|Home | About | Journals | Submit | Contact Us | Français|
Developing neurons deprived of trophic support undergo apoptosis mediated by activation of c-Jun N-terminal kinases (JNK) and c-Jun, induction of the Bcl-2 homology 3 (BH3)-only protein BimEL, Bax-dependent loss of mitochondrial cytochrome c, and caspase activation. However, the mechanisms that regulate each of these events are only partially understood. Here we show that the prolyl isomerase Pin1 functions as a positive regulator of neuronal death through a c-Jun-dependent mechanism. Ectopic Pin1 promoted caspase-dependent death of NGF-maintained neurons that was associated with an accumulation of Ser63-phosphorylated c-Jun in neuronal nuclei and was partially dependent on Bax. Downregulating Pin1 prior to NGF withdrawal suppressed the accumulation of phosphorylated c-Jun, inhibited the release of cytochrome c, and significantly delayed cell death. Pin1 knockdown inhibited NGF deprivation-induced death to a similar extent in Bim (+/+) and Bim (−/−) neurons. The protective effect of Pin1 knockdown was significantly greater than that caused by loss of Bim and nearly identical to that caused by a dominant negative form of c-Jun. Finally, cell death induced by ectopic Pin1 was largely blocked by expression of dominant negative c-Jun. These results suggest a novel mechanism by which Pin1 promotes cell death involving activation of c-Jun.
Depriving neonatal mouse or rat sympathetic neurons of nerve growth factor (NGF) in vitro is a well-established paradigm for studying the intracellular mechanisms that underlie programmed cell death in the developing nervous system (Freeman et al. 2004). A hallmark of this type of neuronal death is the activation of Jun N-terminal kinase (JNK). JNK activation is associated with an increase in the phosphorylation of c-Jun, induction of activator protein-1 (AP-1) transcription factors containing c-Jun, and increased transcription of various AP-1 target genes, including c-Jun itself (Xia et al. 1995; Virdee et al. 1997; Eilers et al. 1998). Numerous studies using pharmacologic, dominant-negative, and gene disruption approaches have established important roles for JNK activation and c-Jun in NGF deprivation-induced apoptosis in sympathetic neurons and in other models of trophic factor deprivation (Estus et al. 1994; Ham et al. 1995; Xia et al. 1995; Virdee et al. 1997; Eilers et al. 1998; Harding et al. 2001; Harris et al. 2002; Palmada et al. 2002).
One target of the JNK/c-Jun pathway during trophic factor deprivation is the Bcl-2 homology 3 (BH3)-only protein BimEL (reviewed by Ham et al. 2005). Ectopic expression of BimEL in sympathetic neurons results in Bax-dependent release of cytochrome c from mitochondria and ultimately cell death, even in the presence of NGF. Conversely, disruption of the Bim gene in mouse sympathetic neurons results in a 12–14 hr delay in cell death caused by NGF withdrawal, which is preceded by a similar delay in the rate of cytochrome c release (Whitfield et al. 2001; Putcha et al. 2001; Coultas et al. 2007). Following NGF withdrawal, Bim transcription is stimulated through a process that depends largely, but not exclusively, on c-Jun and AP-1 (Whitfield et al. 2001; Biswas et al. 2007). In addition, JNK-mediated phosphorylation of BimEL during NGF deprivation enhances its pro-apoptotic activity (Putcha et al. 2003; Becker et al. 2004). In cerebellar granule neurons deprived of survival factors, the activation of BimEL by JNK is partly mediated by the prolyl isomerase Pin1 (Becker and Bonni 2006).
Among peptidyl-prolyl isomerases, Pin1 is unique in its specificity for catalyzing cis-trans isomerization of the proline residue in phosphorylated Ser/Thr-Pro motifs in a growing number of mitogen-activated protein (MAP) kinase substrates (Yaffe et al. 1997; Ranganathan et al. 1997). This allows Pin1 to work in concert with cyclin-dependent protein kinases, extracellular signal-regulated kinases, JNK, and other MAP kinases in the regulation of a wide range of normal cellular processes including cell division, DNA damage response, and gene transcription, and in diseases such as cancer (Lu and Zhou 2007). Pin1 is also implicated in cell survival, and in the nervous system, Pin1 exerts both pro-survival and pro-apoptotic effects. For example, deletion of Pin1 in mice results in increased oligodendrocyte apoptosis after spinal cord injury (Li et al. 2007), and aged Pin1 (−/−) mice generate pathological features and show signs of neurodegeneration reminiscent of those seen in Alzheimer’s disease (Liou et al. 2003). In contrast, Pin1 functions in a pro-apoptotic pathway in cerebellar granule neurons through its ability to bind and stabilize JNK-phosphorylated BimEL (Becker and Bonni 2006). It is not known whether Pin1 functions to promote cell death in other models of neuronal apoptosis and, if so, whether its effects are mediated solely through BimEL.
Here we investigated a role for Pin1 in the death of sympathetic neurons induced by NGF withdrawal, focusing on the apoptotic events that lead to phosphorylation of c-Jun and release of cytochrome c from mitochondria. The results from these experiments provide new evidence for a role for Pin1 in programmed cell death and identify a novel pro-apoptotic pathway for Pin1 that is mediated by c-Jun.
NGF was purchased from Harlan Bioproducts for Science (Indianapolis, IN). Sheep anti-NGF antiserum was from Cedarlane Laboratories (Burlington, Ontario, Canada). Boc-aspartyl(OMe)-fluoromethylketone (BAF) was purchased from MP Biomedicals (Irvine, CA). Cell culture reagents were purchased from Invitrogen Corp. (Carlsbad, CA) and other reagents were obtained from Sigma-Aldrich (St. Louis, MO), unless otherwise indicated.
The human Pin1 open reading frame was amplified by polymerase chain reaction with an N-terminal c-myc tag and inserted into pcDNA3 (Invitrogen) between the BamHI and EcoRI restriction sites. Plasmids that express catalytically inactive forms of Pin1 (H59A and C113A) and plasmids expressing short-hairpin RNA (shRNA) under the control of the U6 RNA Pol III promoter (U6/Pin1 and U6/Ctr) were donated by Azad Bonni (Harvard Medical School, Boston, MA) and are described elsewhere (Becker and Bonni 2006). In some experiments an additional control plasmid (U6/Luc) was used that expresses a luciferase-targeting shRNA under the control of the U6 RNA Pol III promoter in pSHAG-1 (Paddison et al. 2002). Dominant negative c-JunbZIP plasmid is described elsewhere (Leppa et al. 1998) and was a gift from Dirk Bohmann (University of Rochester, Rochester, NY). pBOS-H2BGFP was purchased from BD Biosciences (San Jose, CA).
Bim-deficient (B6.129-Bcl2l11tm1.1Ast/J) mice (Bouillet et al. 1999) and Bax-deficient (B6.129X1-Baxtm1Sjk) mice (Knudson et al. 1995) were obtained from The Jackson Laboratory (Bar Harbor, ME). C57BL/6J mice were used for all other experiments. Primary cultures of sympathetic neurons were established from superior cervical ganglia of newborn (postnatal day 0–1) mice as described previously (Lipscomb et al. 1999) using procedures approved by our institutional committee on the care and use of animal resources. For experiments involving knockout mice, the superior cervical ganglia from each pup were dissociated and plated separately while the corresponding genotype was determined by PCR amplification of tail DNA using primers described in the original publications cited above. Cultures generated from wild type littermates were used as controls.
For microinjection experiments, dissociated cells were plated on polyornithine and laminin-coated 35-mm glass-bottomed dishes (MatTek Corporation, Ashland, MA) in NGF-containing media (90% Eagle’s minimum essential medium, 10% fetal bovine serum, 2 mmol/L L-glutamine, 20 mmol/L uridine, 20 mmol/L fluorodeoxyuridine, 100 U/ml penicillin, 100 mg/ml streptomycin, and 50 ng/ml NGF). For immunoblotting experiments, dissociated cells were first pre-plated for 1 hr on plastic tissue culture dishes (to remove the majority of adhering non-neuronal cells) before plating onto collagen-coated dishes. All cultures were maintained in 5% CO2 incubators at 37°C for 5–6 days before initiating experiments. To initiate NGF deprivation, cells were rinsed twice in NGF-free media prior to addition of fresh media lacking NGF and containing anti-NGF antiserum.
Microinjections of sympathetic neurons were performed essentially as described elsewhere (Sarmiere and Freeman 2001). Injection solutions contained 4 mg/ml rhodamine-conjugated dextran, 50 μg/ml pBOS-H2BGFP [which expresses green fluorescent protein (GFP) fused to histone H2B], and 50–100 μg/ml test plasmids in a buffer consisting of 100 mmol/L potassium chloride and 10 mmol/L potassium phosphate (pH 7.4). Neurons plated on polyornithine/laminin-coated dishes were switched into serum-free L-15 medium (Invitrogen) containing penicillin and streptomycin and microinjected directly into the nucleus. Immediately after the injection, cells were returned to NGF-containing media and incubated overnight to allow time for plasmid expression and for any cells damaged by the microinjection procedure to die. At this point, cells expressing GFP-histone were counted to establish a baseline number of successfully-injected cells.
Cell viability was assessed using epifluorescence microscopy by examining the morphology of the nuclei of injected neurons, visible because of the co-expressed histone H2B-GFP fusion protein. Nuclei with rounded shapes and diffuse GFP-histone fluorescence were scored as healthy. These neurons also exhibited a distinct boundary between nuclear and cytoplasmic compartments, one or more prominent nucleoli, and an intact and refractile cell body, when viewed under phase-contrast microscopy. Cells with condensed, indented, or otherwise irregularly-shaped chromatin and nuclei were considered unhealthy. Cell survival was expressed as the ratio of the number of healthy GFP-histone expressing cells counted at the indicated time points to the initial number of GFP-histone expressing cells determined after injection. All cell counts were performed in a blinded manner.
Immunofluorescence experiments were performed on neurons cultured in the presence of the pan-caspase inhibitor BAF (50 μmol/L), which prevents cell death in these neurons without affecting the signaling events upstream of caspase activation, including c-Jun phosphorylation and release of cytochrome c from the mitochondria (Sarmiere and Freeman 2001). Cells were rinsed twice with ice-cold phosphate-buffered saline, fixed with 3% paraformaldehyde for 15 min, and then incubated in 5% goat serum (Invitrogen) plus 0.3% Triton X-100 in Tris-buffered saline (TBS) for 0.5–1 hr. Cells were then incubated overnight at 4°C with an antibody that recognizes c-Jun phosphorylated at Ser63 (1:500; Cell Signaling Technology, Danvers, MA; Cat. No. 9261S), anti-cytochrome c antibody (1:250; BD Biosciences; Cat. No. 556432), or anti-Pin1 antibody (1:500, Cell Signaling Technology; Cat. No. 3722). After 3–5 washes, cells were incubated with Alexa Fluor-conjugated goat anti-rabbit or anti-mouse secondary antibody (1:500 and 1:400, respectively; Invitrogen) for 2 hr at 4°C in 1% goat serum and 0.3% Triton X-100 in TBS. Cells were rinsed and stained with Hoechst 33,258 before being analysed under epifluorescence with a Nikon Diaphot 300 microscope. The extent of c-Jun phosphorylation was quantified by determining the fraction of injected (GFP-histone-positive) cells exhibiting above background nuclear fluorescence. Cytochrome c localization was assessed by determining the fraction of GFP-histone-expressing cells that lacked the intense, punctate immunofluorescence characteristic of mitochondrial cytochrome c as seen in control NGF-maintained neurons. For each experiment, 100–300 cells were scored per condition and each experiment was repeated at least three times. Images were captured with a Dage-MTI camera and Scion Image software (Scion Corp, Frederick, MD). A Leica TCS SP confocal microscope was used for confocal microscopy.
HEK293 cells were transfected using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. For immunoblotting experiments, HEK293 cells or sympathetic neurons were lysed on ice in a buffer containing 50 mmol/L Tris-HCl (pH 7.5), 120 mmol/L NaCl, 0.5% Nonidet P-40, 1 mmol/L phenylmethyl sulfonyl fluoride, 1 mmol/L Na3VO4, 100 mmol/L NaF, and supplemented with protease inhibitor cocktail (Sigma). Total protein content was assessed using a commercial protein assay reagent (Biorad Laboratories, Hercules, CA). Equal amounts of protein from whole cell lysates were denatured in Laemmli sample buffer, separated by 15% SDS-PAGE, and then transferred to nitrocellulose membranes. Membranes were blocked in 5% fat-free milk and 0.1% Tween-20 in TBS for 1–4 hr at room temperature and then incubated overnight at 4°C with primary antibody diluted in the same buffer containing 1% fat-free milk (anti-Pin1, 1:1000, Cell Signaling Technology, Cat No. 3722; anti-Bim/Bod, 1:1000, Stressgen, Cat No. AAP330; anti-Phospho-c-Jun, 1:1000, Cell Signaling Technology, Cat No. 9261S; anti-c-Jun, 1:1000, Cell Signaling Technology, Cat No. 9162; anti-Bax, 1:1000, BD Biosciences, Cat No. 554106; anti-actin, 1:1000, Sigma). Membranes were rinsed 3–5 times and then incubated with a 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse antibody (Biorad Laboratories) for 1 hr. Proteins were detected with an enhanced chemiluminescence reagent kit (Pierce Biotechnology, Rockford, IL). Densitometry of scanned films was carried out using Scion Image software (Scion Corp. Frederick, MD).
We initially examined whether over-expression of Pin1 was sufficient to promote cell death in neurons maintained in the presence of NGF. Neonatal mouse sympathetic neurons maintained in the presence of NGF for five days were microinjected with a plasmid expressing Pin1 or the empty vector together with pBOS-H2BGFP, which expresses a GFP-histone fusion protein that permits visualization of the overall chromatin organization in injected cells. Examination of the cellular morphology and pattern of GFP-histone fluorescence in the nuclei of injected cells revealed that many of the neurons expressing ectopic Pin1 had undergone or were undergoing cell death (Fig. 1a). At both 72 and 96 hr post injection, expression of Pin1 resulted in significantly greater cell death than did injection of the control plasmid (Fig. 1b). Thus, ectopic Pin1 is sufficient to override NGF-triggered survival signals and promote cell death. These results extend recent findings in cerebellar granule neurons (Becker and Bonni 2006) and raise the possibility that Pin1 acts more generally to promote cell death in developing neurons.
Most biochemical effects of Pin1 are mediated by its cis-trans prolyl isomerase activity. To determine if Pin1 enzymatic activity is necessary for its ability to promote cell death, we took advantage of two previously described point mutants, Pin1(H59A) and Pin1(C113A), which retain the ability to bind substrate proteins but are catalytically inactive (Verdecia et al. 2000; Winkler et al. 2000; Becker and Bonni 2006). In contrast to wild type Pin1, neither mutant Pin1 increased cell death relative to control, suggesting that the prolyl isomerase activity of Pin1 is necessary for inducing death in these neurons.
In NGF-deprived sympathetic neurons, caspase activation occurs through a mitochondria-mediated intrinsic apoptotic pathway that is completely dependent on Bax (Deckwerth et al. 1996). To determine if Pin1-induced death also involves this pathway, we tested the effect of the general caspase inhibitor BAF on Pin1-induced cell death. When used at a concentration that efficiently inhibits NGF deprivation-induced cell death, BAF largely prevented cell death in neurons expressing ectopic Pin1 (Fig 2a). To examine the involvement of Bax, we microinjected Pin1 expression plasmid or control vector into neurons from Bax (+/+) and Bax (−/−) mice. Cell viability was initially assessed at 36 and 60 hr after injection. Consistent with the results in figure 1, expression of Pin1 in Bax (+/+) neurons resulted in significantly greater cell death compared to neurons injected with the control vector (Fig. 2b, c). In contrast, expression of Pin1 in Bax (−/−) neurons resulted in significantly less cell death, such that at 36 hr after injection the survival of these neurons was not different from control vector-injected Bax (+/+) or Bax (−/−) neurons. The results at 60 hr, however, suggested that the dependency on Bax might not be complete. To investigate this further, we also assessed survival at 84 hr after microinjection. Even at this later time, Pin1 expression resulted in more cell death in Bax (+/+) neurons than in Bax (−/−) neurons (51.4 +/− 2.5% survival in Bax (+/+) versus 64.3 +/− 1.1% survival in Bax (−/−), n = 3, p < 0.05). Still, injection of the Pin1 plasmid resulted in more death in Bax (−/−) neurons than did the control vector (64.3 +/− 1.1% survival with Pin1 versus 77.8 +/− 3.9% survival with control, n = 3, p < 0.05). Together these data suggest that Pin1 promotes caspase-dependent cell death in NGF-maintained sympathetic neurons through a mechanism that is partially dependent on Bax.
Pin1 is expressed widely in mouse and human tissues including the central nervous system (Fujimori et al. 1999; Liou et al. 2003). To verify that Pin1 is also expressed in peripheral neurons, whole cell lysates from sympathetic neurons were immunoblotted with a polyclonal antibody raised against a conserved region of Pin1. In contrast to c-Jun and BimEL, Pin1 was readily detected in NGF-maintained neurons and its levels remained unchanged at 7.5 hr after NGF withdrawal, a time when up-regulation of c-Jun, Ser63-phosphorylated c-Jun, and BimEL is readily apparent (Fig. 3a). Similar results were obtained in rat sympathetic neurons, and Pin1 levels did not change as late as 20 hr after removing NGF (data not shown).
To investigate whether Pin1 has a role in regulating cell death induced by NGF deprivation, we utilized a previously characterized shRNA plasmid (U6/Pin1) that targets a highly conserved region of Pin1 mRNA and that has been used effectively to knockdown Pin1 expression in cerebellar granule neurons (Becker and Bonni 2006). As controls for these experiments, we used the empty U6-promoter vector (U6/Ctr) and a second plasmid expressing an shRNA that targets firefly luciferase (U6/Luc). Transient expression of U6/Pin1 in HEK293 cells reduced the endogenous Pin1 protein to approximately 30% the level found in control cells (Fig. 3b). To verify knockdown of Pin1 in mouse sympathetic neurons, we microinjected neurons with U6/Pin1 or the control vectors and then analyzed Pin1 expression by immunofluorescence. While uninjected cells and cells injected with U6/Ctr or U6/Luc (not shown) exhibited similarly bright Pin1 immunofluorescence present throughout the cytoplasm and nucleus, the majority (70–80%) of U6/Pin1-injected neurons showed only faint immunofluorescence that was much less than that of nearby uninjected cells (Fig. 3c). Thus, Pin1 is constitutively expressed in sympathetic neurons and its expression can be suppressed in these cells with the U6/Pin1 plasmid.
To determine if reducing Pin1 expression influences the survival of neurons deprived of NGF, neurons were microinjected with U6/Pin1 or control plasmids and then either maintained in the presence of NGF or subjected to NGF deprivation. When maintained in the presence of NGF, the large majority of microinjected neurons appeared healthy regardless of which plasmid was injected (Fig. 3d). Within 24 hr of removing NGF, approximately half of the neurons injected with the control plasmids had died, consistent with the previously published rates of NGF-deprivation induced death in sympathetic neurons (Deckwerth and Johnson 1993; Edwards and Tolkovsky 1994). However, a much larger fraction (~80%) of U6/Pin1-injected neurons remained healthy at 24 hr after NGF withdrawal. Even after 2 days of NGF deprivation, nearly 50% of U6/Pin1-injected cells were still viable compared to only 25% of control cells (Fig. 3e). These results suggest that normal levels of Pin1 are important for one or more rate-determining processes that promote cell death after neurotrophin withdrawal.
The observation that death was delayed and not completely blocked in neurons expressing Pin1 shRNA was not entirely unexpected and may reflect incomplete knockdown of Pin1 or, perhaps more likely, a rate-determining but non-essential regulatory role for Pin1 during cell death. Future experiments with neurons from Pin1 null mice should help distinguish between these possibilities. As mentioned above, inhibiting JNK, c-Jun, or BimEL in NGF-deprived neurons also delays rather than completely prevents NGF deprivation-induced death.
Within the first few hours after NGF withdrawal, the activation of the JNK/c-Jun pathway together with the quenching of pro-survival pathways sets in motion events that converge to trigger the release of cytochrome c from mitochondria. In NGF-maintained neurons, cytochrome c immunofluorescence is distributed in a punctate pattern that co-localizes with mitochondria. By 24 hr of NGF deprivation, cytochrome c has been released into the cytosol in a majority of neurons, as reflected by a much less intense and more diffuse pattern of immunofluorescence in these cells (Deshmukh and Johnson 1998; Neame et al. 1998). To determine if Pin1 expression is important for the NGF deprivation-induced loss of mitochondrial cytochrome c, neurons injected with U6/Luc or U6/Pin1 were deprived of NGF and then subjected to immunofluorescence for cytochrome c (Fig. 4a). Approximately 65% of neurons injected with the control shRNA vector and deprived of NGF exhibited weak and diffuse immunofluorescence for cytochrome c, indicative of its release from mitochondria (Fig. 4b). In contrast, only 20% of neurons expressing shRNA targeting Pin1 showed evidence of loss of mitochondrial cytochrome c, indicating that Pin1 regulates cell death events upstream of cytochrome c release. A role for Pin1 upstream of cytochrome c release is also consistent with our finding that death induced by Pin1 in NGF-maintained neurons is reduced in the absence of Bax (see figure 2).
The above observations, together with the finding that in survival factor-deprived cerebellar granule neurons Pin1 enhances the pro-apoptotic function of JNK-phosphorylated BimEL (Becker and Bonni 2006), prompted us to examine whether the survival promoting effects of Pin1 knockdown are dependent on BimEL. While sympathetic neurons from Bim (−/−) mice still release cytochrome c from their mitochondria and undergo cell death after NGF withdrawal, they do so only after a 12–14 hr delay compared to Bim (+/+) neurons (Putcha et al. 2001). We reasoned that if the pro-apoptotic effects of Pin1 were mediated entirely through BimEL, then inhibiting Pin1 expression would produce at best the same delay in the rate of death as that caused by loss of Bim. On the contrary, if Pin1 contributes to death-promoting mechanisms other than or in addition to those requiring BimEL, then knockdown of Pin1 in Bim (−/−) neurons might result in even greater protection from cell death. To directly compare the delay in cell death caused by loss of Bim to that produced by inhibiting Pin1 expression, Bim (+/+) and Bim (−/−) neurons were injected with U6/Ctr or U6/Pin1 and then deprived of NGF for varying times. Bim (−/−) neurons died at a slightly slower rate than Bim (+/+) neurons (Fig. 5), consistent with previously published results (Putcha et al. 2001; Coultas et al. 2007). Interestingly, both Bim (+/+) and Bim (−/−) neurons injected with the U6/Pin1 vector died much more slowly than the corresponding control-injected cells. Survival was greatest in Bim (−/−) neurons expressing Pin1 shRNA, although the differences between these cells and wild type neurons expressing Pin1 shRNA were not statistically significant. While these results suggest that Pin1 functions to promote cell death in ways that do not require BimEL, they do not rule out BimEL as an additional target in sympathetic neurons.
The JNK/c-Jun pathway has a prominent role in initiating the cascade of apoptotic events upstream of cytochrome c release in neurons deprived of trophic support. Initial activation of JNK and phosphorylation of c-Jun at Ser63 and Ser73 triggers an autoregulatory loop (Angel et al. 1988) that results in increased c-Jun transcription and marked accumulation of JNK-phosphorylated c-Jun protein in the nuclei of NGF-deprived neurons (Virdee et al. 1997; Eilers et al. 1998). In breast cancer cells, Pin1 has been reported to bind to JNK-phosphorylated c-Jun and to promote its transcriptional activity (Wulf et al. 2001). How Pin1 enhances c-Jun activity and whether this occurs in other cells including neurons remains unknown. To test if Pin1 may contribute to c-Jun activation in NGF-maintained neurons, we performed immunofluorescence on neurons microinjected with the Pin1 expression vector using an antibody that exclusively detects Ser63-phosphorylated c-Jun on western blots of NGF-deprived neurons (Besirli et al. 2005). Pin1 over-expression was accompanied by a 3–4 fold increase in nuclei positive for phosphorylated c-Jun (Fig. 6a, b). In contrast, neurons injected with the control vector did not show any increase in immunofluorescence as compared to neighboring uninjected neurons. Thus, ectopic Pin1 is able to stimulate an increase in phosphorylated c-Jun in neuronal nuclei even in the presence of NGF.
To determine if endogenous Pin1 contributes to the accumulation of phosphorylated c-Jun during NGF deprivation, neurons were microinjected with U6/Pin1 or U6/Luc and then maintained with or without NGF for 20 hr. The cultures were then subjected to immunofluorescence for Ser63-phosphorylated c-Jun (Fig. 6c). NGF-maintained cells injected with either shRNA vector exhibited a low level of mostly cytoplasmic immunofluorescence that was indistinguishable from uninjected neurons (Fig. 6c top panels, and data not shown). In uninjected and U6/Luc-injected neurons, NGF deprivation resulted in intense nuclear immunofluorescence in nearly 60% of the cells (Fig. 6c, d). However, when U6/Pin1-injected neurons were deprived of NGF, only about 25% of the cells exhibited nuclear immunofluorescence above background. The finding that shRNA targeting Pin1 reduced but did not eliminate the nuclear accumulation of phosphorylated c-Jun is not unexpected given our earlier observation that suppressing Pin1 delayed rather than prevented NGF withdrawal-induced death. Because Pin1 can exert pro-apoptotic effects through an interaction with BimEL, we tested whether the reduction in Ser63-phosphorylated c-Jun caused by knocking down Pin1 expression would also occur in the absence of BimEL. As depicted in Fig. 6e, expression of U6/Pin1 in Bim (−/−) neurons inhibited the nuclear accumulation of phosphorylated c-Jun after NGF withdrawal to about the same extent as occurred in wild type neurons (see figure 6d). Together, these observations demonstrate that Pin1 can influence the accumulation of phosphorylated c-Jun in NGF-deprived neurons through a mechanism that is independent of its potential interaction with BimEL.
Since Pin1 can influence the levels of Ser63-phosphorylated and thus activated c-Jun, and blocking c-Jun inhibits death caused by NGF withdrawal, we compared the survival enhancing effects of downregulating Pin1 to those produced by a dominant negative form of c-Jun (c-JunbZIP), which lacks N-terminal sequences including the sites phosphorylated by JNK (Leppa et al. 1998). Death in neurons injected with c-JunbZIP plasmid and deprived of NGF was reduced to nearly the same extent as reported previously using a similar, though not identical, dominant negative c-Jun construct (Ham et al. 1995). At both 24 hr (data not shown) and 48 hr of NGF deprivation, the survival of neurons expressing shRNA targeting Pin1 was nearly identical to that of neurons expressing dominant negative c-Jun (Fig. 7a). Neurons co-injected with U6/Pin1 and c-JunbZIP showed slightly greater survival than those expressing either plasmid alone.
Overall, the results described above support a model in which Pin1 contributes to NGF deprivation-induced death by promoting the accumulation of JNK-phosphorylated c-Jun. A prediction from this model is that c-Jun function would be required for Pin1-induced cell death. To test this, we microinjected NGF-maintained neurons with Pin1, c-JunbZIP, or Pin1 together with c-JunbZIP and assessed survival 36–40 hr after injection. In these experiments, Pin1 expression resulted in approximately 30% cell death at this time, which was significantly greater than in control neurons (Fig. 7b). Co-expressing Pin1 with dominant negative c-Jun, however, almost completely blocked the ability of Pin1 to induce death. Pin1 induced death remained markedly inhibited by c-JunbZIP even when survival was assessed at 60–64 hr after injection (Pin1, 58.6 ± 8.4% cell death; Pin1 + c-JunbZIP, 26.3 ± 2.4% cell death; n = 4). To our knowledge, these results provide the first evidence that induction of cell death by Pin1 requires c-Jun activity.
Here we present new findings that suggest a novel pro-apoptotic role for Pin1 during programmed cell death in developing neurons. First, ectopic expression of catalytically active Pin1 is sufficient to override NGF-derived survival signals and promote caspase-dependent cell death. Second, expression of Pin1 in the presence of NGF results in elevated levels of Ser63-phosphorylated c-Jun. Third, down regulating Pin1 expression prior to NGF withdrawal inhibits the accumulation of phosphorylated c-Jun and the subsequent release of cytochrome c from mitochondria that occurs during trophic factor deprivation. Fourth, knockdown of Pin1 expression inhibits NGF deprivation-induced death to the same extent as a dominant negative form of c-Jun. Finally, Pin1-induced death is blocked by expression of dominant negative c-JunbZIP, indicating that c-Jun functions downstream of Pin1 in these cells. Together, these observations suggest that Pin1 functions during trophic factor deprivation in a pathway that leads to activation of c-Jun.
Precedence for Pin1 as a regulator of c-Jun comes from studies in tumor cells, where the ability of Pin1 to enhance cyclin D1 transcription was shown to be partially dependent on c-Jun (Wulf et al. 2001). In these cells, JNK phosphorylation of c-Jun promotes its binding to Pin1, resulting in greater c-Jun transcription factor activity. Whether Pin1 binds to phosphorylated c-Jun in NGF-deprived neurons remains to be determined. Pin1 could also enhance levels of phosphorylated c-Jun by stimulating the activity of JNK or JNK-activating kinases. Recently, Pin1 was shown to interact with the JNK scaffolding protein JIP3 in neurons (Becker and Bonni 2006) suggesting that a subset of the Pin1 in cells could be in close proximity to JNK. A third possibility is that Pin1 could act on proteins other than c-Jun to enhance AP-1 transcription factor activity, resulting in increased c-Jun transcription and thus c-Jun protein. Along these lines, a role for Pin1 as a transcriptional co-activator has been described in breast cancer cells (Yi et al. 2005).
Becker and Bonni (2006) recently provided the first evidence that Pin1 functions as a positive regulator of apoptosis in neurons. Using a model involving postnatal cerebellar granule neurons deprived of survival factors, they showed that Pin1 can act by binding and stabilizing JNK-phosphorylated forms of BimEL. As outlined in the Introduction, BimEL is also important for cell death in NGF-dependent sympathetic neurons. Furthermore, transcription of the Bim gene after NGF withdrawal is partially regulated by AP-1 (Biswas et al. 2007). However, our finding that inhibiting Pin1 expression results in greater protection from cell death than disruption of Bim indicates the existence of Pin1 targets other than BimEL that contribute to apoptosis caused by NGF deprivation. This is further supported by the observation that Pin1 expression increases c-Jun phosphorylation even in Bim (−/−) neurons. By affecting c-Jun activity, Pin1 would be expected to influence the expression of other pro-apoptotic AP-1 targets in addition to BimEL. These might include cyclin D1, the BH3-only protein Dp5, and the prolyl hydroxylase EGLN3 (Freeman et al. 1994; Park et al. 1997; Imaizumi et al. 1997; Lee et al. 2005; Ma et al. 2007).
While Pin1 has a pro-apoptotic function in developing neurons, it may serve a predominantly protective role in the adult nervous system. A variety of studies including post-mortem analyses in humans suggest that decreased Pin1 function may contribute to Alzheimer’s disease and certain frontotemporal dementias. Strong support for this hypothesis comes from the observation that mice lacking Pin1 develop Tau and Aβ-related neurodegeneration (reviewed by Balastik et al. 2007). A protective role for Pin1 in adult oligodendrocytes was also recently described. In a mouse model of spinal cord injury, the ability of Pin1 to bind and stabilize the anti-apoptotic Bcl-2 family protein Mcl-1 helps serve to protect oligodendrocytes from apoptosis (Li et al. 2007). Ultimately, the importance of Pin1 for cell death and cell survival will likely depend on the developmental stage, cellular context, and type of insult that is under investigation.
In conclusion, our study provides new evidence supporting a role for Pin1 in cell death induced by trophic factor deprivation. Based on these results, we propose a model in which Pin1 stimulates pro-apoptotic signaling in developing neurons at the level of JNK/c-Jun activation.
We thank Azad Bonni for generously providing Pin1 plasmids, Dirk Bohmann for providing plasmids and for critically reading the manuscript, and Elena Marvin for excellent technical assistance. We also thank Richard Libby for providing Bax knockout mice. This work was supported by grants to R.S.F. from the National Institutes of Health (NS034400, NS042224). M.C.B. acknowledges support from the Accademia dei Lincei Postdoctoral fellowship at the University of Rochester.