The neuroprotective effect of nicotine in Parkinson’s disease models is associated with inhibiting PARP-1 and caspase-3 cleavage

Clinical evidence points to neuroprotective effects of smoking in Parkinson’s disease (PD), but the molecular mechanisms remain unclear. We investigated the pharmacological pathways involved in these neuroprotective effects, which could provide novel ideas for developing targeted neuroprotective treatments for PD. We used the ETC complex I inhibitor methylpyridinium ion (MPP+) to induce cell death in SH-SY5Y cells as a cellular model for PD and found that nicotine inhibits cell death. Using choline as a nicotinic acetylcholine receptor (nAChR) agonist, we found that nAChR stimulation was sufficient to protect SH-SY5Y cells against cell death from MPP+. Blocking α7 nAChR with methyllycaconitine (MLA) prevented the protective effects of nicotine, demonstrating that these receptors are necessary for the neuroprotective effects of nicotine. The neuroprotective effect of nicotine involves other pathways relevant to PD. Cleaved Poly (ADP-ribose) polymerase-1 (PARP-1) and cleaved caspase-3 were decreased by nicotine in 6-hydroxydopamine (6-OHDA) lesioned mice and in MPP+-treated SH-SY5Y cells. In conclusion, our data indicate that nicotine likely exerts neuroprotective effects in PD through the α7 nAChR and downstream pathways including PARP-1 and caspase-3. This knowledge could be pursued in future research to develop neuroprotective treatments for PD.


INTRODUCTION
Parkinson's disease (PD) is a neurodegenerative disorder affecting the nigrostriatal dopamine tract that regulates the initiation and fluency of voluntary movement. Patients present with a characteristic set of neurological symptoms that include tremor, muscle rigidity, bradykinesia, stooped posture, shuffling gait and a lack of facial expression (Magrinelli et al., 2016). In the advanced stages, patients may also develop a subcortical dementia and a variety of neuropsychiatric symptoms (Sveinbjornsdottir, 2016). Symptoms can be alleviated temporarily with L-dopa and carbidopa but this does not alter the progression of the illness or the death of nigrostriatal neurons (Connolly & Lang, 2014).
Animal studies have provided useful insights into potential mechanisms for the protective effects of tobacco in PD. Lesioning cholinergic neurons in the pedunculopontine nucleus resulted in loss of substantia nigra dopaminergic neurons (Bensaid et al., 2016), showing that physiological levels of acetylcholine are required for survival of nigrostriatal dopaminergic neurons. In rats with 6-hydroxydopamine (6-OHDA) lesions of the medial forebrain bundle, nicotine or the α7 nAChR agonist ABT-107 improved neurological functioning in conjunction with restoring dopamine transporter levels and dopamine release (Bordia et al., 2015). Treatment with a different α7 nAChR agonist 3-[(2,4dimethoxy)benzylidene]-anabaseine dihydrochloride (DMXBA) or nicotine also protects dopamine neurons in rats injected with 6-OHDA (Costa, Abin-Carriquiry & Dajas, 2001;Suzuki et al., 2013). Blocking nAChR with chlorisondamine prevents the protective effects of nicotine in vivo. Another common animal model for PD relies on 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) that is toxic to nigrostriatal dopamine neurons. Both cigarette smoke and nicotine increased the survival of these neurons in MPTP mice (Parain et al., 2003). Similar results have been reported in non-human primates exposed to MPTP (Quik et al., 2006), and in mice models of PD using methamphetamine to induce dopamine neuron toxicity (Maggio et al., 1997). Nicotine increases the levels of fibroblast growth factor-2 (FGF-2) and brain-derived neurotrophic factor in rat striatum in these models, which could be one mechanism for neuroprotection (Mudo et al., 2007).
Cellular model systems have also been used to investigate specific pathways through which nicotine and other tobacco constituents could protect neurons in PD. Using cultured mouse ventral midbrain neurons that included dopamine neurons, one group used tunicamycin as an endoplasmic reticulum stressor and found that nicotine, at levels comparable to those achieved through smoking cigarettes, attenuated the unfolded protein response (Srinivasan et al., 2016). There is also evidence that dopamine release can be regulated by presynaptic nAChR in rat brain slices (Giorguieff-Chesselet et al., 1979), and mouse striatal synaptosomes (Grady et al., 1992;Rapier, Lunt & Wonnacott, 1990).
Poly (ADP-ribose) polymerase-1 (PARP-1) and caspase have both been implicated in the pathophysiology or etiology of PD. PARP-1 is a DNA-damage sensor that is activated in some PD models such as the MPTP mouse (Wang et al., 2003), and inhibiting PARP-1 reduced dopamine neuron death from MPTP (Iwashita et al., 2004), alpha synuclein and MPP+ (Outeiro et al., 2007). PARP-1 also mediates dopamine neuron degeneration in the 6-OHDA mouse PD model (Kim et al., 2013). Mutations in PARP-1 protect against mitochondrial dysfunction and neurodegeneration in mouse models of PD with mutations in the Parkin gene (Lehmann et al., 2016), and in human clinical populations (Infante et al., 2007).
Caspase-3 has been implicated in cleavage of a proapoptotic kinase protein kinase C delta (PKCdelta) that mediates neuron death in both MPP+ and 6-OHDA cellular PD models (Da Costa, Masliah & Checler, 2003;Kanthasamy et al., 2006;Shimoke & Chiba, 2001). There is also evidence that caspase-1 activation is the final step in apoptotic cell death in PD (Hartmann et al., 2000;Tatton, 2000). Acteoside binding to caspase-3 is neuroprotective in the rotenone rat PD model (Yuan et al., 2016), and caspase-3 activation has been observed to be important in a number of pathways related to PD (Shukla et al., 2014;Zawada et al., 2015). Genetic disruption of caspase-3 is also protective against the effects of MPTP (Yamada et al., 2010). To our knowledge, there have not been attempts to investigate whether the neuroprotective effects of nicotine involve PARP-1 or caspase.
In summary, there is evidence that nicotinic cholinergic drugs may delay progression of PD (Perez, 2015), and thus, the α7 nAChR has been proposed as a target for new medications to treat PD (Quik et al., 2015). However, since the mechanisms underlying the neuroprotective effects remain unclear, we sought to further investigate the role of the α7 nicotinic acetylcholine receptor (α7 nAChR) in mediating the protective effects of nicotine in PD. We used the ETC complex I inhibitor methylpyridinium ion (MPP + ) to induce cell death in SH-SY5Y cells as a cellular model for PD and used 6-hydroxydopamine (6-OHDA) lesions as a mouse model for PD. We investigated the potential involvement of PD-related molecules PARP-1 and caspase in both of these model systems.

Cell culture and treatment
SH-SY5Y cells are derived from a human neuroblastoma and are often used as a cellular model for PD because they express tyrosine hydroxylase, dopamine-beta-hydroxylase, and the dopamine transporter. Xie, Hu & Li (2010) SH-SY5Y cells (American Type Culture Collection (ATCC), Manassas, VA) were maintained as a monolayer in Dulbecco's Modified Eagle Medium (DMEM) (Gibco, ON, Canada) with 10% fetal bovine serum (Gibco, ON, Canada), 100 U/ml penicillin (Sigma-Aldrich, Oakville, ON, Canada), and 100 U/ml streptomycin (Sigma-Aldrich, Oakville, ON, Canada). Cells were cultured in a humidified atmosphere of 5% CO2, at 37 • C. All cells were cultured in 100-mm (diameter) cell culture plates (BD Biosciences, ON, Canada) until ∼80% confluence and then seeded into 24-well plates (BD Bioscience, ON, Canada) to achieve ∼90% confluence 24-28 h prior to treatment. The medium was replaced by DMEM without fetal bovine serum 12 h before treatments.

Drugs
MPP + (methylpyridinium ion) was purchased as MPP + iodide from Sigma-Aldrich, dissolved in water to a stock concentration of 500 mM, and wrapped with foil to protect from light. Choline, nicotine and methyllycaconitine (MLA) were purchased from Tocris Bioscience. Nicotine was used at a concentration of 2 mM for in vitro experiments based on previous reports (Ke et al., 1998;Wang et al., 2011). We used MLA at a concentration of 20 µM based on a previous report that MLA at 5 µM and 10 µM could alleviate amyloidβ peptide-induced cytotoxicity in SH-SY5Y cells, without affecting cell viability (Zheng et al., 2014). At 20 µM, MLA could theoretically interact with α4β2 and α6β2 receptors, but no α4 and α6 receptor subunit mRNA was detected in SH-SY5Y cells (Gould et al., 1992;Lukas, Norman & Lucero, 1993). The α7 acetylcholine receptor subunit has good expression levels in SH-SY5Y cells (Peng et al., 1994).

Propidium iodide (PI) and Hoechst33342 staining
Cultured SH-SY5Y cells were gently rinsed with phosphate-buffered saline (PBS) (prewarmed in 37 • C) twice, incubated with 50 µg/ml PI (Invitrogen, Carlsbad, CA) or double labeling with Hoechst 33342 (20 µg/ml) (Invitrogen, Carlsbad, CA) and PI for 30 min, and then rinsed three times with PBS. Fluorescent intensity was measured by a plate reader (Victor 3; Pekin-Elmer, Waltham, MA). The level of cell death was defined as the ratio of PI: Hoechst 33342. The fraction of dead cells was normalized to the cell toxicity that occurred in the control group.

Protein extraction
Striatial tissues were dissected from mice in 6-OHDA exposure models. Striata were homogenized in ice cold buffer containing (in mmol/L): 50 Tris-Cl, pH 7.4, 150 NaCl, 2 EDTA, 1 PMSF plus 1% Igepal CA-630, 0.5-1% sodium deoxycholate, 1% Triton X-100 and protease inhibitor mixture (5 µL/100 mg of tissue; Sigma-Aldrich, Okaville, ON, Canada) on ice and shaken at 4 • C for 1 h. Striatal tissues dissolved in the lysis buffer was centrifuged at 12,000 g for 10 min at 4 • C to yield the total protein extract in the supernatant. The concentration of protein was measured with the BCA protein assay kit (Pierce Protein Biology, ON, Canada). Equal amounts of samples (50∼100 µg) were denatured and subjected to 10% SDS-PAGE and Western blot analyses.
Seven days prior to 6-OHDA lesion surgeries, animals were given nicotine or saline control by subcutaneous injection (s.c.) (MP Biomedicals, LLC, Santa Ana, CA, USA) twice daily for two weeks. For the first three days animals received a dose of 0.4 mg/kg, which was then increased to 0.8 mg/kg for four days prior to surgery. This dose was continued for one week post-surgery until subjects were sacrificed for tissue collection.

Statistical analysis
Levene's homogeneity test or F test was used to compare the variances between groups. For equal variances, data were analyzed either by t -test, one-way analysis of variance (ANOVA) followed by Tukey's test, or two-way analysis of variance (ANOVA) followed by Bonferroni or Tukey's post-tests (SPSS Statistics, I.B.M Corporation, USA). For groups with unequal variance, data were analyzed either with a t -test with Welch's correction, a one-way analysis of variance (ANOVA), or two-way ANOVA, followed by Dunnett's post hoc test. Data are expressed as mean ± standard error of mean (SEM). The significance levels of p < 0.05, p < 0.01, or p < 0.001 were used for all analyses.

Nicotine inhibits MPP + -induced SH-SY5Y cell death
Tobacco exposure is associated with decreased risk for PD (O'Reilly et al., 2005;Ritz et al., 2007;Searles Nielsen et al., 2012) and nicotine is the most prominent psychoactive component of tobacco. Thus, we first investigated if nicotine could protect against cell death in a cellular model of PD: MPP + -induced SH-SY5Y cell death. As shown in Fig. 1A, using propidium iodide (PI) staining, MPP + treatment (500 µM, 24 h) induced more SH-SY5Y cell death compared to control cells (control: 1.00 ± 0.099; MPP + : 1.40 ± 0.086). Pre-treatment with nicotine (2 mM, 30 min) prior to MPP + treatment, decreased the level of cell death, as compared cells treated with MPP + alone (MPP + : 1.40 ± 0.086; MPP + with nicotine: 0.88 ± 0.068; Fig. 1B). These data show that nicotine can inhibit MPP + -induced SH-SY5Y cell death. as compared to those in control group, n = 5, t -test. (B) Pre-treatment with nicotine (2 mM, 30 min) prior to MPP + exposure in SH-SY5Y cells decreased the level of cell death as compared to those treated with MPP + only. *p < 0.05 as compared to those of control group, ##p < 0.01 as compared to MPP + group, n = 5 for control and MPP + groups, n = 3 for nicotine and MPP + with nicotine groups, two-way ANOVA followed by Bonferroni post-tests. All data are shown as mean ± SEM. The level of cell death was detected using PI (50 µg/ml) and Hoechst33342 (20 µg/ml) double staining, and was defined as the ratio (1 mM, 30 min), a specific nAChR agonist, followed by MPP + treatment (500 µM, 24 hrs) in SH-SY5Y cells decreased the level of cell death as compared to those treated with MPP + only. *p < 0.05 as compared to those of control group, ###p < 0.001 as compared to MPP + group, n = 5, two-way ANOVA followed by Bonferroni post-tests. All data are shown as mean ± SEM. The level of cell death was detected using PI and Hoechst33342 double staining, and was defined as the ratio of fluorescent intensity of PI: Hoechst33342. Full-size DOI: 10.7717/peerj.3933/ fig-2 nAChR is involved in the protective effect of nicotine To determine whether the protective effects of nicotine are mediated by activation of the nAChR, we investigated if activation of nAChR without using nicotine, inhibits MPP +induced SH-SY5Y cell death. Nicotine is an nAChR agonist and previous studies have shown that other nAChR agonists can protect against nigrostriatal dopamine neuron damage in PD animal models (Dajas et al., 2001;Janson et al., 1988;Maggio et al., 1998;Mudo et al., 2007). As shown in Fig. 2, choline (1 mM, 30 min), a nAChR specific agonist, decreased the level of cell death, when administered prior to MPP + treatment (Control: 1.00 ± 0.099; Choline: 0.98 ± 0.10; MPP+: 1.40 ± 0.086; MPP + with choline: 0.81 ± 0.12). These results indicate that activation of nAChR prevents SH-SY5Y cells from MPP + -induced cell death and suggest that nAChR activation is sufficient to protect SH-SY5Y cells against MPP + -induced death.
α7 nAChR mediates the protective effect of nicotine against MPP + -induced SH-SY5Y cell death We hypothesized that the α7 nAChR could be the receptor through which nicotine has neuroprotective effects in PD. Activation of the α7 nAChR has protective effects in other neurodegenerative disorders, and in Alzheimer's disease models (Fan, Gu & Wei, 2015;Hu et al., 2015;Shen & Wu, 2015). To confirm if α7 nAChR mediates the effect of nicotine to protect against MPP + -induced cell death, we pre-treated SH-SY5Y cells with methyllycaconitine (MLA) (20 µM, 30 min), a α7 nAChR-specific antagonist, followed by MPP + as above and either nicotine (2 mM, 30 min) or choline (1 mM, 30 min). As shown in Fig treatments, as compared to those treated with choline and MPP + alone. ***p < 0.001 compared to the control group, #p < 0.05 as compared to the MPP + group, n = 4, one-way ANOVA followed by Tukey's test. All data are shown as mean ± SEM. The level of cell death was detected using PI and Hoechst33342 double staining, and was defined as the ratio of fluorescent intensity of PI: Hoechst33342. Full-size DOI: 10.7717/peerj.3933/ fig-3 to nicotine and MPP + treatments, as compared to those treated with nicotine and MPP + alone (control: 1.00 ± 0.053; MPP + : 1.70 ± 0.119; MPP + with nicotine: 1.37 ± 0.0351; MPP + with nicotine and MLA: 1.81 ± 0.0628). Similarly, Fig. 3B shows that choline can reduce the cell death induced by MPP + and this effect is blocked by MLA (Control: 1.00 ± 0.053; MPP + : 1.70 ± 0.119; MPP + with choline: 1.34 ± 0.0197; MPP + with choline and MLA: 1.85 ± 0.0796). These data indicate that α7 nAChR signaling is necessary for the neuroprotective effect of nicotine.

The neuroprotective effect of nicotine is associated with decreased PARP-1 and caspase-3 cleavage
To expand on the in vitro results above, we performed unilateral 6-hydroxydopamine lesions in mice as an in vivo model of PD. We first confirmed that the 6-OHDA lesion was causing the expected death of dopamine neurons and that nicotine had a neuroprotective effect in vivo, by measuring the amount of tyrosine hydroxylase as a proxy for dopamine neuron survival. Tyrosine hydroxylase is the rate limiting enzyme in the synthesis of catecholamines including dopamine. Figure 5 shows that the lesioned hemisphere of the brain has lost approximately half of the TH-containing neurons, while nicotine treatment protected almost all of these neurons from death (non-lesioned side in 6-OHDA mice: 1.00 ± 0.070; lesioned side in 6-OHDA mice: 0.49 ±0.19, Fig. 5B; and non-lesioned side in 6-OHDA mice with nicotine: 1.00 ± 0.0473; lesioned side in 6-OHDA mice with nicotine: 0.978 ± 0.0565, Fig. 5C).
We also performed a control experiment to examine whether nicotine alone might alter PARP-1 or caspase-3 cleavage, using Western blots to quantify the amount of the intact, full-length protein vs. the cleaved form, in the unlesioned hemisphere of 6-OHDA mice. As shown in Fig. 7, there is no significant effect of nicotine alone on the cleavage of these two of expression levels of full length (B) and cleaved (C) caspase-3. The expression level of caspase-3 was defined as the ratio of the intensity of caspase-3: α-Tubulin, and was normalized as percentage of the control group, **p < 0.01, n = 3, two-way ANOVA. (D) Western blot analysis showing that cleaved PARP-1 was decreased in SH-SY5Y cells pre-treated with nicotine before MPP + treatment, as compared to those treated with MPP + only. α-Tubulin was used as a loading control. (E and F) Densitometric analysis of the intensity of expression levels of full length (E) and cleaved (F) PARP-1. The expression level of PARP-1 was defined as the ratio of the intensity of PARP-1: α-Tubulin, and was normalized as percentage of the control group, ***p < 0.001, n = 3, two-way ANOVA. All data are shown as mean ± SEM.
proteins. These data demonstrate that the neuroprotective effect of nicotine for dopamine neurons in PD models is associated with PARP-1 and caspase-3 cleavage pathways.

DISCUSSION
The data presented above demonstrate that nicotine inhibits MPP + -induced SH-SY5Y cell death through activating α7 nAChR, and inhibits PARP-1 and caspase-3 cleavage in the 6-OHDA mouse model for PD. We first demonstrated that activation of nAChR with either nicotine or choline is sufficient to protect SH-SY5Y cells from MPP + toxicity. Nicotine also inhibits PARP-1 and caspase-3 cleavage in MPP + -treated SH-SY5Y cells. We then showed that α7 nAChR activation is necessary for these neuroprotective effects by using the α7 nAChR antagonist methyllycaconitine, which reduces the number of cells rescued by nicotine. Finally, we used the in vivo 6-OHDA mouse model for PD to demonstrate that nicotine inhibits PARP-1 and caspase-3 cleavage, suggesting a potential downstream molecular mechanism for neuroprotection in PD. This study provides additional knowledge of potential mechanisms to explain the clinical phenomenon of reduced PD incidence in smokers and other people exposed to tobacco. Some have suggested that people who become tobacco users may have an underlying trait that also renders them less susceptible to PD (Barreto, Iarkov & Moran, 2014), a form of ''reverse causation'' rather than nicotine actually being neuroprotective. One group found that PD patients are able to quit smoking more easily than matched population controls (Ritz et al., 2014). However, there are two main arguments against this interpretation. The first is that passive exposure to cigarette smoke is also associated with a dose-dependent decreased risk for PD (Searles Nielsen et al., 2012), and the second is the large body of data from experimental animal and cellular models (Barreto, Iarkov & Moran, 2014;Perez, 2015).
The nAChRs are obvious potential starting points for the mechanism of neuroprotection by nicotine in PD, and these receptors have been investigated in many other studies (Barreto, Iarkov & Moran, 2014;Perez, 2015;Quik et al., 2015). α7nAChR agonists are also likely to have significant impacts in PD via the regulation of the immune system and intestinal permeability (Anderson et al., 2016). However, our study is unique in using SH-SY5Y cells exposed to MPP+ as an in vitro model to investigate the neuroprotective effects of nicotine and nAChR activation. Also novel is our attempt to investigate the effect of nicotine on PARP-1 and caspase-3 cleavage in the 6-OHDA mouse model for PD. These elements Western blot analysis showing that tyrosine hydroxylase expression level decreased in striatal tissues of lesioned hemisphere from mice exposed to 6-OHDA only, but did not change in mice pretreated with nicotine before 6-OHDA exposure. α-Tubulin was used as a loading control. (B) Densitometric analysis of expression levels of tyrosine hydroxylase. The expression level of tyrosine hydroxylase was defined as the ratio of the intensity of tyrosine hydroxylase: α-Tubulin, and was normalized as percentage of non-lesioned hemisphere exposed to 6-OHDA only, n = 3, two-way ANOVA. All data are shown as mean ± SEM. *p < 0.05 as compared to Non-lesioned hemisphere in 6-OHDA mice. showing that cleaved PARP-1 decreased in striatal tissues from mice pre-treated with nicotine before 6-OHDA exposure, as compared to those exposed to 6-OHDA only. α-Tubulin was used as a loading control. (B and C) Densitometric analysis of expression levels of full length (B) and cleaved (C) PARP-1. The expression level of PARP-1 was defined as the ratio of the intensity of PARP-1: α -Tubulin, and was normalized as percentage of the 6-OHDA group, *p < 0.05 compared to the 6-OHDA group, n = 3, t -test. (D) Western blot analysis showing that cleaved caspase-3 was decreased in striatal tissues from mice pre-treated with nicotine prior to 6-OHDA exposure, as compared to those exposed to 6-OHDA only. α-Tubulin was used as a loading control. (E and F) Densitometric analysis of the intensity of expression levels of full length (E) and cleaved (F) caspase-3. The expression level of caspase-3 was defined as the ratio of the intensity of caspase-3: α-Tubulin, and was normalized as percentage of the 6-OHDA group, **p < 0.01 compared to the 6-OHDA group, n = 3, t -test. All data are shown as mean ± SEM. 6-OHDA only. α-Tubulin was used as a loading control. (B and C) Densitometric analysis of expression levels of full length (B) and cleaved (C) caspase-3. The expression level of caspase-3 was defined as the ratio of the intensity of caspase-3: α-Tubulin, and was normalized as percentage of the 6-OHDA group, n = 3, t -test. (D) Western blot analysis showing no difference of cleaved PARP-1 was decreased in striatal tissues of the non-lesioned hemisphere from mice pre-treated with nicotine prior to 6-OHDA exposure, as compared to those exposed to 6-OHDA only. α-Tubulin was used as a loading control. (E and F) Densitometric analysis of the intensity of expression levels of full length (E) and cleaved (F) PARP-1. The expression level of PARP-1 was defined as the ratio of the intensity of PARP-1: α-Tubulin, and was normalized as percentage of the 6-OHDA group, n = 3, t -test. All data are shown as mean ± SEM.
provide insight into molecular mechanisms and potential targets for developing new PD treatments.
Our results do not exclude the involvement of other neuroprotective mechanisms. Several signaling pathways that promote cell survival are enhanced by stimulating nAChR, including the Src family-PI3 K-AKT pathway, with subsequent upregulation of Bcl-2 and Bcl-x, JAK2/STAT3 and MEK/ERK (Kawamata & Shimohama, 2011). Nicotine can protect SH-SY5Y cells from other types of insults, such as beta-amyloid toxicity, through Erk1/2-p38-JNK-dependent signaling pathways (Xue et al., 2014). However, there are no published studies investigating the role of PARP-1 or caspase in the neuroprotective effects of nicotine in PD disease models.
Caspases are a family of proteases that are activated during apopototic cell death (Kroemer & Martin, 2005). There are twelve numbered caspases, some of which initiate or execute apoptosis, but these enzymes also regulate inflammation and cell differentiation (Galluzzi et al., 2016). Caspases are initially synthesized as an inactive pro-caspase, which must undergo dimerization or oligomerization and then cleavage to become active (Shi, 2004). Caspases are involved in the pathophysiology of PD through mediating dopaminergic neuron death from MPTP (Furuya et al., 2004;Qiao et al., 2017;Viswanath et al., 2001), promoting synuclein aggregation (Wang et al., 2016), and cleaving Transactivation response DNA-binding protein 43 (TDP-43), which is a primary component of Lewy bodies in PD (Kokoulina & Rohn, 2010).
A number of neuroprotective compounds that have been studied in PD animal models also affect caspases, such as telmisartan (an angiotensin II type 1 receptor blocker) (Tong et al., 2016), and nerve growth factor (NGF) (Shimoke & Chiba, 2001). Directly blocking a caspase-3 cleavage site on the proapoptotic protein kinase C delta has neuroprotective effects in MPP+ and 6-OHDA PD models (Kanthasamy et al., 2006). Despite the prominence of caspases in neuronal death in PD, they may not be viable targets for treatment since directly blocking caspase-8 resulted in a switch from apoptosis to necrosis (Hartmann et al., 2001), and this may apply to other caspases as well (Kroemer & Martin, 2005). Modulating caspase function in PD through the nicotinic receptors may be a better approach for developing new treatments.
PARP-1 enzymes are involved in a number of neurodegenerative disorders including Alzheimer's disease and PD (Martire, Mosca & d'Erme, 2015). PARP-1 has DNA binding domains that detect DNA damage and facilitate repair. When PARP-1 levels are too high or when DNA damage is too severe, cell death is initiated (Burkle, 2001), and this decision is regulated by NAD+ depletion (Alano et al., 2010). During cell death programs, PARP-1 is cleaved into fragments that are specific to different apoptotic pathways (Chaitanya, Steven & Babu, 2010). The 89 KDa fragment we detected appears during apoptosis, and could have been generated by the action of caspase-3, caspase-7 (Lazebnik et al., 1994) or the lysosomal proteases cathepsin B or D (Gobeil et al., 2001).
Some PARP-1 genetic variants are protective against PD (Infante et al., 2007), and the involvement of PARP-1 in PD pathophysiology includes regulation of alpha-synuclein expression (Chiba-Falek et al., 2005), and modification of p53 in the MPTP model (Mandir et al., 2002). Small molecule inhibitors of PARP-1 reduce cell death induced by alphasynuclein and MPP+ (Outeiro et al., 2007), consistent with our results above. PARP-1 induced depletion of NAD+ could also contribute to decreasing sirtuins and mitochondrial dysfunction in PD (Anderson & Maes, 2014).
In conclusion, we have demonstrated that the neuroprotective effects of nicotine in animal and cellular models of PD is mediated by activation of α7 nAChR and the inhibition of PARP-1 and caspase-3 cleavage. All of these molecules have been previously implicated in the pathophysiology of PD, but until now have not been linked together. This knowledge could be used to aid development of novel treatments for PD, but further work to delineate the molecular pathway linking α7 nAChR to PARP-1 and caspase-3 is required.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
The authors received no funding for this work.

Competing Interests
Albert H.C. Wong is an Academic Editor for PeerJ.

Author Contributions
• Justin Y.D. Lu performed the experiments, analyzed the data, prepared figures and/or tables, reviewed drafts of the paper.
• Ping Su conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, reviewed drafts of the paper.
• James E.M. Barber and Joanne E. Nash contributed reagents/materials/analysis tools, reviewed drafts of the paper.
• Anh D. Le conceived and designed the experiments, contributed reagents/materials/analysis tools, reviewed drafts of the paper.
• Fang Liu and Albert H.C. Wong conceived and designed the experiments, wrote the paper, reviewed drafts of the paper.

Animal Ethics
The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers):