Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The propargylamine ()-deprenyl (DEP) inhibits monoamine oxidase B (MAO-B). DEP was first shown to reduce the death of primate nigrostriatal dopaminergic neurons exposed to 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) (Cohen et al., 1984) and to slow the clinical progress of human Parkinson's disease (Parkinson, 1993). Both actions appeared to depend on MAO-B inhibition. Subsequently, DEP and deprenyl-related propargylamines (DRPs) were demonstrated to reduce neuronal loss independently of MAO-B inhibition in a variety of experimental models including cortical catecholaminergic neurons exposed to N-(-2-chloroethyl)-N-ethyl-2-bromobenzylamine, murine or primate substantia nigra dopaminergic neurons exposed to MPTP, rat facial motoneurons after axotomy, dopaminergic cells treated with the 1-methyl-4-phenylpyridinium ion (MPP⁺) or nitric oxide, and hippocampal neurons exposed to kainate (see Tatton et al., 2000 for details and references). The MAO-B-independent increases in neuronal survival by DRPs were shown to involve decreased apoptosis in a number of the above models (e.g., kainate-exposed hippocampal neurons, nitric oxide, or MPP⁺-treated dopaminergic cells). DRP antiapoptosis also has been found in other models including partially NGF-differentiated PC-12 cells after serum and NGF withdrawal, rat hippocampal neurons after ischemia/hypoxia, neuroblastoma cells treated with rotenone, rat cerebellar granule neurons exposed to cytosine arabinoside, serum-deprived human melanoma cells, rat retinal neurons after hypoxia, ischemia, or serum withdrawal, and rat hippocampal neurons, cerebellar neurons, or neuroblastoma cells exposed to okadaic acid (see Tatton et al., 2000 for details and references). Notably, DRPs do not reduce all forms of apoptosis as has been shown in NGF-naive PC-12 cells (Vaglini et al., 1996), thymocytes treated with dexamethasone (Fang et al., 1995), and cerebellar granule neurons exposed to low-media K⁺ levels (Paterson et al., 1998).

The mechanisms of DRP antiapoptosis are not known. DRPs bind to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Kragten et al., 1998; Carlile et al., 2000) and prevents the GAPDH up-regulation essential to some forms of neuronal apoptosis (see Tatton et al., 2000 for a review). Whether or how DRP binding to GAPDH results in decreased apoptosis is not known. DRP antiapoptosis can require new protein synthesis (Tatton et al., 1994). In keeping with a protein synthesis-dependent mechanism, DEP has been shown to alter the levels, message expression, or activity of a variety of different proteins, each in a different cell type (see Tatton et al., 2000 for references and details). Although some of the protein alterations may contribute to antiapoptosis, the extent of DRP-induced antiapoptotic protein synthesis is not known nor whether any protein synthesis changes are appropriately timed to interrupt apoptosis signaling. DRPs can prevent mitochondrial membrane potential (_M) dissipation in some forms of apoptosis in which _M dissipation results from increases in mitochondrial membrane permeability (Paterson et al., 1998; Wadia et al., 1998; Zhang et al., 1999). In some forms of apoptosis, increases in mitochondrial membrane permeability allow the release of mitochondrial factors that signal for apoptotic degradation (Jacotot et al., 1999). Accordingly, DRPs may alter the synthesis of proteins that influence mitochondrial membrane permeability.

We therefore examined the effects of three DRPs, DEP and desmethyldeprenyl (DES), both of which inhibit MAO-B, and 10-(N-methyl-N-propargyl-amino)methyldibenz[b,f]oxepine (CGP 3466), which does not inhibit MAO-B (see Waldmeier et al., 2000 for a review of CGP 3466) in trophically withdrawn PC-12 cells that were previously exposed to NGF for varying periods. We found that the DRPs reduced apoptosis in PC-12 cells exposed to NGF for 4 days or more but not in NGF-naive PC-12 cells. The decreased apoptosis required new protein synthesis and involved alterations in the expression of a number of proteins. Proteins in the mitochondrial and nuclear fractions were most markedly involved, including proteins known to scavenge oxidative radicals and those that can alter mitochondrial membrane permeability. The time course of the changes in expression was appropriate to antiapoptosis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture, Treatment, and Counting. PC-12 cells (American Type Culture Collection, Manassas, VA) were propagated in minimum essential medium (MEM) containing 10% horse serum, 5% fetal bovine serum, 2 mM L-glutamine, 50 units/ml penicillin, and 50 µg/ml streptomycin (M/S, MEM with serum), all purchased from Invitrogen (Carlsbad, CA). The cells were grown on 24-well plates (8 × 10⁴ cells/well) for counting of intact nuclei as an estimate of survival, poly-L-lysine-treated coverslips (1 × 10⁴ cells/coverslip) for imaging with laser confocal scanning microscopy, or 100-mm dishes (1 × 10⁶ cells/plate) for protein chemistry. The cells were differentiated for up to 9 days in M/S supplemented with 100 ng/ml 7S NGF (Upstate Biotechnology, Lake Placid, NY). MEM with serum and NGF is abbreviated as M/S+N (see Tatton et al., 1994; Wadia et al., 1998; and Carlile et al., 2000 for further details of culture and treatment). Following incubation for 1 to 9 days in M/S+N, cells underwent three successive washes in Hanks' balanced salt solution (HBSS; Invitrogen) to remove NGF and serum-borne trophic agents and then were replaced into M/S+N for controls or into MEM only (M/O) to induce apoptosis by serum and NGF withdrawal. Varying concentrations of DEP (Sigma-Aldrich, St. Louis, MO), DES (Toronto Chemical, Toronto, ON, Canada), CGP 3466 (Novartis AG, Basel, Switzerland), actinomycin D, camptothecin, or cycloheximide (all from Sigma-Aldrich) were added to the M/S+N or M/O cultures at varying times to oppose apoptosis or to inhibit new protein synthesis. The NGF-naive undifferentiated PC-12 cells were maintained in M/S on plates, wells, or coverslips for 6 days prior to washing to induce apoptosis by serum withdrawal. Washed cells were replaced into M/S or M/O with or without identical additives to those above.

Caspase Inhibition of Cells. Two caspase inhibitors, N-benzyloxycarbonyl-Val-Ala-Asp-fluomethylketone (zVAD-fmk; Sigma/RBI, Natick, MA) and Ac-Asp-Glu-Val-Asp-CH (acetyl-DEVD aldehyde; Sigma/RBI) were used with PC-12 cells that were supported with serum and NGF and those that were serum- and NGF-withdrawn. Both inhibitors were applied in MEM with the addition of 0.25% dimethyl sulfoxide. zVAD-fmk has K_i's that show it to be a general caspase inhibitor whereas acetyl-DEVD aldehyde has a K_i for caspase-3 that indicates a strong predilection for that caspase (Garcia-Calvo et al., 1998).

DNA Electrophoresis. Cells grown in medium-sized flasks and trophically withdrawn as above were examined for internucleosomal DNA digestion characteristic of apoptosis. At 6 to 18 h after washing, 2 × 10⁶ cells were rinsed with isotonic PBS, and DNA was extracted and prepared according to the methods of Batistatou and Greene (1993). The samples were incubated with 50 mg/ml DNase free Rnase (Roche Applied Science, Indianapolis, IN) at 37°C for 30 min. The recovered soluble DNA was electrophoresed on a 1.2% agarose gel and blotted onto Gene Screen Plus membrane (PerkinElmer Life Sciences, Boston, MA). Blots were probed with total genomic DNA digested with Sau 3A (Roche Applied Science). ³³P-labeled probe was prepared by the random priming reaction, and hybridization and washings were performed according to the manufacturer's protocol.

Metabolic Labeling of Cells for Two-Dimensional Gel Electrophoresis of Total Cell Protein. After a methionine-free MEM (mfMEM; Sigma-Aldrich) wash, the PC-12 cells were pulse-labeled for 1.5 h with trans-[³⁵S]methionine (75 µCi/ml; Amersham Biosciences, Piscataway, NJ) in the mfMEM at 37°C to label newly synthesized proteins and to estimate overall de novo protein synthesis. Cells undergoing treatment with DRPs were incubated in mfMEM with 10⁹ M concentrations of each DRP. mfMEM was removed from the dishes and replaced with M/S+N, M/O, or MEM with one of the DRPs for up to 4.5 h. At 6 h after the initial washing, all media were removed, and the labeled PC-12 cells were lysed (lysis buffer: 9.5 M urea, 2% 3-[(3-cholamidopropyl)dimethylammonio]propanesulfonate, 1.6% Servalyt, pH 5 to 7, 0.4% Servalyt, pH 2 to 11, and 5% -mercaptoethanol; all from Sigma-Aldrich) directly on tissue culture dishes, scraped, and mechanically homogenized. To remove unincorporated radiolabel, the lysates were precipitated in 1 ml of 7% trichloroacetic acid with deoxycholate (Sigma-Aldrich) (100 µg/ml) and centrifuged for 30 min at 15,000g. The resulting pellets were resuspended in 1 ml of 5% trichloroacetic acid, recentrifuged, and the supernatants discarded. The pellets were finally solubilized in 50 ml of the lysis buffer.

Scintillation Counting and Two-Dimensional Gel Electrophoresis of Subcellular Protein Fractions. PC-12 cells were grown and treated as described above, after washing in HBSS, and cells were metabolically labeled as above with trans-[³⁵S]methionine label in mfMEM for 1.5 h in M/S+N, M/O, or MEM with DRPs as above. At 6 h after the HBSS washing step, cells were harvested and homogenized in MOPS/sucrose/EDTA buffer (pH 7.2) with the addition of phenylmethylsulfonyl fluoride, dithiothreitol, and leupeptin (Sigma-Aldrich) to inhibit protease activity prior to homogenization. The suspension of PC-12 cells in PBS was then centrifuged four times at 200g for 10 min at 4°C, the final wash being in nuclear buffer containing 10 mM 1,4-piperazinediethanesulfonic acid, pH 7.4, 10 mM KCl, 2 mM MgCl₂, 1 mM dithiothreitol, 10 µM cytochalasin B, and 1 mM phenylmethylsulfonyl fluoride (all from Sigma-Aldrich). The cells were allowed to swell for 20 min and homogenized by 30 strokes of a Dounce glass homogenizer. The resulting pellet was layered over 30% sucrose and centrifuged at 800g for 10 min to isolate the nuclear fraction. The final nuclear pellet was suspended in a small volume of nuclear buffer containing 250 mM sucrose and frozen until further analysis was performed. The supernatants above the nuclear pellet were differentially centrifuged to isolate the mitochondrial fraction by a 4°C centrifugation at 14,000g for 10 min. The resulting mitochondrial pellet was resuspended in mitochondrial containing 250 mM mannitol with 0.1% (w/v) bovine serum albumin, pH 7.2, 0.5 mM EGTA, and 5 mM HEPES supplemented with 1 µg/ml leupeptin, 1 µg/ml pepstatin A, 50 µg/ml antipain, and 10 µg/ml chymostatin (all from Sigma-Aldrich) and recentrifuged. The pellet was then suspended in a small volume of mitochondrial buffer and stored until further processing. The supernatant resulting from the mitochondrial isolation containing enriched peroxisomes and soluble cytoplasmic proteins was then centrifuged at 100,000g for 1 h to pellet peroxisomal and cytoplasmic proteins. Finally, the above mitochondrial samples were then applied to metrizamide gradients (Sigma-Aldrich) to further separate the mitochondrial proteins from those of the plasma membrane and the Golgi. Protein concentrations were determined for individual subcellular fractions using a BCA protein assay kit (Pierce Chemical, Rockford, IL), and identical amounts of subcellular proteins were analyzed by scintillation counting or by preparative gels. To demonstrate the enrichment of the subfractions, equal amounts of proteins from each fraction were Western blotted and probed with antibodies for nucleolin, 14-3-3-, and cytochrome oxidase (see Carlile et al., 2000 for details and examples of our use of these methods).

Western Blots for Protein Levels. Alterations in the expression of specific proteins were examined using Western blots of protein lysates that were extracted at various times 3, 6, 9, 12, 18, and 24 h after washing in HBSS and incubation in either M/S+N, M/O, or MEM with DRP added. PC-12 cell proteins were extracted as total soluble lysates or lysates for the nuclear, mitochondrial, and cytosolic fractions (detailed methods and confirmation of fractionation purity are presented in Carlile et al., 2000). Briefly, the cells were treated as above, scraped from Petri dishes in cold PBS, and harvested by centrifugation at 300g for 5 min at 4°C. The pellets were washed twice in cold PBS and resuspended in buffer containing 25 mM Hepes-KOH, pH 7.5, 10 mM KCl, 1.5 mM MgCl₂, 1 mM sodium EDTA, 1 mM sodium EGTA, 1 mM dithiothreitol, and 5 µg/ml leupeptin, 5 µg/ml chymostatin, 5 µg/ml pepstatin A, 5 µg/ml aprotinin, plus 1 mM benzamidine and 250 mM sucrose (all from Sigma-Aldrich). The cells were homogenized by 12 to 15 strokes of a glass Dounce homogenizer and centrifuged at 800g for 10 min at 4°C to pellet the nuclear fraction. The supernatants were again centrifuged at 10,000g for 15 min at 4°C. These pellets contained the mitochondrially enriched fraction, and the supernatant included the cytoplasmic fraction. Both nuclearly and mitochondrially enriched fractions were resuspended in 50 µl of the above buffer and frozen at 30°C. Prior to use, the samples were protein assayed by the BCA protein assay method. The protein fraction lysates (30-40 µg) were electrophoresed on either 10 or 12% SDS-polyacrylamide gels and then transferred to nitrocellulose blotting membranes (Bio-Rad, Hercules CA). Membranes were agitated at room temperature for 2 h with primary antibodies [dilutions being 1:500 for anti-BCL-2 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), 1:2000 for anti-BAX (BD Biosciences, San Jose, CA), 1:800 for anti-c-FOS (Geneka Biotechnology Inc., Montreal, QC, Canada), 1 µg/ml anti-c-JUN (Stressgen Biotechnologies Corp., Collegeville, PA), 1:5,000 anti-CuZn superoxide dismutase (SOD1)/Mn superoxide dismutase (SOD2) (BioDesign International, Abingdon, UK), 1 µg/ml tubulin (Molecular Probes), 1:2000 neurofilament light protein (Sternberger Monoclonals Inc., Lutherville, PA), 1:1000 anti-GAPDH (Chemicon International, Temecula, CA), 1:1000 anti-MAP-2 (Sigma-Aldrich), 1:1200 anti-tyrosine hydroxylase (Zymed Laboratories, South San Francisco, CA), and 1:100 µg/ml anti-glutathione peroxidase (MBL International, Watertown, MA)] in 3% blocking solution (Amersham Biosciences). Following washing in Tris-buffered saline (pH 7.4), membranes were incubated for 1 to 2 h in the appropriate alkaline phosphatase- or horseradish peroxidase-labeled secondary antibody (Amersham Biosciences). Reactions were visualized using either 3,3-diaminobenzidine/0.2% H₂O₂ or nitrotetrazolium blue/5-bromo-4-chloro-3-indolyl-phosphate (all from Sigma-Aldrich) as substrates. All detected bands were then digitized using a charge-coupled device camera and imaged and analyzed using MetaMorph software.

Statistical Evaluation. To statistically evaluate the data, the individual measurements of data from different treatment groups were first analyzed using Statistica software (StatSoft, Tulsa, OK) to carry out two-tailed independent sample t testing. Levene's testing for homogeneity of variances showed that most pairs of samples were not homogeneous and ² evaluation of the distributions showed that most did not fit a normal distribution (see Siegel, 1956). Accordingly, analysis with parametric methods such as the t test may not provide valid results. The data were rank ordered and first tested with the Kruskal-Wallis analysis of variance by ranks (Siegel, 1956) using Statistica. Post hoc analysis was carried out using Mann-Whitney U testing. Both methods do not require homogeneity of variances, that the underlying distributions for the data be known, or that the values are linearly related (Siegel, 1956). Kruskal-Wallis analysis of variance values are presented in the text, and Mann-Whitney U values are only presented where they differ from the Kruskal-Wallis. Asterisks in figures indicate p < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Dependence of PC-12 Survival after Protein Synthesis Inhibition on Duration of Previous NGF Exposure. Previous studies have shown differing effects of protein synthesis inhibitors on PC-12 cell survival. Those studies seemed to suggest that the duration of previous NGF exposure could determine how the inhibitors affect PC-12 cell survival. Protein synthesis inhibitors have decreased survival of serum-supported, NGF-naive PC-12 cells (i.e., cells maintained in medium with serum) (Torocsik and Szeberenyi, 2000) whereas they did not alter the decreased survival induced by serum withdrawal in the NGF-naive cells (Rukenstein et al., 1991). In marked contrast, protein synthesis inhibitors prevented much of the decreased survival caused by serum and NGF withdrawal in PC-12 cells that had been previously exposed to NGF for 12 days (Mesner et al., 1995). Alternatively, protein synthesis inhibitors did not alter the reduced survival caused by serum and NGF withdrawal in PC-12 cells previously exposed to serum and NGF for 6 days (Tatton et al., 1994).

View larger version (57K): [in this window] [in a new window]   Fig. 1.   Effect of protein synthesis inhibitors on PC-12 survival depends on previous NGF exposure. Varying concentrations of actinomycin D (A1, B1, C1, and D1), camptothecin (A2, B2, C2, and D2), and cycloheximide (A3, B3, C3, and D3) were applied to NGF-naive PC-12 cells (A1-A3) or PC-12 cells previously exposed to NGF for 3 (B1-B3), 6 (C1-C3), or 9 days (D1-D3) to determine whether the duration of previous NGF exposure altered the effect of protein synthesis inhibition on cell survival. Cells were either trophically supported by serum or serum and NGF (open circles) or trophically withdrawn (filled circles). Values are means ± S.E.M. in all figures. Means are for 8 to 16 measurements in this and other plots of survival. Distance between dotted lines show the extent of control S.E.M. values for trophically supported (labeled M/S or M/S+N) or trophically withdrawn (M/O) survival without protein inhibitor treatment. , values significantly differ from controls (p < 0.05 relative to M/S, M/S + N, or M/O); , washed and returned to M/S; , washed and placed in M/O, serum- or serum- and NGF-withdrawn.

DRPs Induce Concentration-Dependent Survival Increases in PC-12 Cells Exposed to NGF for 6 Days But Not Those Exposed to NGF for Less Than 4 Days. About 4 days of NGF exposure were required before DRPs increased the survival of serum- and NGF-withdrawn cells. First, we found that DRPs increased the survival of 6-day NGF-differentiated PC-12 cells after serum and NGF withdrawal but did not increase the survival of serum-withdrawn, NGF-naive cells (Fig. 2, A1-A3). DEP (Fig. 2A1) and DES (Fig. 2A2) induced significant increases in survival (p < 0.05 for MEM with DRP addition compared with M/O) over a concentration range of 10⁵ to 10¹¹ M, with the greatest increase at 10⁹ M. CGP 3466 induced a similar survival-concentration relationship (Fig. 3A3) but also significantly increased survival at 10¹³ M (p < 0.05 for MEM with CGP 3466 addition compared with M/O). As shown in the same plots, the DRPs did not alter the survival of NGF-naive PC-12 cells that were maintained in MEM with serum for 6 days prior to undergoing serum withdrawal. DRPs induced similar increases in survival in the 9-day NGF-differentiated cells to those found for the 6-day NGF-differentiated cells (data not shown).

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Concentration relations for DRP increases in survival and duration of NGF exposure necessary for DRP-induced increases in survival after serum and NGF withdrawal. A1, A2, and A3 show that DEP, DES, and CGP 3466 concentrations of 105 to 1011 M increase the survival of 6-day NGF-differentiated PC-12 cells after serum and NGF withdrawal but similar concentrations do not increase the survival of NGF-naive cells after serum withdrawal. B1, B2, and B3 show that approximately 4 days of prior NGF exposure is necessary for the three propargylamines to increase the survival of PC-12 cells after serum and NGF withdrawal. *, significant increases (p < 0.05) in survival relative to those for M/O.

View larger version (89K): [in this window] [in a new window]   Fig. 3.   Apoptotic markers in NGF-naive and 6-day NGF-differentiated PC-12 cells. A1 through D2, each horizontal pair of laser confocal scanning micrographs is for an identical image field. The left-hand member presents an interference contrast image and the right-hand image presents YOYO-1 fluorescence. A1, A2, B1, and B2 are for NGF-naive PC-12 cells maintained in M/S for 6 days and 6-day NGF-differentiated PC-12 cells, respectively, that were washed repeatedly and replaced into M/S or M/S+N to maintain trophic support. C1, C2, D1, and D2 are for NGF-naive PC-12 cells maintained in M/S for 6 days and 6-day NGF-differentiated PC-12 cells, respectively, that were trophically withdrawn by repeated washing and placement into M/O. The interference contrast images for 6-day NGF-differentiated PC-12 cells show apoptotic membrane blebbing and process withdrawal (compare B1 and D1). The filled arrows indicate nuclei with chromatin condensation whereas the open arrow indicates a nucleus undergoing mitosis for comparison. E, DNA electrophoresis gels showing laddering typical of internucleosomal DNA digestion at 6 to 18 h after serum and NGF withdrawal in 6-day NGF-differentiated PC-12 cells. F, acetyl-DEVD aldehyde, a caspase inhibitor with a predilection for caspase-3, and zVAD-fmk, a general caspase inhibitor, induce about 45 and 65% reductions, respectively, in the decreased survival induced by serum and NGF withdrawal in 6-day NGF-differentiated PC-12 cells. *, significant increases in survival (p > 0.05) compared with cells not treated with a caspase inhibitor (0 µM concentration).

9

Decreased Survival after Serum and NGF Withdrawal in the 6-Day NGF-Differentiated PC-12 Cells Involves Apoptosis. We previously showed that partially NGF-differentiated PC-12 cells undergo apoptotic degradation after serum and NGF withdrawal using 1) DNA gel electrophoresis to demonstrate nuclear DNA cleavage, 2) in situ DNA end labeling with Apop Tag (Sigma-Aldrich) or Bodipy dUTP (Serologicals Corp., Norcross, GA) for nuclear DNA cleavage, 3) DNA staining with Hoechst 33258 (Molecular Probes, Eugene, OR) or YOYO-1 for nuclear chromatin condensation, and 4) immunocytochemistry for antibodies against nuclear histones to demonstrate nuclear protein reorganization (Tatton et al., 1994; Wadia et al., 1998; Carlile et al., 2000).

9

View larger version (39K): [in this window] [in a new window]   Fig. 4.   DRP alterations in the time courses of appearance of nuclear chromatin condensation and decreased survival initiated in NGF-differentiated PC-12 cells by serum and NGF withdrawal. A1, A2, and A3 show the time course of the appearance of nuclei with chromatin condensation shown by YOYO-1 staining in 6-day NGF-differentiated PC-12 cells and the time course of reductions in nuclei with chromatin condensation induced by 109 M DEP, DES, or CGP 3466, respectively. B1 and B2 show the time course of decreased 6-day NGF-differentiated PC-12 cell survival initiated by serum and NGF withdrawal and the time course of the increased survival induced by 109 M DEP, DES, or CGP 3466 after serum and NGF withdrawal. C1 and C2 show that delays in the administration of 109 M DEP or DES administration, respectively, of less than 2 h did not alter the increase in survival induced by the DRPs whereas delays in administration of 3 to 5 h progressively reduced the capacity of the DRPs to increase survival of the 6-day NGF-differentiated PC-12 cells. , significant increases in survival (p > 0.05) compared with M/O; , M/S + N; , M/O; , DEP 109 M; , DES 109 M; , CGP 3466 109 M.

Dependence of DRP Antiapoptosis on New Protein Synthesis. Metabolic labeling using [³⁵S]methionine incorporation was determined for the three protein synthesis inhibitors at maximum concentrations (see Fig. 1, C1-C3) that did not reduce survival (3 µg/ml actinomycin D, 10 µg/ml cycloheximide, and 20 µg/ml camptothecin). Scintillation counting showed that new protein synthesis was decreased by 92% or more for each of those concentrations (bar graphs in Fig. 5, A1-A3) when the maximum concentrations of the protein synthesis inhibitors that did not decrease survival were applied. Those concentrations of the protein synthesis inhibitors were applied to the 6-day NGF-differentiated PC-12 cells following serum and NGF withdrawal. Similar to Fig. 1, Fig. 5B1 shows that those concentrations of the protein synthesis inhibitors did not affect the survival of the 6-day NGF- differentiated PC-12 cells when they were serum- and NGF-supported or after serum- and NGF-withdrawn (p > 0.05).

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Transcriptional or translational inhibition prevents DRP antiapoptosis in NGF-differentiated PC-12 cells after serum and NGF withdrawal. A1, A2, and A3 show that 3 µg/ml of actinomycin D, 20 µg/ml of camptothecin, or 8 µg/ml of cycloheximide, respectively, which do not reduce the survival of 6-day NGF-differentiated PC-12 cells, reduce new protein synthesis by 94% or more (black bars, serum- + NGF-supported, gray bars, serum- + NGF-withdrawn). Employment of the three protein synthesis inhibitor at those concentrations did not alter the survival of serum- and NGF-supported or serum- and NGF-withdrawn cells (B1) but prevented the three DRPs administered at 107, 109, and 1011 M (B2, B3, and B4) from increasing survival after serum and NGF withdrawal. Progressive delays in the administration of 3 µg/ml actinomycin D (C1) or 8 µg/ml of cycloheximide (C2) showed that the protein synthesis inhibitors completely blocked the capacity of 109 M DEP from increasing survival for delays of up to 5 h and 109 M DES for delays of up to 3 h. Longer delays in the administration of the two inhibitors allowed for progressive increases in survival. *, significant increases in survival (p > 0.05) compared with M/O; , no inhibitor or DRP addition; , DRP + actinomycin D addition; , DRP + cycloheximide addition; , DRP addition; , DRP + camptothecin addition.

7

11

9

DRPs Alter the New Synthesis of a Number of Proteins during Apoptosis. We examined metabolically labeled 2D protein gels for whole cell lysates at 6 h after washing. Six hours represents the end of the first phase of decreased survival, which corresponded to the onset of increased apoptotic nuclear degradation revealed by chromatin condensation and the time at which delays in the addition of protein synthesis inhibitors allowed for maximal increases in survival to be induced by the DRPs (Fig. 5, B1 and B2). Autoradiograms like those in Fig. 6 represent the incorporation of [³⁵S]methionine into proteins and hence, provided an estimate of the synthesis of new proteins during the first 6 h after washing. The partially NGF-differentiated PC-12 cells maintained in serum and NGF after washing were found to incorporate the ³⁵S label into a relatively large number of proteins as indicated by the distribution and number of punctate autoradiographic densities (Fig. 6A1). In contrast, relatively fewer densities were detected in samples taken from cells that were NGF- and serum-withdrawn (Fig. 6A2). As well as the overall decrease in densities, a number of densities were evident in serum- and NGF-withdrawn cells that did not appear to be present in the serum- and NGF-supported cells (compare Fig. 6, A1 and A2). The autoradiograms also appeared to show that DRP treatment at 10⁹ M increased the number of densities in the serum- and NGF-withdrawn cells (compare Fig. 6A2 to Fig. 6A3). The autoradiograms were repeated for three experiments, each of which offered similar changes in the numbers and distributions of densities for labeled proteins. The DRP-induced increase in densities appeared to result from the recovery of some densities lost with serum and NGF withdrawal together with the appearance of other densities that were not evident in either serum- and NGF-supported or serum- and NGF-withdrawn cells.

View larger version (69K): [in this window] [in a new window]   Fig. 6.   DEP induces widespread changes in new protein synthesis, notably in nuclear and mitochondrial proteins, in 6-day NGF-differentiated PC-12 cells entering apoptosis after serum and NGF withdrawal. Typical autoradiograms for 2D protein gels for metabolically labeled total protein at 6 h after washing and replacement in medium with serum and NGF (A1), washing and placement in medium without serum or NGF (A2), and washing and placement in medium without serum or NGF with 109 M DEP (A3). B1-i to B1-iii, B2-i to B2-iii, B3-i to B3-iii, and B4-i to B4-iii present 2D autoradiograms for the plasma membrane, mitochondrial, cytosolic, and nuclear protein subfractions, respectively, at 6 h after washing and replacement in serum and NGF (i panels), washing and placement in MEM only (ii panels), and washing and placement in MEM with 109 M DEP (iii panels). C presents scintillation counts for metabolically labeled proteins corresponding to those for total protein and each of the subcellular fractions. Pl. M., Mit., Cyt., and Nuc. indicate plasma membrane, mitochondrial, cytosolic, and nuclear subfractions, respectively. *, significant increases in counts (p > 0.05) compared with those for M/O.

9

9

Proteins Affected by DRPs. Western blots for multiple time points were performed to identify some of the proteins of which synthesis was altered by DRP treatment. Examples of typical Western blots are presented in Fig. 7, whereas optical density measurements taken from Western blot bands are presented in Fig. 8. Figure 7, A1, A2, B1, and B2 show typical changes in BCL-2 immunoreaction induced by DES or DEP at 1, 3, 6, and 9 h or 3, 6, 12, and 24 h, respectively, following serum and NGF withdrawal. Levels of BCL-2 immunoreaction were reduced by 3 h following serum and NGF withdrawal, with a further decline in protein levels over subsequent hours (Fig. 7, A2 and B1 and Fig. 8A). DEP (Fig. 7B2), DES (Fig. 7A2), and CGP 3466 (not shown) at 10⁹ M maintained BCL-2 immunoreaction at levels similar to those found in serum- and NGF-supported cells (Fig. 8A). Notably, the DRPs did not alter levels of BCL-2 immunoreaction in cells that were maintained in M/S+N after washing and therefore were not entering apoptosis (see right-hand four bands in Fig. 7A1 for an example).

View larger version (56K): [in this window] [in a new window]   Fig. 7.   Western blots for the immunodensity of selected proteins after serum and NGF withdrawal with and without DRP treatment. BCL-2 immunodensity at 1, 3, 6, and 9 h after washing and replacement in medium with serum and NGF support (A1) or placement into medium without serum and NGF (A2). Immunodensity for BCL-2 (B1-B3), c-JUN (C1-1, C1-2, C2-1 and C2-2), and MAP-2 (D1 and D2). SN6, 6 h after washing and replacement in medium with serum and NGF support; O3, O6, O12, and O24, 3, 6, 12, and 24 h, respectively, after washing and placement into medium without serum and NGF; and d3, d6, d12, and d24, 3, 6, 12, and 24 h, respectively, after washing and placement into medium without serum and NGF with added 109 M DEP. E, BAX immunodensity for nuclear, mitochondrial, and cytosolic subfractions at 3, 6, 9, and 12 h after washing and replacement in medium with serum and NGF support (top panel), after washing and placement into medium without serum and NGF (middle panel), and after washing and placement into medium without serum and NGF with added 109 M DES (bottom panel).

View larger version (41K): [in this window] [in a new window]   Fig. 8.   Plots of average Western blot immunodensity above background for selected proteins at 0, 3, 6, 12, and 24 h after serum and NGF withdrawal. Each point represents the average ± S.E.M. optical density for immunodensity bands from three experiments and is normalized against the average optical density for bands from serum and NGF-supported, 6-day NGF-differentiated PC-12 cells at 6 h after washing. The values for serum- and NGF-supported, 6-day NGF-differentiated PC-12 cells are plotted at time 0. Other points represent the time after washing for serum- and NGF-withdrawn cells. Each plot is labeled for the respective primary antibody. A 9-h time point was substituted for the 12-h time point for GAPDH. , serum + NGF withdrawal; , serum + NGF withdrawal + 109 M DEP. Glut. Perox., glutathione peroxidase; TH, tyrosine hydroxylase; NFL, neurofilament light protein.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

NGF induces multiple changes in PC-12 cells including: 1) differentiation into an action potential-generating, neurite-bearing, sympathetic neuron-like phenotype; 2) a reduction in proliferation; and 3) an increase in survival (Greene and Tischler, 1982). Similar to previous reports, we found that DEP does not decrease apoptosis caused by serum withdrawal from NGF-naive PC-12 cells (Vaglini et al., 1996). We found that 4 days of prior NGF exposure is necessary before DRPs can reduce PC-12 cell apoptosis initiated by serum and NGF withdrawal.

DEP was first shown to prevent decreases in dopaminergic indices of nigrostriatal neuronal survival caused by MPTP exposure in primates (Cohen et al., 1984), which was interpreted to show that DEP protected the neurons by blocking MPTP conversion to MPP⁺ by MAO-B. A number of clinical trials showed that DEP can slow the clinical progression of Parkinson's disease (see Parkinson, 1993 as an example), but it is uncertain whether the slowing represents reduced neuronal death or alterations in dopamine metabolism. It is certain that DEP and other DRPs reduce neuronal death induced in vivo and in vitro by a wide variety of insults in a number of different neuronal models. Those insults have included 6-hydroxydopamine, MPP⁺, MPTP, nitric oxide or peroxynitrite, N-(-2-chloroethyl)-N-ethyl-2-bromobenzylamine, glutathione depletion, peripheral nerve crush or axotomy, optic nerve crush, hypoxia and/or ischemia, cytosine arabinoside, excitotoxins, trophic insufficiency, thiamine deficiency, okadaic acid, and aging. The neurons or neuron-like cells have included mesencephalic or nigral dopaminergic neurons, hippocampal neurons, dentate neurons, cerebellar granule and Purkinje neurons, cerebral cortical neurons, thalamic neurons, retinal ganglion neurons, spinal and facial motoneurons, neuroblastoma cells, and partially differentiated PC-12 cells (reviewed in Tatton et al., 2000). Studies involving a range of models have shown that DRPs can increase neuronal survival without MAO-B inhibition and by reducing apoptosis (Tatton et al., 2000). Since PC-12 cells do not express MAO-B (Youdim et al., 1986) and CGP 3466 does not inhibit MAO-B (Kragten et al., 1998; Waldmeier et al., 2000), DRP antiapoptosis in the serum- and NGF-withdrawn, 6-day NGF-differentiated PC-12 cells is also MAO-B-independent.

As before, we examined multiple indices to determine whether the increases in cell survival result from antiapoptosis. Time course studies of survival versus the percentage of cells with nuclear chromatin condensation showed that about 16% of the cells die during the first 6 h after serum and NGF withdrawal by a process that is unlikely to be apoptotic. The remaining cell loss, occurring after 6 h, meets multiple criteria for apoptosis and is responsive to DRP treatment. zVAD-fmk, a general caspase inhibitor, prevented almost 100% of the apoptosis whereas acetyl-DEVD aldehyde, a caspase-3 inhibitor, prevented less than 50%. This is in keeping with previous studies of NGF-differentiated PC-12 cells, which have shown that general caspase or caspase-2 inhibitors almost completely prevent apoptosis initiated by NGF or serum and NGF withdrawal whereas caspase-3 inhibition only partially prevents the apoptosis (Haviv et al., 1997, 1998; Stefanis et al., 1998).

We found that new protein synthesis is necessary for reductions in apoptosis induced by the three DRPs. In various models, protein synthesis inhibitors can slow or reduce apoptosis, induce apoptosis, or may not affect the extent or timing of apoptosis (Eastman, 1993). Hence, the use of protein synthesis inhibitors to examine antiapoptotic agents can be complicated if the apoptosis itself requires new protein synthesis or if a protein synthesis inhibitor induces apoptosis. Apoptosis initiated by serum withdrawal from NGF-naive PC-12 cells (Rukenstein et al., 1991; Mesner et al., 1995) or serum and NGF withdrawal from partially NGF-differentiated PC-12 cells (Tatton et al., 1994) are not new protein synthesis-dependent. Alternatively, apoptosis initiated in fully differentiated PC-12 cells by NGF withdrawal requires new protein synthesis (Mesner et al., 1995). Some protein synthesis inhibitors induce apoptosis over one concentration range and reduce apoptosis over another (Torocsik and Szeberenyi, 2000). Our findings suggest that the duration of previous exposure to NGF determines how protein synthesis inhibitors influence apoptosis signaling initiated by serum and NGF withdrawal in PC-12 cells.

Our experiments in which protein synthesis inhibitor addition was delayed relative to serum and NGF withdrawal and DRP addition suggested that critical antiapoptotic alterations in transcription/translation induced by DES begin prior to 3 h after DES addition whereas those for DEP continue to 5 h. DES is a principal metabolite of DEP (Baker et al., 1999), and studies using cytochrome P450 inhibitors suggest that DEP antiapoptosis requires DEP metabolism to DES (Tatton and Chalmers-Redman, 1996). The relative prolongation of the protein synthesis inhibitor blockade of DEP antiapoptosis may reflect a period necessary for DEP metabolism to DES. The experiments in which DRP addition was delayed relative to the onset of serum and NGF withdrawal indicate that the apoptosis signaling events that are critical to DRP antiapoptosis occur between 2 and 5 h after serum and NGF withdrawal, which is within the same time domain for critical DRP-induced new protein synthesis suggested by the protein synthesis inhibitor delay experiments.

Metabolic labeling suggested that DRPs induced changes in the new synthesis of a number of proteins, with the most marked changes involving the mitochondrial and nuclear protein subfractions. Previous work used differential display-polymerase chain reaction to identify four genes, c-jun, heat-shock protein 70, phosphoglycerate kinase, and calpactin I heavy chain, of which expression was increased in retinal ganglion neurons initiated into apoptosis by serum deprivation or hypoxia (Xu et al., 1999). In that study, 10⁹ M DEP reversed the increases in c-jun and heat-shock protein 70 gene expression but not that for the other genes. Like with the retinal study, we have found that DRPs alter the synthesis of some proteins but not others and found a transient increase in c-JUN. As well as for c-JUN, we found alterations for a number of proteins that previously were shown to play a role in PC-12 cell apoptosis, including BCL-2, BAX, SOD1, SOD2, c-FOS, glutathione peroxidase, and GAPDH.

Importantly, we found that the timing of alterations in the levels or subcellular distribution of those proteins induced by DRPs were appropriate to antiapoptosis. In this and previous studies (Wadia et al., 1998; Carlile et al., 2000), we found that apoptotic nuclear degradation first became evident at about 6 h after serum and NGF withdrawal. Western blots showed that the levels of antiapoptotic proteins decreased and those of pro-apoptotic proteins increased by 3 h after serum and NGF withdrawal. DRP treatment caused those early alterations to return toward control levels in association with a decrease in the percentage of cells with apoptotic nuclear degradation. The timing for changes in the levels of apoptotic and antiapoptotic proteins appear in accord with that found by the protein synthesis inhibition and DRP addition delay experiments.

Increases in the phosphorylation and levels of c-JUN are induced by NGF withdrawal from sympathetic neurons or NGF-differentiated PC-12 cells and involves the up-regulation of c-Jun N-terminal kinase (JNK) and/or p38 mitogen-activated protein kinase (Xia et al., 1995; Maroney et al., 1999). Pharmacological inhibition of JNK blocks apoptosis in NGF-differentiated PC-12 cells induced by NGF withdrawal but not apoptosis induced in NGF-naive PC-12 cells induced by serum withdrawal (Maroney et al., 1999). NGF withdrawal in 6-day NGF-differentiated PC-12 cells induced an increase in the levels of an endogenous mitogen-activated protein kinase kinase kinase, apoptosis signal-regulating kinase 1 (Kanamoto et al., 2000). Apoptosis signal-regulating kinase 1 up-regulation was necessary for c-JUN up-regulation and apoptosis initiated by NGF withdrawal in the cells (Hatai et al., 2000; Kanamoto et al., 2000). The timing of the c-JUN up-regulation was consistent with the transient c-JUN increase that we found 3 h after serum and NGF withdrawal.

The prevention of decreases in BCL-2 and the decreased mitochondrial BAX localization induced by the DRPs may contribute to the maintenance of _M found with DRP treatment of neurons or neuron-like cells entering apoptosis (Paterson et al., 1998; Wadia et al., 1998). BCL-2 prevents decreases in _M caused by agents that increase mitochondrial membrane permeability and induce apoptosis in PC-12 cells (Dispersyn et al., 1999). Mitochondrial BAX accumulation, similar to our finding in the NGF-differentiated PC-12 cells, has been shown for a number of forms of apoptosis including NGF withdrawal from sympathetic neurons (Putcha et al., 2000). Mitochondrial BAX accumulation increases mitochondrial membrane permeability and decreases _M (Narita et al., 1998). Prevention of increased mitochondrial membrane permeability can be a key step in blocking apoptotic degradation (Jacotot et al., 1999).

DRPs have been shown to bind to GAPDH in association with reductions in apoptosis (Kragten et al., 1998; Carlile et al., 2000). Up-regulation of GAPDH together with the dense nuclear accumulation of GAPDH immunoreactivity is characteristic of apoptosis that can be blocked by GAPDH antisense oligonucleotides (reviewed in Tatton et al., 2000). The tumor suppressor protein, p53, has been shown to up-regulate GAPDH in neuronal apoptosis (Chen et al., 1999). p53 activation is downstream to JNK activation but upstream to BAX (Aloyz et al., 1998; Mielke and Herdegen, 2000) in a variety of apoptosis models, which may suggest that DRP binding to GAPDH could uncouple a JNK-p53-GAPDH apoptosis signaling pathway.