TISSUE & CELL 1980 12 (2) 227-232 (3 1980 Longman Group Ltd
W. P. DE POTTER,
N. H. FRAEYMAN, J. PLUM* and A. F. DE SCHAEPDRYVER
IMMUNOHISTOCHEMICAL EVIDENCE FOR A PLASMA MEMBRANE LOCALIZATION OF DOPAMINE-,+HYDROXYLASE IN THE MOUSE NEUROBLASTOMA ABSTRACT. Dopamine-p-hydroxylase (D/?H), a glycoprotein enzyme which converts dopamine into noradrenaline, was purified from CISOOmouse neuroblastoma and used to raise antibodies in rabbits. Using an indirect immunofluorescence technique the cellular localization of DflH in CKNN,mouse neuroblastoma was compared with that of the superior cervical ganglion. CISOO neuroblastoma D/3H was found to be predominantly localized in the plasma membrane, in contrast to its intracellular localization in the superior cervical ganglion of A/J mice. At least part of the enzyme was found to be associated with the external side of the plasma membrane.
Introduction (E.C. 18.104.22.168; D/3H), the enzyme which converts dopamine (3,4-dihydroxyphenylethylamine) into noradrenaline (NA), is located in catecholamine storage vesicles which have been isolated from the adrenal medulla and peripheral nerves (Geffen and Livett, 1971), pheochromocytoma (Winkler and Smith, 1972) and hypothalamic synaptosomes (Lander and Austin, 1971; Belmar et al., 1974). The enzyme is also present in human neuroblastoma (Winkler and Smith, 1972) and CISOOmouse neuroblastoma (Anagnoste et al., 1972). However, in these tissues, its subcellular distribution is not fully elucidated. Using gradient centrifugation techniques, HGrtnagl et a/. (1972) and De Potter et a/. (1974) showed that in human neuroblastoma tumour tissue, part of the enzyme was located in ‘heavy’ NA-containing vesicles. FurtherDOPA~IINE-/I-HYDROXYLASE
Pharmacology Institute of Heymans and * Department of Bacteriology and Virology, University of Ghent Medical School, De Pintelaan 135. B-9000 Ghent, Belgium. Received 18 July 1979. Revised 18 December 1979. 15
more, De Potter et d. (1974) reported that in some cases appreciable am&& of DPH are associated with lower density cellular components. In a recent study on Cl300 mouse neuroblastoma (Fraeyman and De Potter, 1976) biochemical evidence was obtained that about one-third of the D,8H activity was associated with storage vesicles, whereas the remainder was most probably located in the cell membrane. The present immunohistochemical experiments were performed to confirm this finding. For this purpose homologous antiserum against mouse tumour D/3H was prepared and used to study cryostat sections of the tumour as well as its isolated cells.
Materials and Methods A/J mice (5 weeks old) and CI~OI) murine neuroblastoma were purchased from the Jackson Laboratory (Bar Harbor, Maine). The tumour was maintained in our laboratory by subcutaneous transfer in 5-10 weeks old A/J mice of both sexes. About 2 weeks after inoculation, the mice were killed by 227
cervical dislocation and the tumour dissected free from the surrounding tissue. The antigen was purified by a combination of DEAE-cellulose ion exchange chromatography (Sigma Chemical Company, St. Louis, Missouri) and Con A-sepharose-4B affinity chromatography (Pharmacia, Uppsala, Sweden) as described by Gagnon et trl. (1976). The high speed supernatant (88,000 R for 45 min) of a l/IO diluted homogenate (in cold 0.25 M sucrose buffered with 5 mM Tris-HCI pH 7.3) of the Cl:rr,rr mouse neuroblastoma was used as starting material. An analysis of the purity of the antigen by gel electrophoresis (Clarke, 1964) revealed a single detectable protein band with molecular weight of 77,000 (Grzanna and Coyle, 1976). The immunization of rabbits was achieved by the method of Hartman and Udenfriend (1969) including the gel electrophoretic separation step. Immunodiffusion and immunoelectrophoresis (Kwapinski. 1972) revealed a single precipitation arc against the starting supernatant as well as against the purified antigen. Antigenic cross-reactivity was found in the homogenates of superior cervical ganglia and adrenal medulla from A/J mice using Ouchterlony double diffusion analysis (Kwapinski, 1972). No cross-reactivity could be observed when the antibody was tested against D/3H from bovine adrenal medulla. No further attempts were made to purify the IgG fractions as to isolate the Fab fragments. A suspension of free cells was obtained by gently stirring tumour fragments in ice-cold phosphate buffered saline (PBS) (Comoglio and Filogamo, 1973). In order to remove dead cells this cell suspension was subjected to metrizamide gradient centrifugation (Neyegaard and Co., S/A, Oslo, Norway). The gradient, ranging from 0 to 22’:(1 metrizamide (MAM), was made iso-osmotic with the mouse serum through addition of PBS. and was centrifuged at 500 R for I5 min in the SW40 rotor. The viable cells, which accumulated at the upper fractions of the gradient, were collected and used for immunofluorescence experiments. Frozen sections (5-7 pm) from the tumour, mounted on microscopic slides, were incubated for 1 hr at 30 C with the antiserum. Preliminary trials indicated that a dilution 115 of our antiserum gave optimal results. Hence in
subsequent incubations this dilution was always used. The sections were gently rinsed with PBS and subsequently incubated for I hr at 30 C with fluorescein isothiocyanate (FITC) labelled goat antirabbit immunoglobulins (Nordic, Tilburg, The Netherlands) diluted (l/SO) with PBS. After washing with PBS the slides were mounted in a PBS/ glycerol mixture (8/2) and were viewed under the microscope (Leitz) with oil immersion (magnification x 600). Photographs were taken with a Kodak, High Speed, Ektachrome film (EH 135). Superior cervical ganglia from A/J mice were processed in the same manner as the neuroblastoma tissue, with addition of 0.3 ‘I,, Triton X-100 to the incubation medium (Hartman, 1973). The viable cells were also treated in a similar fashion. Briefly, the cells (approximately 5 x 10”) were washed three times with icecold PBS, resuspended in 200 ~1 of the l/5 diluted (PBS) antiserum and allowed to react for I hr at 4 C. After three washings in PBS the cells were resuspended in 200 ~1 of the l/SO diluted (with PBS) FITC labelled goat antirabbit globulins solution and again incubated for I hr at 4°C. The cells were then washed again three times in PBS and mounted in the glycerol mixture (see above). Conditions for microscopy were the same as for the tissue sections. In order to ascertain the specificity of the antibodies reactive with the tissues or cells, serum specimens obtained from pre-immune rabbits and antisera exhausted by preincubation with purified D/3H were evaluated.
Results Using the immunofluorescent technique, the antigen is demonstrated to be evenly located along the membrane, producing a ribbonlike pattern in the neuroblastoma sections (Fig. I). In the frozen sections of the superior cervical ganglion of A/J mice, the antigen is seen within the cytoplasma after Triton treatment, as observed for the ganglia of other species (Hartman, 1973) (Fig. 2). In the absence of Triton, howe_ver, only a very faint fluorescence was present. On the other hand, when Triton X-100 was included in the medium used for the neuroblastoma sections,
Fig. 1. Fluorescent mouse neuroblastoma fluorescence without
jattern o an anti-D,3H treated cryostat x 600). b ost of the cells show a ribbon4 y clear it racellular localization.
Fig. 2. Fluorescence fan antibody ganglion of the A/J rm se x 600
ection of the Clnoo 2 distribution of the
Fig. 3. Cell membrane (x 600). The fluorescence
fluorescence is restricted
of the isolated Cl300 mouse neuroblastoma to the membrane.
the fluorescence almost compietely disappeared. When the same technique was applied for the isolated cells at 4°C a selective fluorescent staining of the cell membrane occurred in 5-l 5 % of the living cells (Fig. 3). Performing the isolation schedule of the cells at 37°C (see Methods), resulted in a clear-cut decrease of the membrane fluorescence. Independently of the isolation temperature, it is estimated that about 20-25 % of the cells are destroyed during the incubation period. The dead cells are easily recognizable because they show fluorescence over the entire cell bodies. Living cells are characterized by patchy fluorescence and therefore could easily be distinguished from dead cells during counting. Controls treated with pre-immune serum showed only a very faint and nonspecific staining. As a second control, the antiserum was exhausted by pre-incubation with purified
D/3H. In this case all tissues showed only very faint or no fluorescent staining and the ribbon-like pattern of the neuroblastoma was completely absent. All modifications of the incubation conditions are summarized in Table 1. Discussion The results obtained with the immunohistochemical fluorescence technique presented here, showed that in CI~OO mouse neuroblastoma tissue, the D/3H-anti-Dj3H complex is almost exclusively confined to the plasma membrane. A non-specific binding of the antibodies to the cells or tissues can be excluded as the control experiments with either pre-immune antisera or antisera after absorption with purified DPH showed no fluorescent staining. The fact that in isolated cells DPH can combine with its antibody suggests that at least part of the enzyme is
1. Sumn~ary of’ the incubation conditions
Tissue CISIJOmouse neuroblastoma cryostat sections
Incubation conditions No additions, directly mounted for microscopy Addition of goat-antirabbit marked with FITC (D = I /SO) Whole procedure, without Triton X-100
Very faint fluorescence
Ribbon-like fluorescence on the periphery of the cells (Fig. 1) (most or almost all cells show this pattern) Whole procedure, with addition Reduction of the fluorescence, any ribbon formation has of Triton X-100 (0.3 % final disappeared concentration) Faint and non-specific Whole procedure, with preimmune rabbit anti-D/3H serum fluorescence Whole procedure, with adsorbed Faint and non-specific antiserum (see Methods) fluorescence
C1300mouse neuroblastoma isolated cells ..
Cells isolated at 4”C, whole procedure Cells isolated at 37°C whole procedure Cells isolated at 4°C pre-immune or adsorbed anti-D,8H serum
Superior cervical ganglion of the A/J strain mice cryostat sections
Whole procedure, without Triton x-100 Whole procedure, with addition of Triton X-100 (0.3 % final concentration)
associated with the external plasma membrane surface, which is in agreement with the biochemical evidence that D/3H is present in the cell membranes of the tumour cells (Fraeyman and De Potter, 1976), although conclusions about their location on inner or outer surfaces of the plasma membrane were not possible at that time. Of the cells, isolated at 37C, only a limited number showed fluorescent staining and the fluorescence was found to be much diminished. These observations are compatible with the plasma membrane structure for mouse neuroblastoma cells as proposed by Glick (1976). The present results suggest that DfiH is one of the loosely bound glycoproteins and consequently can be clearly visualized in cryostat sections. In isolated cells, where a considerable loss of the weakly bound enzyme is possible, visualization is rather difficult.
Patch formation on the cell surface (5-15 % of the cells living) (Fig. 3) Fluorescence greatly reduced, very faint membrane fluorescence None to faint, non-specific fluorescence Non-specific fluorescence, no recognizable structure Fluorescence of the cytoplasma, nucleus and extracellular space do not show any fluorescence (Fig. 2)
De Potter and Chubb (1977) in their recent studies on the fate of large D/3H containing vesicles in adrenergic nerves suggested that during the exocytotic process, the membranes of these vesicles, after fusion with the plasma membrane, are likely to be retrieved in the form of small DBH containing vesicles. Although it is not known whether release of noradrenaline and D/3H by exocytosis of large vesicles can occur in mouse neuroblastoma cells, the presence of the enzyme in the plasma membrane suggests such a possibility. Assuming that exocytosis does occur, it seems unlikely that D/IH is retrieved from the plasma membrane in the form of small vesicles, as no such small noradrenaline and DfiH containing vesicles are found in this type of neuroblastoma cells. There is always the possibility that, instead of being recaptured by an endocytotic process, D/3H could be extruded in the extracellular space, by a process of membrane shedding, a
phenomenon which would at least explain the high Dj3H levels found in the sera of neuroblastoma bearing mice (Anagnoste
et al., 1972).
certain properties of the cell membranes of the tumour (such as cell recognition) which, among other properties, distinguishes tumour cells from non-tumour cells (Burger, 1975).
The significance of the association of DPH with the plasma membrane is at present difficult to fully evaluate, although as a glycoprotein it might well contribute to
This work was supported by a grant from the A.S.L.K. Cancer Fund, Belgium.
References ANAGNOSTE, B., FREEDMAN,L. S., GOLDSTEIN, M., BROOME,J.and FUXE, K. 1972. Dopamine-j3-hydroxylase activity in mouse neuroblastoma tumors and in cell cultures. Proc. natn. Acad. Sci. USA, 69, 1883-1886. BELMAR,J., DE POTTER, W. P. and DE SCHAEPDRYVER,A. F. 1974. Subcellular distribution of noradrenaline and dopamine-&hydroxylase in the hypothalamus of the rat. Evidence for the presence of two populations of noradrenaline storage particles. J. Neurochem., 23, 607-609. BURGER, M. M. 1975. Surface properties of neoplastic cells. In Cell Membranes, Biochemistry, Cell Biology and Pathology (ed. G. Weissmann and R. Claiborne), pp. 215-222. HP Publishing Co. Inc., New York. CLARKE, J. T. 1964. Simplified ‘disc’ (Polyacrylamide gel) electrophoresis. Ann. N. Y. Acad. Sci., 121, 428436.
COMOGLIO, P. M. and FILOGAMO, G. 1973. Plasma membrane fluidity and surface motility of mouse Cl300 neuroblastoma cells. J. Cell Sci., 13, 415420. DE POTTER, W. P. and CHUBB, I. W. 1977. Biochemical observation on the formation of small noradrenergic vesicles in the splenic nerve of the dog. Neuroscience, 2, 167-174. DE POTTER, W. P., DE SCHAEPDRYVER,A. F., DE SMET. F., DELBEKE,M. J. and HOOFT, C. 1974. Subcellular distribution of catecholamines and enzymes in human neuroblastoma. Experienfia, 30, 1323-1324. FRAEYMAN,N. H. and DE POTTER, W. P. 1976. Subcellular distribution of noradrenaline and enzymes in Cl300 mouse neuroblastoma. Arch.7 inf. Physiol., 84, 1074-1075. GAGNON, C., OTTEN,U. and THOENEN, H. 1976. Increased synthesis of dopamine @-hydroxylase in cultured rat adrenal medullae after in viva administration of reserpine. J. Neurochem., 27, 259-265. GEFFEN, L. and LIVETT. B. G. 1971. Synaptic vesicles in sympathetic neurons. Physiol. Rev., 51, 98-157. GLICK, M. C. 1976. Glycoproteins on the surface of neuroblastoma cells. J. nafn. Cancer Inst., 57, 653-658. GRZANNA, R. and COYLE, J. T. 1976. Rat adrenal DfiH: purification and immunological characteristics. J. Neurochem., 27, 1091-1096. HARTMAN, B. K. 1973. Immunofluorescence of dopamine-j3-hydroxylase. Application of improved methodology to the localization of the peripheral and central noradrenergic nervous system. J. Hisfochem. Cytochem.,
HARTMAN,B. and UDENPRIEND, S. 1969. A method for immediate
visualization in acrylamide gels and its use for preparation of antibodies to enzymes. Analyt. Biochem., 30, 391-394. HBRTNAGL, H., HGRTNAGL, H., WINKLER, H., ASAMER, H., FBDISCH, H. J. and KLIMA, I. 1972. Storage of catecholamines in neuroblastoma and ganglioneuroma. A biochemical, immunologic, and morphologic study. Lab. Invest., 27, 613-619. KWAPINSKI, J. B. G. 1972. Merhodology o~immunochemical and Immunological Research. Wiley-lnterscience, New York. LANDER, J. and AUSTIN, L. 197 I. Localization of dopamine-p-hydroxylase in nerve ending particles of sheep hypothalamus. Proc. Aust. Biochem. Sot., 4, 89P. WINKLER, H. and SMITH, A. D. 1972. Phaechromocytoma and other catecholamine producing turnours. In Catecholomines (ed. H. Blaschko and E. Muscholl), pp. 900-924. Springer-Verlag, New York.