twenty-four hours following substratum attachment, cultures were supplemented with 10 pg/ml of soluble laminin. Differentiated cells were harvested 10 days after laminin treatment. Control cells were grown in 50% non-conditioned medium and were treated similarly to differentiated cells. Catecholamines
Concentrations of catecholamines, indolealkyl amines and their metabolites were measured in three different groups of Y79 cell cultures: Undifferentiated suspension cultures grown in 50% non-conditioned medium for 2 weeks; Stimulated - suspension cultures grown in 50% RPE-CM for 2 weeks; Differentiated - cells stimulated as in stimulated cells, then seeded onto a poly-Dlysine substratum and further cultured for 10 days. High pressure liquid chromatography Cells from each of the 3 culture conditions
were harvested, washed with 0.1 M phosphate
209 Table 1. Modulations in the or without RPE-CM treatment.
concentrations of 5-HT, 5-HIAA, HVA and HMPLA (pmol/mg Results are the average
of two separate
protein) in Y79 cells with
experiments.
Y79 Cells
5-HT
5-HIAA
3-MTY
HVA
HMPLA
Non-differentiated Stimulated Differentiatiated
9.3 0 0
35.5 0 0
0 130.8 637.8
0 10.2 857.8
0 35.4 137.3
buffered saline (pH 7.4)) counted and pelleted by centrifugation. Cells were resuspended at about 8 x lo7 cells/ml in ice-cold 0.1 M trichloroacetic acid, 0.1 mM EDTA and 0.01 M sodium acetate (pH 3.85, aqueous 0.1 M TCA) and sonicated. Sonicates were centrifuged at 10 000 x g for 5 min. The supernatants were stored at -8OOC (for 2-3 weeks) for HPLC analyses. Pellets were stored at - 80°C for subsequent protein assay. Supernatants were analyzed for catecholamines, indolealkyl amines and their metabolites by reverse phase, ion pair HPLC. The HPLC system was fitted with a C-18 reverse
0
lo
20
30
phase analytical column (250 x 4.6 mm) and an electrochemical detector [3]. Compounds were separated using a variety of mobile phase conditions. The mobile phases consisted of mixtures of methanol (MeOH) and aqueous 0.1 M TCA, which were varied with respect to MeOH content and pH to obtain appropriate separations of the compounds. Mobile phases consisting of 2% MeOH (pH 3.8)) 5% MeOH (pH 3.7) or 7% MeOH (pH 3.85) were used for the separation of serotonin (5-HT), 5hydroxyindoleacetic acid (5-HIAA), 3methoxytyrosine (3-MTY), homovanillic acid (HVA), 3-methoxy-4-hydrophenylacetic acid
40 llME((min)
50
60
70
Fig. 1. Upper: Chromatogram of indolealkyl amines in undifferentiated Y79 cells (cultured in non-conditioned medium containing 15% fetal calf serum), indicating the presence of 5-HT which is peak 7(9.3 pmol/mg protein), 8.12 x 10’ cells/ml. Lower: Coinjection of 5-HT standard with the sample above showing coelution at peak 7. Chromatographic conditions: 7% MeOH, 93% 0.1 M TCA (pH 3.85).
(HMPLA) and an unidentified compound (Compound 3). Quantitation of compounds depended on the use of the internal standard, 5,6_dihydroxybenzylamine (DBZ) . Electron pammagnetic resonance spectrometry (EPR) Spectra were recorded at 23’ and 100 kHz with a Bruker ECS 106 spectrometer. Other parameters were modulation amplitude 1.0 G, microwave power 20 mW and time constant 1.3 s.
I
0
*.
.
.
.
10
.
.
1
.
.
.
.
.
.
*
.
20
.
30
TIME(min)
Fig. 2.
Upper: Chromatogram of indolealkyl amines in undifferentiated Y79 cells (cultured in non-conditioned medium containing 15% fetal calf serum), indicating the presence of 5-HIAA which is peak 5 (35.5 pmol/mg protein), 8.12 x lo7 cells/ml. Lower: Coinjection of 5HIAA standard with the sample above indicating coelution of 5-HIAA with peak 5. Chromatographic conditions: 7% MeOH, 93% 0.1 M TCA (pH 3.85).
Y79 cells cultured in medium containing serum typically grew in suspension as aggregates of relatively undifferentiated, round cells. However, when cultured in 50% RPECM for 2 weeks and seeded onto pOly-Dlysine substratum, a high frequency of cellular differentiation and attachment was observed. !
I
2
k---
2 1
I
3
1 I
L, 0 Fig. 3.
Upper:
t
0 20 TIME (min)
Chromatogram of chatecholamines in RPE-CM stimulated Y79 cells, indicating the presence of 3-MTY which is peak 2 (130.8 pmol/mg protein) and an unidentified compound (Compound 3) which is peak 3, 8.12 x lo7 cells/ml. Lower: Coinjection of 3-MTY standard with the sample above showing coelution with peak 2. Chromatographic conditions: 2% MeOH, 98% 0.1 M TCA (pH 3.8).
IO
20
30
TIME(min)
30
Fig. 4.
Upper: Chromatogram of catecholamines in differentiated Y79 cells indicating the presence of 3-MTY which is peak 2 (637.8 pmol/mg protein) and an unidentified compound (Compound 3) which is peak 3, 5 x lo6 cells/ml. Lower: Coinjection of 3-MTY standard with the sample from differentiated Y79 cells incoelution at peak 2. Chromatographic dicating conditions: 2% MeOH, 98% 0.1 M TCA (pH 3.8).
211
a
IO
20
30
40
50
TIME (min)
Fis. 5. Upper: Chromatogram of catecholamines in RPE-CM stimulated Y79 cells, indicating the presence of HMPLA which is peak 4 (35.4 pmol/mg protein), an unidentified compound (Compound 3) which is peak 3 and HVA which is peak 6 (10.2 pmol/mg protein), 8.12 x lo7 cells/ml. Lower: Coinjection of HMPLA and HVA standards with the sample above indicating coelution at peak 4 (HMPLA) and peak 6 (HVA). Chromatographic conditions: 5% MeOH, 95% 0.1 M TCA (pH 3.7).
The process of differentiation was also associated with a marked change in neurotransmitter phenotype (Table I). The process of stimulation and differentiation of Y79 cells involved a switch from serotonergic cells to a phenotype dominated by dopa metabolites. Each of the compounds listed in Table I was carefully identified by coinjection with synthetic standards as demonstrated in Figs. l-6. Dopamine and dopa were not present to a measurable extent in any cell fraction or in the medium. An unidentified compound, Compound 3, was present as a peak with a retention time longer than 30 min, under some conditions, and was present only in stimulated and differentiated cells (Figs. 3 -5). Compound 3 does not coelute with dopamine, dopa, epinephrine, norepinephrine or any of their available metabolites.
EPR of dried cell lysates demonstrated a one-line spectrum identical to sythetic dopa melanin and identical to published spectra [7]. The spectrum exhibited the following characteristics: g = 2.004 and line width = 8 G. This may demonstrate that dopa is a precursor for melanin in these cells, perhaps analogous to retinal pigmented epithelial cells. Discussion As Y79 cells differentiate, they change from serotonergic cells to neuron-like cells containing the metabolites of dopa and melanin. 5-HT and its metabolite, 5-HIAA, are present only in non-differentiated cells. 3-MTY is a catechol0-methyltransferase metabolite of dopa that is present in stimulated and differentiated cells. HVA and HMPLA are also metabolites of
IO
20
30 TIME( min)
40
50
8
Fig. 6. Upper: Chromatogram of catecholamines in differentiated Y79 cells, indicating the presence of HMPLA which is peak 4 (137.3 pmol/mg protein) and HVA which is peak 6 (857.8 pmol/mg protein). Compound 3 is not resolved from the elution front under these conditions, 5 x lo6 cells/ml. Lower: Coinjection of HMPLA and HVA standards with the sample above indicating co-elution at peak 4 (MMPLA) and peak 6 (HVA) . Chromatographic conditions: 5% MeOH, 95% 0.1 M TCA (pH 3.7).
dopa and are present in stimulated and differentiated cells. HVA can be a metabolite of dopamine, unlike 3-MTY and HMPLA. Neither dopa nor dopamine were detected in the cells at any time, which demonstrates that neither compound is stored by the cells for potential use as a neurotransmitter. Dopa, and perhaps dopamine, is probably synthesized in the cells, but turns over rapidly with the accumulation of metabolic products and melanin. This alteration in cellular phenotype is not likely to be associated with subcloning of a particular cell type, in that very little cell death is seen during the stimulation and differentiation process. Dopa is not known to be a neurotransmitter and is probably not functioning as a neurotransmitter in Y79 cells. The data suggest that Y79 cell phenotype is altered by a factor(s) secreted by RPE cells.
The differentiation factor may function in a number of ways. It may influence the uptake systems for dopa and serotonin precursors. It may also alter the expression of genes that code for neurotransmitter synthetic or metabolic enzymes. Determination of the degree of cellular differentiation in retinoblastoma tumors could be a valuable prognostic indicator in patients. It could be that more aggressive tumors will be composed of serotonergic cells, whereas the more differentiated dopanergic tumor cells may be less metastatic. In addition, analysis of urinary serotonin or dopa metabolite levels in retinoblastoma patients may help determine if metastases exist and may provide clues about the nature of the metastases. It may be of interest to examine serotonergic or dopanergic drugs to see if they can alter the survival or
213
differentiation of Y79 cells, and ultimately be of use in the treatment of retinoblastoma.
current 5
Acknowledgements 6
The authors thank Jean Foster and Juana Roy for expert technical assistance. This work was supported, in part, by a grant from the National Eye Institute to LVJ.
7
References 8 Abramson, D.H., Notterman, R.B., Ellsworth, R.M. and Kitchin, F.D. (1983) Retinoblastoma treated in infants in the first six months of life. Arch. Ophthalmol., 101, 1362 - 1366. Ehringer, B. (1976) Biogenic monoamines as transmitters in the retina. In: Transmitters in the Visual Process, pp. 145- 163. Editor: S. Bonting. Pergamon Press, Oxford. Kalivas, P. W. (1985) Sensitization to repeated enkephalin administration into the ventral tegmental area of the rat. II. Involvement of the mesolimbic dopamine system. J. Pharmacol. Exp. Ther., 235, 544- 550. Kyritsis, A.P., Tsokos, M. and Chader, G.J. (1986) Control of retinoblastoma cell growth by differentiation agents:
9
10
11
work and future
directions.
Anticancer
Res.,
6,
465-473. Kyritsis, A.P., Tsokos, M. and Chader, G. J. (1987) Behavior of human retinoblastoma cells in tissue culture. Prog. Ret. Res., 6, 245-274. Kyritsis, A.P., Tsokos, M., Triche, T.J. and Chader, G.J. (1984) Retinoblastoma: origin from a primitive neuroectodermal cell? Nature, 307, 471-473. Scaly, R.C., Felix, C.C., Hyde, J.S. and Swartz, H.M. (1980) Structure and reactivity of melanins: influence of free radicals and metal ions. In: Free Radicals in Biology, Vol. 4, pp. 209-259. Editor: W. A. Pryor. Academic Press, New York. Tombran-Tink, J. and Johnson, L.V. (1989) Neuronal differentiation of retinoblastoma cells induced by medium conditioned by human RPE cells. Invest. Ophthalmol. Vis. Sci., 30, 1700- 1707. Tombran-Tink, J. and Johnson, L.V. (1989) RPE-54 - A unique RPE product with neuronal differentiating activity. Invest. Ophthalmol. Vis. Sci., 30, 414. Tombran-Tink, J. and Johnson, L.V. (1991) PEDF: a pigment epithelium derived factor with potent neuronal differentiation activity. Exp. Eye Res., 53, 411-414. Yorek, M.A., Strom, D.K. and Spector, A.A. (1987) Synthesis and high affinity uptake of serotonin and dopamine by human Y79 retinoblastoma cells. J. Neurochem., 49, 1316- 1323.
Cancer Letters, 68 (1993) 207 - 213 Elsevier Scientific Publishers Ireland Ltd.
Effects of medium conditioned by retinal pigmented epithelial cells on neurotransmitter phenotype in retinoblastoma cells Lori K. Klaidmanb, Joyce Tombran-Tink”, Lincoln V. Johnsona ‘Department Toxicology,
of Anatomy
and Cell Biology,
School of Pharmacy,
(Received
18 November
1992)
(Accepted
16 December
1992)
School of Medicine,
Uniuersity of Southern
Summary We previously reported that medium conditioned by retinal pigmented epithelial cells can induce cellular differentiation in human retinoblastoma cells. Extensive neurite outgrowth, of neuronal marker increased expression molecules and decreased expression of glial marker molecules are characteristic of the dijjerentiated phenotype. In the studies described here, we examine whether modulations in the expression of potential neurotransmitter molecules, catecholamines and indolealkyl amines, might be associated with the differentiation of retinoblastoma cells. Concentrations of seroacid, S-methtonin, Shydroxyindoleacetic homovanillic acid, and 3oxytyrosine, methoxy-4-hydroxyphenylacetic acid in extracts of differentiated and undifferentiated retinoblastoma cells were assessed by HPLC. The results show that serotonin and its metabolite, Shydroxyindoleacetic acid, are characteristically present in undifferentiated cells. Dopa metabolites, 3-methoxytyrosine, homovanillic acid and 3-methoxy-4-hydroxyphenylacetic acid, are uniquely present in dijCorrespondence to: James D. Adams, Jr., Dept. Molecular Pharmacology and Toxicology, School of Pharmacy, University of Southern California, 1985 Zonal Ave PSC 510, Los Angeles, CA 90033, U.S.A.
0304-3835/93/$06.00 Printed and Published
James
0 1993 Elsevier Scientific Publishers in Ireland
D. Adams,
and bDepartment
California,
Los Angeles,
Jr.b and
of Molecular
Pharmacology
and
CA (U.S.A.)
jerentiated cells. It appears that differentiation of retinoblastoma cells induced by factors secreted by retinal pigmented epithelial cells involves a switch from a serotonergic phenotype to one dominated by metabolites of dopa. These findings may provide clues about the factors that control retinoblastoma cells and metastasis.
pigmented Keywords: retinoblastoma; epithelium; serotonin; dopa; differentiation
Introduction Retinoblastoma is a tumor of the eye that is seen most commonly in children under the age of 5 years old. Metastasis of the tumor is uncommon in infants [l] . Malignant spread of the tumor is more common in children of ages between 1 and 5 years old and has a very poor prognosis. Therefore, it is important for the neoplasm to be diagnosed early and treated without delay. Retinoblastoma is probably derived from a mixture of primitive, multipotential retinoblastic cells [6]. Cultured human retinoblastoma cells (Y79) can be induced to differentiate into cells with neuronal, glial or photoreceptor cell characteristics [4,5]. Undifferentiated Y79 cells Ireland Ltd.
208
have many biochemical characteristics typical of neuronal cells. They express neuronal marker molecules [5] and have uptake systems for dopamine and serotonin [ 111. In the retina, dopamine and serotonin have been implicated as neurotransmitters in a subclass of amacrine cells [2]. In a recent report, we showed that medium conditioned by cultured human retinal pigmented epithelial cells (RPE-CM) induces neuronal differentiation in Y79 cells [8]. A critical period of stimulation with RPE-CM prior to substratum attachment promotes a high frequency of neuronal differentiation in Y79 cells. Extensive neurite outgrowth, increased expression of neuronal marker molecules and decreased expression of glial marker molecules are associated with phenotypic changes and differentiation of Y79 cells [8]. The differentiation inducing activity present in RPE-CM is associated with a 50 -55 kDa pigmented epithelial differentiating factor (PEDF) that can be isolated and purified from RPE-CM 19,lo]. A childhood tumor related to retinoblastoma is neuroblastoma, which can be monitored in its metastatic form due to the elaboration of norepinephrine by the cancer cells. When neuroblastoma becomes apparent in children under the age of 1, regression is common, which involves either spontaneous cytolysis of the tumor cells or differentiation into ganglioneuroma or ganglioneuroblastoma cells. Retinoblastoma behaves similarly, being much more metastatic when present in children older than the age of 1, as mentioned above. The retina is not completely formed at birth, but requires a few months of development postnatalIy. It is possible that a differentiation factor that becomes less abundant in more differentiated retina leads to the maturation of aggressive retinoblastoma tumors. A differentiation factor, such as PEDF, that is present in immature retina may protect the retina from retinoblastoma by inducing cellular differentiation of retinoblastoma cells. We therefore became interested in the ability of PEDF to alter the neurotransmitter phenotype in retinoblastoma
cells. Possible alterations in neurotransmitter
phenotype may provide a valuable clinical marker of the differentiation of the retinoblastoma, which could correlate with the potential of the tumor to metastasize. Materials and Methods Y79 retinoblastoma cells were obtained from the American Type Culture Collection in Rockville, MD, and maintained in culture as described previously [8]. Human RPE-CM and control non-conditioned media were kindly provided by Dr. Dean Bok of the Jules Stein Eye Institute at UCLA. All other reagents were obtained from Sigma Chemical Co. Cell culture
Human Y79 retinoblastoma cells were induced to differentiate into neuron-like cells by ,human RPE-CM as described previously [8]. The cells were stimulated by growth in 50% RPE-CM for 2 weeks, then seeded onto glass coverslips coated with 250 pg/ml of poly-Dlysine. Twenty-four hours following substratum attachment, cultures were supplemented with 10 pg/ml of soluble laminin. Differentiated cells were harvested 10 days after laminin treatment. Control cells were grown in 50% non-conditioned medium and were treated similarly to differentiated cells. Catecholamines
Concentrations of catecholamines, indolealkyl amines and their metabolites were measured in three different groups of Y79 cell cultures: Undifferentiated suspension cultures grown in 50% non-conditioned medium for 2 weeks; Stimulated - suspension cultures grown in 50% RPE-CM for 2 weeks; Differentiated - cells stimulated as in stimulated cells, then seeded onto a poly-Dlysine substratum and further cultured for 10 days. High pressure liquid chromatography Cells from each of the 3 culture conditions
were harvested, washed with 0.1 M phosphate
209 Table 1. Modulations in the or without RPE-CM treatment.
concentrations of 5-HT, 5-HIAA, HVA and HMPLA (pmol/mg Results are the average
of two separate
protein) in Y79 cells with
experiments.
Y79 Cells
5-HT
5-HIAA
3-MTY
HVA
HMPLA
Non-differentiated Stimulated Differentiatiated
9.3 0 0
35.5 0 0
0 130.8 637.8
0 10.2 857.8
0 35.4 137.3
buffered saline (pH 7.4)) counted and pelleted by centrifugation. Cells were resuspended at about 8 x lo7 cells/ml in ice-cold 0.1 M trichloroacetic acid, 0.1 mM EDTA and 0.01 M sodium acetate (pH 3.85, aqueous 0.1 M TCA) and sonicated. Sonicates were centrifuged at 10 000 x g for 5 min. The supernatants were stored at -8OOC (for 2-3 weeks) for HPLC analyses. Pellets were stored at - 80°C for subsequent protein assay. Supernatants were analyzed for catecholamines, indolealkyl amines and their metabolites by reverse phase, ion pair HPLC. The HPLC system was fitted with a C-18 reverse
0
lo
20
30
phase analytical column (250 x 4.6 mm) and an electrochemical detector [3]. Compounds were separated using a variety of mobile phase conditions. The mobile phases consisted of mixtures of methanol (MeOH) and aqueous 0.1 M TCA, which were varied with respect to MeOH content and pH to obtain appropriate separations of the compounds. Mobile phases consisting of 2% MeOH (pH 3.8)) 5% MeOH (pH 3.7) or 7% MeOH (pH 3.85) were used for the separation of serotonin (5-HT), 5hydroxyindoleacetic acid (5-HIAA), 3methoxytyrosine (3-MTY), homovanillic acid (HVA), 3-methoxy-4-hydrophenylacetic acid
40 llME((min)
50
60
70
Fig. 1. Upper: Chromatogram of indolealkyl amines in undifferentiated Y79 cells (cultured in non-conditioned medium containing 15% fetal calf serum), indicating the presence of 5-HT which is peak 7(9.3 pmol/mg protein), 8.12 x 10’ cells/ml. Lower: Coinjection of 5-HT standard with the sample above showing coelution at peak 7. Chromatographic conditions: 7% MeOH, 93% 0.1 M TCA (pH 3.85).
(HMPLA) and an unidentified compound (Compound 3). Quantitation of compounds depended on the use of the internal standard, 5,6_dihydroxybenzylamine (DBZ) . Electron pammagnetic resonance spectrometry (EPR) Spectra were recorded at 23’ and 100 kHz with a Bruker ECS 106 spectrometer. Other parameters were modulation amplitude 1.0 G, microwave power 20 mW and time constant 1.3 s.
I
0
*.
.
.
.
10
.
.
1
.
.
.
.
.
.
*
.
20
.
30
TIME(min)
Fig. 2.
Upper: Chromatogram of indolealkyl amines in undifferentiated Y79 cells (cultured in non-conditioned medium containing 15% fetal calf serum), indicating the presence of 5-HIAA which is peak 5 (35.5 pmol/mg protein), 8.12 x lo7 cells/ml. Lower: Coinjection of 5HIAA standard with the sample above indicating coelution of 5-HIAA with peak 5. Chromatographic conditions: 7% MeOH, 93% 0.1 M TCA (pH 3.85).
Y79 cells cultured in medium containing serum typically grew in suspension as aggregates of relatively undifferentiated, round cells. However, when cultured in 50% RPECM for 2 weeks and seeded onto pOly-Dlysine substratum, a high frequency of cellular differentiation and attachment was observed. !
I
2
k---
2 1
I
3
1 I
L, 0 Fig. 3.
Upper:
t
0 20 TIME (min)
Chromatogram of chatecholamines in RPE-CM stimulated Y79 cells, indicating the presence of 3-MTY which is peak 2 (130.8 pmol/mg protein) and an unidentified compound (Compound 3) which is peak 3, 8.12 x lo7 cells/ml. Lower: Coinjection of 3-MTY standard with the sample above showing coelution with peak 2. Chromatographic conditions: 2% MeOH, 98% 0.1 M TCA (pH 3.8).
IO
20
30
TIME(min)
30
Fig. 4.
Upper: Chromatogram of catecholamines in differentiated Y79 cells indicating the presence of 3-MTY which is peak 2 (637.8 pmol/mg protein) and an unidentified compound (Compound 3) which is peak 3, 5 x lo6 cells/ml. Lower: Coinjection of 3-MTY standard with the sample from differentiated Y79 cells incoelution at peak 2. Chromatographic dicating conditions: 2% MeOH, 98% 0.1 M TCA (pH 3.8).
211
a
IO
20
30
40
50
TIME (min)
Fis. 5. Upper: Chromatogram of catecholamines in RPE-CM stimulated Y79 cells, indicating the presence of HMPLA which is peak 4 (35.4 pmol/mg protein), an unidentified compound (Compound 3) which is peak 3 and HVA which is peak 6 (10.2 pmol/mg protein), 8.12 x lo7 cells/ml. Lower: Coinjection of HMPLA and HVA standards with the sample above indicating coelution at peak 4 (HMPLA) and peak 6 (HVA). Chromatographic conditions: 5% MeOH, 95% 0.1 M TCA (pH 3.7).
The process of differentiation was also associated with a marked change in neurotransmitter phenotype (Table I). The process of stimulation and differentiation of Y79 cells involved a switch from serotonergic cells to a phenotype dominated by dopa metabolites. Each of the compounds listed in Table I was carefully identified by coinjection with synthetic standards as demonstrated in Figs. l-6. Dopamine and dopa were not present to a measurable extent in any cell fraction or in the medium. An unidentified compound, Compound 3, was present as a peak with a retention time longer than 30 min, under some conditions, and was present only in stimulated and differentiated cells (Figs. 3 -5). Compound 3 does not coelute with dopamine, dopa, epinephrine, norepinephrine or any of their available metabolites.
EPR of dried cell lysates demonstrated a one-line spectrum identical to sythetic dopa melanin and identical to published spectra [7]. The spectrum exhibited the following characteristics: g = 2.004 and line width = 8 G. This may demonstrate that dopa is a precursor for melanin in these cells, perhaps analogous to retinal pigmented epithelial cells. Discussion As Y79 cells differentiate, they change from serotonergic cells to neuron-like cells containing the metabolites of dopa and melanin. 5-HT and its metabolite, 5-HIAA, are present only in non-differentiated cells. 3-MTY is a catechol0-methyltransferase metabolite of dopa that is present in stimulated and differentiated cells. HVA and HMPLA are also metabolites of
IO
20
30 TIME( min)
40
50
8
Fig. 6. Upper: Chromatogram of catecholamines in differentiated Y79 cells, indicating the presence of HMPLA which is peak 4 (137.3 pmol/mg protein) and HVA which is peak 6 (857.8 pmol/mg protein). Compound 3 is not resolved from the elution front under these conditions, 5 x lo6 cells/ml. Lower: Coinjection of HMPLA and HVA standards with the sample above indicating co-elution at peak 4 (MMPLA) and peak 6 (HVA) . Chromatographic conditions: 5% MeOH, 95% 0.1 M TCA (pH 3.7).
dopa and are present in stimulated and differentiated cells. HVA can be a metabolite of dopamine, unlike 3-MTY and HMPLA. Neither dopa nor dopamine were detected in the cells at any time, which demonstrates that neither compound is stored by the cells for potential use as a neurotransmitter. Dopa, and perhaps dopamine, is probably synthesized in the cells, but turns over rapidly with the accumulation of metabolic products and melanin. This alteration in cellular phenotype is not likely to be associated with subcloning of a particular cell type, in that very little cell death is seen during the stimulation and differentiation process. Dopa is not known to be a neurotransmitter and is probably not functioning as a neurotransmitter in Y79 cells. The data suggest that Y79 cell phenotype is altered by a factor(s) secreted by RPE cells.
The differentiation factor may function in a number of ways. It may influence the uptake systems for dopa and serotonin precursors. It may also alter the expression of genes that code for neurotransmitter synthetic or metabolic enzymes. Determination of the degree of cellular differentiation in retinoblastoma tumors could be a valuable prognostic indicator in patients. It could be that more aggressive tumors will be composed of serotonergic cells, whereas the more differentiated dopanergic tumor cells may be less metastatic. In addition, analysis of urinary serotonin or dopa metabolite levels in retinoblastoma patients may help determine if metastases exist and may provide clues about the nature of the metastases. It may be of interest to examine serotonergic or dopanergic drugs to see if they can alter the survival or
213
differentiation of Y79 cells, and ultimately be of use in the treatment of retinoblastoma.
current 5
Acknowledgements 6
The authors thank Jean Foster and Juana Roy for expert technical assistance. This work was supported, in part, by a grant from the National Eye Institute to LVJ.
7
References 8 Abramson, D.H., Notterman, R.B., Ellsworth, R.M. and Kitchin, F.D. (1983) Retinoblastoma treated in infants in the first six months of life. Arch. Ophthalmol., 101, 1362 - 1366. Ehringer, B. (1976) Biogenic monoamines as transmitters in the retina. In: Transmitters in the Visual Process, pp. 145- 163. Editor: S. Bonting. Pergamon Press, Oxford. Kalivas, P. W. (1985) Sensitization to repeated enkephalin administration into the ventral tegmental area of the rat. II. Involvement of the mesolimbic dopamine system. J. Pharmacol. Exp. Ther., 235, 544- 550. Kyritsis, A.P., Tsokos, M. and Chader, G.J. (1986) Control of retinoblastoma cell growth by differentiation agents:
9
10
11
work and future
directions.
Anticancer
Res.,
6,
465-473. Kyritsis, A.P., Tsokos, M. and Chader, G. J. (1987) Behavior of human retinoblastoma cells in tissue culture. Prog. Ret. Res., 6, 245-274. Kyritsis, A.P., Tsokos, M., Triche, T.J. and Chader, G.J. (1984) Retinoblastoma: origin from a primitive neuroectodermal cell? Nature, 307, 471-473. Scaly, R.C., Felix, C.C., Hyde, J.S. and Swartz, H.M. (1980) Structure and reactivity of melanins: influence of free radicals and metal ions. In: Free Radicals in Biology, Vol. 4, pp. 209-259. Editor: W. A. Pryor. Academic Press, New York. Tombran-Tink, J. and Johnson, L.V. (1989) Neuronal differentiation of retinoblastoma cells induced by medium conditioned by human RPE cells. Invest. Ophthalmol. Vis. Sci., 30, 1700- 1707. Tombran-Tink, J. and Johnson, L.V. (1989) RPE-54 - A unique RPE product with neuronal differentiating activity. Invest. Ophthalmol. Vis. Sci., 30, 414. Tombran-Tink, J. and Johnson, L.V. (1991) PEDF: a pigment epithelium derived factor with potent neuronal differentiation activity. Exp. Eye Res., 53, 411-414. Yorek, M.A., Strom, D.K. and Spector, A.A. (1987) Synthesis and high affinity uptake of serotonin and dopamine by human Y79 retinoblastoma cells. J. Neurochem., 49, 1316- 1323.