Journal ofhreurochemisrry Raven Press, Ltd., New York 0 1990 International Society for Neurochemistry
Changes in Glycosaminoglycans During the Neuritogenesis in PC 12 Pheochromocytoma Cells Induced by Nerve Growth Factor Ritsuko Katoh-Semba, *Atsuhiko Oohira, and Shigeo Kashiwamata Departments of Perinatology and *Embryology, Institute for Developmental Research, Aichi Prefecture Colony,Aichi, Japan
Abstract: Previously, we had suggested that heparan sulfate (HS) makes some contribution to a flat-shaped morphology of PCl2D cells. Therefore, we carried out quantitative and qualitative analyses of glycosaminoglycans (GAGs), the polysaccharide moiety of proteoglycans, during neuritogenesis in PC12 cells that is induced by nerve growth factor (NGF). (a) In PC12 cells, NGF induced a flat-shaped morphology with a few short processes after 3 days of culture, and then it elicited short and long neurites after 6 (in -30% of cells) and 9 (in 60-70%) days of culture, respectively. (b) HS and chondroitin sulfate (CS) were detected in the cell layer at all times. Only CS was found in the medium at 3 and 6 days, whereas a low level of HS, in addition to CS, was detectable on day 9. (c) In the NGF-treated cultures, the amounts of cell-associated HS per cell were two to three times as high as those in the respective nontreated cultures at all times, whereas the amount based on phospholipidwas about twofold higher after 3 days of culture. (d) The levels of HS labeled with [35S]sulfateduring the last 48 h of the culture were 1.5to twofold higher in the NGF-treated cultures than in the respective controls at any time. (e) The amount of cell-associated CS per cell (or per unit of phospholipid),but not of labeled CS per cell, was transiently enhanced at 3 days in
culture with or without NGF. At all times, NGF treatment caused an increasein the levels of total and [35S]sulfate-labeled CS associated with the cells and released into the medium. (f) NGF enhanced the amount ofN-sulfation ofglucosamine residues of HS at all times, but it did not change the ratio of 4-sulfate units to 6sulfate units in CS. (g) At 3 days in culture, the uptake of ["S]sulfate by PC12 cells was lower in the NGFtreated culture than in the nontreated control. (h) In chase experiments, the percentage of unrecovered CS was about twofold higher in the NGF-treated culture than in the nontreated control. These results suggest that the enhanced synthetic activity and the accumulation of GAGs as well as the structural change of HS induced by NGF occur preceding the neurite elongation from PC12 cells. Also, it is suggested that the increase in content of HS is closely correlated with the morphological change from round to flat in PC12 cells. Key Words: Chondroitin sulfate-GlycosaminoglycanHeparan sulfate-Nerve growth factor-Neuritic growthPC12 cells. Katoh-Semba R. et al. Changes in glycosaminoglycans during the neuritogenesis in PC12 pheochromocytoma cells induced by nerve growth factor. J. Neurochem. 55, 1749-1757 (1990).
During the development of the nervous system, complex and reproducible changes in cell morphology occur in response to information from cellular and extracellular environments in which they develop. Such changes in cell morphology are thought to be accompanied by changes in cellular functions and gene expression. Several recent observations suggest that extracellular matrix components influence the morphol-
ogy of neuronal cells and their cellular functions. (a) It has been reported that the precoating of dishes with extracellular matrix components causes cultured PC 12 cells to become flattened, and the release of dopamine from the cells is significantly enhanced (Bethea et al., 1987). (b) Some constituents of extracellular matrices promote the neurite extension from various neurons in nervous tissues in vitro (Lander et al., 1985; Davis
Received September 8, 1989; revised manuscript received April 30, 1990; accepted May 3, 1990. Address correspondence and reprint requests to Dr. R. KatohSemba at Department of Pennatology, Institute for Developmental Research, Aichi Prefecture Colony, Aichi 480-03, Japan. Abbreviafionswed:CHase, chondroitinase;CS,chondroitinsulfate;
CTAB, cetyltrimethylarnmonium bromide; DMEM, Dulbecco's modified Eagle's medium containing 44.4 mM sodium bicarbonate, 2 Mglutamine, and 100 U/mt ofpenicillin;FBS, fetal bovine serum; GAG, glycosaminoglycan; HRS, horse serum; HS, heparan sulfate; HSase, heparitinase; NGF, nerve growth factor; PG,proteogiycan; TCA, trichloroacetic acid.
1749
I750
R. KATOH-SEMBA ET AL.
et al., 1987; Hantaz-Ambroise et al., 1987). Thus, increasing attention is being paid to the effects of extracellular matrix components on the development of the nervous system. Proteoglycans (PGs), a major constituent of extracellular matrices, are also believed to influence the behavior of cells during their participation in cell adhesion, intracellular recognition, migration, proliferation, and differentiation (Hascall and Hascall, 1981; Hook et al., 1984). Recently, it has been demonstrated in vitro that heparan sulfate (HS) PGs play an important role in the attachment of the cells to the substratum at the points of contact, which have been designated “footpads” (Rosen and Culp, 1977). The formation of “footpads” seems to cause the cells to flatten. However, little is known about changes in the property of PGs during the development of the nervous system, because there are complex heterogeneities in the structures of PGs and their core proteins, as well as in the structures of glycosaminoglycans (GAGs), which are polysaccharide moieties of PGs. At least four types of chondroitin sulfate (CS) PG have been reported to exist in the rat brain (Oohira et al., 1988). It has also been shown that the relative amount of CS increases during the development of the rat brain, whereas that of hyaluronate decreases (Oohira et al., 1986). These results are representative of a mixed system of neurons and glial cells. PGs that are synthesized by either neuronal or glial cells are not fully characterized. Furthermore, the search for the changes in the properties of PGs that occur during the maturation of individual cells has not progressed significantly. Because neurons seem to produce only 10% of the amount of PGs synthesized by other types of cells, it is essential to use a pure system comprising a single kind of cells to characterize PGs synthesized by cells, particularly by neurons. A clonal cell line of PC12 pheochromocytoma cells, which was established by Greene and Tischler (1976), is thought to be useful for this purpose. In these cells, a dramatic alteration in morphology is induced by nerve growth factor (NGF), that is, the cells become flattened, followed by elongation of neurites. It has also been suggested that there are considerable differences between PC 12 cells with and without long neurites in the amounts and in the structure of synthesized GAGs (Margolis et al., 1983, 1987). Furthermore, our previous report showed that the amounts of GAGs were higher in PC12D cells, a flat-shaped variant of PC12 cells, than in round-shaped conventional PC12 cells (Katoh-Semba et al., 1989~). These findings suggest that PGs and/or GAGs may exert some influence on the morphology of PC12 cells. It remains obscure when and how the changes in GAGS occuf during the NGF-induced neuritogenesis in PC12 cells. In this communication, we report that enhancement in the amount of GAGs and the structural change of HS are observed preceding the neurite elongation from PC12 cells, namely, in the time when the morphological change of the cells from round to flat occurs. J. Neurochem.. Vol. 55. No.5. 1990
MATERIALS AND METHODS Materials Dulbecco’s modified Eagle’s medium (catalogue no. 4301600) was purchased from GIBCO (Grand Island, NY, U.S.A.). Fetal bovine serum (FBS) and horse serum (HRS) were obtained from Whittaker M.A. Bioproducts (Walkersville, MD, U.S.A.). Cetyltrimethylammonium bromide (CTAB) and Hanks’ balanced salt solution were from N a d a i Tesque (Kyoto, Japan). Alcian blue, cellulose acetate membranes (Sepraphore III), Cellulofine GCL-90m, Pronase E, and Aquasol-2 were from Chroma Gesellschaft Schmid & Co. (Stuttgart, F.R.G.), Gelman Sciences (Ann Arbor, MI, U.S.A.), Seikagaku Kogyo Co., Ltd. (Tokyo, Japan), Kaken Seiyaku (Tokyo), and NEN Research Products (Boston, MA, U.S.A.), respectively. All other enzymes and standard GAGs used were the products of Seikagaku Kogyo. [35S]Sulfuric acid was purchased from the Japan Radioisotope Association (Tokyo). NGF was purified according to the method of Varon et al. (1967). Chemicals of the highest grade available were used for all experiments.
Cell culture Stock cultures of PC12 cells were expanded on 100-mmdiameter plastic tissue culture dishes in Dulbecco’s modified Eagle’s medium containing 44.4 M s o d i u m bicarbonate, 2 mM glutamine, and 100 U/ml of peniciliin (DMEM) supplemented with 5% FBS and 10% HRS. When the cell density reached 5-7 X lo6 dish, cells were harvested in the same medium as mentioned above. Quantification and characterization of GAGs. For the quantification and characterization ofGAGs, PC12 cells were seeded on 100-mm-diameter dishes precoated with 0.1 mg/ ml of polylysine at a density of 0.5 X lo6 cells/plate in DMEM supplemented with 5% FBS and 10% HRS, in the absence or presence of NGF. The seeding medium was replaced every other day with fresh medium that contained the same components, with the exception that, when the cells were harvested after 3 and 9 days of culture, the medium was switched first after 24 h. NGF was added at a concentration of 5 biological units (1 unit is defined as the smallest concentration of NGF that exhibits maximal activity for the extension of neurites from PC12 cells) throughout the experiments. The cells were labeled with 925 kBq/ml of Nat[35S]04 for 48 h after I , 4, and 7 days of culture. GAGS were prepared from the cell layer and the medium at 3,6, and 9 days in culture. Cell cultures under floating-cell conditions were maintained in exactly the same way except for the use of Petri dishes in the place of tissue culture dishes. Under these conditions, the viability of the cells was 88%. Chase experiments with labeled GAGs. For the chase experiments with labeled GAGs, cells were seeded in the presence or absence of NGF on polylysine-coated 100-mmdiameter tissue culture dishes at the same density as mentioned above. The labeling with Na2[35S]04(925 kBq/ml) was performed for 24 h after a 24-h preculture. The cells in some cultures were harvested after labeling for 24 h and subjected to analysis of GAGS. Other cultures were washed once with nonlabeled DMEM supplemented with 5% FBS and 10% HRS, and then they were incubated for another 24 h in the same medium. GAGs were prepared separately from the cell layer and the medium. Uptake of [3sS]sulfate. For the experiments that involved the uptake of [3SS]sulfate,cells were seeded on polylysinecoated 60-mm-diameter dishes at a density of 2 X 105/dish.
1751
GL YCOSAMINOGLYCAN IN PC12 CELLS On day 3, 6, or 9, the cultures were labeled with 925 kBq/ ml of Na2[35S]04for 90 min. Quantificationof total and neurile-bearingcells. Total and neurite-bearing cells in three 0.5- X 16-mm strips per culture were counted under phase-contrast microscopy just before harvest. The cells with processes of two soma1 diameters or longer were defined as neurite-bearing cells.
Preparation of crude GAGs GAGs were prepared as described previously (KatohSemba et al., 1989~).In brief, crude PGs from the medium were precipitated with ammonium sulfate,and the precipitate was dissolved in the buffer, followed by treatments with Pronase E, NaOH, and trichloroaceticacid (TCA). After removal of insoluble materials by centrifugation, the supernatant was treated with CTAB to make the CTAB-GAG complex. Crude GAGs were obtained by dissociating GAGs from the complex using potassium acetate. PGs in the cell layer were solubilized with sodium dodecyl sulfate in Tris-HC1 buffer (pH 7.5) containing EDTA after the dishes had been washed once with phosphate-buffered saline. Sodium dodecyl sulfate-solubilized PGs were stirred for 5 h at room temperature, and then insoluble materials were removed by centrifugation. Sequential treatments with Pronase, NaOH, and TCA followed.
Quantificationof individual GAGs As described in our previous article (Katoh-Semba et al., 1989a), individual GAGs were separated on a cellulose acetate membrane by two-dimensional electrophoresis, visualized by staining the membranes with Alcian blue, and then quantified spectrophotometrically by measuring the absorbance of 6 15 nm after dissolution of the pigment associated with GAGs in the CTAB solution. The radioactivity of [35S]sulfateincorporated into individual GAGs was measured by cutting out the spots on the membrane whose positions had been confirmed by exposure of the membrane to an x-ray film. Radioactivity was counted in a toluene scintillator.
Characterization of GAGs To characterize the structure of CS, the crude GAGs from the cell layer and the medium were digested with chondroitinase (CHase) AC (Saito et al., 1968). Subsequently, paper chromatography was camed out to separate the disaccharide materials from CS. The amounts of 4-sulfated and 6-sulfated disaccharides were determined by measuring their associated radioactivity. To characterize the structure of HS, crude GAGs were degraded with heparitinase (HSase) (Linker and Hovingh, 1972) or nitrous acid (Lindahl et al., 1973) and then chromatographed on a column of Cellulofine GCL-90m ( 1 X 90 cm) buffered with 50 mM Tris-HCI (pH 7.5) containing 2 mM EDTA, 0.2 M NaCl, 0.2% Nonidet-40, and I mM NaN,. The relative amounts of CS and the sulfation at the amino group (N-sulfate)or the hydroxy group (O-sulfate) of the constituent sugars of HS were estimated from the areas under the peaks of radioactivity due to [35S]sulfateon the chromatogram. The peak areas shown by the bars in Fig. 4 were cut separately out of the chromatograms, and their areas were determined by weight.
Uptake of [3sS]sulfateby cells At the end of a 90-min labeling, dishes were washed three times with 5 ml of ice-cold Hanks’ balanced salt solution. Incorporated [35S]sulfatewas extracted with 0.75 ml of 5% TCA for 30 min. Dishes were washed with 0.25 ml of TCA. The wash was combined with the first extract. The radioactivity of neutralized extracts was measured in Aquasol-2 with
a Beckman LS-9OOO liquid scintillation counter. [35S]Sulfatelabeled materials detected in the TCA-soluble fraction after a 90-min labeling were not GAGs, because the extracted, labeled materials were not precipitated by 75% ethanol containing 1.3% potassium acetate.
Other methods Lipids were extracted by the method of Bligh and Dyer (1959). Inorganic phosphorus in crude lipids was assayed according to the method of Bartlett (1959), and the weight of phospholipids was calculated from the amount of phosphorus as described by Dickerson (1968). DNA was quantified by the method of Burton (1956). The amount of cell-associated CS was corrected by subtracting the amount of CS attached to dishes, as described in our previous article (Katoh-Semba et al., 1989~).
RESULTS Proliferation and neuritic growth of PC12 cells The number of PC12 cells in the absence of NGF increased linearly up to 9 days of culture with a doubling time of 2 days. In the presence of NGF, proliferation proceeded at the same rate as that in its absence until -5 days of culture, and then the rate leveled off (data not shown). Neurite-bearing PC 12 cells were not observed in the absence of NGF at any time (Fig. 1A). In the presence of NGF, the cells at 3 days were flat, slightly enlarged, and had a few short processes. However, they did not yet have any process that could be defined as neurites (Fig. 1B). After 6 days of culture, -30% of cells developed neurites, although they were short, and some neurites possessed growth cones (Fig. 1C). Very long neurites were observed in PC12 cells cultured for 9 days with NGF (Fig. 1D). The percentage of neurite-bearing cells in the total cell population reached 60-70% after 9 days of culture. Cell proliferation under the present conditions was almost the same as reported previously (Katoh-Semba et al., 1984). However, the maximal percentage of neurite-bearing cells appeared to be slightly lower. This difference may be explained by the fact that the initial density of seeded cells was much higher in the present study. Characterization and quantification of GAGs In the absence of NGF, two kinds of GAG, corresponding to HS and CS, were detected by cellulose acetate electrophoresis in the cells cultured for 3, 6, and 9 days (data not shown). In contrast, a GAG spot having the same mobility as CS appeared on the electrophoretograms of the medium sampled at 3 and 6 days in culture. A faint spot corresponding to HS, in addition to CS, was found in the medium at 9 days. All GAGs detected with Alcian blue were labeled with [35S]sulfate(data not shown). The radioactive, pigmented spots corresponding to HS and CS disappeared after treatment with nitrous acid and CHase, respectively. This result demonstrates that HS and CS are synthesized in PC12 cells. The species of GAGs detected in the cell layer and the medium did not differ between the NGF-treated and control cultures. When J. Neurochem., Val. 55. No. 5. 1990
1752
R. KATOH-SEMBA ET AL.
FIG. 1. Photomicrographsof PC12 cells in the absence (A) or presence (8-D) of NGF: (A and D) at 9 days, (B) at 3 days, and (C) at 6 days. Bar = 20 pm. Arrows indicate growth cones.
the cells were cultured under floating-cell conditions without NGF, both HS and CS were detected in the cell layer and the medium. HS and CS were also labeled with [35S]sulfate(data not shown). Table 1 summarizes changes in the amounts of GAGS associated with the cells and released into the medium during the NGF-induced neuritogenesis in PC12 cells. The cells just before reseeding contained 0.06 pg of HS/ 1O6 cells, which was lower than the levels of HS detected in dish-attached cells under any culture condition. When the cells were seeded to the dishes, in the absence of NGF, the amounts of cell-associated HS per cell remained unchanged for the first 6 days and then decreased by 50% after 9 days of culture. In the presence of NGF, the levels of HS at 3, 6, and 9 days were almost equal to each other and were about two to three times as high as those in its absence. Because the membrane surfaces of neurite-bearing cells J. Neuroehem., Vol. 55, No. 5 , 1990
were considered to be increased in area, the amounts of phospholipids were determined. As shown in Fig. 2, the amounts of phospholipids per cell were increased with the neurite extension, and, after 9 days of culture, they reached 2.4 times the level in cells without neurites. In the absence of NGF, changes in the amounts of cell-associated HS per microgram of phospholipids were similar in pattern to those of HS per cell (Table 1). In contrast, in the presence of NGF, the level was transiently increased at 3 days and then decreased. The NGF-treated cells at 3 and 9 days had about twofold more HS than the nontreated controls. Irrespective of the presence of NGF, changes in the amounts of [35S]sulfate-labeledHS per cell were similar to those of HS per cell (Fig. 3A and Table 1). Radioactivity incorporated into cell-associated HS was 1.5- to twofold higher in the NGF-treated culture than in the nontreated culture at 3, 6, and 9 days.
1753
GLYCOSAMINOGLYCAN IN PC12 CELLS TABLE 1. Changes in levels of cell-associated and medium-released GAGs during neuritogenesis in PCl2 cells
cs
HS Culture (days) Cell layer 0" 3
6
9
NGF
~ gI O6 f cells
ng/pg of PL
&lo6 cells
n&pg of PL
-
0.06 k 0.01 0.19 f 0.05 0.38 f 0.05' 0.22 f 0.04 0.43 f 0.07' 0.1 1 k 0.03 0.39 f 0.02
0.83 f 0. I5 2.7 k 0.6 4.2 f 0.5' 3.1 ? 0.5 2.8 f 0.4 1.5 f 0.4 2.3 f 0.1'
0.03 f 0.00 0.19 f 0.07 0.37 ? 0.10' 0.05 f 0.01 0.14 f 0.04' 0.03 f 0.01 0.15 f 0.04'
0.33 f 0.02 2.6 f 0.8 4.2 f 1.0' 0.66 f 0.09 0.94 f 0.23' 0.42 f 0.08 0.89 2 0.21
-
+ + +
'
'
Medium Od
3 6 9
0.08 f 0.03 ND ND ND ND Trace Trace
-
-
++ -
+
0.52 f 0.1 I 0.73 f 0.12 1.16 f 0.31' 0.70 f 0.16 1.10 f 0.16' 0.46 f 0.02 0.96 f 0.27'
Data are average f SD values of four determinations from two separate experiments. Because serumderived CS attached to dishes, the amounts of cell-associated CS were corrected as described previously (Katoh-Semba et al., 19890). PL, phospholipid; ND, not detectable. a Determined using the cells just before reseeding. ' p < 0.01, ' p < 0.05 compared with the level in the absence of NGF. Determined using the culture medium under the floating-cellconditions.
The amount of cell-associated CS per cell or per unit of phospholipid was transiently enhanced after 3 days of culture and thereafter decreased in the absence of NGF (Table 1). The level of CS at 9 days was very close to that of CS associated with the cells just before seeding. In the presence of NGF, the time course of changes in the levels of cell-associated CS showed a profile similar to that in its absence, but the levels of CS were two to three times as high as those in the respective nontreated controls. On the other hand, in the absence of NGF, the amounts of [35S]sulfate-labeled CS per cell in the cell layer were almost the same at any day in culture as that in the floating cells, although they were slightly lower on day 9 (Fig. 3B). Treatment with NGF produced an apparent increase in the in-
FIG. 2. Changes in levels of phospholipids during neuritogenesis in PC12 cells inducedby NGF: no NGF (hatched cdumn) and with NGF, measured after 3. 6, and 9 days of culture (open columns). Data are average f SD (bars) values of four determinationsfrom two separate experiments.
g 8
5
gloo
3z 0
corporation of [35S]sulfate,and the level attained was maintained for 9 days without any significant change. The changes in the amounts of CS released into the medium per cell were not very marked during culture with or without NGF (Table 1). However, the level of CS released from the NGF-treated cells into the medium on each day was significantly higher than those released from the nontreated attached cells and from the nontreated floating cells. The amounts of [35S]sulfate-labeledCS released from NGF-treated and nontreated cells into the medium also remained unchanged during culture (Fig. 3C). However, the levels of labeled CS in the medium were lower in the absence of NGF and higher in its presence, compared with the level of labeled CS released from the nontreated floating cells, resulting in a distinct difference in the level of labeled CS between the NGF-treated and nontreated cultures. Changes in structure of GAGs during neuritic growth Because CS is known to be composed of chondroitin 4- and 6-sulfate units, changes in the composition of these [35S]sulfate-labeled disaccharides during the NGF-induced neuritogenesis in PC 12 cells were examined. As far as crude fractions of CS were concerned, >95% of CS in the cell layer or the medium was detected as a 4-sulfate unit, in the presence or absence of NGF (data not shown). There were also no changes in J. Neurochem.. Vol. 55, No. 5. 1990
R. KATOH-SEMBA ET AL.
I754
C
B
100
: '
FIG. 3. Changes in amounts of [36S]sulfatelabeled GAGs during neuritogenesis in PC12 cells induced by NGF: (A) HS in the cell layer, (6)CS in the cell layer, and (C) CS in the medium. Cells were cultured without NGF (O), with NGF (0).and under floatingcell conditions in the absence of NGF (A).
5 0 . ,/'
Changes in Glycosaminoglycans During the Neuritogenesis in PC 12 Pheochromocytoma Cells Induced by Nerve Growth Factor Ritsuko Katoh-Semba, *Atsuhiko Oohira, and Shigeo Kashiwamata Departments of Perinatology and *Embryology, Institute for Developmental Research, Aichi Prefecture Colony,Aichi, Japan
Abstract: Previously, we had suggested that heparan sulfate (HS) makes some contribution to a flat-shaped morphology of PCl2D cells. Therefore, we carried out quantitative and qualitative analyses of glycosaminoglycans (GAGs), the polysaccharide moiety of proteoglycans, during neuritogenesis in PC12 cells that is induced by nerve growth factor (NGF). (a) In PC12 cells, NGF induced a flat-shaped morphology with a few short processes after 3 days of culture, and then it elicited short and long neurites after 6 (in -30% of cells) and 9 (in 60-70%) days of culture, respectively. (b) HS and chondroitin sulfate (CS) were detected in the cell layer at all times. Only CS was found in the medium at 3 and 6 days, whereas a low level of HS, in addition to CS, was detectable on day 9. (c) In the NGF-treated cultures, the amounts of cell-associated HS per cell were two to three times as high as those in the respective nontreated cultures at all times, whereas the amount based on phospholipidwas about twofold higher after 3 days of culture. (d) The levels of HS labeled with [35S]sulfateduring the last 48 h of the culture were 1.5to twofold higher in the NGF-treated cultures than in the respective controls at any time. (e) The amount of cell-associated CS per cell (or per unit of phospholipid),but not of labeled CS per cell, was transiently enhanced at 3 days in
culture with or without NGF. At all times, NGF treatment caused an increasein the levels of total and [35S]sulfate-labeled CS associated with the cells and released into the medium. (f) NGF enhanced the amount ofN-sulfation ofglucosamine residues of HS at all times, but it did not change the ratio of 4-sulfate units to 6sulfate units in CS. (g) At 3 days in culture, the uptake of ["S]sulfate by PC12 cells was lower in the NGFtreated culture than in the nontreated control. (h) In chase experiments, the percentage of unrecovered CS was about twofold higher in the NGF-treated culture than in the nontreated control. These results suggest that the enhanced synthetic activity and the accumulation of GAGs as well as the structural change of HS induced by NGF occur preceding the neurite elongation from PC12 cells. Also, it is suggested that the increase in content of HS is closely correlated with the morphological change from round to flat in PC12 cells. Key Words: Chondroitin sulfate-GlycosaminoglycanHeparan sulfate-Nerve growth factor-Neuritic growthPC12 cells. Katoh-Semba R. et al. Changes in glycosaminoglycans during the neuritogenesis in PC12 pheochromocytoma cells induced by nerve growth factor. J. Neurochem. 55, 1749-1757 (1990).
During the development of the nervous system, complex and reproducible changes in cell morphology occur in response to information from cellular and extracellular environments in which they develop. Such changes in cell morphology are thought to be accompanied by changes in cellular functions and gene expression. Several recent observations suggest that extracellular matrix components influence the morphol-
ogy of neuronal cells and their cellular functions. (a) It has been reported that the precoating of dishes with extracellular matrix components causes cultured PC 12 cells to become flattened, and the release of dopamine from the cells is significantly enhanced (Bethea et al., 1987). (b) Some constituents of extracellular matrices promote the neurite extension from various neurons in nervous tissues in vitro (Lander et al., 1985; Davis
Received September 8, 1989; revised manuscript received April 30, 1990; accepted May 3, 1990. Address correspondence and reprint requests to Dr. R. KatohSemba at Department of Pennatology, Institute for Developmental Research, Aichi Prefecture Colony, Aichi 480-03, Japan. Abbreviafionswed:CHase, chondroitinase;CS,chondroitinsulfate;
CTAB, cetyltrimethylarnmonium bromide; DMEM, Dulbecco's modified Eagle's medium containing 44.4 mM sodium bicarbonate, 2 Mglutamine, and 100 U/mt ofpenicillin;FBS, fetal bovine serum; GAG, glycosaminoglycan; HRS, horse serum; HS, heparan sulfate; HSase, heparitinase; NGF, nerve growth factor; PG,proteogiycan; TCA, trichloroacetic acid.
1749
I750
R. KATOH-SEMBA ET AL.
et al., 1987; Hantaz-Ambroise et al., 1987). Thus, increasing attention is being paid to the effects of extracellular matrix components on the development of the nervous system. Proteoglycans (PGs), a major constituent of extracellular matrices, are also believed to influence the behavior of cells during their participation in cell adhesion, intracellular recognition, migration, proliferation, and differentiation (Hascall and Hascall, 1981; Hook et al., 1984). Recently, it has been demonstrated in vitro that heparan sulfate (HS) PGs play an important role in the attachment of the cells to the substratum at the points of contact, which have been designated “footpads” (Rosen and Culp, 1977). The formation of “footpads” seems to cause the cells to flatten. However, little is known about changes in the property of PGs during the development of the nervous system, because there are complex heterogeneities in the structures of PGs and their core proteins, as well as in the structures of glycosaminoglycans (GAGs), which are polysaccharide moieties of PGs. At least four types of chondroitin sulfate (CS) PG have been reported to exist in the rat brain (Oohira et al., 1988). It has also been shown that the relative amount of CS increases during the development of the rat brain, whereas that of hyaluronate decreases (Oohira et al., 1986). These results are representative of a mixed system of neurons and glial cells. PGs that are synthesized by either neuronal or glial cells are not fully characterized. Furthermore, the search for the changes in the properties of PGs that occur during the maturation of individual cells has not progressed significantly. Because neurons seem to produce only 10% of the amount of PGs synthesized by other types of cells, it is essential to use a pure system comprising a single kind of cells to characterize PGs synthesized by cells, particularly by neurons. A clonal cell line of PC12 pheochromocytoma cells, which was established by Greene and Tischler (1976), is thought to be useful for this purpose. In these cells, a dramatic alteration in morphology is induced by nerve growth factor (NGF), that is, the cells become flattened, followed by elongation of neurites. It has also been suggested that there are considerable differences between PC 12 cells with and without long neurites in the amounts and in the structure of synthesized GAGs (Margolis et al., 1983, 1987). Furthermore, our previous report showed that the amounts of GAGs were higher in PC12D cells, a flat-shaped variant of PC12 cells, than in round-shaped conventional PC12 cells (Katoh-Semba et al., 1989~). These findings suggest that PGs and/or GAGs may exert some influence on the morphology of PC12 cells. It remains obscure when and how the changes in GAGS occuf during the NGF-induced neuritogenesis in PC12 cells. In this communication, we report that enhancement in the amount of GAGs and the structural change of HS are observed preceding the neurite elongation from PC12 cells, namely, in the time when the morphological change of the cells from round to flat occurs. J. Neurochem.. Vol. 55. No.5. 1990
MATERIALS AND METHODS Materials Dulbecco’s modified Eagle’s medium (catalogue no. 4301600) was purchased from GIBCO (Grand Island, NY, U.S.A.). Fetal bovine serum (FBS) and horse serum (HRS) were obtained from Whittaker M.A. Bioproducts (Walkersville, MD, U.S.A.). Cetyltrimethylammonium bromide (CTAB) and Hanks’ balanced salt solution were from N a d a i Tesque (Kyoto, Japan). Alcian blue, cellulose acetate membranes (Sepraphore III), Cellulofine GCL-90m, Pronase E, and Aquasol-2 were from Chroma Gesellschaft Schmid & Co. (Stuttgart, F.R.G.), Gelman Sciences (Ann Arbor, MI, U.S.A.), Seikagaku Kogyo Co., Ltd. (Tokyo, Japan), Kaken Seiyaku (Tokyo), and NEN Research Products (Boston, MA, U.S.A.), respectively. All other enzymes and standard GAGs used were the products of Seikagaku Kogyo. [35S]Sulfuric acid was purchased from the Japan Radioisotope Association (Tokyo). NGF was purified according to the method of Varon et al. (1967). Chemicals of the highest grade available were used for all experiments.
Cell culture Stock cultures of PC12 cells were expanded on 100-mmdiameter plastic tissue culture dishes in Dulbecco’s modified Eagle’s medium containing 44.4 M s o d i u m bicarbonate, 2 mM glutamine, and 100 U/ml of peniciliin (DMEM) supplemented with 5% FBS and 10% HRS. When the cell density reached 5-7 X lo6 dish, cells were harvested in the same medium as mentioned above. Quantification and characterization of GAGs. For the quantification and characterization ofGAGs, PC12 cells were seeded on 100-mm-diameter dishes precoated with 0.1 mg/ ml of polylysine at a density of 0.5 X lo6 cells/plate in DMEM supplemented with 5% FBS and 10% HRS, in the absence or presence of NGF. The seeding medium was replaced every other day with fresh medium that contained the same components, with the exception that, when the cells were harvested after 3 and 9 days of culture, the medium was switched first after 24 h. NGF was added at a concentration of 5 biological units (1 unit is defined as the smallest concentration of NGF that exhibits maximal activity for the extension of neurites from PC12 cells) throughout the experiments. The cells were labeled with 925 kBq/ml of Nat[35S]04 for 48 h after I , 4, and 7 days of culture. GAGS were prepared from the cell layer and the medium at 3,6, and 9 days in culture. Cell cultures under floating-cell conditions were maintained in exactly the same way except for the use of Petri dishes in the place of tissue culture dishes. Under these conditions, the viability of the cells was 88%. Chase experiments with labeled GAGs. For the chase experiments with labeled GAGs, cells were seeded in the presence or absence of NGF on polylysine-coated 100-mmdiameter tissue culture dishes at the same density as mentioned above. The labeling with Na2[35S]04(925 kBq/ml) was performed for 24 h after a 24-h preculture. The cells in some cultures were harvested after labeling for 24 h and subjected to analysis of GAGS. Other cultures were washed once with nonlabeled DMEM supplemented with 5% FBS and 10% HRS, and then they were incubated for another 24 h in the same medium. GAGs were prepared separately from the cell layer and the medium. Uptake of [3sS]sulfate. For the experiments that involved the uptake of [3SS]sulfate,cells were seeded on polylysinecoated 60-mm-diameter dishes at a density of 2 X 105/dish.
1751
GL YCOSAMINOGLYCAN IN PC12 CELLS On day 3, 6, or 9, the cultures were labeled with 925 kBq/ ml of Na2[35S]04for 90 min. Quantificationof total and neurile-bearingcells. Total and neurite-bearing cells in three 0.5- X 16-mm strips per culture were counted under phase-contrast microscopy just before harvest. The cells with processes of two soma1 diameters or longer were defined as neurite-bearing cells.
Preparation of crude GAGs GAGs were prepared as described previously (KatohSemba et al., 1989~).In brief, crude PGs from the medium were precipitated with ammonium sulfate,and the precipitate was dissolved in the buffer, followed by treatments with Pronase E, NaOH, and trichloroaceticacid (TCA). After removal of insoluble materials by centrifugation, the supernatant was treated with CTAB to make the CTAB-GAG complex. Crude GAGs were obtained by dissociating GAGs from the complex using potassium acetate. PGs in the cell layer were solubilized with sodium dodecyl sulfate in Tris-HC1 buffer (pH 7.5) containing EDTA after the dishes had been washed once with phosphate-buffered saline. Sodium dodecyl sulfate-solubilized PGs were stirred for 5 h at room temperature, and then insoluble materials were removed by centrifugation. Sequential treatments with Pronase, NaOH, and TCA followed.
Quantificationof individual GAGs As described in our previous article (Katoh-Semba et al., 1989a), individual GAGs were separated on a cellulose acetate membrane by two-dimensional electrophoresis, visualized by staining the membranes with Alcian blue, and then quantified spectrophotometrically by measuring the absorbance of 6 15 nm after dissolution of the pigment associated with GAGs in the CTAB solution. The radioactivity of [35S]sulfateincorporated into individual GAGs was measured by cutting out the spots on the membrane whose positions had been confirmed by exposure of the membrane to an x-ray film. Radioactivity was counted in a toluene scintillator.
Characterization of GAGs To characterize the structure of CS, the crude GAGs from the cell layer and the medium were digested with chondroitinase (CHase) AC (Saito et al., 1968). Subsequently, paper chromatography was camed out to separate the disaccharide materials from CS. The amounts of 4-sulfated and 6-sulfated disaccharides were determined by measuring their associated radioactivity. To characterize the structure of HS, crude GAGs were degraded with heparitinase (HSase) (Linker and Hovingh, 1972) or nitrous acid (Lindahl et al., 1973) and then chromatographed on a column of Cellulofine GCL-90m ( 1 X 90 cm) buffered with 50 mM Tris-HCI (pH 7.5) containing 2 mM EDTA, 0.2 M NaCl, 0.2% Nonidet-40, and I mM NaN,. The relative amounts of CS and the sulfation at the amino group (N-sulfate)or the hydroxy group (O-sulfate) of the constituent sugars of HS were estimated from the areas under the peaks of radioactivity due to [35S]sulfateon the chromatogram. The peak areas shown by the bars in Fig. 4 were cut separately out of the chromatograms, and their areas were determined by weight.
Uptake of [3sS]sulfateby cells At the end of a 90-min labeling, dishes were washed three times with 5 ml of ice-cold Hanks’ balanced salt solution. Incorporated [35S]sulfatewas extracted with 0.75 ml of 5% TCA for 30 min. Dishes were washed with 0.25 ml of TCA. The wash was combined with the first extract. The radioactivity of neutralized extracts was measured in Aquasol-2 with
a Beckman LS-9OOO liquid scintillation counter. [35S]Sulfatelabeled materials detected in the TCA-soluble fraction after a 90-min labeling were not GAGs, because the extracted, labeled materials were not precipitated by 75% ethanol containing 1.3% potassium acetate.
Other methods Lipids were extracted by the method of Bligh and Dyer (1959). Inorganic phosphorus in crude lipids was assayed according to the method of Bartlett (1959), and the weight of phospholipids was calculated from the amount of phosphorus as described by Dickerson (1968). DNA was quantified by the method of Burton (1956). The amount of cell-associated CS was corrected by subtracting the amount of CS attached to dishes, as described in our previous article (Katoh-Semba et al., 1989~).
RESULTS Proliferation and neuritic growth of PC12 cells The number of PC12 cells in the absence of NGF increased linearly up to 9 days of culture with a doubling time of 2 days. In the presence of NGF, proliferation proceeded at the same rate as that in its absence until -5 days of culture, and then the rate leveled off (data not shown). Neurite-bearing PC 12 cells were not observed in the absence of NGF at any time (Fig. 1A). In the presence of NGF, the cells at 3 days were flat, slightly enlarged, and had a few short processes. However, they did not yet have any process that could be defined as neurites (Fig. 1B). After 6 days of culture, -30% of cells developed neurites, although they were short, and some neurites possessed growth cones (Fig. 1C). Very long neurites were observed in PC12 cells cultured for 9 days with NGF (Fig. 1D). The percentage of neurite-bearing cells in the total cell population reached 60-70% after 9 days of culture. Cell proliferation under the present conditions was almost the same as reported previously (Katoh-Semba et al., 1984). However, the maximal percentage of neurite-bearing cells appeared to be slightly lower. This difference may be explained by the fact that the initial density of seeded cells was much higher in the present study. Characterization and quantification of GAGs In the absence of NGF, two kinds of GAG, corresponding to HS and CS, were detected by cellulose acetate electrophoresis in the cells cultured for 3, 6, and 9 days (data not shown). In contrast, a GAG spot having the same mobility as CS appeared on the electrophoretograms of the medium sampled at 3 and 6 days in culture. A faint spot corresponding to HS, in addition to CS, was found in the medium at 9 days. All GAGs detected with Alcian blue were labeled with [35S]sulfate(data not shown). The radioactive, pigmented spots corresponding to HS and CS disappeared after treatment with nitrous acid and CHase, respectively. This result demonstrates that HS and CS are synthesized in PC12 cells. The species of GAGs detected in the cell layer and the medium did not differ between the NGF-treated and control cultures. When J. Neurochem., Val. 55. No. 5. 1990
1752
R. KATOH-SEMBA ET AL.
FIG. 1. Photomicrographsof PC12 cells in the absence (A) or presence (8-D) of NGF: (A and D) at 9 days, (B) at 3 days, and (C) at 6 days. Bar = 20 pm. Arrows indicate growth cones.
the cells were cultured under floating-cell conditions without NGF, both HS and CS were detected in the cell layer and the medium. HS and CS were also labeled with [35S]sulfate(data not shown). Table 1 summarizes changes in the amounts of GAGS associated with the cells and released into the medium during the NGF-induced neuritogenesis in PC12 cells. The cells just before reseeding contained 0.06 pg of HS/ 1O6 cells, which was lower than the levels of HS detected in dish-attached cells under any culture condition. When the cells were seeded to the dishes, in the absence of NGF, the amounts of cell-associated HS per cell remained unchanged for the first 6 days and then decreased by 50% after 9 days of culture. In the presence of NGF, the levels of HS at 3, 6, and 9 days were almost equal to each other and were about two to three times as high as those in its absence. Because the membrane surfaces of neurite-bearing cells J. Neuroehem., Vol. 55, No. 5 , 1990
were considered to be increased in area, the amounts of phospholipids were determined. As shown in Fig. 2, the amounts of phospholipids per cell were increased with the neurite extension, and, after 9 days of culture, they reached 2.4 times the level in cells without neurites. In the absence of NGF, changes in the amounts of cell-associated HS per microgram of phospholipids were similar in pattern to those of HS per cell (Table 1). In contrast, in the presence of NGF, the level was transiently increased at 3 days and then decreased. The NGF-treated cells at 3 and 9 days had about twofold more HS than the nontreated controls. Irrespective of the presence of NGF, changes in the amounts of [35S]sulfate-labeledHS per cell were similar to those of HS per cell (Fig. 3A and Table 1). Radioactivity incorporated into cell-associated HS was 1.5- to twofold higher in the NGF-treated culture than in the nontreated culture at 3, 6, and 9 days.
1753
GLYCOSAMINOGLYCAN IN PC12 CELLS TABLE 1. Changes in levels of cell-associated and medium-released GAGs during neuritogenesis in PCl2 cells
cs
HS Culture (days) Cell layer 0" 3
6
9
NGF
~ gI O6 f cells
ng/pg of PL
&lo6 cells
n&pg of PL
-
0.06 k 0.01 0.19 f 0.05 0.38 f 0.05' 0.22 f 0.04 0.43 f 0.07' 0.1 1 k 0.03 0.39 f 0.02
0.83 f 0. I5 2.7 k 0.6 4.2 f 0.5' 3.1 ? 0.5 2.8 f 0.4 1.5 f 0.4 2.3 f 0.1'
0.03 f 0.00 0.19 f 0.07 0.37 ? 0.10' 0.05 f 0.01 0.14 f 0.04' 0.03 f 0.01 0.15 f 0.04'
0.33 f 0.02 2.6 f 0.8 4.2 f 1.0' 0.66 f 0.09 0.94 f 0.23' 0.42 f 0.08 0.89 2 0.21
-
+ + +
'
'
Medium Od
3 6 9
0.08 f 0.03 ND ND ND ND Trace Trace
-
-
++ -
+
0.52 f 0.1 I 0.73 f 0.12 1.16 f 0.31' 0.70 f 0.16 1.10 f 0.16' 0.46 f 0.02 0.96 f 0.27'
Data are average f SD values of four determinations from two separate experiments. Because serumderived CS attached to dishes, the amounts of cell-associated CS were corrected as described previously (Katoh-Semba et al., 19890). PL, phospholipid; ND, not detectable. a Determined using the cells just before reseeding. ' p < 0.01, ' p < 0.05 compared with the level in the absence of NGF. Determined using the culture medium under the floating-cellconditions.
The amount of cell-associated CS per cell or per unit of phospholipid was transiently enhanced after 3 days of culture and thereafter decreased in the absence of NGF (Table 1). The level of CS at 9 days was very close to that of CS associated with the cells just before seeding. In the presence of NGF, the time course of changes in the levels of cell-associated CS showed a profile similar to that in its absence, but the levels of CS were two to three times as high as those in the respective nontreated controls. On the other hand, in the absence of NGF, the amounts of [35S]sulfate-labeled CS per cell in the cell layer were almost the same at any day in culture as that in the floating cells, although they were slightly lower on day 9 (Fig. 3B). Treatment with NGF produced an apparent increase in the in-
FIG. 2. Changes in levels of phospholipids during neuritogenesis in PC12 cells inducedby NGF: no NGF (hatched cdumn) and with NGF, measured after 3. 6, and 9 days of culture (open columns). Data are average f SD (bars) values of four determinationsfrom two separate experiments.
g 8
5
gloo
3z 0
corporation of [35S]sulfate,and the level attained was maintained for 9 days without any significant change. The changes in the amounts of CS released into the medium per cell were not very marked during culture with or without NGF (Table 1). However, the level of CS released from the NGF-treated cells into the medium on each day was significantly higher than those released from the nontreated attached cells and from the nontreated floating cells. The amounts of [35S]sulfate-labeledCS released from NGF-treated and nontreated cells into the medium also remained unchanged during culture (Fig. 3C). However, the levels of labeled CS in the medium were lower in the absence of NGF and higher in its presence, compared with the level of labeled CS released from the nontreated floating cells, resulting in a distinct difference in the level of labeled CS between the NGF-treated and nontreated cultures. Changes in structure of GAGs during neuritic growth Because CS is known to be composed of chondroitin 4- and 6-sulfate units, changes in the composition of these [35S]sulfate-labeled disaccharides during the NGF-induced neuritogenesis in PC 12 cells were examined. As far as crude fractions of CS were concerned, >95% of CS in the cell layer or the medium was detected as a 4-sulfate unit, in the presence or absence of NGF (data not shown). There were also no changes in J. Neurochem.. Vol. 55, No. 5. 1990
R. KATOH-SEMBA ET AL.
I754
C
B
100
: '
FIG. 3. Changes in amounts of [36S]sulfatelabeled GAGs during neuritogenesis in PC12 cells induced by NGF: (A) HS in the cell layer, (6)CS in the cell layer, and (C) CS in the medium. Cells were cultured without NGF (O), with NGF (0).and under floatingcell conditions in the absence of NGF (A).
5 0 . ,/'
