Early low-dose ghrelin intervention via mini-osmotic pumps could protect against the progressive dopaminergic neuron loss in Parkinson’s disease mice

Ghrelin has been identied as a multifunctional peptide that has a potential applications for the treatment Parkinson’s disease (PD). Method: ELISA was used for detecting plasma total and active ghrelin levels, dopamine (DA) content was measured by HPLC-ECD, immunouorescence staining and Western blot were used to detect protein expressions, and cytokine was tested by Bio-PlexPro™ assay. Here, we reported a subcutaneous administration of low-dose ghrelin via mini-osmotic pumps to PD mice. The decreased levels of total and active ghrelin in plasma were rescued by ghrelin administration. Interestingly, ghrelin had no effect on weight gain in wild-type mice but improved weight loss in PD mice. We observed the attenuation of dopaminergic (DAergic) neuron loss in substantia (SN) and low level of dopamine content in the striatum in PD mice with ghrelin treatment. Ghrelin administration could improve the environment of DAergic neuron by inhibiting microglia proliferation and pro-inammatory cytokine expression, and could enhance cell survival by upregulating Bcl-2/Bax ratio and superoxide dismutase1 (SOD1) protein level in SN in PD mice.

symptom reported by nearly 90% of all PD patients, with rates rising as the disease progresses. We found that plasma total ghrelin and active ghrelin levels were signi cantly decreased both in early PD patients, which indicated that plasma ghrelin levels might be irrelevant to PD progression [7]. For ghrelin's multifunction, it might be used as a biomarker and endogenous protective factor for PD diagnosis and therapy in future. Our previous studies also showed that the non-motor symptoms such as colon motility, anxiety-like and depressive-like behaviors, and motor coordination also changed from 3 months age, whereas no motor symptoms exhibited in homozygous mice which expressed human mutant A53T alpha synuclein (A53T mice) [8]. Thus, we wanted to know whether the progressive dopaminergic neuron loss were prevented if given protective administration in the early stage of PD.
In this work, we rst tested the plasm ghrelin level in homozygous A53T mice. We showed that both total and active ghrelin in plasma were decreased in A53T mice at 3 month age. Because of the potential therapeutic application of ghrelin, we administered low and continuous doses of ghrelin to sustain its concentration using spontaneous Alzet mini-osmotic pumps before the onset of plasma ghrelin decrease, and then detected whether recovered ghrelin could prevent PD progression. The ghrelin treatment were divided into two types: one was administrating for 4 weeks, and another was administrating for 8 weeks.
Finally, we observed the oxidant, apoptosis and in ammatory levels after ghrelin administration to evaluate ghrelin's neuroprotective effects. Our results revealed that subcutaneous administration of lowdose ghrelin could prevent the progression of PD and provided a possible method of ghrelin application for PD treatment. 83Vle/J) were originally acquired from the Jackson Laboratory (004479) in breeding pairs to generate a stable breeding colony. Thirty homozygous A53T mice and thirty wild-type littermates aged 4 weeks were raised on a 12-hr light-dark cycle at room temperature (22 ± 2℃) with natural food and water. Genotyping was conducted according to the protocol used in our previous studies. Mice were randomly assigned to four groups for treatment: (1) P-WT group: wild-type mice with administered non-sense peptide (a misorder ghrelin peptide; amino acid sequences: G-L-S-F-E-H-Q-S-P-Q-Q-R-A-K-E-K-K-S-P-K-L-P-A-Q-P-R-K); (2) P-A53T group: homozygous A53T mice with administered non-sense peptide; (3) G-WT group: wildtype mice with administered ghrelin (Sigma, USA) (G-S-S-F-L-S-P-E-H-Q-K-A-Q-Q-R-K-E-S-K-K-P-P-A-K-L-Q-P-R), which contained acylated ghrelin as physiological level; (4) G-A53T group: homozygous A53T mice with administered ghrelin. Ghrelin and non-sense peptide were administered subcutaneously in saline via Alzet mini-osmotic pumps (Alzet, USA). There were three time points: (1) mice that were administered peptide starting at the age of 1 month, lasting for 4 weeks, and that were harvested at the age of 3 months (3 m (4 w); Fig. 1); (2) mice that were administered peptide starting at the age of 1 month, lasting for 8 weeks, and that were harvested at the age of 3 months (3 m (8 w), Fig. 1); (3) mice that were administered peptide starting at the age of 1 month, lasting for 8 weeks, and that were harvested at the age of 6 months (6 m (8 w) Fig. 1).
All mice were fasted for approximately 8 hrs before treatment but had free access to water. At the end of treatment, mice were put under 45 mg/kg sodium pentobarbitone anaesthesia, and blood plasma was aliquoted into polypropylene tubes and stored at -80 °C for enzyme-linked immunosorbent assay (ELISA) without undergoing any additional freeze-thaw cycles. Brains were isolated from mice and divided into two hemispheres. The right hemisphere SN was used to assess some protein levels by western blot, and the Str was isolated for high-performance liquid chromatography electrochemical detection (HPLC-ECD); the remaining hemisphere was used for immuno uorescence (IF) staining.

ELISA for plasma total and active ghrelin levels
The plasma sample was acidi ed with HCl (0.05 mol/L nal concentration). Total and active ghrelin levels in plasma were measured by corresponding commercially available kits; mouse ghrelin (total) or mouse ghrelin (active) enzyme-linked immunosorbent assay kits (Millipore, USA) were applied according to the provided instructions.

HPLC-ECD for DA content
Str sample was homogenized in 300 µL liquid A (0.4 M perchloric acid). And the details for HPLC-ECD was described as our previous study [5,9].

IF staining for TH or Iba1-positive cells
The fresh left SN was post xed in 4% paraformaldehyde (PFA) over 2 days at 4 °C, followed by dehydration in 20%, and then 30%, sucrose at 4 °C for approximately 2 days. The tissue was sectioned into 20-µm-thick slices. The sections were pre-treated with 0.1% Triton X-100 and blocked with 10% goat serum for 1 hr, and were incubated with rabbit monoclonal anti-TH (Millipore, USA) and anti-Iba1 (Wako, JPN) overnight at 4 °C. Secondary antibodies of Alexa Fluro®-555 conjugate H + L (Invitrogen, USA) were incubated for 1 hr, and images were taken with microscopy (Zeiss, Germany).

Western blot for protein levels
Total protein was extracted from the right SN with lysis buffer (RIPA (strong), Beyotime, CN) and a protease inhibitor cocktail (Roche, USA). Subsequently, protein was subjected to 10% SDS-PAGE gels and transferred onto PVDF membranes (Invitrogen, USA). After blocking with 10% skimmed milk for 1 hr at room temperature, the PVDF membranes were incubated with primary antibodies, including rabbit monoclonal anti-β-actin (Santa Cruz, USA), anti-TH (Millipore, USA), anti-Bcl-2, anti-Bax, anti-SOD1 and anti-interleukin-6 (IL-6) (all from CST, USA), respectively, overnight at 4 °C. The membranes were further probed with HRP-conjugated secondary antibodies (Santa Cruz, USA). Last, the membranes were exposed to ECL (Thermo Fisher Scienti c, USA) for 1 min and visualized using a Tanon image system. 2.6 Bio-PlexPro™ mouse cytokine assays Total protein was extracted from the SN with lysis buffer (RIPA (weak and without SDS), Beyotime, CN).
Concentrations of TNFα, IL-6, and IL-10 were measured by Bio-Plex Pro™ mouse cytokine assays (Bio-Rad, USA) according to the provided instructions.

Statistical analysis
All data are shown as the mean ± SEM. Two-way analysis of variance (ANOVA) followed by the Student-Newman-Keuls test was used to detect differences between two groups using SPSS 19.0 (SPSS Inc, Chicago, IL, USA); P < 0.05 was considered statistically signi cant.

Ghrelin recovered the decrease of plasma ghrelin levels and weight loss in homozygous A53T mice
A53T homozygous mice exhibited decreased plasma total and active ghrelin levels in the P-A53T group both at 3 months and 6 months of age ( Fig. 2A-B). After 4 weeks of ghrelin administration, there was no signi cant difference in plasma total ghrelin levels in 3 month-old G-A53T mice (4.82 ± 1.49 ng/mL) ( Fig. 2A). When the duration was doubled, the plasma total ghrelin levels of the G-A53T group at 3 months of age increased by 20% ( Fig. 2A); however, at 6 months of age, no signi cant increase in ghrelin levels was found in the G-A53T group. Unexpectedly, the plasma active ghrelin levels of the G-A53T group at 3 m (4 w), 3 m (8 w) and 6 m (8 w) showed increases of 26%, 33% and 14%, respectively, compared with levels in the control group (Fig. 2B). Conversely, plasma active ghrelin levels in the P-A53T group decreased by 38% (3 m (4 w)), 27% (3 m (8 w)) and 26% (6 m (8 w)) ( Fig. 2B). No signi cant differences were detected in WT mice in either plasma total or active ghrelin levels at any of these time points (Fig. 2). These results indicate that ghrelin administration could recover decreased plasma ghrelin in A53T homozygous mice.
However, at the beginning of this study, we were afraid that obesity would follow the recovery of ghrelin levels because ghrelin is considered to play an important role in food intake and metabolism, but beyond our expectations, the body weight of the mice in the A53T homozygous group decreased at the 3-month and 6-month time points (Fig. 2C). Meanwhile, body weight in G-A53T group mice increased compared with the control group (Fig. 2C). Furthermore, WT mice did not show signi cant changes at either of these time points (Fig. 2C). Our results suggested that ghrelin administration for 8 weeks could reverse weight loss in homozygous A53T mice.
3.2 Ghrelin restricted both the progressive loss of dopaminergic neurons in the SN and the depletion of DA in the Str of homozygous A53T mice After 4 weeks of ghrelin administration, no signi cant changes in the number of TH-positive neurons or TH protein levels in the SN were detected in the G-A53T group at the age of 3 months ( Fig. 3A-C).
Interestingly, DA content (11.71 ± 1.37 ng/mg) in the Str of G-A53T mice at the age of 3 months was signi cantly increased (Fig. 3E). The A53T homozygous mice showed a 12% decrease in the number of TH-positive neurons in the SN (Fig. 3A-B). Correspondingly, TH protein levels in the SN were decreased, and DA content in the Str of P-A53T group mice decreased from 11.56 ± 2.46 ng/mg to 5.57 ± 2.48 ng/mg, with a 51% decrease at the age of 3 months (Fig. 3C-E).
To verify whether this kind of ghrelin administration is neuroprotective in A53T homozygous mice, we doubled the duration of ghrelin administration. After treatment with ghrelin for 8 weeks, the number of TH-positive neurons in the SN of G-A53T group mice increased at 3 months and 6 months by 13% and 15%, respectively ( Fig. 3F-G). Correspondingly, TH protein levels in the SN and DA content in the Str of G-A53T group mice at the age of 3 months and 6 months both increased (Fig. 3H). The number of THpositive neurons in the SN of P-A53T group mice at the ages of 3 months and 6 months decreased by 14% and 16%, respectively (

Ghrelin inhibited in ammation in the SN of homozygous A53T mice
After treatment with ghrelin for 4 weeks, no signi cant differences were observed in the number of microglia ( Fig. 4A-B) and the protein level of IL-6 ( Fig. 4C-D) in the SN of G-A53T group mice at the age of 3 months. When the duration of ghrelin administration was doubled, the number of microglia in the SN of G-A53T group mice decreased by 22% and 45% at the ages of 3 months and 6 months, respectively ( Fig. 4A-B). Correspondingly, IL-6 protein levels in the SN of G-A53T group mice were decreased at the age of 3 months and 6 months, as well (Fig. 4C-D). In addition, A53T homozygous mice displayed more activated microglia and increased release of IL-6 at all of these time points. Furthermore, WT mice did not show changes at any of these time points.
Furthermore, we used Bio-Plex Pro™ mouse cytokine assays for mice administered ghrelin for 8 weeks. As expected, the concentration of IL-6 signi cantly increased, whereas the concentration of IL-10 decreased (Fig. 4E-F). There was no change in the concentration of TNFα in the SN of homozygous A53T mice at the age of 3 months. In addition, no difference was found between A53T homozygous and WT mice at the age of 6 months. Similarly, ghrelin administration reversed the changes in IL-6 and IL-10 levels at the age of 3 months. These ndings demonstrated that ghrelin administration has potent anti-in ammatory properties.

Ghrelin boosted decreased SOD1 levels and the Bcl-2/Bax ratio in the SN of homozygous A53T mice
No signi cant differences were observed in the protein levels of SOD1 in the SN among the four groups at the ages of 3 m (4 w) and 3 m (8 w). After ghrelin administration for 8 weeks, the protein levels of SOD1 in the G-A53T group mice increased at the age of 6 months (Fig. 5A-B). In addition, the SOD1 protein levels of A53T homozygous mice decreased at the age of 6 months (Fig. 5A-B). Nevertheless, there was no signi cant change in WT mice (Fig. 5A-B). These results illustrated that ghrelin administration boosts decreased levels of SOD1 in the SN of homozygous A53T mice.
No signi cant differences were observed in the Bcl-2/Bax ratio in the SN among the four groups at the ages of 3 m (4 w) and 3 m (8 w). After treatment for 8 weeks, the Bcl-2/Bax ratio in G-A53T group mice increased at the age of 6 months (Fig. 5C-D). A53T homozygous mice showed a signi cant decrease in the Bcl-2/Bax ratio at the age of 6 months (Fig. 5C-D). Furthermore, no signi cant difference was observed in WT mice (Fig. 5C-D). These results illustrated that ghrelin administration promotes a decreased Bcl-2/Bax ratio in the SN of homozygous A53T mice.

Discussion
In this study, we found that subcutaneous administration of low-dose ghrelin for 8 weeks provided a neuroprotective effect in a mouse model of PD. These protective effects may stem from three major functions of this peptide. First, ghrelin enhanced energy intake that induced the increase of body weight. Second, ghrelin blocked microglial activation and inhibited the release of IL-6, both of which can mitigate neuro-in ammation and promote neuronal survivability. Last, ghrelin elevated the activity of SOD1 and increased the Bcl-2/Bax ratio, which prevented neurons from oxidative damage and inhibited apoptosis during the course of the progressive degeneration of dopaminergic neurons, respectively. PD patients exhibit non-motor symptoms, including hyposmia, sleep disturbances and gastrointestinal dysfunction, 10-20 years before they experience motor impairment [10]. Indeed, approximately 98.6% of PD patients suffered at least one non-motor symptom [11]. Moreover, PD patients usually have a lower body weight [12]. Ghrelin, the "hunger hormone", functions as a neuropeptide in the central nervous system [13]. In addition to regulating appetite, ghrelin increases body weight and fat mass by triggering receptors in the arcuate nucleus [3,14,15]. In PD patients, plasma total and active plasma ghrelin levels were decreased, and postprandial ghrelin suppression and preprandial peak responses were both attenuated [7]. In this study, we also found homozygous A53T mice plasma ghrelin were decreased at 3 months age when non-motor symptom exhibited. Injections of ghrelin in both humans and rodents have been shown to increase plasma ghrelin level, and thus increase food intake and motivation to seek out food as well [16]. Here, we showed that the plasma total and active ghrelin levels were both recovered at the age of 3 months after treatment with ghrelin for 8 weeks. Ghrelin injections increased food intake in a dose-dependent manner, and ghrelin could only stimulate food intake when it was elevated much higher than normal levels. In our study, we rst reported that nearly normal ghrelin levels rescued weight loss in homozygous A53T mice 3-6 months old. However, the plasma ghrelin level was lower in obese individuals than leaner individuals [16], suggesting that ghrelin did not contribute to obesity. In line with this research, ghrelin did not increase the body weight of normal mice, and there was "ghrelin resistance" in mice with adiposis [17]. Therefore, it could explain why normal physiological ghrelin dose treatment had no effect on WT mice body weight. Report has found that ghrelin increased the content of DA in the SN through its receptor [18]. Even though all ghrelin administrations were began at the age of 1 month, the administration process with 8 weeks rather than 4weeks, could recover the decreased plasma ghrelin and restrict both the progressive loss of dopaminergic neurons in the SN and the depletion of DA in the Str. The administration time was refer to Therefore, our results indicated that early intervention of ghrelin may delay the progression of PD or even prevent the onset of PD, and also suggested that the lasting time of drug delivery was an important factor for ghrelin clinical application.
Accumulating evidence has suggested a strong link between neurodegeneration and chronic in ammation resulting in the activation of microglia and astrocytes and the increased release of proin ammatory cytokines [19]. Elevated pro-in ammatory cytokines were observed in the brain, cerebrospinal uid and serum of PD patients [20][21][22][23]. Lipopolysaccharides (LPS) or viral pathogens could also induce PD-like neuronal death in rodents [24][25][26]. This study found increased activation of microglia and increased release of IL-6 in the SN of homozygous A53T mice. Ghrelin markedly inhibited oligodendrocyte cell death in oligodendrocytes and LPS-stimulated BV-2 cell co-culture systems, which imitated microglial activation in vitro [27]. Ghrelin was reported to reduce microglial activation in an MPTP-induced PD model, inhibiting the increased expression of TNFα and IL-1β mRNA and inducible nitric oxide synthase in the SN [28]. Nevertheless, the anti-in ammatory effects of ghrelin may originate from the attenuation of the release of in ammatory cytokines, such as IL-6, TNFα, IL-1β and cyclooxygenase-2 [29][30][31]. However, no signi cant difference was found in the concentration of TNFα between the homozygous A53T and WT mice. In addition, in vitro and in vivo studies have revealed the neuroprotective activity of IL-10 on dopaminergic neurons when exposed to LPS and in a 6-OHDA PD model [32][33][34]. This nding is in accordance with results of the inhibition of microglia, decreased proin ammatory cytokine IL-6 and increased anti-in ammatory cytokine IL-10 in the SN of G-A53T mice, suggesting anti-in ammatory and neuroprotective roles of ghrelin on dopaminergic neurons.
Increased levels of oxidation and its products, including nucleic acids, proteins and lipids, were detected in post-mortem PD brains [35,36], suggesting that oxidative stress plays a vital role in the loss of dopaminergic neurons. Superoxide dismutase Cu-Zn, also known as SOD1, can bind copper and zinc ions and act as a homodimer to destroy naturally occurring, but harmful, superoxide radicals in the body [37]. The upregulation of malonaldehyde and the downregulation of SOD1 and catalase have been found in MPTP/MPP + -induced PD models [38,39]. Correspondingly, this study found that homozygous A53T mice exhibited lower SOD1 protein levels in the SN at the age of 6 months. Nevertheless, exogenous ghrelin could antagonize MPTP-induced oxidative stress by reversing the reduced SOD1 [39]. This study observed that ghrelin administration for 8 weeks boosts the decreased levels of SOD1 in the SN of homozygous A53T mice, indicating that ghrelin may exert a neuroprotective effect on dopaminergic neurons against oxidative stress. Additionally, ghrelin pre-treatment could reduce ROS generation and oxidative stress by increasing the activity of uncoupling protein 2 (an important mitochondrial protein in control of ROS production that acts as a sensor for mitochondrial oxidative stress), resulting in decreased dopaminergic neuron loss in the MPTP-induced PD model [18].
The present study found that homozygous A53T mice exhibited a lower Bcl-2/Bax ratio. Emerging evidence indicates that ghrelin has anti-apoptotic effects in multi-pathophysiological conditions, including ischaemia, in ammation, and nutrient deprivation [40][41][42]. Recently, it has been reported that the anti-apoptotic effects of ghrelin in neurons occur via multiple signalling pathways, including the PI3K/Akt, ERK1/2, GSK-3β, JNK and p38 pathways [4]. Report have shown that ghrelin could increase Bcl-2/Bax ratio, prevent of cytochrome c release and inhibit activation of caspase-3, thus promoting the survival of cortical neuronal cells [19]. Moreover, exogenous ghrelin also inhibited the activation of caspase-3, c-Jun and p-38 in oxygen-glucose deprivation-induced apoptosis [19,43]. Our previous study demonstrated that ghrelin could also attenuate Bax expression and caspase-3 activation to promote the survival of dopaminergic neurons in a MPTP-induced PD mouse model [5]. Additionally, ghrelin administration for 8 weeks promoted a decreased Bcl-2/Bax ratio in the SN of homozygous A53T mice. These ndings suggest that ghrelin exerts its neuroprotective effect on dopaminergic neurons by antiapoptosis processes.
In summary, with the progression of disease in homozygous A53T mice, the plasma ghrelin levels decreased, and continuous ghrelin administration could restore the decreased plasma ghrelin to normal levels. Continuous dose of ghrelin did not lead to obesity in WT mice. With the recovery of the plasma ghrelin, the degeneration of dopaminergic neurons in homozygous A53T mice was improved signi cantly in terms of increased numbers of TH-positive cells and TH protein levels in the SN. Treatment with ghrelin for 4 weeks had no neuroprotective effects on dopaminergic neurons at the age of 3 months; however, treatment with ghrelin for 8 weeks had a neuroprotective effect on dopaminergic neurons in the SN at the age of 3 months, and this effect lasted through 6 months of age. This early intervention with ghrelin may delay the progression of disease in homozygous A53T mice by inhibiting apoptosis decreasing oxidative stress and in ammation. Therefore, ghrelin can improve the progression of neurodegenerative diseases, which will have important potential clinical value for preventing the onset of PD, especially familial PD.

Conclusions
In the present study, we found endogenous brain-gut peptide "ghrelin" could recover the decrease of plasma ghrelin levels and weight loss of PD mice. Ghrelin administration could restrict the degeneration of dopaminergic neurons in the SNpc and prevent the onset of PD. Anti-antioxidant, anti-in ammatory and anti-apoptosis effects of ghrelin, which might make ghrelin as a new possible peptide for preventing PD.

Consent to publication
Informed consent for publication was obtained from all participants.

Availability of data and material
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
There is no con ict of interest or commercial relationship and/or support from pharmaceutical or other companies.