Med Electron Microsc (1998) 31:24-30

9 The Clinical Electron MicroscopySocietyof Japan 1998

Jojiro Nakada 9 Nozomu Furuta 9 Motoshi Kawahara Shoichi Onodera 9 Yukihiko Oishi 9 Masaaki Kawada Tomokazu Matsuura 9 Satoshi Hasumura Seisbi N a g a l n o r i

Effects of nerve growth factor and glucocorticoid on cultured human

pheochromocytoma cells

Received: September 25, 1997 / Accepted: November 21, 1997 A b s t r a c t Primary cell cultures of two human pheochro-

mocytomas (PC) that were associated with high serum levels of adrenaline and noradrenaline were developed to study the effects of nerve growth factor (NGF) and dexamethasone on the morphology and function of PC cells in vitro. By phase-contrast microscopy, cultured cells were small and hyperchromatic on the first day of culture; neurite-like processes that extended to other cells developed several days later and were maintained for more than 3 months. NGF (100ng/ml), dexamethasone (10-5M), or NGF + dexamethasone were added to the culture media 2 weeks after the cultured cells had stabilized. Catecholamine concentrations in the medium were maintained at higher levels after addition of NGF, dexamethasone, or NGF + dexamethasone as compared to control cells. In the presence of NGF, extension of neurite-like processes was clearly accelerated, while high levels of dexamethasone inhibited growth of processes. These in vitro studies showed that the addition of NGF or the removal of dexamethasone induces differentiation of adrenal neurons present in pheochromocytomas, suggesting that adrenocortical steroid hormones influence the morphological control of adrenal medullary cells. K e y w o r k s Human . Pheochromocytoma - Nerve growth

developmental process and that nerve growth factor (NGF) is a protein involved as a neurotropic factor) In a study of NGF, Unsicker et al. 2 found that cultured chromaffin cells obtained from young rats extended their axons in the presence of NGF and that corticosteroids disturbed process formation. He concluded that chromaffin cells of the adrenal medulla lack neuronal processes in vivo because of the effect of cortical hormones. On the other hand, Najoks et al. 3 isolated chromaffin cells from bovine fetuses and both young and adult bovine animals to study the elongation of neurite-like processes in cultured cells; they reported that only the chromaffin cells obtained from 3-month-old fetuses responded to NGF. Chromaffin cells obtained from adult rats did not appear to exhibit NGF-dependent growth of axons, 4 suggesting that in rats and cattle, the ability for NGF-dependent axonal growth is lost as chromaffin cells mature. In humans, extension of axons in response to NGF has been noted using chromaffin cells derived from adult adrenal glands s and human pheochromocytomas (PC). 6-s However, studies at the subcellular level using pheochromocytes derived from human PC have not been performed. In this study, NGF was added to human PC cells and its effect on neuronal growth was studied. Both morphology and function were specifically examined.

factor 9 Cultured cell

Materials and methods Introduction Materials It is well known that the intricate network which exists between nerve cells is established at an early stage in the

J. Nakada (~4) . N. Furuta 9M. Kawahara 9S. Onodera. Y. Oishi Department of Urology, Jikei University School of Medicine, 3-25-8 Nishishimbashi, Minato-ku, Tokyo 105, Japan Tel. +81-3-3433-1111,ext.3561; Fax +81-3-343%2389 M. Kawada 9T. Matsuura. S. Hasumura 9S. Nagamori First Department of Internal Medicine, Jikei University School of Medicine, Tokyo, Japan

Cells obtained from two resected PC were cultured. The frst tumor was from a 37-year-old man with blood pressure within the normal range. Endocrinological testing revealed his adrenaline was 383pg/ml (normal, <100pg/ml), noradrenaline, 1309pg/ml (normal, 100-450pg/ml), and dopamine, 20pg/ml (normal, <20pg/ml). Abdominal computed tomography (CT) revealed a round tumor approximately 3 cm in diameter with a clear border, probably arising from the left adrenal gland.

25

The other tumor was from a 48-year-old man with a paroxysmally high systolic blood pressure, 200-220mmHg, and elevated serum catecholamines. Endocrinological testing revealed an adrenaline of 323 pg/ml, a noradrenaline of 1486pg/ml, and a dopamine of 18pg/ml. Abdominal CT scan revealed a 2 • 3 cm tumor on the right adrenal gland. On the basis of the clinical findings, both patients were diagnosed with PC and abdominal tumor resections were performed. The resected tumors measured 2 • 3 • 3 cm (case 1) and 2.7 X 3 x 3 cm (case 2), and both were yellow brown and solid. Light microscopic findings in both cases were similar, showing comparatively large polygonal or round parenchymal cells containing bright cytoplasmic microgranules in an alveolar or trabecular pattern. The cellular structure and the nuclear features were not considered malignant by histology. Methods Tumor cells were cultured as previously described. 9'1~After sectioning the resected tumors with a scalpel, cells were dispersed by shaking for about an hour in Williams' E medium containing 1000U/ml dispase (Godo Shusei, Tokyo, Japan), fltered through a stainless steel mesh, and centrifuged at 60 • g for 3 min. The isolated cells in suspension were then recovered. Cells were diluted in Williams' E medium containing 10% fetal bovine serum (FBS) and dispersed on a plastic dish coated with type 1 collagen (Sigma Chemical, St. Louis, MO, USA). Cells were incubated in 5% COg and 95% air at 37~ and the medium was changed every 2-3 days. After about 2 weeks of incubation, the stability of the cultured cells was confirmed and t00ng/ml human recombinant NGF-beta (Sigma), 10 5M dexamethasone (Sigma), or 100ng/ml NGF + 10-5M dexamethasone were added and the functional and morphological features of the cultured cells were observed. Specifically, three catecholamines (adrenaline, noradrenaline, and dopamine) present in the medium were assayed throughout incubation by enzyme immunoassay. Cells were observed under a phase-contrast microscope and an electron microscope. For electron microscopy, cultured cells were fixed in 2.5% glutaraldehyde and postfixed in 1% osmium tetroxide. After dehydrating with graded ethanols, sections were embedded in Epon. Ultrathin sections stained with uranyl acetate were then observed (JEOL 1200EX).

However, in case 1, the medium concentrations of adrenaline (unmeasurable) and dopamine (65pg/ml) were quite low at 3 months, while noradrenaline maintained a high concentration (1330pg/ml). Noradrenaline levels fell to 46pg/ml at 4 months and to 20pg/ml by 6 months. About 10 months later, noradrenaline was still detectable at 18pg/ml (Table 1). In case 2, adrenaline and dopamine were both unmeasurable after 36 days of culture, while noradrenaline remained elevated. Noradrenaline was measured at 24pg/ ml on day 42 of culture but became undetectable after 60 days. The cell culture was then suspended at about 5 months (Table 2). As a preliminary study, catecholamine levels were measured on the 5th, 8th, and 15th days of culture after adding NGF to the PC cells of case 1. Table 3 shows levels of the measured catecholamines for each culture day. We could not compare catecholamine formation in cultures with and without N G F because cell density was not controlled. However, catecholamine production in the NGF-supplemented culture tended to be maintained at a higher level than in the control culture. The medium was exchanged 3 days after adding agents to PC-cultured cells from case 2. Figure 1 shows changes in catecholamine concentrations after adding NGF, dexamethasone, or N G F + dexamethasone to cultured ceils. The abscissa shows days after addition of reagents. The ordinates were noted as 1000, based on control values as plotted

Table 1. Catecholamine concentrations in cultured media of case 1 pheochromocytoma cells Culture days

Adrenaline (pg/ml)

Noradrenaline (pg/ml)

Dopamine (pg/ml)

1 3 6 14 28 60 90 120 180 300

1593000 155000 16300 11200 1200 ND ND ND ND ND

3015000 642300 27000 36200 14200 4640 1330 46 20 18

10800 2800 390 2680 2060 258 65 ND ND ND

ND, not detected.

Table 2. Catecholamine concentrations in cultured media of case 2 pheochromocytoma cells

Results Effects of NGF and dexamethasone on catecholamine concentrations The changes in the levels of the three catecholamines measured in the medium over time are shown in Tables 1 and 2. PC cells from both cases showed very high levels of adrenaline and noradrenaline on the first and third days of culture.

Culture days

1 3

7 21

36 42 60 ND, not detected.

Adrenaline (pg/ml)

Noradrenaline (pg/ml)

Dopamine (pg/ml)

1340800 10136 2421 704 ND ND ND

3015000 9227 8751 9630 218 24 ND

11620 37 28 871 ND ND ND

26 Table 3. Catecholamine concentrations of cultured cell media of case 1 after adding NGF Culture days

Adrenaline (pg/ml/day)

Noradrenaline (pg/mVday)

on semilogarithmic paper. Compared with the untreated

Dopamine (pg/ml/day)

Control

5 8 15

760 583 96

1020 1087 586

7.6 3.7 10

NGF (100ng/ml)

5 8 15

1860 969 486

3160 2709 6471

104 89.3 746

Morphological observation of h u m a n PC cells by phasecontrast microscopy Small r o u n d cells with dark cytoplasm were found on the first day of incubating PC cells from case 1 (Fig. 2A). The r o u n d cells a p p e a r e d polygonal or stellate in shape on days 5 to 7 (Fig. 2B). A f t e r 10-14 days incubation in Williams' E m e d i u m containing 10% FBS, the cells slightly e x t e n d e d their neurite-like processes. These cells continued to extend neurite-like processes even after 6 months of culture. A f t e r 10 days in culture with N G F , extension of neurite-like processes was clearly identified in the N G F - t r e a t e d group. By phase-contrast microscopy, similar findings were noted in the cultured PC cells of case 2. In the fourth week of incubation, the neurite-like processes were slightly elongated. N G F was a d d e d in the second week of incubation,

NGF, nerve growth factor.

I0000!

100(~ 1 O0 10

~

I

0 I

,

0

~

2

I

i

4

l

control culture, addition of either N G F or d e x a m e t h a s o n e m a i n t a i n e d catecholamine levels. A f t e r adding N G F , adrenaline and d o p a m i n e showed their highest levels as c o m p a r e d to initial samples on the 3rd day of incubation.

~ i

6

I

,

10

8

10000

Noradrenaline

10001

IO0

10

2

0

4

6

8

10

10000

Dopamine 1 O001 1 O0 0 10 1

,

0

control NGF |

2

~

l,

h

4

I

6

*

I

8

i

10

DAY Fig. 1. Catecholamine concentrations following addition of NGF (solid circles), dexamethasone (open squares), or NGF + dexamethasone (solid squares) to cultured cells (case 2). The abscissa shows days after adition of reagents; ordinates are noted as 1000 on the basis of control values as plotted on semilogarithmicpaper. Compared with the control untreated cells (open circles), catecholamine concentrations were maintained with addition of NGF (100ng/ml), dexamethasone (10-s M), or NGF + dexamethasone. NGF, nerve growth factor; DEX, dexamethasone

Fig. 2A,B. Morphological evaluation of PC cells (case 1) using phasecontrast microscopy. The small, round cells with dark cytoplasm were obtained on the first day of incubation (A). The round cells appeared polygonal and stellate in shape after 5 days (B). PC, pheochromocytoma

27 and 2 weeks later the neurite-like processes appeared to extend further than in the untreated group. On the other hand, growth of neurite-like processes decreased after dexamethasone was added. When both dexamethasone and NGF were added to cultures, neurite-like processes grew slowly (Fig. 3). Morphological observations of human PC cells by electron microscopy Electron microscopy of the original resected tumor from case 1 revealed tumor cells that contained numerous cytoplasmic granules and a round, centrally placed nucleus. At higher magnifications, numerous granules 100-300nm in diameter with a centrally placed electron-dense core and a peripheral halo under the limiting membrane were noted

Fig. 3A-D. Morphologicalevaluation of PC cells (case 2) using phasecontrast microscopy.In the 4th week of incubation, neurite-like processes are seen to extend slightly in control cells (A). When NGF (100ng/ml) was added during the second week of incubation, extending neurite-like processes were identified 2 weeks later (B). On addition of dexamethasone (10-SM), growth of neurite-like processes was inhibited (C). In the dexamethasone and NGF-treated group, neuritelike processes were present, but few in number (D). NGF, nerve growth factor; DEX, dexamethasone; PC, pheochromocytoma

(Fig. 4). After about 4 weeks of culture, cells with granules similar to those seen in the primary tumor were found. These granules were similar in size to those seen in the primary tissue and also contained electron-dense cores. However, an enveloping membrane surounding the granules was not clearly observed in cultured cells (Fig. 5). Two weeks after adding NGF to cell cultures, the neuritelike processes were more clearly observed in the NGFtreated group than the untreated group by electron microscopy. Scattered granules associated with microtubules or smooth endoplasmic reticulum were noted in the neurite-like processes (Fig. 6).

Discussion In this study, cultures of two human PC cells were maintained for 10 months and 5 months, respectively. Production of catecholamines was observed for 10 months in the first case and for 42 days in the second case. All catecholamine concentrations were very high in both cultures on the first and third days of incubation, possibly resulting from the release of catecholamines when cells were isolated from the resected tumors for cell culture. 6 In case 1, noradrenaline was also noted at a level of 1330pg/ml in the third month, which then declined to 20pg/ml by 6 months, a finding similar to that reported by Peragner and Walser. 6 Both tumors were adrenaline- and noradrenaline-producing PC. However, when the ratio of adrenaline to noradrenaline in the medium was maintained at 1:2, adrenaline became unmeasurable earlier, suggesting that adrenaline may not be produced during culture. Tischler et al. 4performed a detailed study of cultured chromaffin cells obtained from the adrenal medulla of young adult rats. On addition of dexarnethasone to the culture medium, catecholamine formation, including adrenaline, was maintained. The specific activity of tyrosine hydroxylase (TH), an enzyme that converts tyrosine to dopa, is also enhanced in the presence of dexamethasone. Dexamethasone also activates phenylethanolamine-N-methyltransferase (PNMT), an enzyme that converts noradrenaline to adrenaline. 11 Thus, it was speculated that adrenaline became unmeasurabie earlier in both our cases. There are few reports describing the changes in medium catecholamine concentrations over time with the addition of NGF. 7'8 In both our patients, catecholamine concentrations were maintained up to 12-15 days after adding NGF, although at reduced levels. It has been reported, in a chromaffin cell culture from mature rats, that TH activity is selectively induced on addition of NGF. 1~In chromaffin cell cultures from bovine fetuses, both TH activity and the specific activity of dopamine-[3-hydroxylase (DBH), an enzyme that converts dopamine to noradrenaline, are enhanced. 3 All three catecholamines measured were maintained by adding NGF to both of our PC cultures, wherein increased TH-specific activity is suggested. However, using a cell culture of human PC, Tischler et al. s found no differences in catecholamine concentrations in cell cultures treated with

Fig. 4. Electron photomicrograph of the original resected tumor (case 1) shows electrondense deposits in the secretory vesicles, numerous granules measuring 100-300nm with a centrally placed electron-dense core, and a peripheral halo under the limiting membrane. Bar 5 bum

Fig. S. Electron photomicrograph of the PC cultured cells (case 1) shows cells with granules similar to those seen in the primary tumor. However, the granule-enveloping membrane was not clearly observed in cultured cells. Bar 5 btm

29 Fig. 6. Electron photomicrograph of the PC cultured cells (case 1) after 2 weeks incubation with NGF (100ng/ml) shows that the neurite-like processes are more clearly observed with the addition of NGF. Scattered granules seen in the neurite-like processes appear to retain catecholamines associated with microtubules or smooth endoplasmic reticulum. Bar 5 ~tm

100ng/ml NGF. Evaluations were performed on the 21st day of culture. In our study, the effects of NGF on catecholamine production were limited to 15 days. Differences between Tischler's and our studies may be related to the sampling times of catecholamine measurements. Peragner et al. 7 also added 50ng/ml NGF to cultured cells obtained from two human PC and observed that the number of cytoplasmic granules increased, as medium catecholamine concentrations declined, when compared to non-NGF-treated cells because NGF acted to retain cellular catecholamines. We could not reproduce these findings. In our second case, all groups that were exposed to NGF, dexamethasone, or NGF + dexamethasone showed higher levels of catecholamines than control cells. In the NGF + dexamethasone-treated group, no additive effect on catecholamine concentrations was noted. These results suggest that TH activity was enhanced in the presence of NGF or dexamethasone] '4'~2In addition, the ratio of adrenaline to noradrenaline in the dexamethasone group was higher than that in the vehicle (data not shown). It is likely that dexamethasone contributes to the maintenance of PNMT activity. Phase-contrast microscopy of cultured cells revealed cell bodies with hyperchromatic, small, round nuclei on the first day and extending neurite-like processes with morphological cellular changes on the seventh day of incubation

in media lacking NGF. In support of this view, Jaques and Tobes 13 have speculated that the appearance of the neurite-like processes is caused by differences in cell density, number of passages, interactions with the different substrates (plastic, collagen, or polylysine), NGF synthesis by fibroblasts, or the morphological potential of the cells themselves. In both cases, NGF was added during the second week of cell culture. Growth of neurite-like processes was apparently accelerated by the addition of NGF. Using chromaffin cells obtained from human PC 68 and human adrenal medulla, 5 similar results were obtained by adding purified NGF isolated from mouse submandibular glands. In this study, the influence on the axons was confirmed using recombinant human NGF. This study probably represents the first trial of recombinant NGF using cultured cells from a human PC. In case 2, growth of neurite-like processes was inhibited when dexamethasone was added to the culture medium. It has been reported that NGF extends neurite-like processes in human PC and adult chromaffin cultured cells, but the simultaneous addition of dexamethasone inhibits the growth of these processes] ~8Our results compare favorably with these reports. Blood is supplied to the adrenal medulla by the suprarenal arteries, which run through the adrenal cortex and supply the capillaries that descend along the

30 cortical cell columns. M e d u l l a r y cells are in close contact with blood that contains adrenocortical h o r m o n e s at high concentrations. 141s O u r in vitro studies also suggest that high concentrations of adrenocortical steroid h o r m o n e s influence the morphological and functional formation of the adrenal medulla, a'4'11 Electron microscopy was only p e r f o r m e d on cells obtained from t u m o r patient 1, revealing a b u n d a n t cytoplasmic granules with a d i a m e t e r of 100-300 nm. Electron-dense deposits of the secretory vesicles a p p e a r to contain cate c h o l a m i n e s ] 6 The cultured cells contained granules of nearly the same size as those seen in the primary tumor. However, the m e m b r a n e envelopes of the granules were not clearly observed in cultured cells. The differences in p r i m a r y tissue and cultured cells m a y result from the conditions of p r e p a r i n g the tissue blocks and cell monolayers. In another electron microscopic study of the effect of adding N G F to cell cultures, Tischler et al. 5 have r e p o r t e d , using h u m a n chromaffin cells, that granules p r e s u m e d to be catecholamines are distributed with microtubules or smooth endoplasmic reticulum in neurite-like processes. That finding was r e p r o d u c e d in this study.

Conclusion Primary cultured cells can be used to help clarify the effect of N G F and d e x a m e t h a s o n e on the function and morphology of PC. Functionally, N G F and d e x a m e t h a s o n e were effective in p r o m o t i n g secretion of cathecholamines by PC cells in vitro. O n the other hand, by morphology, d e x a m e t h a s o n e inhibited the growth of neurite-like processes, while N G F induced process formation. These results suggest that adrenocortical steroid h o r m o n e s influence the morphological control of cells in the a d r e n a l medulla. Acknowledgment The authors gratefully acknowledge the expert technical assistance of Miss N. Takahashi.

References 1. Thoenen H, Barde YA (1980) Physiology of nerve growth factor. Physiol Rev 60:1284-1335

2. Unsickcr K, Krisch B, Otten U, Thoenen H (1978) Nerve growth factor-induced fiber outgrowth from isolated rat adrenal chromaffin cells: impairment by glucocorticoids. Proc Natl Acad Sci USA 75:3498-3502 3. Najoks KW, Korsching S, Rohrer H, Thoenen H (1982) Nerve growth factor-mediated induction of tyrosine hydroxylase and of neurite outgrowth in cultures of bovine adrenal chromaffin cells: dependence on developmental stage. Dev Biol 92:365379 4. Tischler AS, Perlman RL, Nunnemacher G, Morse GM, DeLellis RA, Wolfe HJ, Sheard BE (1982) Long-term effects of dexamethasone and nerve growth factor on adrenal medullary cells cultured from young adult rats. Cell Tissue Res 225:525542 5. Tischler AS, DeLelli RA, Biales B, Nunnemacher G, Carabba V, Wolfe HJ (1980) Nerve growth factor-induced neurite outgrowth from normal human chromaffin cells. Lab Invest 43:399409 6. Peragner R, Walser V (1980) Long-term tissue culture of human pheochromocytomas. Exp Pathol (Jena) 18:423-429 7. Peragner R, Sadjak A, Walser V (1984) The effect of nerve growth factor (NGF) on the catecholamine contents of two human pheochromocytomas in tissue culture. Exp Pathol (Jena) 26:2131 8. Tischler AS, Lee YC, Perlman RL, Costopoulus D, Slayton VW, Bloom SR (1984) Production of "ectopic" vasoactive intestinal peptide-like and neurotensin-like immunoreactivity in human pheochromocytoma cell cultures. J Neurosci 4:1398-1404 9. Matsuura T, Nagamori S, Fujise K, Hasumura S, Homma S, Sujino H, Shimizu K, Niiya M, Kameda H (1989) Morphological characteristics of human fat-storing cells fractionated and cultured by new techniques. Hum Cell 2:181-189 10. Kawada M, Aizaki H, Fukuya K, Niiya M, Matsuura T, Sujino H, Hasumura S, Nagamori S, Toda G (1994) Morphological changes in human gall bladder carcinoma cell line-NOZ due to epirubicin and doxorubicin. Electron Microsc 27:175-177 11. Wan D e c , Livett BG (1989) Induction of phenylethanolamine N-methyltransferase mRNA expression by glucocorticoids in cultured bovine adrenal chromaffin cells. Eur J Pharmacol 172:107-115 12. Otten U, Schwab W, Gagnon C, Thoenen H (1977) Selective induction of tylosine hydroxylase and dopamine [3-hydroxylase by nerve growth factor: comparison between adrenal medulla and sympathetic ganglia of adult and newborn rats. Brain Res 133: 291-303 13. Jaques S Jr, Tobes MC (1986) Morphologic and biochemical variability of tissue and cultured cells from human pheochromocytoma. J Cell Physiol 128:261-270 14. Sasano N, Takezawa Y, Sato H, Horikawa N (1971) Microangiography of normal and pathologic human adrenals in prenatal and aging course. Tohoku J Exp Med 104:129-141 15. Sasano N, Sasano H (1990) The adrenal cortex. In: Kovacs K, Asa SL (eds) Functional endocrine pathology. Blackwell, Boston, pp 546-584 16. Wrzolkowa T, Mrozawicz A, Lewinski A, Pryczkowski J (1975) Phaeochromocytoma. Electron microscopic study on catecholamine storage. Pathol Eur 10:179-191