Therapeutic effect of hydrogen sulfide in a parkinsons disease model

ACS84 reduced the oxidative stress induced by 6-OHDA .... ACS84 significantly reversed 6-OHDA-induced oxidative stress in SH-SY5Y cells.. Therapeutic effect of ACS84 on 6-OHDA-induced Pa

THERAPEUTIC EFFECT OF AN HYDROGEN

SULFIDE-RELEASING COMPOUND IN A PARKINSON’S

DISEASE MODEL

XIE LI

(B.SC) FUDAN UNIVERSITY

DEPARTMENT OF PHARMACOLOGY

YONG LOO LIN SCHOOL OF MEDICINE

NATIONAL UNIVERSITY OF SINGAPORE

2012

Acknowledgements

I would like to express my heartfelt gratitude to my supervisor, Prof Bian Jin-Song, for giving me the opportunity to work on this research project I would like to thank him for his generous instruction and support, in both

my research and my life

I am also grateful to my seniors, Dr Hu Li Fang, Dr Lu Ming, Ms Tiong Chi Xin and all other laboratory members for their encouragement, technical help and critical comments I would like to thank Shoon Mei Leng for her technical support With the presence of these adorable colleagues, my experiences in research for the past three years have been enjoyable I would also like to thank my family and friends for their constant support and encouragement

Table of Contents

Acknowledgements   i 

Publications   iv 

Summary   v 

List of Tables   vi 

List of Figures   vii 

Abbreviations   viii 

1.  Introduction   1 

1.1.  Parkinson’s disease   1 

1.1.1.  Genetics   1 

1.1.2.  The pathogenesis of Parkinson’s disease   4 

1.1.3.  Treatments of Parkinson’s disease   7 

1.1.4.  Experimental models for Parkinson’s disease   10 

1.2.  Hydrogen sulfide (H2S)   13 

1.2.1.  Endogenous production of H 2 S   13 

1.2.2.  The physiological role of H2S in CNS   14 

1.2.3.  H 2 S-releasing compound   17 

1.3.  Research objectives   19 

2.  Materials and Methods   20 

2.1.  Chemicals and reagents   20 

2.2.  Cell culture and treatment   20 

2.3.  Cell viability assay   20 

2.4.  Lactate dehydrogenase (LDH) release assay   21 

2.5.  Reactive oxygen species (ROS) measurement   21 

2.6.  Superoxide Dismutase (SOD) activity Determination   21 

2.7.  Reverse Transcription-PCR   22 

2.8.  Western blot   23 

2.9.  Nuclear and cytoplasmic protein fractionation   23 

2.10.  6-OHDA induced PD rat model   24 

2.11.  Behavioural test   24 

2.12.  Immunohistofluorescence staining   25 

2.13.  Lipid peroxidation assessment   25 

2.14.  Concentration determination of dopamine and its metabolites   26 

2.15.  Statistical analysis   26 

3.  Results   27 

3.1.  Protective effect of ACS84 on 6-OHDA-induced cell injury   27 

3.2.  ACS84 reduced the oxidative stress induced by 6-OHDA   27 

3.3.  ACS84 promoted anti-oxidative stress associated gene expression   29 

3.4.  ACS84 ameliorated behaviour symptom in the unilateral 6-OHDA rat model   32 

3.5.  ACS84 attenuated the degeneration of dopaminergic neurons in both SN and striatum   33 

3.6.  ACS84 reversed the declined dopamine level in the 6-OHDA-injured striatum   33 

3.7.  ACS84 suppressed the oxidative stress in the injured striatum   35 

4.  Discussion   36 

4.1.  ACS84 significantly reversed 6-OHDA-induced oxidative stress in SH-SY5Y cells   ……….36 

4.2.  ACS84 suppressed pathological progresses and improved symptoms in unilateral 6-OHDA rat models   37 

4.3.  Limitations of the study and future directions   38 

4.4.  Conclusion   40 

References   42 

Publications

Xie L, Tiong CX, Bian JS Hydrogen sulfide protects SH-SY5Y cells

against 6-hydroxydopamine-induced endoplasmic reticulum (ER) stress

Am J Physiol Cell Physiol 2012 (In Press)

Xie L, Hu LF, Tiong CX, Sparatore A, Del Soldato P, Dawe GS, Bian JS

Therapeutic effect of ACS84 on 6-OHDA-induced Parkinson’s disease rat model (Ready for submission)

Summary

Parkinson’s disease (PD), characterized by loss of dopaminergic neurons in the substantia nigra, is a neurodegenerative disorder of the central nervous system The present study was designed to investigate the effect of ACS84, an H2S-releasing L-Dopa derivate, in a 6-hydroxydopamine (6-OHDA)-induced PD model ACS84 protected SH-SY5Y cells against 6-OHDA-induced cell injury and oxidative stress The protective effect resulted from the stimulation of Nrf-2 nuclear translocation and the promotion of anti-oxidant enzymes expression In the 6-OHDA-induced PD model, intragastric administration of ACS84 relieved the movement dysfunction of the model rats Immunohistochemistry and HPLC analysis showed that ACS84 reversed the loss of tyrosine-hydroxylase positive (TH+) neurons in the substantia nigra and striatum, and the decline of dopamine concentration in the injured striatums

of the 6-OHDA-induced PD model Moreover, ACS84 reversed the elevated malondialdehyde level in model animals

In conclusion, ACS84 may prevent neurodegeneration via the anti-oxidative mechanisms and has potential therapeutic values for Parkinson’s disease

Fig 3.5 Effect of ACS84 on 6-OHDA-induced TH+ neuronal degeneration……… 34

Fig 3.6 Effect of ACS84 on oxidative stress in the striatum of unilateral 6-OHDA-lesioned

PD rat model ……… 35

CMA Chaperon-medicated autophagy

CNS Central nervous system

DAT Dopamine active transporter

DBS Deep brain stimulation

DOPAC 3,4-Dihydroxyphenylacetic acid

ER Endoplasmic reticulum

FBP-1 Far upstreamelement-binding protein 1

Gclc Glutamate-cysteine ligase catalytic subunit

GclM Glutamate-cysteine ligase modulatory subunit

GPi globuspallidusinterna

GSH Glutathione

GWAS Genome wide association study

H 2 S Hydrogen sulfide

HPRT Hypoxanthine-guanine phosphoribosyltransferase

HVA Homovanillic acid

iNOS Inducible form of nitric oxide synthase

K ATP ATP-sensitive potassium channel

NaHS Sodium hydrosulfide

NET Norepinephrine transporter

PPARγ peroxisomeproliferator-activated receptor gamma

ROS Reactive oxygen species

SN Substantia nigra

SOD Superoxide Dismutase

SQR Sulfidequinonereductase

STN Subthalamic nucleus

UPR Unfolded protein response

UPS Ubiquitin-proteasome system

1 Introduction

1.1 Parkinson’s disease

Parkinson’s disease (PD) is an age-related progressive degenerative movement disorder, which was firstly described by James Parkinson in 1817 [1] As the second most common neurodegenerative disease, PD affects nearly 1% of the population aged above 65 years [2-5] PD patients suffer from symptoms such as bradykinesia, resting tremor, rigidity, and postural instability, which is associated with the loss of dopaminergic neurons and the decrease of dopamine (DA) in the substantia nigra (SN) [6] One hallmark of PD pathology is the presence of Lewy Bodies (LBs) in the dopaminergic neurons, which is the inclusion of misfolding proteins [7]

1.1.1 Genetics

Although most PD cases are sporadic and largely influenced by environmental factors,

PD has already been recognised as a disorder with a significant genetic component [6, 8-11] As listed in Table 1.1, more than ten loci have been identified associated with different types of PD and parkinsonism

Table 1.1

Loci associated with PD

Locus Mode of inheritance Chromosomal

location

Gene Reference

PARK1 (4) Autosomal dominant 4q21–q23 SNCA [12, 13] PARK2 Autosomal recessive 6q25.2–q27 parkin [14]

PARK6 Autosomal recessive 1p35–p36 PINK1 [17]

PARK8 Autosomal dominant 12p11.2–q13.1 LRRK2 [19]

Besides that, recently Genome-wide Association Studies (GWAS) and analysis also provided a huge amount of information indicating the suspicious loci associated with PD [22-29] All these investigations contributed greatly to the understanding of molecular mechanisms of PD pathogenesis Here we will discuss the roles of several genes as listed above

meta-α-Synuclein

α-Synuclein is encoded by gene SNCA, which is the first gene found to be linked to

PD Three mutations (A53T [12], A30P [30], E46K [31]) and genome triplication of

SNCA[13] have been identified in familial PD patients The physiological role of

α-synuclein remains unknown, though it is highly expressed in the brain α-Synuclein is mainly located in the presynaptic terminal and is involved in the maintenance of membrane structures [32, 33] Some scientists speculated that α-synuclein might be involved in the DA neurotransmission and synaptic vesicle recycling [34] α-Synuclein is the main component of the LBs The mutations and over-expressions of α-synuclein are believed to promote the formation of LBs It has been shown that compared to wild-type α-synuclein, A53T and A30P mutants exhibit increased

propensity to form oligomers and fibrils in vitro [35] Moreover, the A30P mutant

expressed in transgenic mice or flies indicated inclusions formation as well as

neurodegeneration [36, 37] However, the mechanism of wild-type α-synuclein accumulation in LBs inclusion is less elucidated It was speculated to be associated with the mitochondria complex-I malfunction [38-41], tyrosine nitration [42] and the impairment of proteasome function [43, 44] It is also worth noting that α-synuclein may directly suppress proteasome function in cells Reports had suggested that α-synuclein filaments and oligomers were resistant to proteasome degradation and inhibit proteasome activity by directly binding to 20/26S proteasomal subunits [45-47] Overexpression of mutant α-synuclein was also proved to induce proteasome impairment in cells [48, 49] The impairment of proteasome function induced by α-synuclein may be a crucial pathological process in PD

Parkin and PINK1

Parkin is a ubiquitin E3 ligase, which is responsible for tagging proteins for

proteasome degradation The function of Parkin can be disrupted by parkin mutations

[50, 51] as well as the nitrosative and oxidative stress in sporadic PD [52] The dysfunction of Parkin leads to the accumulation of its substrates, including aminoacyl-tRNA-synthetase-interacting multifunctional protein type 2 (AIMP2) [53, 54], far upstream element-binding protein 1 (FBP-1) [55] and most importantly, PARIS (parkin-interacting substrate) [56] In conditional parkin knock-out mice, PARIS accumulated in the brain and suppressed the expression of peroxisome proliferator-activated receptor gamma (PPARγ) coactivator-1α (PGC-1α), leading to the degeneration of DA neurons [56]

Like Parkin, PINK1 mutations are also associated with familial PD PINK1

is a protein kinase with a mitochondria-targeting domain [57], which was believed to

be involved in mitochondria quality control with Parkin [58] Flies with PINK1 or

parkin deficits suggested the vulnerability of DA neurons to oxidative stress [59, 60]

It has been recognised that PINK1 and Parkin may play crucial roles in the turnover

of damaged mitochondria PINK1 is cleaved during mitochondria depolarization, leading to the recruitment of Parkin and proceeding to mitophagy [61-63]

LRRK2

Leucine-rich repeat kinase 2 (LRRK2) is a serine/threonine kinase with a GTPase modulation domain Mutations on LRRK2 had been isolated from familial PD patients, which would lead to the late-onset of PD symptoms [64] The G2019S mutation is the most common mutation in familial PD, and it is also recognised as a significant risk factor in sporadic PD patients [65] Several pathogenic mutations on LRRK2 promote the formation of dimers and LRRK2 kinase activity is dependent on the dimer formation [66] Evidences had also suggested that the applications of compounds which blocked LRRK2 kinase reversed LRRK2 toxicity in neurons [67] Recent studies indicated that LRRK2 was involved in the modulation of neurite outgrowth in neurons development [68-70], and the regulation of protein translation via protein-microRNA interaction [71]

1.1.2 The pathogenesis of Parkinson’s disease

Although the specific molecular mechanisms for PD are still uncertain, scientists have concluded several theories, including mitochondria dysfunction and oxidative stress, ubiquitin-proteasome system malfunction and autophagy failure, and neuroinflammation to explain the pathogenesis of PD

Mitochondria dysfunction and oxidative stress

Oxidative damage in sporadic PD brains has been observed in post-mortem studies, and the source of oxidative stress might be induced by mitochondria dysfunction and

DA metabolism [72] In order to maintain the oxidation phosphorylation, there is a highly oxidative environment inside mitochondria During mitochondria dysfunction, especially the defects in complex-I, the production of ATP is reduced and the release

of reactive oxygen species (ROS) is elevated in the cells, resulting in oxidative stress

in PD brains [6] This speculation has been supported by the observation that complex-I activity was decreased in the SN of sporadic PD patients [73] The cytoplasmic hybrid cells containing mitochondria DNA (mtDNA) from PD patients, which displayed the deficits of complex-I and increased ROS generation [74, 75], also indicated the role of complex-I deficits in PD pathogenesis Moreover, some neurotoxins like MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) and rotenone, which are the inhibitors of mitochondria complex-I, are used to induce Parkinson mimetic symptoms in animal models [38, 76-78] Another source of ROS generation

in dopaminergic neurons is the metabolism of DA Under physiology condition, DA can be degraded non-enzymatically into quinone by oxygen and enzymatically into 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA) by monoamine oxidase (MAO) and Catechol-O-methyl transferase (COMT), respectively Both ways of degradation would generate H2O2 [79-82]

Ubiquitin-proteasome system malfunction and autophagy failure

Some scientists have also focused their research on the ubiquitin-proteasome system (UPS) and autophagy, which are the main intracellular degradation methods [83-85]

As the existence of LBs is a major clinical hallmark for sporadic PD and some familial PD, it is believed that UPS impairment may be a crucial process in PD pathology [6] Both structural and functional deficits of 20/26S proteasome have been

observed in sporadic PD patients [43, 86] Besides, animals treated with proteasome inhibitors displayed PD-like symptoms, which includes DA neurons degeneration and LB-like inclusion formation [87] Moreover, overexpression of molecular chaperones

by transgenic or pharmacological methods reversed the pathological progresses in

Drosophila models [88, 89], which further indicated the importance of UPS activity in

PD pathogenesis

Autophagy has emerged to be a hot-spot in neurodegenerative diseases research It is the pathway by which cells degrade the long-lived, stable proteins and recycle the organelles [85] Three types of autophagy have been introduced: marcoautophagy, microautophagy and chaperon-medicated autophagy (CMA) [90] Autophagy is believed to be closely related to PD pathogenesis Numerous investigations have demonstrated that α-synuclein could also be cleared by autophagy

in addition to the UPS [91-93] More evidences also presented that mutations on

ATP13A2, which encodes a lysosomal ATPase, led to autophagy failure and

α-synuclein aggregation [20, 94] Moreover, both UPS and autophagy activity are reduced during aging [95-98] Therefore, it is understandable that age is one of the key risk factors in PD

Glial activation and neuroinflammation

The activation of glia cells and the neuroinflammation have been recognised as a keynote contributor in the processes of neurodegeneration [99, 100] Activated microglia cells [101-103] and the increment of astrocyte density [104] have been observed in SN of PD patients in post-mortem studies Alongside with these findings,

it was also reported that the concentrations of cytokines such as TNFα, interleukins 1β,

6, and 2, β2-microglobulin, TGFα and β1, and interferon γ were upregulated in

striatum [105-109] and SN [110] of PD patients Moreover, enzymes which are involved in neuroinflammation, including inducible form of nitric oxide synthase (iNOS), NADPH oxidase, cyclo-oxygenase 2 (COX2), andmyeloperoxidase (MPO), were found to be upregulated in PD patients and PD models [111-114] All these lines

of evidences suggested the crucial role of neuroinflammation in the pathogenesis of

PD Some scientists believed that the release of protein aggregates from neurons [115, 116] or even nitrated extracellular α-synuclein [117] triggered microglia activation during the progress of PD Others suggested the possible influences of environmental factors on neuroinflammation Animals exposed to neurotoxins such as MPTP and rotenone were observed to exhibit glia activation and neuroinflammation [118, 119] Apart from that, although the role of infection in neuroinflammation still remains unclear, injection of Lipopolysaccharide (LPS) intracranially would induce PD-like symptom in rodents [120]

1.1.3 Treatments of Parkinson’s disease

There is no cure for PD so far However, numerous medications had been developed

to supplement the DA deficit and to improve the life qualities of the patients Clinically, there are pharmacologic and surgical treatments being adopted to relieve

PD symptoms

Levodopa (L-Dopa)

L-Dopa is the most widely used treatment for PD since its first development about 30 years ago L-Dopa is able to pass through the blood-brain-barrier (BBB) and is uptaken by dopaminergic neurons to transform into DA by dopa-decarboxylase to compensate for the decline of DA in the brain [121] Administration of L-Dopa efficiently reverses the motor dysfunction in the patients However, only 1-5% of L-

Dopa is distributed to the centre nerves system (CNS), and the rest of the L-Dopa would induce side-effects peripherally In clinical practise, L-Dopa is administrated with carpidopa, which is a BBB impermeable dopa-decarboxylase inhibitor to block the L-Dopa metabolism in peripheral systems

Although L-Dopa is effective in relieving the PD symptoms in patients, chronic treatment with L-Dopa would lead to the suppression of endogenous synthesis

of DA and the disruption of DA system Patients would experience the wear-off effects when the effective period of the drug begins to reduce Half of the patient may even develop dyskinesia after years of medication [122, 123] Moreover, L-Dopa does not arrest the progression of PD and long-term treatment accelerates the neuron degeneration due to oxidative stress [124-127]

Etilevodopa, which is an L-Dopa derivative, has also been developed for PD treatment However, the clinical trial reports suggested that little advantages were observed in patients with motor fluctuations [128]

Dopamine agonist

DA agonist is designed to activate DA receptors, which can be a supplementary treatment for L-Dopa medication and used to treat early PD patients The most commonly prescribed DA agonists are pramipexole, ropinirole and rotigotine Clinical trials have suggested that initial treatment of PD with pramipexole would reduce the incidence of dopaminergic motor complications like dyskinesia compared with L-Dopa [129-131] Rotigotine is also reported to relief symptoms in early PD patients in clinical research [132, 133] However, DA agonists also produce similar side effects compared with L-Dopa, although they might postpone the occurrence of involuntary movements [134, 135]

Monoamine oxidase-B (MAO-B) inhibitor

MAO-B is the main enzyme in dopaminergic neurons which breaks down dopamine Therefore, the inhibition of MAO-B would increase the level of dopamine in the brain Two MAO-B inhibitors had been developed, namely selegiline and rasagiline Numerous clinical researches have revealed that monotherapy of rasagiline or combined with L-Dopa have effectively improved the motor function decline in early

PD patients [136-139] Experimental investigations also indicated that rasagiline protected neurons against injuries via maintenance of mitochondria integrity and induction of neurotropic factors [140] Based on these observations, rasagiline has been recognized as a promising potential therapy for PD, although more information about the safety and further side effects are still required

Catechol-O-methyl transferase (COMT) inhibitor

COMT is also an enzyme involved in the degradation of DA in the dopaminergic neurons The usage of COMT inhibitor is to prolong the effects of L-Dopa The adjunction of entacapone, which is a COMT inhibitor, used in combination with L-Dopa in PD patients with motor fluctuation, although did not significantly reverse the symptoms, but it improved the life quality of the patients [141] However, one adverse effect of COMT inhibitors is that they may enhance the dyskinesia induced by L-Dopa

Deep brain stimulation (DBS)

DBS is a surgical treatment using implanted electrodes to give electrical pulses to specific brain regions In PD patients, DBS would manage PD symptoms and improve patients’ life quality, as well as reverse the side effects of PD medication

Subthalamic nucleus (STN) and globuspallidusinterna (GPi) are two major stimulation site for PD, but other sites like caudal zonaincerta and pallidofugal fibers are also reported to be effective [142] However, it should be noted that DBS would induce psychiatric dysfunction in the patients, although this adverse effect was reported to be reversible [143]

1.1.4 Experimental models for Parkinson’s disease

Animal models would always be the powerful tools to understand the disease mechanisms and to seek the effective potential medications in biomedical research For PD, both non-genetic and genetic models have been established However, none

of those models would be capable to represent the pathogenesis of human PD Here,

we will discuss the advantages and imperfections of those widely used models

6-hydroxydopamine (6-OHDA)

6-OHDA-induced PD model is the most widely used animal models for PD research When injected intracerebrally, 6-OHDA is selectively taken up by dopamine transporter (DAT) and norepinephrine transporter (NET) into the dopaminergic neurons Consequently, 6-OHDA undergoes catalytic processes and releases reactive oxygen species (ROS) which induces cell injury in neurons [144] 6-OHDA-induced

PD model displays similar clinical features of human PD, including dopamine depletion, dopaminergic neuron loss, and neurobehavioral deficits [145] However, the pathological protein aggregations and the deposition of LBs are neglected in this model Moreover, the acute lesion of dopaminergic nerve system in this model might not represent the slow progress of clinical PD pathogenesis In the present study, unilateral 6-OHDA rat model was used to test the anti-oxidative effects of compound

ACS84 The severity of the lesion can be monitored by amphetamine or apomorphine-induced turning behaviour

1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)

MPTP animal model is a widely accepted PD model which mimics a majority of PD features including oxidative stress, mitochondria dysfunction and neuroinflammation MPTP is BBB permeable and it is transformed into active form MPP+ in astrocytes by MAO-B Following that, MPP+ enters neurons through DAT Once inside the neurons, MPP+ blocks mitochondria complex-I activity, leading to the release of ROS and ATP deficiency This animal model would display akinesia and rigidity after MPTP administration, although protein aggregation is rare in this model [146]

Rotenone

Rotenone is a pesticide which inhibits the mitochondria respiration chain in cells Although the effect is not selective, rotenone application exhibits almost all the characteristic of human PD symptoms, especially the aggregation of α-synuclein and the formation of LB-like inclusions [41, 147] Moreover, rotenone also induces microglia activation in animal models [148-150], indicating that rotenone models are capable of mimicking the neuroinflammatory features of PD Interestingly, chronically administration of rotenone suggested a highly selectivity to nigrostriatal neurons [38], while few theories could explain this selective vulnerability Some scientists purposed that rotenone might also inhibit the microtubule stability The microtubule malfunction further disrupted the transport of dopamine vesicles in the dopaminergic neurons, leading to elevation of dopamine oxidation in the cells [151] Genetic models

As α-synuclein is intimately associated with PD pathogenesis, some researchers attempted to establish transgenic mice which express mutant α-synuclein in the brain However, no model has been found to perfectly replicate the clinical and pathological features of PD Only one model, mPrP-A53T mice displayed α-synuclein pathology including α-synuclein aggregation and age-dependent progressive DA neurodegeneration [152, 153], despite that this degeneration was not L-Dopa responsive [154]

LRRK2 mutation is also a major risk factor in late-onset PD However, Bacterial artificial chromosome (BAC) transgenic mice expressing R1441G or G2019S mutants of LRRK2, and conditional knock-in of the R1141C mutation did not exhibit significant dopaminergic neurodegeneration, although all of these models displayed some abnormalities in the nigrostriatal system [155-157]

As discussed above, Parkin and PINK1 are involved in the mitochondria maintenance, and mutations on Parkin and PINK1 would lead to familial PD The

knockouts of parkin or pink1 in Drosophila lead to significant motor deficit and mitochondria dysfunction [60, 158-160] In contrast, parkin or pink1 knockout mice

did not show any substantial dopaminergic or behavioural abnormalities [161-166] However, overexpression of mutant human parkin in mice induced progressive degeneration of DA neurons [167]

Interestingly, disruption of some other genes which are not suggested to associate with PD also leads to PD-like symptoms in mice models For example, the deficiency of transcription factor Pitx3 and conditional knockout of mitochondrial

transcription factor Tfam in dopaminergic neurons in mice produced progressive loss

of dopaminergic neurons and displayed PD-like phenotypes [168, 169] These investigations provided novel insights into the molecular mechanisms of PD

1.2 Hydrogen sulfide (H 2 S)

H2S, which is a flammable, water soluble and colourless gas with an unfavourable odour, was traditionally thought to be a toxic gas but recently is recognized as one of the gas-transmitter followed by NO and CO In the last decade, numerous investigations have focused on the physiological and pathological functions of H2S in the body systems, especially in the centre nervous system and cardiovascular system

1.2.1 Endogenous production of H 2 S

The identification of endogenous H2S was inspired by the detection of sulfide levels

in the brains from rats, humans and bovine [170-172] as well as in blood samples [173] and hearts [174] Although the exact concentration of H2S is quite controversial due to the high variety of measurement methods, there is no doubt that H2S is endogenously produced in many tissues

There are two kinds of enzymes which are responsible for H2S production: pyridoxal-5’-phosphate (PLP)-dependent enzymes including cystathionine-synthase (CBS) and cystathionine-lyase (CSE) [175-178] and a PLP-independent enzyme, called 3-mercaptopyruvate sulfurtransferase (3MST) The main substrates of CBS and CSE are L-cysteine and/or homocysteine [179, 180], while 3MST facilitates the transfer of thiol group from L-cysteine to -ketoglutarate, in combination with cysteine aminotransferase (CAT) [181]

However, CBS is the predominant enzyme for H2S production in CNS, suggested by the results from western and northern blots detecting the protein and

mRNA expression levels in the rat brains [182] Further investigations localized CBS

to astrocytes [183, 184], while 3MST was found to be expressed in neurons [185] The endogenous levels of H2S in CNS are still controversial nowadays Originally, it was reported that H2S levels in brain is around 47-166 µM [170-172, 182, 186, 187] However, with novel methods, this value had been reconsidered to be as low as few nano molars [188, 189] Recently, some scientists suggested that the intracellular half-life of H2S was as short as few seconds [190, 191] They indicated the enzyme, sulfidequinone reductase (SQR), oxidised H2S and transferred the electron to the mitochondria respiration chain [191] However, SQR is absent in neurons, which may suggest a unique role of H2S in neurons

1.2.2 The physiological roles of H 2 S in CNS

Neurophysiology modulation

In 1996, it was first reported that physiological concentration (≤130 µM) of H2S selectively upregulated the N-methyl-D-aspartate (NMDA) receptor-mediated responses and improved the induction of the hippocampal long-term potentiation (LTP), which indicated the potential role of H2S in neuromodulation [182] Further investigation revealed that the enhancing of the NMDA receptor activity was dependent on the H2S-induced increment of cyclic-AMP (cAMP) [192]

Other investigations also indicated that H2S elevated intracellular Ca2+ and induced Ca2+ waves in astrocytes, via mechanisms which modulated neuron functions [193] This observation had been confirmed by an independent investigation which suggested that H2S induced both Ca2+ influx and the release of Ca2+ from intracellular stores, and this effect was cAMP/PKA dependent [194]

Suppression of neuroinflammation

H2S was originally recognized as a proinflammatory modulator in acute pancreatitis, endotoxin-induces global inflammation, and polymicrobial sepsis-associated lung

injury [195-199] However, Hu et al first demonstrated that H2S attenuated neuroinflammation induced by lipopolysaccharide (LPS) in microglia cells [200] Further investigation also indicated that H2S suppressed rotenone- and Aβ-induced inflammation in microglia cells and animal models [201, 202] It was suggested that the anti-inflammation effects of H2S involved the inhibition of p38 mitogen-activated protein kinase [200]

Suppression of oxidative stress

The anti-oxidative stress effects of H2S in neurons were first reported by Kimura’s group They described that H2S improved the activity of γ-glutamylcysteine synthetase and elevated cystine transport to boost the glutathione levels in primary cultured neurons [203] Furthermore, they also suggested that in supplement to up-regulating glutathione levels, H2S also activated ATP-dependent K+ (KATP) and Cl- channels [204] Our group also demonstrated the anti-oxidative effects of H2S in cellular models In neuroblastoma cell line SH-SY5Y cells, H2S protected cells against cell injuries-induced by neurotoxins including rotenone and 6-OHDA via increasing mitochondria stability and upregulating PKC/Akt pathways [205, 206] Works on glia cells suggested that H2S rescued astrocytes from H2O2-induced injury

by enhancing glutamate uptaking [207] More direct evidence was obtained from animal models that H2S ameliorated symptoms in 6-OHDA unilateral rat model and rotenone treated rat model The mechanism involved was concluded to be the suppression of NADPH oxidase activity and oxygen consumption in the neurons by

H2S [201] Other observations showed that H2S was also capable of reversing MPP+induced apoptosis in PC12 cells by maintaining mitochondrial membrane potential

-and reducing intracellular ROS generation [208] However, although the suppression

of oxidative stress by H2S was confirmed by independent researches, the exact mechanism of this phenomenon was still unclear In a recent animal-based investigation, the scientists suggested that the protective effects of H2S was not associated with KATP channels but involved uncoupling protein 2 in mice administrated with MPTP [209], which was contradicting to the previous understanding of H2S functions Therefore, investigation to identify the exact action site of H2S in the neurons is still worthwhile

Suppression of endoplasmic reticulum (ER) stress

ER is an organelle whereby the secreting proteins or membrane proteins are synthesized, folded, modified and transported Stress conditions such as oxidative stress, nutrition deprivation, aberrant Ca2+ regulation, and viral infection, would lead

to the disturbance of protein processing in ER, and induced the unfolded protein response (UPR) [210] Overwhelming and persisting ER stress would induce apoptosis, and the existence of ER stress had been identified in PD models [211]

The role of H2S in ER stress is quite controversial recently It was reported that H2S suppressed the cardiomyocytic ER stress induced by hyperhomocysteinemia (HHcy), thapsigargin or tunicamycin [212] However, in INS-1E cell, which is an insulin-secreting beta cell line, upregulation of H2S induced ER stress and stimulated apoptosis [213] In CNS, evidences support the protective effects of H2S against ER stress Our group has reported that H2S relieved 6-OHDA-induced ER stress in SH-SY5Y cells via upregulation of Hsp90 expression [214] Similar results have also been obtained in animal models as H2S treatment significantly reduced the expression

of UPR related proteins in MPTP mice [209] Moreover, H2S has been shown to

sulfhydrate phosphatase PTP1B, and therefore suppressed the ER stress processes [215]

1.2.3 H 2 S-releasing compound

Although H2S has been proven to be a potential therapeutic agent in PD treatment, the challenges remain in clinical practises as how to give the H2S treatments accurately and safely Some scientists started to seek compounds which would release H2S in the body

One group had recognized morpholin-4-ium 4 methoxyphenyl (morpholino) phosphinodithioate (GYY4137) as a H2S-slow releasing compound Despite its slow-releasing features, GYY4137 achieved almost every physiological characteristics compared with sodium hydrosulfide (NaHS), including smooth muscle relaxation and blood pressure reduction [216] Moreover, GYY4137 relieved LPS-induced inflammation responses in macrophage RAW264.7 and in rats [217, 218], and these effects were comparable with low-dose of NaHS treatment [218] These researchers also believed that the slow-releasing pattern of GYY4137 made GYY4137 a better representative model for H2S investigations [217]

Another candidate of H2S-releasing compound, anetholedithiolethione (ADT), has also drawn attentions recently ADT is a compound with a unique thiol group and is originally developed as a choleretic and sialogogue [219] Since the last two decades, it had emerged that ADT was an effective anti-oxidant, suppressing oxidative damage in astrocytes [220], Jurkat T cells [221], and endothelial cells[222] Evidence has also revealed that ADT significantly suppressed the MAO-B activity but not MAO A in astrocytes [223] Recently, some researches started to consider that the

effects of ADT come from the H2S released from its thiol group, and they combined ADT with other widely-accepted drugs to enhance the therapeutic effects For instances, the ADT-diclofenac hybrid (ACS 15), suggested anti-inflammatory effects, protected hearts against ischemia-reperfusion injury, suppressed vascular smooth muscle cell proliferation, as well as inhibited breast cancer-induced osteoclastogenesis and preventedosteolysis [224-227] ACS14, the hybrid of ADT and aspirin, has also been found to modulate thiol homeostasis in cells, protected the heart from ischemia/reperfusion, suppressed microglia activation and neuroinflammation, and suppressed breast cancer cellsproliferation [228-232] Overall, all of these observations indicated the potential application of ADT as an H2S-releasing agent for disease treatments

It was speculated that the combination of L-Dopa and H2S may have potential therapeutic value [233, 234] ACS84, which is also a family member among ACS14 and 15, is a hybrid compound derived from L-Dopa and ADT, and is permeable to BBB and release H2S in cells [233] Although the effect of ACS84 on PD is not known yet, ACS84 and other H2S-releasing L-Dopa derivatives have been proven to suppress neuroinflammation and inflammation-induced cell injury, and elevate glutathione (GSH) level while inhibit MAO B activity [233] Further investigation also suggested that ACS84 protected cells against Aβ-induced cell injury via attenuation of inflammation and preservation of mitochondrial function, a similar effect like ACS14 [235] As discussed previously, L-Dopa treatment in PD would enhance the oxidative stress and lead to the worsening of dopaminergic neuron loss It

is worthwhile to investigate whether the combination of L-Dopa and H2S ameliorates the oxidative stress and improve the therapeutic effect of L-Dopa

1.3 Research objectives

L-Dopa is still the most widely used treatment for PD since 1970s Although L-Dopa efficiently compensates for the dopamine deficit in the brain, it fails to reverse the disease progresses Long-term treatment of L-Dopa would enhance the oxidative stress and promote the neurodegeneration

H2S, which is a powerful anti-oxidant and neuroprotector, has been proven to reverse the cell injuries and disease progresses in PD cell and animal models Evidences also suggest that H2S-releasing compound, including ADT and ACS84, may also protect the cells against neural damages

ACS84 is a hybrid compound derivate from L-Dopa and ADT moiety Reports have indicated that it suppressed glia cells activation and relieved inflammation-induced cell injury However, the anti-oxidative stress effects of ACS84 have not been investigated Therefore, in this present study, we will examine the suppression of oxidative stress in 6-OHDA cell model and the amelioration of disease progress in unilateral 6-OHDA rat model

Fig 1.1 Chemical structure of ACS84 ACS84 is a hybrid of L-Dopa (left part) and ADT (right

part) The dithiol thione group on ADT moiety is believed to release H 2 S in cells

2 Materials and Methods

2.1 Chemicals and reagents

All chemicals, antibodies for detecting tyrosine hydroxylase (TH) and LDH assay kit were purchased from Sigma (Sigma, St Louis, MO) Antibodies for detecting Nrf-2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) The Glutathione Assay Kit, TBARS Assay Kit and Superoxide Dismutase Assay Kit were purchased from Cayman Chemical (Ann Arbor, Michigan) ACS84 was prepared as previously described [233]

2.2 Cell culture and treatment

The human neuroblastoma cell line, SH-SY5Y, was obtained from the American Type Culture Collection (Manassas, VA, USA) Cells were maintained in Dulbecco’s modified Eagle’s Medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and 0.05 U•mL-1 penicillin and 0.05 mg•mL-1 streptomycin at 37°C in a humidified atmosphere containing 5% CO2/95% air Cells were plated onto 96-well plates for viability tests and ROS generation assay, or 35 mm dishes and incubated overnight Regular medium was replaced with low-serum medium (0.5% FBS/DMEM) before treatment Note that for Nrf-2 translocation, medium was changed to non-serum medium and incubated for another 12 h Cells were treated with ACS84, L-Dopa or NaHS for 1 – 8 h

2.3 Cell viability assay

Cell viability was measured using the MTT reduction assay as described previously [205] At the end of each treatment, MTT was added to each well at a final concentration of 0.5 mg·mL-1 and the cells were further incubated at 37°C for 4 h

Then, the insoluble formazan was dissolved in dimethyl sulphoxide (DMSO) Colorimetric determination of MTT reduction was measured at 570 nm with a reference wavelength of 630 nm

2.4 Lactate dehydrogenase (LDH) release assay

At the end of treatment, cell culture medium was collected and briefly centrifuged The supernatants were transferred into wells in 96-well plates Equal amounts of lactate dehydrogenase assay substrate, enzyme and dye solution were mixed A half volume of the above mixture was added to one volume of medium supernatant After incubating at room temperature for 30 min, the reaction was terminated by the addition of 1/10 volume of 1N HCl to each well Spectrophotometrical absorbance was measured at a wavelength of 490 nm and reference wavelength of 690 nm

2.5 Reactive oxygen species (ROS) measurement

Formation of reactive oxygen species (ROS) was evaluated using non-fluorescent dye 2’, 7’- dichlorofluorescindiacetate (DCFH-DA), which freely penetrates cells and yields the highly fluorescent product dichlorofluorescein (DCF) by ROS oxidation Following ACS84, L-Dopa or NaHS treatment, cells were rinsed with PBS solution and incubated with Hank's Buffered Salt Solution (HBSS) containing DCFH-DA dye (10 μM final concentration) 30 minutes in the dark 6-OHDA was added then and fluorescence was read immediately for 1 h, at an excitation wavelength (Ex) of 490

nm and an emission wavelength (Em) of 520 nm

2.6 Superoxide Dismutase (SOD) activity Determination

SOD activity was measured in cells using the Cayman Chemical Superoxide Dismutase Assay Kit (Cayman Chemicals, Inc, Ann Arbor, MI) Briefly, cells were sonicated in 20 mM HEPES buffer, pH 7.2, containing 1 mM EGTA, 210