The neuroprotective effect of BDNF on oxidative DNA damage in rat cortical neurons : evaluation by comet assay

UNIVERSIDADE DE LISBOA

The Neuroprotective Effect of BDNF on Oxidative DNA Damage in

Rat Cortical Neurons: Evaluation by Comet Assay

Rúben Balau Delgado Gonçalves Mestrado em Neurociências

UNIVERSIDADE DE LISBOA

The Neuroprotective Effect of BDNF on Oxidative DNA Damage in

Rat Cortical Neurons: Evaluation by Comet Assay

Rúben Balau Delgado Gonçalves

Orientador: Professora Doutora Ana Sebastião, FMUL Co-Orientador: Professor Doutor João Ferreira, FMUL

Todas as afirmações contidas neste trabalho são da exclusiva responsabilidade do candidato, não cabendo à Faculdade de Medicina da Universidade de Lisboa qualquer responsabilidade.

Mestrado em Neurociências Lisboa, 2012

Index

Abstract ... i

Resumo ... ii

Abbreviations List ... iii

Figures Index ...viii

Agradecimentos ... xi

1. Introduction... 1

1.1 - BDNF and Other Neurotrophins ... 1

1.2 - Oxidative Stress, ROS and DNA Damage ... 10

1.3 - Antioxidant Defences ... 15

1.4 - Ageing ... 17

1.5 - Neurodegenerative Disorders and Oxidative Stress in Neurons ... 19

1.6 - Alzheimer’s Disease and Aβ Peptide ... 20

2. Objectives ... 25

3. Methodology ... 26

3.1 - Primary Cell Culture ... 26

3.1.1 - Culture Maintenance ... 27

3.2 - Immunocytochemistry ... 27

3.3 - Cell Viability ... 28

3.4 - Comet Assay ... 29

3.4.2 Experimental Design... 33

3.4.3 Technique Description ... 37

3.4.3.1 Electrophoretic Force (V/cm) Calculation ... 38

3.4.3.2 Standard Cells ... 39

4. Results and Discussion ... 40

4.1 - The Oxidative Comet Assay: Implementation and Optimization ... 40

4.2 - Cell Culture Characterization: Are the cultured neurons viable at the times of experiment? What percentage of glial cells is present in culture?... 49

4.3 - Do neurons accumulate DNA damage, either structural or oxidative, as they mature in culture? ... 55

4.4 - Does Aβ25-35 peptide induce structural and/or oxidative DNA damage at sub-lethal concentrations? ... 59

4.5 - Does BDNF protect against H2O2 induced DNA damage? ... 63

5. Conclusions ... 68

6. Future Work... 69

7. References ... 71

8. Annexes ... 76

Annex I – Comet Assay Protocol ... 76

Abstract

Alzheimer’s disease is an increasing prevalent disease nowadays although its underlying mechanisms are not fully understood. In addition to other factors, it is known that the presence of amyloid-β peptide in the senile plaques of patients with Alzheimer disease, oxidative stress and neurotrophic deprivation are involved in the onset of the disease.

In vitro studies refer that incubation with BDNF prevented Aβ-induced apoptosis and also increased the activity of antioxidant enzymes. However, most studies only refer oxidative stress and the loss of cell viability and do not focus in oxidative damage particularly to the nuclear DNA.

In this study, the structural and oxidative DNA damage of primarily cultured rat cortical neurons was measured by Comet Assay. The assay was performed after incubation with the Aβ25-35 peptide at sub-lethal concentrations for a 24 hour period and after a 10 minutes exposure to 10 µM H2O2 in the presence or absence of BDNF. Although no DNA damage was observed after incubation with the Aβ25-35 peptide, the H2O2 incubation induced both structural and oxidative damage to the nuclear DNA, while a 48 hour pre-incubation with BDNF decreased the induced oxidative damage. This study thus provided a useful insight into BDNF’s protective effects on DNA damage. However, further studies are necessary to evaluate if the Aβ25-35 peptide induces DNA damage at higher concentrations.

Keywords: Cortical neurons, BDNF, Aβ25-35 peptide, H2O2, Comet Assay, DNA damage.

Resumo

A doença de Alzheimer é cada vez mais prevalente actualmente devido ao envelhecimento da população mundial, no entanto os mecanismos biológicos que levam ao seu aparecimento ainda não são conhecidos. Entre outros factores, sabe-se que a presabe-sença de agregados do péptido β-amilóide assim como o stress oxidativo e deprivação neurotrófica estão envolvidos no desenvolvimento da doença. Estudos in vitro demonstraram que a incubação com BDNF evita a apoptose de neurónios induzida pelo péptido β-amilóide assim como aumenta a actividade de enzimas antioxidantes. No entanto, a maioria dos estudos foca-se na presença de

stress oxidativo e na perda de viabilidade celular, não estudando os níveis de dano

ao DNA nuclear, especialmente de natureza oxidativa.

Neste estudo, o dano estrutural e oxidativo ao DNA nuclear de culturas primárias de neurónios corticais de rato foi medido utilizando a técnica de Comet Assay. A técnica foi realizada após incubação com o péptido Aβ25-35 em concentrações sub-letais por um período de 24 horas e após exposição a 10 µM de H2O2 na presença ou ausência de BDNF.

Apesar de nenhum dano ao DNA ter sido observado após incubação com o péptido Aβ25-35, a incubação com H2O2 induziu dano estrutural e oxidativo ao DNA nuclear sendo que uma pré-incubação com BDNF reduziu os níveis medidos de dano oxidativo. O presente estudo permitiu então determinar que o BDNF tem um efeito protector em condições de dano ao DNA. No entanto, são necessários mais estudos para avaliar se o péptido Aβ25-35 induz dano ao DNA quando presente em concentrações mais elevadas.

Palavras-Chave: Neurónios Corticais, BDNF, Péptido Aβ25-35, H2O2, Ensaio Comet, Dano ao DNA.

Abbreviations List

8-oxoG - 8-oxoGuanine AD - Alzheimer’s Disease

Akt – also known as PKB (Protein Kinase B) ANOVA - Analysis of Variance

AP sites - Apurinic/Apyrimidinic sites APP - β-Amyloid Precursor Protein araC - Cytosine Arabinoside

ATP – Adenosine Tri-Phosphate Aβ peptide - Amyloid-β peptide Aβ1-40 - Amyloid-β peptide 1-40 Aβ1-42 - Amyloid-β peptide 1-42 Aβ25-35 - Amyloid-β peptide 25-35

BAD - Bcl-2-associated death promoter (BAD) Bax - Bcl-2–associated X protein

Bcl2 - B-cell lymphoma 2

Bcl-xL- B-cell lymphoma-extra large

BDNF - Brain-Derived Neurotrophic Factor BER - Base Excision Repair

CDK2 - Cyclin-dependent kinase 2 CDK3 - Cyclin-dependent kinase 3 DAG - Diacylglycerol

DAPI - 4’, 6’ – Diamidino-2-Phenylindole DIV - Days In Vitro

DMEM - Dulbecco's Modified Eagle Medium DMSO - Dimethyl sulfoxide

DSBs - Double-Strand DNA Breaks EDTA - Ethylenediaminetetraacetic acid EndoIII - Endonuclease III

ERK - Extracellular Signal-Regulated Kinase EtBr - Ethidium bromide

Fapy - Formamidopyrimidines FBS - Fetal Bovine Serum

FISH - Fluorescent in situ hybridization FKHRL-1 - Forkhead Transcription Factor Fpg - Formamidopyrimidine DNA Glycosylase Gab1 - GRB2-associated-binding protein 1

GC-MS - Gas Chromatography-Mass Spectrometry GDP - Guanosine Di-Phosphate

GPx - Glutathione Peroxidase GR - Glutathione Reductase

Grb2 - Growth factor receptor-bound protein 2 GSH - Glutathione

GSSG - Glutathione Disulfide GTP - Guanosine Tri-Phosphate H2O2 – Hydrogen Peroxide

HBSS - Hanks’ Balanced Salt Solution

HEPES - 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HPLC - High-Pressure Liquid Chromatography

IP3 - Inositol 1,4,5-Trisphosphate IκB - Inhibitor of κB

KSR1 - Kinase Suppressor of Ras 1 LMPA - Low Melting Point Agarose

LY-294002 - 2-(4-Morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one (inhibitor of PI-3 kinase)

MAP2 - Microtubule Associated Protein 2 MAPK - Mitogen-Activated Protein Kinase MEK - MAPK-ERK Kinase

Mn-SOD – Manganese Superoxide Dismutase

NADPH – reduced form of Nicotinamide Adenine Dinucleotide Phosphate NF-κB - Nuclear factor kappa-light-chain-enhancer of activated B cells NGF - Nerve Growth Factor

NMA - Normal Melting Agarose NMDA - N-Methyl-D-aspartate NT-3 - Neurotrophin-3 NT-4/5 - Neurotrophin-4/5 NT-6 - Neurotrophin-6 NT-7 - Neurotrophin-7 1 O2 - Singlet Oxygen O2•– - Superoxide Anion

Ogg1 - 8-oxo-Guanine-DNA Glycosylase OH• - Hydroxyl Radical

OHdG - 8-hydroxy-2-deoxyguanosine PBS - Phosphate Buffer Saline

PD - Parkinson’s Disease PDL - Poly-D Lysine

Pen/Strep - Penicillin/Streptomycin PI3-K - Phosphatidylinositol 3 - Kinase PKC - Protein Kinase C

pRb - Retinoblastoma Protein PTB - Phosphotyrosine-Binding

PtdIns (4,5)P2 -Phosphatidylinositol 4,5-Bisphosphate RNAi - Interference RNA

ROS - Reactive Oxygen Species

SCGE - Single Cell Gel Electrophoresis SH2 - Src Homology 2

SOD - Superoxide Dismutase SOS - Son Of Sevenless

SSBs - Single-Strand DNA Breaks TG - Thymine Glycol

TNF - Tumour Necrosis Factor Trk - Tropomyosin-Related Kinase

U-0126 - 1,4-Diamino-2,3-dicyano-1,4-bis(2-aminophenylthio)butadiene (inhibitor of MAPK)

U-73122 - 1-[6-[[(17b)-3-Methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-1H-pyrrole-2,5-dione (inhibitor of Phospholipase C)

FiguresIndex

Figure 1 – Neurotrophin interaction with Trk and p75NTR receptors. ... 2

Figure 2 – Neurotrophin signalling through Trk receptor activation. ... 4

Figure 3 – BDNF induced action in several signalling pathways after exposure to NMDA or H2O2. ... 8

Figure 4 – Hydroxyl radical (OH•) production in mitochondria through Fenton’s reaction. ... 11

Figure 5 – Stochastic causes of ageing. ... 18

Figure 6 – APP cleavage by secretase enzymes. ... 21

Figure 7 – Diagrams explaining the comet assay technique and mechanisms. ... 31

Figure 8 – Representation of the analysis process used by comet analysis software ... 32

Figure 9 – Schematic representation of the comet assay timeline. ... 33

Figure 10 – Schematic representation of the comet assay timeline with a 24 hour exposure to Aβ25-35. ... 34

Figure 11 – Schematic representation of the comet assay timeline with a 24 hour exposure to Aβ25-35 and a 48 hour incubation with BDNF... 35

Figure 12 – Schematic representation of the comet assay timeline with a 48 hour incubation with BDNF and exposed to H202. ... 36

Figure 13 – Comet assay using the THP-1 cell line after incubation with etoposide for one hour at several concentrations. ... 42

Figure 14 – Graphic representations of the comet assay tail length data using the THP-1 cell line after incubation with etoposide for one hour at several

concentrations. ... 43 Figure 15 – Graphic representations of the comet assay percentage of DNA in tail

data using the THP-1 cell line after incubation with etoposide for one hour at several concentrations ... 44 Figure 16 – Image of comets obtained from primary cultured cortical neurons

exposed to H2O2. ... 46 Figure 17 – Graphic representations of the standard cells tail length and percentage

of DNA in tail data mean values. ... 47 Figure 18 – Graphics of the fluorescence and absorbance data obtained at various

DIVs. ... 50 Figure 19 – Overview of a neuronal culture ... 51 Figure 20 – Detail of an astrocyte at the centre of the image. ... 51 Figure 21 – Graphic representation of structural and oxidative DNA damage as the

neuronal culture matures. ... 55 Figure 22 – Graphic representation of structural and oxidative DNA damage as the

neuronal culture matures. ... 56 Figure 23 – Graphic representation of structural and oxidative DNA damage caused

in the presence of the Aβ25-35 and in the presence or absence of the B27

supplement. ... 57 Figure 24 – Graphic representation of structural and oxidative DNA damage caused

in the presence of the Aβ25-35 and in the presence or absence of the B27

supplement. ... 58 Figure 25 – Graphic representation of structural and oxidative DNA damage caused

Figure 26 – Graphic representation of structural and oxidative DNA damage caused

by sub-lethal concentrations of Aβ25-35 ... 62

Figure 27 – Graphic representation of structural and oxidative DNA damage after exposure to H2O2 with or without a pre-incubation with BDNF ... 64

Figure 28 – Graphic representation of structural and oxidative DNA damage after exposure to H2O2 in the presence or absence of BDNF. ... 65

Figure A – Outline of the slide cleaning and coating first steps ... 81

Figure B – Outline of the slide cleaning and coating last steps. ... 81

Figure C – Outline of the agarose embedment and cell lysis steps ... 84

Agradecimentos

Quero agradecer à minha mãe e avó que me apoiaram imenso neste período da minha vida. Por toda a paciência, carinho e amor, sem os quais não me teria sido possível superar mais esta etapa.

Agradeço igualmente ao meu pai, pelos conselhos práticos e pela perspectiva que proporcionou sobre o que verdadeiramente importa na vida profissional.

Estou profundamente grato à minha namorada pelo apoio, carinho e amor. Por sempre me conseguir animar, mesmo nos dias de maior frustração e por me ter aturado e ajudado a concentrar durante o processo de escrita, sem cujo apoio esta dissertação não existiria.

Quero agradecer ao Prof. Doutor Alexandre Ribeiro, por me ter acolhido no laboratório e me ter proporcionado esta oportunidade.

A todos os colegas de laboratório da UFN, e em especial ao Jorge Valadas, André Santos e Pedro Pereira (UBCR), pelo apoio e conselhos práticos, o meu obrigado. Um agradecimento especial à minha orientadora, Prof. Doutora Ana Sebastião, pela liberdade e possibilidade de desenvolver trabalho com uma técnica não previamente utilizada no laboratório, pela acessibilidade e disponibilidade e pelos conselhos e orientação sempre pertinentes.

Agradeço igualmente ao Prof. Doutor João Ferreira, meu co-orientador, pelo rigor e conselhos experimentais aquando da montagem da técnica.

Um agradecimento final ao Prof. Doutor Andrew Collins (U. Oslo) pela enzima FPG e pelas sugestões referentes ao protocolo de comet sem as quais não teria sido possível realizar o ensaio em neurónios.

1. Introduction

1.1 BDNF and Other Neurotrophins

Neurotrophins are a highly conserved family of secreted proteins that regulate brain functions, particularly development, differentiation, survival and plasticity [23,55]. The neurotrophin family includes the nerve growth factor (NGF), neurotrophin-3 (NT-3), neurotrophin-4/5 (NT-4/5), neurotrophin-6 (NT-6), neurotrophin-7 (NT-7) and the brain-derived neurotrophic factor (BDNF) [20, 50]. NT-6 and NT-7 were isolated from fish and have no mammalian or bird orthologous but appear to interact with the same receptors as the related proteins in mammals [44]. The effect of neurotrophins is exerted through the binding to two different classes of cell surface receptors, the tropomyosin-related kinase (Trk) receptor family composed by TrkA, TrkB and TrkC receptors and the neurotrophin receptor p75NTR, which is a member of the tumour necrosis factor (TNF) receptor family. While the p75NTR receptor allows the binding of all mature neurotrophins, the Trk receptor family exhibits ligand specificity to each receptor, namely NGF preferentially binds to TrkA, BDNF and NT-4 to TrkB and NT-3 to TrkC [25]. The p75NTR receptor is known to regulate the Trk receptors response to neurotrophins. In the presence of this receptor, NT-3 has a diminished affinity to TrkA and together with NT-4 is also much less effective at activating TrkB. The TrkA and TrkB receptors ligand specificity is then enhanced by the p75NTR receptor which leads to a higher affinity of these receptors to their primary ligands, NGF and BDNF, respectively [23].

The BDNF protein is processed from a pro-protein to a mature form and it is the mature form that binds specifically to the TrkB receptor leading to cell survival or

differentiation. Pro-BDNF and other proneurotrophins, on the other hand, can bind with high affinity to the p75NTR receptor, activating signalling pathways that often result in apoptosis [44].

As can be seen in Figure 1, the Trk receptors are highly conserved and present an extracellular domain composed by a cluster rich in cysteine residues. These are

Figure 1 – Neurotrophin interaction with Trk and p75NTR receptors. Conserved protein

domains are present in Trk receptors (TrkA, TrkB and TrkC). NGF, NT-3, NT-4 and BDNF neurotrophins have an equal low affinity to the p75NTR receptor, but their pro-protein form binds with high affinity to this receptor. NGF binds preferentially to the TrkA receptor. BDNF and NT-4 specifically bind to the TrkB receptor. NT-3 favourably binds to TrkC but in certain situations, it also has a low affinity to the other Trk receptors. CR1-CR4: cysteine-rich motifs; C1/C2: cysteine-rich clusters; LRR1–3: leucine-rich repeats: Ig1/Ig2: immunoglobulin-like domains [44].

followed by three leucine-rich repeats, another cysteine-rich cluster and two Ig-like domains. The transmembrane region of these receptors contains a terminal tyrosine kinase domain that is surrounded by many tyrosine residues which allows the binding of cytoplasmic adaptors and enzymes in a phosphorylation dependent way [50].

BDNF, as a member of the family of neurotrophins, then influences cell survival, differentiation and death [5] by the activation of its high-affinity receptor, the tyrosine kinase receptor B (TrkB) [20]. So far, three different TrkB receptors have been described with different signalling capabilities: TrkB.FL, a full-length catalytic receptor and two truncated isoforms of the latter, TrkB.T1 and TrkB.T2. The TrkB.FL is a transmembrane tyrosine kinase receptor with a conserved intracellular domain that interacts with several signalling pathways including the Ras/MAPK, PI3K and PLC-γ1 pathways. The TrkB.T1 and TrkB.T2 receptors allow a signal transduction but lack an intracellular tyrosine kinase activity. The gene expression of BDNF and TrkB varies, depending on the development stage, age and cognitive performance. During development, the TrKB.FL protein levels increase and there is a reduction of the expression of this receptor in the hippocampus of aged rats. A decreasing expression of BDNF and its receptors in dendrites could then partially explain memory impairment in aged animals [55].

The activation of the Trk receptors is stimulated by neurotrophin binding that induces dimerization of these membrane proteins and ultimately leads to the transphosphorylation of their tyrosine kinase terminal domain. Of note that this activation only occurs if the neurotrophin bound to the receptor is in its mature form and not in the pro-form. The several tyrosine residues present in the terminal domain

of the Trk receptors are also subject to additional phosphorylation by the tyrosine kinase receptor domain. The phosphorylated tyrosines then allow the recruitment of several adaptor proteins and enzymes that propagate the signal transduction. The proteins that bind to these phosphotyrosines and surrounding amino acid residues usually contain phosphotyrosine-binding (PTB) or Src homology 2 (SH2) domains [50].

One of the major Trk receptor activating pathways is the Ras pathway which is required for normal differentiation of neurons as well as promoting neuronal survival. The adaptor protein Shc is recruited by a phosphotyrosine residue in the Trk receptor by interacting with its PTB domain. The Trk receptor then mediates the

Figure 2 – Neurotrophin signalling through Trk receptor activation. The main activated

signalling pathways are depicted. Neurotrophin (NT) binding to the Trk receptor leads to its dimerization and consequently to autophosphorylation. The Ras/Raf/MEK pathway promotes the expression of prodifferentiation genes whereas the PI3-K/Akt pathway leads to the expression of prosurvival genes. Adapted from Skaper (2012) [50].

phosphorylation of a tyrosine residue in Shc which leads to the recruitment of another adaptor protein, Grb2. Grb2 in turn bounds to the Ras exchange factor SOS (Son Of Sevenless) which activates Ras by replacing GDP with GTP [44]. The activated Ras protein then interacts with the serine-threonine kinase Raf which sequentially leads to activation of MEK (MAP kinase- ERK (Extracellular signal-regulated) kinase) that further activates MAPK (mitogen-activated protein kinase). The translocation of activated MAPK to the nucleus leads to the phosphorylation of transcription factors that promote neurons to differentiate [50] (see Figure 2). There is a feedback in the MAPK signalling cascade that attenuates and terminates signal responses through the phosphorylation of intermediates and activation of phosphatases. For example, ERK mediates SOS phosphorylation resulting in the dissociation of the SOS-Grb2 complex [44].

Activation of the PI3-K (phosphatidylinositol 3 - kinase) pathway can occur in a Ras-dependent way or in a Ras-independent way which involves the Gab1 factor. Both activating pathways lead to the promotion of neuronal survival and growth [50]. As seen in Figure 2, the PI3-kinase Ras-independent activation is initiated by the recruitment of Gab1 by the phosphorylated form of Grb2 which subsequently leads to the binding and activation of PI3-K. Activated PI3-K generates P3-phosphorylated phosphoinosides, neuron survival essential lipids that are substract of many phosphoinositide-dependent kinases that together with PI3-K activate the Akt protein kinase. Through phosphorylation, Akt protein then controls the activity of several proteins that are important in promoting neuronal survival. These proteins include BAD and Bcl2-family members that are substrates that directly regulate the caspase

signal transduction pathway by binding to Bcl-xL, thus preventing this factor from inhibiting the pro-apoptotic activity of the Bax protein. Phosphorylation of BAD leads to its sequestering by the 14-3-3 proteins which prevents its pro-apoptotic actions. Akt also regulates the activity of the forkhead transcription factor (FKHRL-1) transcription factor by phosphorylation leading to the cytoplasmic sequestering of its phosphorylated form by 14-3-3 proteins. This, in turn, prevents FKHRL-1 from activating the transcription of several genes, which had the purpose of promoting apoptosis. Furthermore, the degradation of IκB is mediated by phosphorylation induced by the Akt protein, which leads to the liberation of the NF-κB factor that promotes the transcription of genes involved in the sensory promotion of neuronal survival [44].

PLC-γ1 (phospholipase C – γ1) activation results in the activation of two pathways: the Ca2+-regulated pathway and the protein kinase (PKC)-regulated pathway. Both ways lead to the promotion of synaptic plasticity [50]. The phosphorylated tyrosine residues in the Trk receptor recruit PLC-γ, activating it through Trk-mediated phosphorylation which then leads to the hydrolysis of PtdIns (4,5)P2, generating IP3 and DAG (see Figure 2). The presence of IP3 induces the release of Ca2+ from cytoplasmic stores while DAG stimulates the DAG-regulated isoform of PKC. These two signalling molecules can potentially activate many intracellular enzymes, including the majority of PKC isoforms, Ca2+ - calmodulin-dependent protein kinases and other Ca2+ - calmodulin-regulated targets [44].

The activation of the MAPK signalling cascade and increased levels of ERK1/2, has been shown to occur both by pro-survival stimulus such as BDNF and toxic

stimuli including oxidative stress and are, as such, not predictive of the effect upon neuronal survival [15,52]. This effect has, however, been found to be dependent upon ERK1/2 localization, possibly explaining a biphasic effect of MAPK activation following oxidative injury in which ERK1/2 has a protective effect upon the DNA through the glutathione metabolism in early stages and later, following long exposure to oxidative agents, contributes to cellular toxicity and death [15,29]. These findings are consubstantiated by Hetman et al (2007) that reports a BDNF dependent activation of MAPK involving co-localization of ERK1/2 with Kinase suppressor of Ras 1 (KSR1) in membrane vesicles resembling the Golgi. Despite this, ERK1/2 also suffers a initial rapid nuclear translocation though to be necessary for cell cycle re-entry [15,54].

A recent study by N. Boutahar et al (2010) has concluded that BDNF protects cortical neurons from oxidative stress through the Ras/MAPK pathway and proteins E2F1 and Rb (Retinoblastoma protein). As post-mitotic cells, neurons pause in the G0 phase of the cell cycle. To re-enter it, cell cycle proteins must be activated so that neurons can exit G0 and enter G1 phase. However, protein activation may also potentiate a cell death mechanism in which the re-entry in the cell cycle ends in mitotic aberration and ultimately cell death. Proteins Rb and from the E2F protein family are included in the group of factors that regulate the cell cycle progression. In quiescent cells, protein Rb is mainly hypo-phosphorylated and sequesters transcriptions factors from the E2F family, hence inhibiting the cell from re-entering the cell cycle. In contrast, the hyper-phosphorylation of Rb leads to its dissociation from the E2F1 transcription factor, allowing the activation of E2F1-responsive genes, S phase progression and neuronal death [5].

The image in Figure 3 depicts the action of BDNF in signalling pathways after

exposure to NMDA or H2O2. The H2O2 exposure induced ERK activation and cell death which was prevented by BDNF [5]. These results, together with the fact that the PI3-K inhibitor tested in the paper had no effect on ERK activation, lead to the conclusion that the PI3-K pathway was probably not involved in the activation of the Ras/MAPK pathway when the cortical neurons were exposed to H2O2 [5]. Cortical neurons treated with NMDA or H2O2 had an increase in phosphorylation of protein Rb and E2F1 expression and this effect was completely abolished when these neurotoxins were combined with BDNF, leading to the previously reported conclusion that neuroprotection induced by BDNF also involves proteins Rb and the E2F1 transcription factor [5].

Figure 3 – BDNF induced action in several signalling pathways after exposure to NMDA or H2O2. The PI3-K pathway was activated by BDNF. On the other hand, BDNF stopped both the NMDA and H2O2 activated MEK pathway and the activation of proteins Rb and E2F1 induced by both compounds. NMDA induced expression of ER stress was not modified by BDNF [5].

The conclusion that BDNF prevents ERK activation induced by toxic agents while capable to induce ERK activation on its own is of particular interest. One must consider that these kinases can be activated by several extracellular signals transducing them to various effector mechanisms [54]. This includes the ability to mediate suppression of apoptosis by neurotrophins [54] but ERK kinases also play a major role in promoting cell cycle progression and particularly ERK1 and ERK2 are involved in DNA damage response [63]. In cortical neurons, the activation of ERK1/2 by BDNF suppresses apoptosis induced by DNA damaging agents [54]. Although the way DNA damage induces ERK activation is still poorly understood, the general consensus is that MEK mediates ERK activation in DNA damage responses. The use of MEK inhibitors lead to the inhibition of ERK activation induced by several genotoxic agents but whether DNA damage activates MEK through Raf remains to be elucidated. Of note that the impact of the MEK/ERK pathway on checkpoint activation in DNA damage response is dependent of the cell type [63].

BDNF is known to protect neurons from excitotoxicity through a signalling pathway that activates NF-κB, a transcription factor that not only induces the expression of antioxidant enzymes such as Mn-SOD but also induces the expression of Bcl-2, an anti-apoptotic protein [31]. Regarding the increase in the activity of antioxidant enzymes, previous studies have indicated that BDNF increased the GPx (glutathione peroxidase) and GR (glutathione reductase) activities. Neurotrophic factors can protect neurons against oxidative insults and it has indeed been reported the ability of BDNF to protect mesencephalic neurons against

1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) toxicity and hippocampal neurons against the accumulation of ROS (reactive oxygen species) induced by glutamate exposure [32].

1.2 Oxidative Stress, ROS and DNA Damage

Oxidative stress is a state of imbalance in which cells might suffer oxidative damage. The equilibrium between the biochemical processes that lead to the production of reactive oxygen species (ROS) and the ones that dispose of such oxidative compounds is thus impaired [53].

The brain is a relative small organ but its oxygen consumption can be as high as 20% of the body’s total basal oxygen consumption due to its high demand on the ATP molecule. These high levels of oxygen intake then lead to high levels of ROS and as such, neurons are exposed to a more oxidative environment than any other cell type [53].

Reactive oxygen species (ROS) are a group of highly reactive molecules derived from oxygen. These free radicals contain one or more unpaired electrons which allow them to act as oxidizing agents. These molecules are more reactive than their corresponding non-radicals and their presence leads to oxidative aggressions towards any cellular biomolecules [33].

There is a continuous formation of ROS derived from the normal cellular metabolism and as the result of some extracellular processes. Peroxisome metabolism, enzymatic synthesis of nitric oxide, phagocytic leukocytes, heat,

ultraviolet (UV) light, therapeutic drugs, ionizing radiation and redox-cycling compounds are some of the pathways and processes that produce ROS [3]. These comprise a group composed mainly by hydrogen peroxide (H2O2), superoxide anion (O2•–), singlet oxygen (1O2) and hydroxyl radical (OH•) [3,33].

In mitochondria, oxygen is mostly converted to water. However, during the oxidative metabolism almost 2% of this oxygen can be converted into ROS and one of the reactive species produced is the superoxide anion (O2•–). The superoxide anion is relatively weak in aqueous media but can be converted to H2O2 either

spontaneously or catalysed by superoxide dismutase (SOD). The H2O2 can in turn be completely reduced to water or partially reduced to hydroxyl radical (OH•), a very powerful oxidizing agent. Although it is not as reactive as other ROS, H2O2 plays an important part in the cellular oxidative damage and carcinogenesis processes as it is a particularly stable molecule and diffuses easily across biological membranes. Thus, H2O2 allows other cellular compartments to suffer oxidative damage, further increasing the cellular injury, particularly if it is converted to the highly reactive hydroxyl radical (OH•) [33].

Fe2+ + H2O2
Fe3+ + OH- + OH• Equation 1 Fe3+ + O2•–
Fe2+ + O2 Equation 2

Figure 4 – Hydroxyl radical (OH•) production in mitochondria through Fenton’s

reaction. Equation 1 resumes Fenton reaction. Equation 2 resumes the reduction reaction of

During oxidative stress, the mitochondria can thus produce OH• through the Fenton reaction. This reaction is depicted in Figure 4 as Equation 1 where low molecular ions and iron ligands present in mitochondria react with H2O2 to form Fe3+ and OH -ions as well as the hydroxyl radical (OH•). Fe3+ ions can then interact with another ROS, the superoxide anion (O2•–) and reduce iron to Fe2+ ions resulting in molecular oxygen (O2) being produced (Equation 2). The reaction in Equation 2 thus provide for more Fe2+ ions that can be reused through Fenton’s reaction leading to a OH• influx that can cause significant biological damage [56].

Despite the cells tendency to eliminate ROS, it is important to refer that these molecules also play significant roles in the regulation of numerous physiological processes such as platelet adhesion, neurotransmission and vascular permeability. The free radical nitric oxide acts as a second messenger in these processes and it is a highly diffusible molecule derived from L-arginine. In itself, nitric oxide is not highly reactive to macromolecules, but it can react with the superoxide anion (O2•–) and produce the strong oxidant peroxynitrate (ONOO–) [33].

The level of ROS in cells is thus inevitably increased as a consequence of the metabolic stress. If the amount of ROS present in a cell is higher than their ability to dispose of such reactive molecules, oxidative stress can occur and lead to several damaging effects namely modifications of proteins, lipids and DNA which in turn lead to mitochondria and cellular dysfunction [6].

Although ROS can cause damage to any cellular biomolecule, these can usually be replaced and if their turnover is high, damage may not even accumulate. The DNA

molecule however, is the prime information molecule of the cell and nuclear DNA in particular must last throughout the lifetime of the cell. As such, damage in the vital molecule seriously threatens the cell function and must be repaired immediately [14]. Any modification to DNA changes its coding properties, the transcription and replication processes and is thus considered as DNA damage. The lesions that the DNA molecule can endure include apurinic/apyrimidinic (AP) sites, adducts, single-strand DNA breaks (SSBs), double-single-strand DNA breaks (DSBs), crosslinks between proteins and DNA and also mismatches by insertion or deletion of nucleotides [30].

AP sites are frequent DNA lesions in which the DNA bases are freed from the deoxyribose backbone. They can be formed both spontaneously and as intermediates during the process of repairing oxidized, deaminated or alkylated bases. These AP sites are also one of the major types of damage produced by ROS in DNA molecules. Approximately 50,000 to 200,000 AP sites can be found in a mammalian cell induced by ROS action and in fact brain cells comprise most of these AP sites [30].

DNA damage caused by ROS also includes common lesions such as 8-oxoguanine (8-oxoG) and thymine glycol (TG). The levels of these lesions are increased in cells treated with UV light, ionizing irradiation or chemical mutagens that generate oxygen radicals. 8-oxoG lesions adopt a syn conformation and base pairs with adenine leading to transversion mutations that can potentially play a role in the process of ageing and the development of cancer. On the other hand, TG lesions strongly block DNA replication and transcription and have to be efficiently removed and repaired to

maintain a genetic stability. The main DNA repair system that removes these types of lesions is the base excision repair (BER) system [3].

These 8-oxoG lesions are included in the group of the most frequently studied oxidized base lesions that also comprise the 8-hydroxy-2-deoxyguanosine (OHdG) and 5-hydroxyuracil lesions. OHdG lesions can be generated from hydroxyl radicals (OH•) which arise trough the hemolytic cleavage of H2O2 through the Fenton reaction or by the decomposition of peroxynitrite (ONOO-), formed by the combination of superoxide with nitric oxide. OHdG lesions accumulate in specific DNA sequence sites and it is also postulated that in the genome of the human brain, DNA damage occurs preferentially in some promotor regions. As previously mentioned, 8-oxoG and OHdG DNA lesions are mutagenic as they mispair with adenine during the replication and transcription processes. The accumulation of OHdG lesions in preapoptotic neurons during retrograde degeneration and in prenecrotic neurons during ischemic neurodegeneration have also been detected in well-characterized animal models. 5-hydroxyuracil arises from cytosine oxidation leading to unstable cytosine glycol which undergoes deamination. This lesion is also premutagenic as it gives rise to C to T transitions resulting in base transversions [30].

Cells have developed several mechanisms to identify and repair DNA lesions, which include the use of several DNA repair enzymes such as Endonuclease III (EndoIII), formamidopyrimidine DNA glycosylase (Fpg) and 8-oxo-guanine-DNA glycosylase (Ogg1) [13]. Of note that Ogg1 protein in eukaryotes is a DNA glycosylase that removes 8-oxoG and other oxidized guanine bases from nuclear and mitochondrial DNA and that is a function analogue of the Fpg protein in bacteria [39].

The DNA glycosylase Fpg has AP-lyase activity and specifically recognizes several oxidized DNA bases including 8-oxoG, 8-oxo-adenine, formamidopyrimidine (fapy)-guanine, fapy-adenine, 5-hydroxycytosine and 5-hdroxyuracil [64].

1.3 Antioxidant Defences

Antioxidants convert the oxidative compounds to less reactive species and constitute the first line of defence against the ROS induced cellular damage. As the ratio of steady-state concentration of oxidants to antioxidants increases, so does oxidative stress. Cellular response to this type of stress includes DNA repair, cell cycle arrest and apoptosis [33].

Cellular antioxidants include molecules such as glutathione, α-tocopherol (vitamin E), carotenoids and ascorbic acid and antioxidant enzymes such as catalase and glutathione peroxidase. If the levels of reactive oxygen species surpass the capacity of these cellular antioxidants, then the cell is under the effect of oxidative stress [53].

Neurons, as any other cell type, contain specific enzymes to eliminate ROS from the cytoplasm. As mentioned above, this enzyme group includes catalase, glutathione peroxidase (GPx) and superoxide dismutase (SOD) [19].

There are several types of superoxide dismutase (SOD) enzymes depending on the ion they contain. Copper-zinc-SODs are stable enzymes which are present in the cytosol, particularly in lysosomes and the nucleus. Manganese-SODs (MnSODs) are present in yeast and animal mitochondrias while iron-SODs have yet to be discovered in animal cells [10]. SOD is an enzyme that converts superoxide anion

(O2•–) into hydrogen peroxide which, as previously mentioned can readily diffuse through neural membranes [19].

Catalase activity is mostly located in peroxisomes in all cell types but it is particularly concentrated in hepatocytes. The catalase enzyme reduces H2O2 to O2 and H2O, particularly when this reactive species is present in high concentrations. In fact, catalase requires two H2O2 molecules to carry out this reduction reaction and as such, when the H2O2 concentration is low, its enzymatic activity is reduced. For the same reason, catalase activity is augmented as the H2O2 concentration increases and it is very difficult to saturate the catalase enzyme [10].

The GPx enzyme is distributed throughout animal cells and its levels are higher in the kidney, liver and whole blood. Its substract is the reduced form of glutathione (GSH) which functions as an electron donor in the removal of H2O2 from the cell. GPx is considered to be the main peroxide removing enzyme in human cells, presenting a high specificity to GSH but not to H2O2, as it also reacts with other peroxides [10]. Thus, GPx is reduced, leading to the oxidation of GSH to GSSG, its disulfide redox partner, ultimately leading to the removal of H2O2 from the cell. The glutathione reductase (GR) enzyme is then responsible to recycle GSH from its oxidized form (GSSG) by NADPH oxidation. The glutathione conversion GSH/GSSG is commonly used as a biomarker of oxidative stress in biological systems. A decrease in GSH/GSSG ratio could induce mitochondrial membrane structural damage, activity changes in mitochondrial enzymes and membrane potential leading to mitochondrial dysfunction and possibly affecting excitability in neurons [61].

1.4 Ageing

All organisms are subject to the ageing of its cells and tissues and eventually death. The study of ageing is a complex area of research and is the result of various interconnected cellular processes. Many theories have been proposed to explain ageing, however many of them are not mutually exclusive – see Figure 5.

Two major groups exist within the theories of aging: the genetic and the stochastic theories. The first are based on the finite number of population doublings of mitotic cells mediated by the telomeres, while the latter proposes ageing as a consequence of the accumulation of cellular damage owing to environmental exposure [43].

The somatic mutation theory was formulated by Medawar in 1952, soon after the discovery of the DNA structure. This theory proposes that the gradual accumulation of genomic DNA alterations is in the basis for ageing. Mutation in somatic cells that occur before the reproductive age are select against, however, those that occur afterwards are not as much as they do not impair the ability to produce descendants. The rate of somatic mutations would then increase with age correlating to the increase in mis-repaired DNA damage. These mutations could lead to the production of abnormal proteins but there is no direct evidence of an age-dependent protein mis-synthesis, although the rate of protein synthesis is known to be altered with aging [43,59]_.

Although the somatic mutation theory has not been disproven and it is quite robust, there is another theory which can better explain the source of cellular damage in aging. This theory is termed oxidative theory of ageing which refers that aging occurs as a consequence of the action of free radicals, namely ROS, produced by mitochondria during the cellular respiration. ROS would then inflict damage upon various cell organelles and biomolecules leading impaired functions. Most importantly however, they would generate damage in the mitochondrial DNA leading to positive feedback in ROS production [43,59].

Figure 5 – Stochastic causes of ageing. Illustration of various factors involved in cellular

1.5 Neurodegenerative Disorders and Oxidative Stress in Neurons

As neurons consume substantial amounts of oxygen, have a low glutathione content and a high proportion of polyunsaturated fatty acids, they are particularly vulnerable to oxidative stress. Being postmitotic cells also implies that they cannot be replaced if irreversible damage occurs. Accumulation of nuclear DNA damage in neurons has been suggested as one of the main forms of brain aging and neurodegeneration [6].

Neurodegenerative diseases onset and progression is thought to be dependent in genetic factors, oxidative stress and the complex interactions between the individual genetic background and environmental factors. The risks associated with each of these factors are still poorly understood as well as the relation between neuronal death and the diseases clinical expression. Neurodegenerative diseases are characterized by site specific premature and slow death of determined neuronal population. For example, in Alzheimer’s disease (AD), degeneration of neurons occurs mainly in the nucleus basalis while, in Parkinson’s disease (PD), the affected neurons are in the substantia nigra. It is also known that the neuronal populations affected by neurodegeneration are usually synaptically interconnected, although the mechanisms associated with the specificity that leads these neurons to cell death is still not elucidated. Increasing evidence has been suggesting that this specificity may be involved with alterations in the energy status of degenerative neurons, ubiquitin-proteasome system defects, presence of aggregates of abnormal proteins (like β-amyloid and tau proteins in AD), trophic factor deficiency, alterations in cytokine levels and disruption of both ionic gradient and signal transduction processes [19].

1.6 Alzheimer’s Disease and Aβ Peptide

One of the most common neurodegenerative diseases is Alzheimer’s disease (AD), characterized by mild cognitive impairments at onset and multiple cortical function deficiencies in the later stages of the disease. Numerous senile plaques and neurofibrillary tangles accompanied by neuronal death can be observed in the dementia stages. These plaques are mostly composed of amyloid β peptide (Aβ peptide). The Aβ peptide is a fragment of the β-amyloid precursor protein or APP, containing 40 to 42 amino acid residues [2]. APP is a ubiquitously expressed transmembrane glycoprotein and its secreted form, sAPP, is considered as a cell survival signal with extensive influence on neuronal development. sAPP is released in an activity-dependent manner promoting neurite outgrowth and preventing cell death in hippocampal neurons [20].

As seen in Figure 6, the APP can be cleaved by three different secretase enzymes. The α-secretase leads to the release of sAPP by cleaving APP in the centre of the β amyloid domain whereas the β-secretase and γ-secretase action lead to the release of the Aβ peptide. Following its release, the Aβ peptide can thus form aggregates. Mutations in the APP gene inhibit the action of the α-secretase enzyme, consequently allowing the β-secretase cleavage. Mutations in components of the γ-secretase complex, the presenilin-1 and presenelin-2 (PSEN-1 and PSEN-2) genes, also increase cleavage by the γ-secretase enzyme. The Aβ peptide production is thus excessive and it is suggested that the soluble oligomers can impair synaptic function between neurons. On the other hand, the Aβ aggregates may trigger a local inflammatory response [40].

The APP cleavage by secretase enzymes can thus give rise to two main forms of the Aβ peptide: Aβ1-40 which is the more soluble form of the Aβ peptide and Aβ1-42 which is the primary component of senile plaques and more neurotoxic than Aβ1-40[8]. On the other hand, the neurofibrillary tangles are composed of paired helical filaments aggregated, which in turn are formed by the hyper-phosphorylated form of tau protein [38].

The amyloid β peptide is derived from APP through an initial β-secretase cleavage which is followed by an intramembraneous cut by γ-secretase. An early onset of AD has been associated with an autosomal dominant mutation in APP that results in increased formation of the Aβ peptide. It is also known that, in primary neuronal cultures, β-amyloid induces cell death. The tau protein is a microtubule-associated Figure 6 – APP cleavage by secretase enzymes. The α-secretase cleaves APP in the

middle of the amyloid β domain leading to the release of the normal secreted form of APP. When APP is cleaved by the β-secretase and γ-secretase it leads to the release of the Aβ peptide which can thus form aggregates [40].

protein that contributes to the stability of these cell structures. The hyper-phosphorylation of tau leads to the formation of neurofibrillary tangles, which destabilizes the microtubule network, while its hypo-phosphorylated state has a high microtubule affinity. This change in function of the tau proteins has been linked to the disruption of neuronal function and alterations in the distribution of several organelles including mitochondria [38].

Oxidative stress, inflammation and neurotoxicity are some of the results of Aβ peptide accumulation in neurons. These processes can then lead to the deterioration of the neurotransmission systems and ultimately to apoptosis [2]. In AD, the formation and resulting effects of the senile plaques and neurofibrillary tangles has been associated with oxidative stress and mitochondrial defects. The oxidation of proteins often leads to altered protein solubility and hence, an increase in the formation of protein aggregates. The cytotoxicity induced by multimeric Aβ peptide aggregates can be explained by ROS production of copper ions which seem to be essential, in

vitro, for the Aβ1-42 inhibition of cytochrome C oxidase terminal complex. Tau protein is also influenced by oxidative stress as its phosphorylation and aggregation processes can be modulated by oxidative signals. Moreover, in primary hippocampal neurons, tau phosphorylation is modified by iron-induced oxidative stress. In primary neuronal cultures, oxidation by H2O2 induced the dephosphorylation of the tau protein, but this is thought to have occurred as the concentration of H2O2 used was high and it might have increased the over-oxidation of fatty acids that is known to inhibit the polymerization of tau. In fact, another study using a lower H2O2 concentration reached the conclusion that H2O2 increases tau phosphorylation in

primary cortical neuron cultures, but the way the oxidative stress alters tau phosphorylation is yet to be determined. The balance between kinases and phosphatases does regulate the phosphorylated state of the tau protein and oxidative changes in either of these enzymes are thought to create an imbalance between their antagonist activities leading to tau hyper-phosphorylation [38].

The mechanisms by which the Aβ peptide causes cellular death are not fully understood, however, as previously mentioned, oxidative stress is present in the early onset of the Alzheimer’s disease (AD) pathology [53]. In in vitro studies, both Aβ1-42 and Aβ25-35 peptides are known to play a role in oxidative stress leading to the oxidation of several biomolecules [9,58]. The methionine in position 35 of the wild type Aβ1-42 is necessary for the oxidative stress and neurotoxic proprieties of this peptide and it is known to induce the reduction of copper ions (Cu2+) leading to the production of H2O2 in the absence of cells [24]. The artificial peptide Aβ25-35 has similar oxidative and neurotoxic proprieties but the mechanisms by which it acts are different. Aβ25-35 induces cell death much faster and its effects are more pronounced. Although this peptide also displays the methionine 35 residue, its presence in the C-terminal alters its proprieties and leads to no effect on the copper ion Cu2+, suggesting that the Aβ 25-35 induces oxidative damage by other mechanisms. Also, the presence of this residue in the peptide is necessary for its oxidative proprieties as the truncated Aβ25-34 peptide, which lacks this terminal methionine, is known to be toxic and non-oxidative [8,9].

A recent study by Arancibia et al has determined that BDNF induces a neuroprotective effect against Aβ peptide toxicity, both in vivo and in vitro. The

collected data reported a dose-dependent toxic effect of β amyloid in cortical neurons. Aβ25-35 was more toxic for cell survival than Aβ1-42 and although the protection effect of BDNF had a dose-response result, it differed depending on the peptide used for β amyloid toxicity. BDNF completely reversed the toxic action of Aβ1-42 on neuronal survival whereas with Aβ25-35 this reversion was only partial [2]. As previously mentioned, this neurotrophin can act as an antioxidant factor by increasing the level of activity of several antioxidant enzymes [31]. As the neuronal loss after amyloid β exposure, both in vivo and in vitro, can be explained by oxidative damage caused by Aβ25-35, and that oxidative stress is present at the onset of AD, the antioxidant defence provided by BDNF could explain its protective effect in Aβtoxicity [55]

.

It has also been determined that BDNF and/or its TrkB receptor are impaired in aging and in AD patients. Moreover, the production and signalling of BDNF in vivo and in vitro is threatened by Aβ peptide which could ultimately lead to neurodegeneration. On the other hand, neurons can be rescued from death by the exogenous addition of this neurotrophin, as BDNF prevents Aβ-induced neurodegeneration, both in vivo and in vitro [55].

2. Objectives

This study aims to determine if BDNF can protect neurons from structural and oxidative DNA damage. Damage was induced by the Aβ25-35 peptide or by H2O2, a known DNA damaging agent. As such, the following questions were posed:

o Do neurons accumulate DNA damage as they mature in culture?

o Does Aβ peptide induce structural and/or oxidative DNA damage at

sub-lethal concentrations?

o If Aβ_25-35 has an effect on DNA integrity, can BDNF protect neurons

from this oxidative and/or structural damage?

o If Aβ_25-35 has no effect, does BDNF protect neurons against the action

3. Methodology

3.1 – Primary Cell Culture

Pregnant Sprague-Dawley rats, provided by a commercial company (Harlan Interfauna Iberia, Barcelona, Spain), were anesthetized in a halothane chamber and sacrificed by decapitation in accordance to the Portuguese animal handling laws. Under sterile conditions, the 18 days of gestation embryos (E18) were removed from the uterus and their brains harvested in cold dissection media (HBSS – Hanks’ Balanced Salt Solution – Ca2+ and Mg2+ free supplemented with 0.37% glucose). After brain dissection and white matter removal, the cerebral cortex was placed in 2.7 mL of fresh dissection media. The cortexes were minced, Trypsin was added and the minced tissue was incubated at 37ºC for 15 minutes. Cells were then precipitated by a centrifuging step (188g, 5 minutes), the supernatant removed and cells ressuspended in Neurobasal® (Gibco®) media supplemented with 10% FBS (Fetal Bovine Serum), 25 U/mL Pen/Strep (Penicillin/Streptomycin), 0.5 mM glutamine, 2% B27 supplement and 25 µM Glutamic Acid. The cells were further dissociated by passing the solution through a pipette several times followed by filtration through a 70 µm cell strainer and ressuspension in Neurobasal® Media supplemented with 25 U/mL Pen/Strep, 0.5 mM glutamine, 2% of B27 supplement and 25 µM Glutamic Acid. Cellular density was determined in a haemocytometer and neurons were plated at 3x105 cells / mL (cellular density of 7x104 cells / cm2) in 6 or 24 well plates previously coated with PDL (poly-D lysine). The cultures were maintained in a 37ºC incubator with a 5% CO2 atmosphere.

3.1.1 – Culture Maintenance

The media used for cultured neurons was supplemented with B27, Pen/Strep and Glutamine, but without Glutamic Acid. For the maintenance of cells in culture, the media was changed twice a week, removing half of the culture media and substituting it with fresh media supplemented with B27. Two days before the experiment, all culture media is removed and fresh media is added in the absence of B27 supplement. All cultures were maintained in a 37ºC incubator with a 5% CO2 atmosphere.

3.2 – Immunocytochemistry

For the immunocytochemistry experiments, 24 well plates with coverslips, previously coated in PDL, were used. At the time of the experiment, the culture media was removed and the coverslips transferred to another 24 well plate. The coverslips were then washed twice with PBS (Phosphate Buffer Saline) for 5 minutes to completely remove Neurobasal® media. Fixation of the DIV 5 and DIV 9 cultures followed, being performed with a 4% paraformaldehyde solution in PBS for 15 minutes at room temperature. The fixation solution was completely removed by aspiration and a solution containing 0.05% Triton X-100 in PBS added for 5 minutes at room temperature. This permeating solution was then removed and substituted by the blocking solution containing 0.25% gelatine in PBS for 40 minutes at room temperature. After the removal, the coverslips were washed twice with PBS for 5 minutes. This was followed by 1 hour incubation with the monoclonal primary antibody solution containing 0.05% Tween 20 and 0.1% gelatine in PBS. The primary

antibodies used were the anti-MAP2 (Microtubule Associated Protein 2) mouse antibody (1:200) and the anti-GFAP (Glial Fibrillary Acidic Protein) rabbit antibody (1:100). Coverslips were then washed twice with a PBS solution containing 0.05% Tween 20 for 5 minutes, following which, incubation with the secondary antibody solution for 1h, containing 0.05% Tween 20 and 0.1% gelatine in PBS was performed. The anti-mouse secondary antibody (1:500) was conjugated with the Alexa 488 fluorescent dye (green) and the anti-rabbit secondary antibody (1:500) to the Alexa 568 fluorescent dye (red). Coverslips were again washed with a PBS solution containing 0.05% Tween 20 and incubated with DAPI (4’, 6’ – diamidino-2-phenylindole) (1:15000) for 5 minutes. This is followed by a washing step using the 0.05% Tween 20 in PBS solution and a final washing step using PBS. The coverslips were then mounted in Mowiol® (Sigma®) before observation in the inverted fluorescence microscope Axiovert® 200 (Zeiss®) and images captured using the AxioCamMR3® (Zeiss®) digital camera.

3.3 – Cell Viability

To determine cell viability, a commercial kit named Alamar Blue® (Invitrogen®) was utilized and the protocol was followed as instructed in the product’s manual. The DIV 1, DIV 5, DIV 9 and DIV 10 cultures were prepared in 24 well plates. The cells were incubated for 1 hour with a 10% Alamar Blue® solution directly dissolved in the culture media. All plates were automatically analysed in the Infinite M200® (Tecan®) plate reader, using a gain of 85 and 25 flashes. Fluorescence values were obtained using an excitation wavelength of 550 nm and an emission wavelength of 590 nm.

The absorbance values were obtained at 570 nm using the 600 nm wavelength as reference.

3.4 – Comet Assay

3.4.1 – General Considerations

The comet assay, also called single cell gel electrophoresis (SCGE) was described for the first time in 1984 by Ostling and Johanson. The basis of the assay, however, relies on work published by Cook et al in 1976 as well as the work from Rydberg and Johanson in 1978. In 1988 Singh et al described a version of the assay with increased damage detection range by performing the electrophoresis in highly alkaline conditions (pH>13) [16]. This version, the alkaline comet assay, quickly gained popularity as a cheap, adaptable and highly sensitive technique to assess DNA damage [42].

The comet assay is a versatile technique regarding the source of the biological material. In fact, almost all types of cells can be used including cell lines, cultured cells, frozen samples, tissue samples, plant cells and even sperm and prokaryotic cells [42]. This technique is also very versatile regarding the type of DNA damage measured as different versions of the assay can be used to observe different kinds of lesions to the nuclear DNA. The original comet assay, with alterations from Olive et al in 1990, labelled neutral comet assay, is able to measure both DSB and SSB, while the alkaline version also measures AP sites in the genome. The comet assay can still be further extended to measure even more types of DNA damage which is done by incubating the DNA with lesion specific enzymes such as FPG, endonuclease III, T4

endonuclease V and Alk A. Other variants of the assay exist, such as the Bromodeoxyuridine labelling version to detect replicating DNA, the version using DNA synthesis inhibitors to detect intermediates of DNA repair and even Fluorescent

in situ hybridization (FISH) Comet to detect gene specific damage and repair [16,41]. A popular version is the oxidative comet assay, performed under alkaline conditions (pH>13), in which the experimental conditions to be tested are performed in duplicates with one of the two being incubated with FPG. Since this enzyme recognizes and removes 8-oxoG lesions as well as ring-opened purines also called Fapy (Formamidopyrimidines), this assay allows for the simultaneous determination of structural and oxidative damage to the nuclear DNA [16].

The assay is also a very sensitive and cheap technique to perform, being able to observe the DNA damage directly without having to rely on expensive antibodies. It also obviates the need for special equipment such as for gas chromatography-mass spectrometry (GC-MS) or high-pressure liquid chromatography (HPLC) while, in addition, being free from these techniques oxidation artefacts which may result in inaccurate measurements especially when measuring basal levels of oxidative DNA damage [16]. As the assay is not standardized, its sensitivity and range differs between different groups, however, Collins et al (2008) estimates this to be from 0.06 to 3 breaks per 109 Daltons of cellular DNA, corresponding to approximately 200 to 10000 breaks per diploid mammalian cell [18].

What experimental procedures comprise the alkaline comet assay however, and by which mechanisms can the DNA damage be measured? The alkaline comet assay entails the DNA electrophoresis of a population of cells embedded in a thin layer of LMPA (Low Melting Point Agarose), which is liquid at 37ºC. This embedded

population of cells is subjected to lysis with a non-ionic detergent and high salt concentrations, removing membranes, cytoplasm, nucleoplasm, disrupting the A

B

Figure 7 – Diagrams explaining the comet assay technique and mechanisms. A –

Representation of main steps of the comet assay protocol [35]. B – Representation of the comet assay comet formation process [48].

nucleosomes and removing almost all histones. What is left is the nucleoid, a scaffold of proteins and supercoiled DNA as depicted in Figure 7 [16].

The supercoiled DNA may optionally be incubated with enzymes, such as FPG, which will recognize and remove specific DNA damage, introducing a SSB. The comet slides are then incubated in an alkaline solution causing DNA unwinding which allows the relaxation of the supercoiled DNA. The electrophoresis then creates the characteristic comet shapes as the more damaged the DNA is, greater DNA migration into the comet tail will be observed. After staining with a fluorescent dye, such as ethidium bromide, comet images are acquired and image analysis is performed, subtracting the background fluorescence and analysing multiple parameters such as the tail length and the percentage of DNA in tail as depicted in Figure 8[16,27].

Figure 8 – Representation of the analysis process used by comet analysis software. A

– Image of an acquired comet. B – Representation of the analysis performed by the image analysis software on an acquired comet image [27].

3.4.2 – Experimental Design

To answer the questions posed in the present work objectives, the experiments were designed as follows:

o Do neurons accumulate DNA damage as they mature in culture?

To answer this question, the comet assay technique was performed on primary cultured cortical neurons at DIV 5, DIV 9 and DIV 10. As described in Figure 9, at DIV 3, the culture media was removed and replaced by fresh media without B27 supplement for the DIV 5 comet assay. Alternatively, half the culture media was

replaced by culture media with B27 supplement for the remaining experiments. At DIV 7 and DIV 8, the culture media was completely removed and replaced by fresh media without B27 supplement, and the comet assay was performed at DIV 9 and DIV 10, respectively.

Figure 9 – Schematic representation of the comet assay timeline. Culture media

o Does Aβ_25-35 peptide induce structural and/or oxidative DNA damage at

sub-lethal concentrations?

For this second question, the same neuronal cultures were subjected to the comet assay technique at DIV 5 and DIV 9 after a 24 hour exposure to sub-lethal concentrations (0.3 µM, 1 µM and 3 µM) of Aβ25-35. As seen in the scheme of Figure 10, at DIV 3 the culture media was completely removed and replaced by fresh media without B27 supplement. At DIV 4 the cultured neurons were exposed to the

sub-lethal concentrations of Aβ25-35 and at DIV 5 the comet assay was performed. Alternatively, at DIV3, half the culture media was removed and replaced by media with B27 supplement. At DIV 7 the media was completely removed, replaced by fresh media without B27 supplement and at DIV 8 the cultured neurons were also exposed Figure 10 – Schematic representation of the comet assay timeline with a 24 hour exposure to Aβ25-35. Culture media replacement, Aβ25-35 exposure and comet assay experiment dates.

to sub-lethal concentrations of Aβ25-35. The comet assay was then performed at DIV 9.

As this second experiment could have had one of two outcomes, either DNA damage was induced or not by the Aβ25-35 peptide, the following sub-experiments were designed:

o If the Aβ_25-35 has an effect on DNA integrity, can BDNF protect neurons

from this oxidative and/or structural damage?

As illustrated in Figure 11, the experimental conditions were the same as described for the previous experiment with the exception of the incubation with BDNF. The cultured neurons were incubated with this neurotrophin at 20 ng / mL for 48 hours Figure 11 – Schematic representation of the comet assay timeline with a 24 hour

exposure to Aβ25-35 and a 48 hour incubation with BDNF. Culture media replacement,

and the exposure to the Aβ25-35 peptide was performed 24 hours prior to the comet assay.

o If the Aβ_25-35 has no effect, does BDNF protect neurons against the

action of another oxidative agent, such as H2O2?

In the event that the Aβ25-35 induced no oxidative or structural damage, the cultured rat cortical neurons were still incubated with BDNF at 20 ng / mL for 48 hours prior to the comet assay. However, the cells were exposed to H2O2 instead of the Aβ25-35. As seen in Figure 12, at DIV 3 half the culture media was removed and replaced by

culture media with B27 supplement. At DIV 8 the media was completely removed and replaced by fresh media without B27 supplement in the presence or absence of Figure 12 – Schematic representation of the comet assay timeline with a 48 hour incubation with BDNF and exposed to H202. Culture media replacement, BDNF